From mark@klomp.org Tue Oct 02 18:35:05 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15oY81-0000kf-00
	for <classpath@gnu.org>; Tue, 02 Oct 2001 18:35:05 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15oY7F-0003zL-00; Wed, 03 Oct 2001 00:34:17 +0200
Date: Wed, 3 Oct 2001 00:34:17 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Fixed small bug in InheritableThreadLocal
Message-ID: <20011003003417.A15196@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

when adding InheritableThreadLocal to libgcj I found a small bug.
When you would call get() from within an overridden childValue()
method the heritage set would be modified while in the process of
being iterated in the newChildThread() method. This has been
fixed by the following patch which I have checked in.

    * java/lang/InheritableThreadLocal.java (addToHeritage): check if this
      local is already contained in the heritage before adding it.

Cheers,

Mark

-- classpath/classpath/java/lang/InheritableThreadLocal.java	2000/05/30
22:02:34	1.1
+++ classpath/classpath/java/lang/InheritableThreadLocal.java	2001/10/02
22:27:22	1.2
@@ -122,7 +122,9 @@
 				threadMap.put(currentThread, heritage);
 			}
 		}
-		heritage.add(this);
+		if (!heritage.contains(this)) {
+			heritage.add(this);
+		}
 	}
 	
 	/**

-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Wed Oct 03 16:27:17 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15osbs-0001lS-00
	for <classpath@gnu.org>; Wed, 03 Oct 2001 16:27:17 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15osb5-0004q9-00; Wed, 03 Oct 2001 22:26:27 +0200
Date: Wed, 3 Oct 2001 22:26:26 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Merged SequenceInputStream and StringBufferInputStream with libgcj
Message-ID: <20011003222626.A18486@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

I merged the SequenceInputStream and StringBufferInputStream with libgcj.
The code was almost identical but I took the libgcj version of both and
added the Classpath comments. The main differences were that the Classpath
version of SequenceInputStream used more variables and tried to fill the
whole buffer on a read(). Both InputStream classes in Classpath had some
synchronized methods, the new code has no synchronized methods. The original
Classpath version was not really thread safe and the spec does not say
that the classes have to be thread safe.


2001-10-03  Mark Wielaard  <mark@klomp.org>

	* java/io/SequenceInputStream.java: Merge with libgcj
	* java/io/StringBufferInputStream.java: Ditto

I am checking this in.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Thu Oct 04 20:00:17 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pIPY-0005T1-00
	for <classpath@gnu.org>; Thu, 04 Oct 2001 20:00:16 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15pIOm-0007X8-00; Fri, 05 Oct 2001 01:59:28 +0200
Date: Fri, 5 Oct 2001 01:59:27 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Package and Modifier
Message-ID: <20011005015927.A18242@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

I reindented java/lang/Package.java so a diff with the libgcj version
is more clear about the real differences.

--- java/lang/Package.java	Thu Oct  4 23:48:02 2001
+++ ../gcc-3-1/libjava/java/lang/Package.java	Wed Apr 25 17:45:12 2001
@@ -236,5 +236,6 @@
   {
     // Get the caller's classloader
-    Class c = VMSecurityManager.getClassContext()[1];
+    SecurityManager sm = System.getSecurityManager();
+    Class c = sm.getClassContext()[1];
     ClassLoader cl = c.getClassLoader();
 
@@ -252,5 +253,6 @@
   {
     // Get the caller's classloader
-    Class c = VMSecurityManager.getClassContext()[1];
+    SecurityManager sm = System.getSecurityManager();
+    Class c = sm.getClassContext()[1];
     ClassLoader cl = c.getClassLoader();
 
Note that the libgcj version is clearly wrong since getSecurityManager() can
return null when no SecurityManager is installed. But I am not sure if the
Classpath version really works correctly. Will investigate this.

I also merged java/lang/reflect/Modifier.java with the libgcj version.
No surprises here since the class is so simple. The libgcj version defined
the STRICT constant and isStrict() method. I also took the toString()
implementation from libgcj since it was a little bit more elegant.

2001-10-04  Mark Wielaard  <mark@klomp.org>

	* java/lang/Package.java: Reindent and fix comments
	* java/lang/reflect/Modifier.java: Merge with libgcj

I have committed it to Classpath CVS.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Fri Oct 05 07:39:14 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pTJy-0006mn-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 07:39:14 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15pTJB-0002YY-00; Fri, 05 Oct 2001 13:38:25 +0200
Date: Fri, 5 Oct 2001 13:38:25 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Partial java.util.zip merge with libgcj
Message-ID: <20011005133825.A9770@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

The following classes from java.util.zip were easy to merge with libgcj.

2001-10-05  Mark Wielaard  <mark@klomp.org>

	* java/util/zip/Adler32.java: Merge with libgcj
	* java/util/zip/CRC32.java: Merge with libgcj
	* java/util/zip/CheckedInputStream.java: New file from libgcj
	* java/util/zip/CheckedOutputStream.java: Ditto
	* java/util/zip/Checksum.java: Merge with libgcj
	* java/util/zip/DataFormatException.java: Ditto
	* java/util/zip/ZipException.java: Ditto
	* java/util/zip/Makefile.am: add new classes

We can probably share some more classes which will make it easier to find
and fix bugs. Does anybody know about a (g)zip test suite?

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Fri Oct 05 10:52:29 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pWKy-0000p3-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 10:52:28 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15pWKB-0001At-00; Fri, 05 Oct 2001 16:51:39 +0200
Date: Fri, 5 Oct 2001 16:51:39 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Look I played with DoxyGen
Message-ID: <20011005165139.A4245@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

I played a bit with DoxyGen <http://www.doxygen.org/> to generate
documentation for Classpath. You can see that the program was originally
written for C++ but the results on our java files is already very nice.
It generates (clickable!) class and collaboration diagrams, seems to
correctly parse javadoc comments, generates color coded and hyperlinked
source files, all kinds of indexes. Very very nice.

Take a look at <http://www.klomp.org/mark/classpath/html/>

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From a_rueckert@gmx.net Fri Oct 05 11:26:43 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15pWs7-0003C6-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 11:26:43 -0400
Received: (qmail 4170 invoked by uid 0); 5 Oct 2001 15:26:33 -0000
Received: from krsn-d93363d3.pool.mediaways.net (HELO krsn-d93363d3) (217.51.99.211)
  by mail.gmx.net (mp002-rz3) with SMTP; 5 Oct 2001 15:26:33 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: classpath@gnu.org
To: classpath@gnu.org
Subject: Re: Look I played with DoxyGen
Date: Fri, 5 Oct 2001 17:17:44 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <20011005165139.A4245@klomp.org>
In-Reply-To: <20011005165139.A4245@klomp.org>
MIME-Version: 1.0
Message-Id: <01100517203502.00773@krsn-d93363d3>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Fre, 05 Okt 2001 Mark Wielaard wrote:
> Hi,
> 
> I played a bit with DoxyGen <http://www.doxygen.org/> to generate
> documentation for Classpath. You can see that the program was originally
> written for C++ but the results on our java files is already very nice.
> It generates (clickable!) class and collaboration diagrams, seems to
> correctly parse javadoc comments, generates color coded and hyperlinked
> source files, all kinds of indexes. Very very nice.
> 
> Take a look at <http://www.klomp.org/mark/classpath/html/>

Looks nice, but it's a oneway solution, right? I hoped that ArgoUML would be up
to the task at some point, but I'll still take a while. Java reverse
engineering should work more or less and the new code generation merges model
changes and existing sources, so roundtrip engineering is within reach...

Ciao,
Andreas


From bcox@virtualschool.edu Fri Oct 05 11:48:10 2001
Received: from avocet.mail.pas.earthlink.net ([207.217.121.50])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pXCs-0004xF-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 11:48:10 -0400
Received: from dialup-209.244.86.174.dial1.washington2.level3.net ([209.244.86.174] helo=linux.virtualschool.edu)
	by avocet.mail.pas.earthlink.net with smtp (Exim 3.32 #2)
	id 15pXCp-0002u3-00; Fri, 05 Oct 2001 08:48:08 -0700
Date: Fri, 5 Oct 2001 11:54:28 -0400
From: Brad Cox <bcox@virtualschool.edu>
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Look I played with DoxyGen
Message-Id: <20011005115428.34f78f76.bcox@virtualschool.edu>
In-Reply-To: <20011005165139.A4245@klomp.org>
References: <20011005165139.A4245@klomp.org>
Reply-To: bcox@virtualschool.edu
Organization: Virtual School
X-Mailer: Sylpheed version 0.6.1claws (GTK+ 1.2.10; i686-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Fri, 5 Oct 2001 16:51:39 +0200
Mark Wielaard <mark@klomp.org> wrote:

> I played a bit with DoxyGen <http://www.doxygen.org/> to generate
> documentation for Classpath. Very very nice.

You bet. Doxygen is one nice piece of work. Rock solid, amazingly useful,
great aesthetics, yada yada. I just wish more java projects used it. See
http://virtualschool.edu/jwaa/index.html 
---
Brad J Cox, Ph.D. bcox@virtualschool.edu, 703 361 4751
For industrial age goods there were checks and credit cards
For everything else there is http://virtualschool.edu/mybank
Java Web Application Architecture: http://virtualschool.edu/jwaa


From mark@klomp.org Fri Oct 05 16:49:14 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pbuD-0007pW-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 16:49:14 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15pbtK-0002rD-00; Fri, 05 Oct 2001 22:48:18 +0200
Date: Fri, 5 Oct 2001 22:48:18 +0200
From: Mark Wielaard <mark@klomp.org>
To: Tom Tromey <tromey@redhat.com>
Cc: java@gcc.gnu.org, classpath@gnu.org
Subject: Re: Partial java.util.zip merge with Classpath
Message-ID: <20011005224817.A10585@klomp.org>
References: <20011005134933.A9852@klomp.org> <87r8siyuxz.fsf@creche.redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87r8siyuxz.fsf@creche.redhat.com>
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

[Discussion on java-patches@gcc.gnu.org]
On Fri, Oct 05, 2001 at 09:20:56AM -0600, Tom Tromey wrote:
> >>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:
> 
> Do we want to use all of jazzlib in libgcj or do we just want to
> continue using zlib?
We could merge more classes then I have done now. Whether we want to
replace zlib completely with java code is a seperate question. As Per
Bothner already pointed out zlib is in use with a lot of programs and
Jazzlib is only of interest to the pure java crowed and is fairly new.

But only Deflater and Inflater use native code from zlib. We should
really look at merging the other classes that are already written in
java in both versions. I am CCing this to the Classpath list so people
can comment on how much code can be merged.

> Switching to jazzlib doesn't mean we can delete zlib from the
> repository -- we will still need it for gcj.  So there isn't a
> significant maintenance win there.
But it does mean one less library to link against for an application
compiled with gcj. But I don't know if that is such a great benefit.
 
> Even so I think there is a maintenance benefit to merging with
> Classpath.  For one thing it means more potential bug-fixers for the
> code.
That is the reason I am merging classes. Some differences will remain
but there are lots of classes that don't have to be maintained seperately.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Fri Oct 05 17:03:29 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pc81-0008JE-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 17:03:29 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15pc7A-0002vc-00; Fri, 05 Oct 2001 23:02:36 +0200
Date: Fri, 5 Oct 2001 23:02:36 +0200
From: Mark Wielaard <mark@klomp.org>
To: Tom Tromey <tromey@redhat.com>
Cc: classpath@gnu.org
Subject: VM* classes (Was: Package and Modifier)
Message-ID: <20011005230236.B10585@klomp.org>
References: <20011005015927.A18242@klomp.org> <87eloivzpx.fsf@creche.redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87eloivzpx.fsf@creche.redhat.com>
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Fri, Oct 05, 2001 at 10:06:02AM -0600, Tom Tromey wrote:
> 
> I'd like to see the VM* classes moved out of java.lang and into some
> other package, say gnu.classpath.  In libgcj we've adopted a policy of
> only putting public classes into java.lang -- implementation classes
> go somewhere else.  This seems like a good policy to me.

The reason you want the (package local) VM* classes in java.lang is because
you only want the core java.lang classes to be able to call the VM interface
methods. Moving the VM* classes to another package would mean that the
classes and methods have to be made public so they could be used from
the java.lang package. But that also means that other classes can call
those methods and manipulate the VM directly.

We have to make sure that the VM* classes define a really minimal interface
to the VM but, we cannot make them public and move them to another package
without giving arbitrary classes access to things they should not use
directly. And the classes are not that big, they only define a couple of
native methods.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From bryce@waitaki.otago.ac.nz Fri Oct 05 19:10:30 2001
Received: from 210-55-118-246.adsl.xtra.co.nz ([210.55.118.246] helo=localhost.localdomain)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pe6v-0008IG-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 19:10:30 -0400
Received: from kiteless ([192.168.2.2])
	by localhost.localdomain (8.11.2/8.11.2) with ESMTP id f95MtTh14209;
	Sat, 6 Oct 2001 10:55:29 +1200
Date: Sat, 6 Oct 2001 11:10:14 +1200
Subject: Re: Partial java.util.zip merge with Classpath
Content-Type: text/plain; charset=US-ASCII; format=flowed
Mime-Version: 1.0 (Apple Message framework v472)
Cc: java@gcc.gnu.org, classpath@gnu.org
To: Mark Wielaard <mark@klomp.org>
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
In-Reply-To: <20011005224817.A10585@klomp.org>
Message-Id: <2234603E-B9E6-11D5-94CC-003065F97F7C@waitaki.otago.ac.nz>
Content-Transfer-Encoding: 7bit
X-Mailer: Apple Mail (2.472)
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Saturday, October 6, 2001, at 08:48  AM, Mark Wielaard wrote:

>> Do we want to use all of jazzlib in libgcj or do we just want to
>> continue using zlib?
> We could merge more classes then I have done now. Whether we want to
> replace zlib completely with java code is a seperate question. As Per
> Bothner already pointed out zlib is in use with a lot of programs and
> Jazzlib is only of interest to the pure java crowed and is fairly new.

We ought to do some performance tests. I'd certainly be in favour of 
switching to Jazzlib provided it doesn't turn out to be significantly 
slower.

regards

Bryce.



From tromey@creche.cygnus.com Fri Oct 05 11:52:59 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15pXHX-0005HR-00
	for <classpath@gnu.org>; Fri, 05 Oct 2001 11:52:59 -0400
Received: from creche.cygnus.com (ta0200.peakpeak.com [204.144.244.200])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id JAA05626;
	Fri, 5 Oct 2001 09:52:55 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id KAA12667;
	Fri, 5 Oct 2001 10:06:02 -0600
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Package and Modifier
References: <20011005015927.A18242@klomp.org>
X-Zippy:  YOU'D cry too if it happened to YOU!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 05 Oct 2001 10:06:02 -0600
In-Reply-To: Mark Wielaard's message of "Fri, 5 Oct 2001 01:59:27 +0200"
Message-ID: <87eloivzpx.fsf@creche.redhat.com>
Lines: 23
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:

Mark> I reindented java/lang/Package.java so a diff with the libgcj
Mark> version is more clear about the real differences.
Mark> [ patch omitted ]
Mark> Note that the libgcj version is clearly wrong since
Mark> getSecurityManager() can return null when no SecurityManager is
Mark> installed. But I am not sure if the Classpath version really
Mark> works correctly. Will investigate this.

I imagine the libgcj version was done this way because libgcj doesn't
have a VMSecurityManager class.  Adding one would be fine -- but
there's no easy way to implement getClassContext in libgcj right now,
so this code is simply going to fail anyway.  I don't know exactly how
much work it would be to change libgcj to allow getClassContext to
function, but I think it is significant.

I'd like to see the VM* classes moved out of java.lang and into some
other package, say gnu.classpath.  In libgcj we've adopted a policy of
only putting public classes into java.lang -- implementation classes
go somewhere else.  This seems like a good policy to me.

Tom


From mark@klomp.org Sat Oct 06 18:29:12 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15pzwU-0007D2-00
	for <classpath@gnu.org>; Sat, 06 Oct 2001 18:29:11 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15pzvh-00013c-00; Sun, 07 Oct 2001 00:28:21 +0200
Date: Sun, 7 Oct 2001 00:28:21 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Merging BufferedInputStream with libgcj
Message-ID: <20011007002821.A3676@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

Today I looked at BufferedInputStream from Classpath and libgcj.
A lot of code was almost the same except for the support for mark()/reset().
The libgcj version just uses the normal buffer array to store the bytes
that had to be remembered for the mark(). When the end of the buffer is
reached a new buffer is created which can hold all extra bytes needed for
the mark(). The Classpath version uses an extra byte buffer when more
bytes have to be remembered then can fit into the normal buffer. But it
uses 4 extra fields to do the bookkeeping required and has a lot of extra
code to get this right.

I decided to take most of the code of the libgcj version and added the
javadoc from Classpath in the merged version. This does mean that when a
large marklimit is given the array buffer will stay extra big even after
the mark expires.

If people think this is a problem then we could add one private field to
remember the original size of the buffer in the constructor and reset the
buffer when the mark is expired in the refill() method. I think that is
not necessary because my feeling is that normally one does not set a mark
much larger then the buffer size given to BufferedInputStream and when
someone does set such a large mark then it is likely that the same large
limit is used again later. But I can only rely on my own experience so
please correct me if this is a silly assumption.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sun Oct 07 16:00:52 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15qK6U-0001q0-00
	for <classpath@gnu.org>; Sun, 07 Oct 2001 16:00:51 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15qK5f-0001cm-00; Sun, 07 Oct 2001 21:59:59 +0200
Date: Sun, 7 Oct 2001 21:59:59 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Reimplemented URLEncoder/Decoder
Message-ID: <20011007215959.A5997@klomp.org>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="vkogqOf2sHV7VnPd"
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

--vkogqOf2sHV7VnPd
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

Hi,

While trying to merge the URLEncoder/Decoder with libgcj I noticed that the
latest 1.4 spec defines these classes a little bit different then in past
versions. It is now possible to specify which character encoding to use
to en/decode the hex-encoded unsafe characters. And by default UTF-8 is used.
I have reimplemented both classes to actually take the encoding into account.

Older specs did not clearly specify which character encoding would be used
or implied that only the lower 8 bits would be used when encoding (which
works fine if only 'normal' iso-8859-1 characters are used.

This does mean that the default behaviour is now slightly different from the
old one. But since it was never clearly defined in the past I don't think
that is a real problem. And a friend (hi Erwin) told me that the Sun
implementations also didn't do the same thing in different versions or on
different platforms.

2001-10-07  Mark Wielaard  <mark@klomp.org>

    * java/net/URLDecoder.java: Reimplement to take character encoding
    into account.
    * java/net/URLEncoder.java: Ditto

If nobody objects I will check these in. The new versions are attached.

Even though these version do more things some of my tests were two times
faster when compiled with gcj. I think the use of a StringBuffer in the
URLEncoder made a large difference.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore

--vkogqOf2sHV7VnPd
Content-Type: text/x-java; charset=us-ascii
Content-Disposition: attachment; filename="URLDecoder.java"

/* URLDecoder.java -- Class to decode URL's from encoded form.
   Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc.

This file is part of GNU Classpath.

GNU Classpath is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.
 
GNU Classpath is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
General Public License for more details.

You should have received a copy of the GNU General Public License
along with GNU Classpath; see the file COPYING.  If not, write to the
Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
02111-1307 USA.

As a special exception, if you link this library with other files to
produce an executable, this library does not by itself cause the
resulting executable to be covered by the GNU General Public License.
This exception does not however invalidate any other reasons why the
executable file might be covered by the GNU General Public License. */

package java.net;

import java.io.UnsupportedEncodingException;

 /**
  * This utility class contains one static method that converts a 
  * string encoded in the x-www-form-urlencoded format to the original
  * text.  The x-www-form-urlencoded format 
  * replaces certain disallowed characters with
  * encoded equivalents.  All upper case and lower case letters in the
  * US alphabet remain as is, the space character (' ') is replaced with
  * '+' sign, and all other characters are converted to a "%XX" format
  * where XX is the hexadecimal representation of that character.  Note
  * that since unicode characters are 16 bits, and this method encodes only
  * 8 bits of information, the lower 8 bits of the character are used.
  * <p>
  * This method is very useful for decoding strings sent to CGI scripts
  *
  * Written using on-line Java Platform 1.2 API Specification.
  * Status:  Believed complete and correct.
  *
  * @since 1.2
  *
  * @author Warren Levy <warrenl@cygnus.com>
  * @author Aaron M. Renn (arenn@urbanophile.com) (documentation comments)
  * @author Mark Wielaard (mark@klomp.org)
  */
public class URLDecoder
{
 /**
  * This method translates the passed in string from x-www-form-urlencoded
  * format using the default encoding "UTF-8" to decode the hex encoded
  * unsafe characters.
  *
  * @param s the String to convert
  *
  * @return the converted String
  */
  public static String decode(String s)
  {
    try
      {
	return decode(s, "UTF-8");
      }
    catch (UnsupportedEncodingException uee)
      {
	// Should never happen since UTF-8 encoding should always be supported
	return s;
      }
  }

 /**
  * This method translates the passed in string from x-www-form-urlencoded
  * format using the given character encoding to decode the hex encoded
  * unsafe characters.
  *
  * @param s the String to convert
  * @param encoding the character encoding to use the decode the hex encoded
  *        unsafe characters
  *
  * @return the converted String
  *
  * @since 1.4
  */
  public static String decode(String s, String encoding)
    throws UnsupportedEncodingException
  {
    StringBuffer result = new StringBuffer();

    // First convert all '+' characters to spaces.
    String str = s.replace('+', ' ');
    
    // Then go through the whole string looking for byte encoded characters
    int i;
    int start = 0;
    byte [] bytes = null;
    while ((i = str.indexOf('%', start)) >= 0)
      {
	// Add all non-encoded characters to the result buffer
	result.append(str.substring(start, i));
	start = i;

	// Get all consecutive encoded bytes
	int length = str.length();
	while ((i+2 < length) && (str.charAt(i) == '%'))
	  i += 3;

	// Decode all these bytes
	if ((bytes == null) || (bytes.length < ((i-start)/3)))
	  bytes = new byte[((i-start)/3)];

	int index = 0;
	while (start < i)
	  {
	    String sub = str.substring(start + 1, start + 3);
	    try
	      {
		bytes[index] = (byte)Integer.parseInt(sub, 16);
		index++;
	      }
	    catch (NumberFormatException nfe)
	      {
	      	// Ignore badly encoded char
	      }
	    start += 3;
	  }

	// Add the bytes as characters according to the given encoding
	result.append(new String(bytes, 0, index, encoding));

	// Make sure we skip to just after a % sign
	start = i+1;
      }

    // Add any character left
    if (start < str.length())
      result.append(str.substring(start));

    return result.toString();
  }

} // class URLDecoder

--vkogqOf2sHV7VnPd
Content-Type: text/x-java; charset=us-ascii
Content-Disposition: attachment; filename="URLEncoder.java"

/* URLEncoder.java -- Class to convert strings to a properly encoded URL
   Copyright (C) 1998, 1999, 2001 Free Software Foundation, Inc.

This file is part of GNU Classpath.

GNU Classpath is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.
 
GNU Classpath is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
General Public License for more details.

You should have received a copy of the GNU General Public License
along with GNU Classpath; see the file COPYING.  If not, write to the
Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
02111-1307 USA.

As a special exception, if you link this library with other files to
produce an executable, this library does not by itself cause the
resulting executable to be covered by the GNU General Public License.
This exception does not however invalidate any other reasons why the
executable file might be covered by the GNU General Public License. */

package java.net;

import java.io.UnsupportedEncodingException;

/**
 * Written using on-line Java Platform 1.2/1.4 API Specification, as well
 * as "The Java Class Libraries", 2nd edition (Addison-Wesley, 1998).
 * Status:  Believed complete and correct.
 */

/**
  * This utility class contains static methods that converts a 
  * string into a fully encoded URL string in x-www-form-urlencoded
  * format.  This format replaces certain disallowed characters with
  * encoded equivalents.  All upper case and lower case letters in the
  * US alphabet remain as is, the space character (' ') is replaced with
  * '+' sign, and all other characters are converted to a "%XX" format
  * where XX is the hexadecimal representation of that character in a
  * certain encoding (by default "UTF-8".
  * <p>
  * This method is very useful for encoding strings to be sent to CGI scripts
  *
  * @author Aaron M. Renn (arenn@urbanophile.com)
  * @author Warren Levy <warrenl@cygnus.com>
  * @author Mark Wielaard (mark@klomp.org)
  */
public class URLEncoder
{
  /**
   * This method translates the passed in string into x-www-form-urlencoded
   * format using the standard "UTF-8" character encoding to hex-encode the
   * unsafe characters.
   *
   * @param s The String to convert
   *
   * @return The converted String
   */
  public static String encode(String s)
  {
    try
      {
        return encode(s, "UTF-8");
      }
    catch (UnsupportedEncodingException uee)
      {
        // Should never happen since UTF-8 should always be supported
	return s;
      }
  }

  /**
   * This method translates the passed in string into x-www-form-urlencoded
   * format using the character encoding to hex-encode the unsafe characters.
   *
   * @param s The String to convert
   * @param encoding The encoding to use for unsafe characters
   *
   * @return The converted String
   *
   * @since 1.4
   */
  public static String encode(String s, String encoding)
    throws UnsupportedEncodingException
  {
    StringBuffer result = new StringBuffer();
    int length = s.length();
    int start = 0;
    int i = 0;

    while (true)
    {
      while ( i < length && isSafe(s.charAt(i)) )
	i++;

      // Safe character can just be added
      result.append(s.substring(start, i));

      // Are we done?
      if (i >= length)
	return result.toString();
      else if (s.charAt(i) == ' ')
        {
	  result.append('+');  // Replace space char with plus symbol.
	  i++;
	}
      else
	{
	  // Get all unsafe characters
	  start = i;
	  char c;
	  while ( i < length && (c = s.charAt(i)) != ' ' && !isSafe(c) )
	    i++;

	  // Convert them to %XY encoded strings
	  String unsafe = s.substring(start,i);
	  byte bytes[] = unsafe.getBytes(encoding);
	  for (int j = 0; j < bytes.length; j++)
	    {
	      result.append('%');
	      result.append(Integer.toHexString(((int) bytes[j]) & 0xFF));
	    }
	}
      start = i;
    }
  }

  /**
   * Private static method that returns true if the given char is either
   * a uppercase or lowercase letter from 'a' till 'z', or a digit froim
   * '0' till '9', or one of the characters '-', '_', '.' or '*'. Such
   * 'safe' character don't have to be url encoded.
   */
  private static boolean isSafe(char c)
  {
    return  ((c >= 'a' && c <= 'z') ||
	     (c >= 'A' && c <= 'Z') ||
	     (c >= '0' && c <= '9') ||
	     c == '-' || c == '_' || c == '.' || c == '*');
  }

  /**
   * Private constructor that does nothing. Included to avoid a default
   * public constructor being created by the compiler.
   */
  private URLEncoder() { }

} // class URLEncoder

--vkogqOf2sHV7VnPd--


From tromey@creche.cygnus.com Sun Oct 07 16:40:20 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15qKii-0003Xj-00
	for <classpath@gnu.org>; Sun, 07 Oct 2001 16:40:20 -0400
Received: from creche.cygnus.com (ta0205.peakpeak.com [204.144.244.205])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id OAA13128;
	Sun, 7 Oct 2001 14:40:16 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id OAA07539;
	Sun, 7 Oct 2001 14:53:59 -0600
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: VM* classes (Was: Package and Modifier)
References: <20011005015927.A18242@klomp.org> <87eloivzpx.fsf@creche.redhat.com> <20011005230236.B10585@klomp.org>
X-Zippy:  Is this BOISE??
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 07 Oct 2001 14:53:59 -0600
In-Reply-To: Mark Wielaard's message of "Fri, 5 Oct 2001 23:02:36 +0200"
Message-ID: <87zo73dvdk.fsf@creche.redhat.com>
Lines: 9
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:

Mark> The reason you want the (package local) VM* classes in java.lang
Mark> is because you only want the core java.lang classes to be able
Mark> to call the VM interface methods.

Thanks, that makes sense.

Tom


From tromey@creche.cygnus.com Sun Oct 07 19:30:58 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15qNNq-0003fy-00
	for <classpath@gnu.org>; Sun, 07 Oct 2001 19:30:58 -0400
Received: from creche.cygnus.com (ta0203.peakpeak.com [204.144.244.203])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id RAA27465;
	Sun, 7 Oct 2001 17:30:51 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id RAA08207;
	Sun, 7 Oct 2001 17:44:36 -0600
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Merging BufferedInputStream with libgcj
References: <20011007002821.A3676@klomp.org>
X-Zippy:  Hey, I LIKE that POINT!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 07 Oct 2001 17:44:35 -0600
In-Reply-To: Mark Wielaard's message of "Sun, 7 Oct 2001 00:28:21 +0200"
Message-ID: <873d4vdnh8.fsf@creche.redhat.com>
Lines: 13
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:

Mark> I think that is not necessary because my feeling is that
Mark> normally one does not set a mark much larger then the buffer
Mark> size given to BufferedInputStream and when someone does set such
Mark> a large mark then it is likely that the same large limit is used
Mark> again later.

I think this is fine.  If it turns out to be a problem we can always
change it later.  Anybody really interested could also see what the
JDK does.

Tom


From tromey@creche.cygnus.com Sun Oct 07 19:43:44 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15qNaB-0004Pd-00
	for <classpath@gnu.org>; Sun, 07 Oct 2001 19:43:43 -0400
Received: from creche.cygnus.com (ta0203.peakpeak.com [204.144.244.203])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id RAA28505;
	Sun, 7 Oct 2001 17:43:40 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id RAA08278;
	Sun, 7 Oct 2001 17:57:25 -0600
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Reimplemented URLEncoder/Decoder
References: <20011007215959.A5997@klomp.org>
X-Zippy:  This MUST be a good party -- My RIB CAGE is being painfully
 pressed up against someone's MARTINI!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 07 Oct 2001 17:57:25 -0600
In-Reply-To: Mark Wielaard's message of "Sun, 7 Oct 2001 21:59:59 +0200"
Message-ID: <87zo73c8be.fsf@creche.redhat.com>
Lines: 37
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:

Mark>     * java/net/URLDecoder.java: Reimplement to take character encoding
Mark>     into account.
Mark>     * java/net/URLEncoder.java: Ditto

This is great.  Thanks for doing it.  I do have a couple nits to pick
though.

Mark> Even though these version do more things some of my tests were
Mark> two times faster when compiled with gcj.

Cool.

Mark>   * where XX is the hexadecimal representation of that character.  Note
Mark>   * that since unicode characters are 16 bits, and this method encodes only
Mark>   * 8 bits of information, the lower 8 bits of the character are used.

This part of the comment is now wrong.


Mark> 	    try
Mark> 	      {
Mark> 		bytes[index] = (byte)Integer.parseInt(sub, 16);
Mark> 		index++;
Mark> 	      }
Mark> 	    catch (NumberFormatException nfe)
Mark> 	      {
Mark> 	      	// Ignore badly encoded char
Mark> 	      }

The 1.4 docs say that the implementation can either leave illegal
characters alone or it can throw IllegalArgumentException, but this
piece of code seems to do neither.  Maybe I'm misunderstanding the
docs?  Or maybe we should throw IllegalArgumentException here?

Tom


From cbj@gnu.org Mon Oct 08 09:03:21 2001
Received: from mail7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15qa41-00022y-00
	for <classpath@gnu.org>; Mon, 08 Oct 2001 09:03:21 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Mon, 8 Oct 2001 09:03:20 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f98D3K729472;
	Mon, 8 Oct 2001 09:03:20 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: volunteer status
From: Brian Jones <cbj@gnu.org>
Date: 08 Oct 2001 09:03:20 -0400
Message-ID: <m3lmimqo6f.fsf@lyta.haphazard.org>
Lines: 12
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

All,

I went through all the paperwork that I can see is registered at
gnu.org yesterday and found that serveral people who have signed
paperwork for past and future contributions are not listed as
developers with Savannah.  I'll be contacting each person individually
to make sure this was not a mistake made when the project was created.

Thanks,
Brian
-- 
Brian Jones <cbj@gnu.org>


From mark@klomp.org Mon Oct 08 16:55:05 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15qhQW-0005Pq-00
	for <classpath@gnu.org>; Mon, 08 Oct 2001 16:55:05 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15qhPa-0001WJ-00; Mon, 08 Oct 2001 22:54:06 +0200
Date: Mon, 8 Oct 2001 22:54:06 +0200
From: Mark Wielaard <mark@klomp.org>
To: Tom Tromey <tromey@redhat.com>
Cc: classpath@gnu.org
Subject: Re: Reimplemented URLEncoder/Decoder
Message-ID: <20011008225406.A29884@klomp.org>
References: <20011007215959.A5997@klomp.org> <87zo73c8be.fsf@creche.redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87zo73c8be.fsf@creche.redhat.com>
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Sun, Oct 07, 2001 at 05:57:25PM -0600, Tom Tromey wrote:
> >>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:
> 
> Mark>   * where XX is the hexadecimal representation of that character.  Note
> Mark>   * that since unicode characters are 16 bits, and this method encodes only
> Mark>   * 8 bits of information, the lower 8 bits of the character are used.
> 
> This part of the comment is now wrong.
Yes. I forgot to update the class javadoc. Fixed.

> Mark> 	    try
> Mark> 	      {
> Mark> 		bytes[index] = (byte)Integer.parseInt(sub, 16);
> Mark> 		index++;
> Mark> 	      }
> Mark> 	    catch (NumberFormatException nfe)
> Mark> 	      {
> Mark> 	      	// Ignore badly encoded char
> Mark> 	      }
> 
> The 1.4 docs say that the implementation can either leave illegal
> characters alone or it can throw IllegalArgumentException, but this
> piece of code seems to do neither.  Maybe I'm misunderstanding the
> docs?  Or maybe we should throw IllegalArgumentException here?
I thought that leave alone might also mean skip.
But you do have a point since we already allow 'unsafe' characters into
the decoded result string. So it might make more sense to also add malformed
hex decodings to the result and not just silently skip them.

I have moved the try catch block over the complete while statement so that
a NumberFormatException ends the while loop. Then after I have decoded all
the chars I explicitly check if there is still a % left in which case the
% char is added to te result and decoding resumes at the char just after
that % char.

The javadoc for decode now has the following paragraph added to explain
this:

  * This implementation will decode the string even if it contains
  * unsafe characters (characters that should have been encoded) or if the
  * two characters following a % do not represent a hex encoded byte.
  * In those cases the unsafe character or the % character will be added
  * verbatim to the decoded result.

Thanks for your review.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Mon Oct 08 18:10:09 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15qibB-0000VX-00
	for <classpath@gnu.org>; Mon, 08 Oct 2001 18:10:09 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15qiaM-0005i7-00; Tue, 09 Oct 2001 00:09:18 +0200
Date: Tue, 9 Oct 2001 00:09:18 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: DatagramSocketImpl merge with libgcj
Message-ID: <20011009000918.A20794@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

I merged java.net.DatagramSocketImpl with the libgcj version.
Nothing exciting since it is a purely abstract class.
Added our javadoc to the libgcj version and used their indentation.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Tue Oct 09 17:36:07 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15r4Xn-0004Md-00
	for <classpath@gnu.org>; Tue, 09 Oct 2001 17:36:07 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15r4Xl-0004IQ-00; Tue, 09 Oct 2001 23:36:05 +0200
Date: Tue, 9 Oct 2001 23:36:05 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Re: [Datagram]SocketImpl merge with libgcj
Message-ID: <20011009233604.A16503@klomp.org>
References: <20011009000918.A20794@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20011009000918.A20794@klomp.org>
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Tue, Oct 09, 2001 at 12:09:18AM +0200, Mark Wielaard wrote:
> I merged java.net.DatagramSocketImpl with the libgcj version.
> Nothing exciting since it is a purely abstract class.
> Added our javadoc to the libgcj version and used their indentation.
Did the same for java.net.SocketImpl.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Fri Oct 12 10:59:36 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15s3mg-0007tY-00
	for <classpath@gnu.org>; Fri, 12 Oct 2001 10:59:35 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15s3md-0006ou-00; Fri, 12 Oct 2001 16:59:31 +0200
Date: Fri, 12 Oct 2001 16:59:31 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Partial Double/Float merge with libgcj
Message-ID: <20011012165931.A26185@klomp.org>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="C7zPtVaVf+AK4Oqc"
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

--C7zPtVaVf+AK4Oqc
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

Hi,

Brian Jones had merged java.lang.Double/Float a couple of months ago.
But there were still a couple of comment/indentation differences.
I reindented/reformatted and renamed a method variable here and there.
The actual code changes are minimal:

        * java/lang/Double.java: Partial merge with libgcj
        (serialVersionUID): new private field
        (byteValue): removed, already defined in superclass Number
        (shortValue): likewise
        * java/lang/Float.java: Partial merge with libgcj
        (serialVersionUID): new private field

There are still a few differences between the code and I did not look
at the native code at all. If my patch is accepted by the libgcj people
then the actual diff between the Classpath and libgcj version looks like
the attached diff. Maybe someone who knows a bit about float/doubles in
java can comment on these differences.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore

--C7zPtVaVf+AK4Oqc
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="Double.diff"

--- java/lang/Double.java	Fri Oct 12 16:08:07 2001
+++ ../gcc-3-1/libjava/java/lang/Double.java	Fri Oct 12 16:10:50 2001
@@ -248,7 +248,11 @@
    */
   public static boolean isNaN (double v)
   {
-    return (doubleToLongBits (v) == 0x7ff8000000000000L);
+    long bits = doubleToLongBits (v);
+    long e = bits & 0x7ff0000000000000L;
+    long f = bits & 0x000fffffffffffffL;
+
+    return e == 0x7ff0000000000000L && f != 0L;
   }
 
   /**
@@ -273,7 +277,10 @@
    */
   public static boolean isInfinite (double v)
   {
-    return (v == POSITIVE_INFINITY || v == NEGATIVE_INFINITY);
+    long bits = doubleToLongBits (v);
+    long f = bits & 0x7fffffffffffffffL;
+
+    return f == 0x7ff0000000000000L;
   }
 
   /**
@@ -494,18 +501,12 @@
    * @see #NEGATIVE_INFINITY
    * @since 1.2
    */
-  public static double parseDouble (String s) throws NumberFormatException
-  {
-    String t = s.trim ();
-      return parseDouble0 (t);
-  }
-
-  private static native double parseDouble0 (String s)
+  public native static double parseDouble (String s) 
     throws NumberFormatException;
 
   /**
    * Initialize JNI cache.  This method is called only by the 
    * static initializer when using JNI.
    */
-  private static native void initIDs ();
+  private static void initIDs () { /* Not used in libgcj */ };
 }

--C7zPtVaVf+AK4Oqc
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="Float.diff"

--- java/lang/Float.java	Fri Oct 12 14:37:56 2001
+++ ../gcc-3-1/libjava/java/lang/Float.java	Fri Oct 12 14:33:38 2001
@@ -221,7 +221,12 @@
    */
   public boolean equals (Object obj)
   {
-    return (obj instanceof Float && ((Float) obj).value == value);
+    if (!(obj instanceof Float))
+      return false;
+
+    Float f = (Float) obj;
+
+    return floatToIntBits (value) == floatToIntBits (f.floatValue ());
   }
 
   /**
@@ -359,7 +364,11 @@
    */
   public static boolean isNaN (float v)
   {
-    return (floatToIntBits (v) == 0x7fc00000);
+    int bits = floatToIntBits (v);
+    int e = bits & 0x7f800000;
+    int f = bits & 0x007fffff;
+
+    return e == 0x7f800000 && f != 0;
   }
 
   /**
@@ -384,7 +393,10 @@
    */
   public static boolean isInfinite (float v)
   {
-    return (v == POSITIVE_INFINITY || v == NEGATIVE_INFINITY);
+    int bits = floatToIntBits (v);
+    int f = bits & 0x7fffffff;
+
+    return f == 0x7f800000;
   }
 
   /**

--C7zPtVaVf+AK4Oqc--


From tromey@creche.cygnus.com Fri Oct 12 13:09:17 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15s5oC-0007dQ-00
	for <classpath@gnu.org>; Fri, 12 Oct 2001 13:09:16 -0400
Received: from creche.cygnus.com (ta0195.peakpeak.com [204.144.244.195])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id LAA18751;
	Fri, 12 Oct 2001 11:09:12 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id LAA31238;
	Fri, 12 Oct 2001 11:24:14 -0600
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Partial Double/Float merge with libgcj
References: <20011012165931.A26185@klomp.org>
X-Zippy:  On SECOND thought, maybe I'll heat up some BAKED BEANS and
 watch REGIS PHILBIN..  It's GREAT to be ALIVE!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 12 Oct 2001 11:24:14 -0600
In-Reply-To: Mark Wielaard's message of "Fri, 12 Oct 2001 16:59:31 +0200"
Message-ID: <878zeg22m9.fsf@creche.redhat.com>
Lines: 94
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:

Mark> Maybe someone who knows a bit about float/doubles in java can
Mark> comment on these differences.

I don't know if I qualify, but ...

Mark>    public static boolean isNaN (double v)
Mark>    {
Mark> -    return (doubleToLongBits (v) == 0x7ff8000000000000L);
Mark> +    long bits = doubleToLongBits (v);
Mark> +    long e = bits & 0x7ff0000000000000L;
Mark> +    long f = bits & 0x000fffffffffffffL;
Mark> +
Mark> +    return e == 0x7ff0000000000000L && f != 0L;
Mark>    }

I think this is a remnant of when libgcj's doubleToLongBits would
return the actual raw value instead of turning all NaNs into a single
value.  Both implementations are correct, but the Classpath
implementation is more desirable.

Mark>    public static boolean isInfinite (double v)
Mark>    {
Mark> -    return (v == POSITIVE_INFINITY || v == NEGATIVE_INFINITY);
Mark> +    long bits = doubleToLongBits (v);
Mark> +    long f = bits & 0x7fffffffffffffffL;
Mark> +
Mark> +    return f == 0x7ff0000000000000L;
Mark>    }

These implementations are equivalent.

Mark> -  public static double parseDouble (String s) throws NumberFormatException
Mark> -  {
Mark> -    String t = s.trim ();
Mark> -      return parseDouble0 (t);
Mark> -  }
Mark> -
Mark> -  private static native double parseDouble0 (String s)
Mark> +  public native static double parseDouble (String s) 
Mark>      throws NumberFormatException;

Here the Classpath code seems a bit wasteful.  Can't the native code
do the trimming for us?  Then we avoid allocating a new String.

Mark> -  private static native void initIDs ();
Mark> +  private static void initIDs () { /* Not used in libgcj */ };

Feel free to add a native initIDs() method to libgcj.  It can just do
nothing, or maybe throw InternalError.

Mark>    public boolean equals (Object obj)
Mark>    {
Mark> -    return (obj instanceof Float && ((Float) obj).value == value);
Mark> +    if (!(obj instanceof Float))
Mark> +      return false;
Mark> +
Mark> +    Float f = (Float) obj;
Mark> +
Mark> +    return floatToIntBits (value) == floatToIntBits (f.floatValue ());
Mark>    }

Here I think the libgcj implementation is correct.
`==' on float values will decide 0.0 == -0.0.
Comparing the bits decides 0.0 != -0.0, which is required for
equals().
Similarly, `==' and equals() disagree on whether NaNs are equal.

Mark>    public static boolean isNaN (float v)
Mark>    {
Mark> -    return (floatToIntBits (v) == 0x7fc00000);
Mark> +    int bits = floatToIntBits (v);
Mark> +    int e = bits & 0x7f800000;
Mark> +    int f = bits & 0x007fffff;
Mark> +
Mark> +    return e == 0x7f800000 && f != 0;
Mark>    }

This is the same as the Double case; the Classpath implementation is
better.

Mark>    public static boolean isInfinite (float v)
Mark>    {
Mark> -    return (v == POSITIVE_INFINITY || v == NEGATIVE_INFINITY);
Mark> +    int bits = floatToIntBits (v);
Mark> +    int f = bits & 0x7fffffff;
Mark> +
Mark> +    return f == 0x7f800000;
Mark>    }

These are equivalent.

Tom


From tromey@creche.cygnus.com Fri Oct 12 18:05:02 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15sAQP-0004Zb-00
	for <classpath@gnu.org>; Fri, 12 Oct 2001 18:05:01 -0400
Received: from creche.cygnus.com (ta0195.peakpeak.com [204.144.244.195])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id QAA06422;
	Fri, 12 Oct 2001 16:04:58 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id QAA11749;
	Fri, 12 Oct 2001 16:20:04 -0600
To: Classpath Project <classpath@gnu.org>
Subject: FYI: RMIC
X-Zippy:  ..  My pants just went on a wild rampage through
 a Long Island Bowling Alley!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 12 Oct 2001 16:20:04 -0600
Message-ID: <87r8s8pkkr.fsf@creche.redhat.com>
Lines: 10
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I just checked in some code that lets the rmic program actually invoke
the Java compiler.  It is based on the interface already exposed in
RMIC.java; I didn't read the kaffe implementation to write this one.
I also changed it a little bit.

Right now I've only added a concrete implementation for gcj.  Adding a
new one is very easy; I encourage others to add the compilers they
care about.

Tom


From a_rueckert@gmx.net Sat Oct 13 13:02:24 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15sSB6-00020Z-00
	for <classpath@gnu.org>; Sat, 13 Oct 2001 13:02:24 -0400
Received: (qmail 3522 invoked by uid 0); 13 Oct 2001 17:02:21 -0000
Received: from krsn-d933644f.pool.mediaways.net (HELO krsn-d933644f) (217.51.100.79)
  by mail.gmx.net (mp005-rz3) with SMTP; 13 Oct 2001 17:02:21 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Classpath classes
Date: Sat, 13 Oct 2001 18:31:32 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
MIME-Version: 1.0
Message-Id: <01101318325705.00769@krsn-d9336483>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

Can you use the classpath classes with orp as a zipfile? It seems they only
work unpacked here.

Ciao,
Andreas


From michal.cierniak@intel.com Sat Oct 13 13:26:57 2001
Received: from fmfdns01.fm.intel.com ([132.233.247.10] helo=calliope1.fm.intel.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15sSYr-00034D-00
	for <classpath@gnu.org>; Sat, 13 Oct 2001 13:26:57 -0400
Received: from fmsmsxvs041.fm.intel.com (fmsmsxv041-1.fm.intel.com [132.233.48.109])
	by calliope1.fm.intel.com (8.9.1a+p1/8.9.1/d: relay.m4,v 1.44 2001/10/01 19:10:43 root Exp $) with SMTP id RAA28643
	for <classpath@gnu.org>; Sat, 13 Oct 2001 17:26:56 GMT
Received: from fmsmsx27.fm.intel.com ([132.233.42.27])
 by fmsmsxvs041.fm.intel.com (NAVGW 2.5.1.6) with SMTP id M2001101310282318740
 for <classpath@gnu.org>; Sat, 13 Oct 2001 10:28:23 -0700
Received: by fmsmsx27.fm.intel.com with Internet Mail Service (5.5.2653.19)
	id <4Y5P7D94>; Sat, 13 Oct 2001 10:27:33 -0700
Message-ID: <794826DE8867D411BAB8009027AE9EB90FE77D23@FMSMSX38>
From: "Cierniak, Michal" <michal.cierniak@intel.com>
To: "'Classpath list'" <classpath@gnu.org>
Subject: RE: Classpath classes
Date: Sat, 13 Oct 2001 10:28:30 -0700
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Andreas Rueckert wrote:
> Can you use the classpath classes with orp as a zipfile?
> It seems they only work unpacked here.

You're right.  Currently ORP works only with unzipped Classpath, but we are
working on fixing that problem.

FYI, a mailing list for ORP is archived at http://groups.yahoo.com/group/orp

Michal


From a_rueckert@gmx.net Sun Oct 14 06:10:04 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15siDb-00080e-00
	for <classpath@gnu.org>; Sun, 14 Oct 2001 06:10:03 -0400
Received: (qmail 8948 invoked by uid 0); 14 Oct 2001 10:10:01 -0000
Received: from krsn-d9336479.pool.mediaways.net (HELO krsn-d9336479) (217.51.100.121)
  by mail.gmx.net (mp020-rz3) with SMTP; 14 Oct 2001 10:10:01 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath Project <classpath@gnu.org>
To: Classpath Project <classpath@gnu.org>
Subject: ORP + Classpath: AWT apps
Date: Sun, 14 Oct 2001 12:06:07 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
MIME-Version: 1.0
Message-Id: <01101412073900.01135@hun.bre.de>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

Is it correct, that you cannot exec AWT apps with the current CVS version? It
seems the native Gtk peers are missing?

Ciao,
Andreas


From mark@klomp.org Sun Oct 14 06:42:48 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15sijH-00014Z-00
	for <classpath@gnu.org>; Sun, 14 Oct 2001 06:42:47 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15sijE-0002Qb-00; Sun, 14 Oct 2001 12:42:44 +0200
Date: Sun, 14 Oct 2001 12:42:44 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Classpath home page gives 403 error
Message-ID: <20011014124244.A9323@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

The Classpath home page at <http://www.classpath.org/> which redirects
to <http://www.gnu.org/software/classpath/> does no longer work.
It gives me a 403 error "You don't have permission to access
/software/classpath/ on this server."

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sun Oct 14 08:43:08 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15skbk-0001ET-00
	for <classpath@gnu.org>; Sun, 14 Oct 2001 08:43:08 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15skbh-0002XR-00; Sun, 14 Oct 2001 14:43:05 +0200
Date: Sun, 14 Oct 2001 14:43:05 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Re: Classpath home page works fine
Message-ID: <20011014144305.A9739@klomp.org>
References: <20011014124244.A9323@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20011014124244.A9323@klomp.org>
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Sun, Oct 14, 2001 at 12:42:44PM +0200, Mark Wielaard wrote:
> The Classpath home page at <http://www.classpath.org/> which redirects
> to <http://www.gnu.org/software/classpath/> does no longer work.
> It gives me a 403 error "You don't have permission to access
> /software/classpath/ on this server."

Plese ignore me. Everything is working again.

Strange,

Mark

-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Sun Oct 14 10:27:00 2001
Received: from mail7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15smEG-0005PQ-00
	for <classpath@gnu.org>; Sun, 14 Oct 2001 10:27:00 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sun, 14 Oct 2001 10:26:59 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9EEQwo09017;
	Sun, 14 Oct 2001 10:26:58 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Classpath home page works fine
References: <20011014124244.A9323@klomp.org> <20011014144305.A9739@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 14 Oct 2001 10:26:58 -0400
In-Reply-To: Mark Wielaard's message of "Sun, 14 Oct 2001 14:43:05 +0200"
Message-ID: <m3669ithzh.fsf@lyta.haphazard.org>
Lines: 20
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> Hi,
> 
> On Sun, Oct 14, 2001 at 12:42:44PM +0200, Mark Wielaard wrote:
> > The Classpath home page at <http://www.classpath.org/> which redirects
> > to <http://www.gnu.org/software/classpath/> does no longer work.
> > It gives me a 403 error "You don't have permission to access
> > /software/classpath/ on this server."
> 
> Plese ignore me. Everything is working again.
> 
> Strange,

I think that the update finally went through and the new symbolic
links that were needed had not been created yet.

Brian
-- 
Brian Jones <cbj@gnu.org>


From a_rueckert@gmx.net Sun Oct 14 11:04:38 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15smog-0007nw-00
	for <classpath@gnu.org>; Sun, 14 Oct 2001 11:04:38 -0400
Received: (qmail 17301 invoked by uid 0); 14 Oct 2001 15:04:34 -0000
Received: from krsn-d9336497.pool.mediaways.net (HELO krsn-d9336497) (217.51.100.151)
  by mail.gmx.net (mp007-rz3) with SMTP; 14 Oct 2001 15:04:34 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Missing AWT peer
Date: Sun, 14 Oct 2001 16:59:59 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
MIME-Version: 1.0
Message-Id: <01101417020801.01309@krsn-d9336412>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

I'm trying to run a simple AWT app, but get this error:
=======================================
Dispatcher created
Starting dispatchher run loop
Looping....
No Java handlers found
Uncaught exception:
java.lang.UnsatisfiedLinkError
        at gnu/java/awt/peer/gtk/GtkMainThread.run
        at java/lang/Thread.run        
=======================================
How can I get more info on the problem? The libgtkpeer is in the dynamic linker
path, so that should work.

Ciao,
Andreas


From cbj@gnu.org Sun Oct 14 18:37:45 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15sttB-0002SN-00
	for <classpath@gnu.org>; Sun, 14 Oct 2001 18:37:45 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sun, 14 Oct 2001 18:37:39 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9EMbND01202;
	Sun, 14 Oct 2001 18:37:23 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
References: <01101417020801.01309@krsn-d9336412>
From: Brian Jones <cbj@gnu.org>
Date: 14 Oct 2001 18:37:17 -0400
In-Reply-To: Andreas Rueckert's message of "Sun, 14 Oct 2001 16:59:59 +0200"
Message-ID: <m3d73pvof6.fsf@lyta.haphazard.org>
Lines: 27
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Andreas Rueckert <a_rueckert@gmx.net> writes:

> Hi!
> 
> I'm trying to run a simple AWT app, but get this error:
> =======================================
> Dispatcher created
> Starting dispatchher run loop
> Looping....
> No Java handlers found
> Uncaught exception:
> java.lang.UnsatisfiedLinkError
>         at gnu/java/awt/peer/gtk/GtkMainThread.run
>         at java/lang/Thread.run        
> =======================================
>
> How can I get more info on the problem? The libgtkpeer is in the
> dynamic linker path, so that should work.

I don't see a System.loadLibrary in that class.  This is a "maybe"
problem in a number of the gtk classes.  Obviously something is going
to load it and the rest don't really need to do so but maybe putting
this in all the relevant classes is just safer?

Brian
-- 
Brian Jones <cbj@gnu.org>


From a_rueckert@gmx.net Mon Oct 15 04:33:13 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15t3BR-0002xT-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 04:33:13 -0400
Received: (qmail 5909 invoked by uid 0); 15 Oct 2001 08:33:11 -0000
Received: from krsn-d93364be.pool.mediaways.net (HELO krsn-d93364be) (217.51.100.190)
  by mail.gmx.net (mp010-rz3) with SMTP; 15 Oct 2001 08:33:11 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
Date: Mon, 15 Oct 2001 10:03:02 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <01101417020801.01309@krsn-d9336412> <m3d73pvof6.fsf@lyta.haphazard.org>
In-Reply-To: <m3d73pvof6.fsf@lyta.haphazard.org>
MIME-Version: 1.0
Message-Id: <01101510104100.00934@hun.bre.de>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mon, 15 Okt 2001 Brian Jones wrote:

--<snip>--

> > How can I get more info on the problem? The libgtkpeer is in the
> > dynamic linker path, so that should work.
> 
> I don't see a System.loadLibrary in that class.  This is a "maybe"
> problem in a number of the gtk classes.  Obviously something is going
> to load it and the rest don't really need to do so but maybe putting
> this in all the relevant classes is just safer?

Sorry, I'm not really into Classpath, so I'm a bit slow here. You think, that

  static native void gtkInit();
  native void gtkMain();

in gnu/java/awt/peer/gtk/GtkMainThread.java are the problem. These methods are
part of the Gtk library? 

BTW: At this time, there's no configure switch to compile Classpath for ORP, so
you have to use the japhar or jdk includes for jni? Will you add a with-orp
switch at some point or is there a good reason not to do so?

Ciao,
Andreas


From cbj@gnu.org Mon Oct 15 06:34:52 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15t55A-000483-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 06:34:52 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Mon, 15 Oct 2001 06:34:50 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9FAYiu02417;
	Mon, 15 Oct 2001 06:34:44 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
References: <01101417020801.01309@krsn-d9336412> <m3d73pvof6.fsf@lyta.haphazard.org> <01101510104100.00934@hun.bre.de>
From: Brian Jones <cbj@gnu.org>
Date: 15 Oct 2001 06:34:44 -0400
In-Reply-To: Andreas Rueckert's message of "Mon, 15 Oct 2001 10:03:02 +0200"
Message-ID: <m38zedur7f.fsf@lyta.haphazard.org>
Lines: 28
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Andreas Rueckert <a_rueckert@gmx.net> writes:

> > I don't see a System.loadLibrary in that class.  This is a "maybe"
> > problem in a number of the gtk classes.  Obviously something is going
> > to load it and the rest don't really need to do so but maybe putting
> > this in all the relevant classes is just safer?
> 
> Sorry, I'm not really into Classpath, so I'm a bit slow here. You think, that
> 
>   static native void gtkInit();
>   native void gtkMain();
> 
> in gnu/java/awt/peer/gtk/GtkMainThread.java are the problem. These
> methods are part of the Gtk library?

No, there is a library that should be created during compilation
called libgtkpeer.so.  There is no System.loadLibrary("gtkpeer") in
that class.  Your error could also be because LD_LIBRARY_PATH
(Solaris/Linux) did not point to a directory with libgtkpeer.so in it.

> BTW: At this time, there's no configure switch to compile Classpath
> for ORP, so you have to use the japhar or jdk includes for jni? Will
> you add a with-orp switch at some point or is there a good reason
> not to do so?

Going to try to take out the VM switches completely.
-- 
Brian Jones <cbj@gnu.org>


From a_rueckert@gmx.net Mon Oct 15 07:09:38 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15t5co-0006Jv-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 07:09:38 -0400
Received: (qmail 31825 invoked by uid 0); 15 Oct 2001 11:09:24 -0000
Received: from krsn-d9336147.pool.mediaways.net (HELO krsn-d9336147) (217.51.97.71)
  by mail.gmx.net (mp006-rz3) with SMTP; 15 Oct 2001 11:09:24 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
Date: Mon, 15 Oct 2001 12:49:26 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <01101417020801.01309@krsn-d9336412> <01101510104100.00934@hun.bre.de> <m38zedur7f.fsf@lyta.haphazard.org>
In-Reply-To: <m38zedur7f.fsf@lyta.haphazard.org>
MIME-Version: 1.0
Message-Id: <01101512565101.00769@krsn-d93364a2>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mon, 15 Okt 2001 Brian Jones wrote:
> Andreas Rueckert <a_rueckert@gmx.net> writes:
> 
> > > I don't see a System.loadLibrary in that class.  This is a "maybe"
> > > problem in a number of the gtk classes.  Obviously something is going
> > > to load it and the rest don't really need to do so but maybe putting
> > > this in all the relevant classes is just safer?
> > 
> > Sorry, I'm not really into Classpath, so I'm a bit slow here. You think, that
> > 
> >   static native void gtkInit();
> >   native void gtkMain();
> > 
> > in gnu/java/awt/peer/gtk/GtkMainThread.java are the problem. These
> > methods are part of the Gtk library?
> 
> No, there is a library that should be created during compilation
> called libgtkpeer.so.  There is no System.loadLibrary("gtkpeer") in
> that class.

And you think it should be there? If you have to load the library, shouldn't
the call rather be done in GtkGenericPeer ? GtkMainThread extends that class.

> Your error could also be because LD_LIBRARY_PATH
> (Solaris/Linux) did not point to a directory with libgtkpeer.so in it.

Don't think so...
=======================================
[andreas@hun andreas]$ /sbin/ldconfig -p | grep -i gtk
        libobgtk.so.1 (libc6) => /usr/lib/libobgtk.so.1
        libobgtk.so (libc6) => /usr/lib/libobgtk.so
        libgtkxmhtml.so.1 (libc6) => /usr/lib/libgtkxmhtml.so.1
        libgtkxmhtml.so (libc6) => /usr/lib/libgtkxmhtml.so
        libgtkpeer.so.0 (libc6) => /usr/local/OSS_JDK/classpath/lib/classpath/libgtkpeer.so.0
        libgtkpeer.so (libc6) => /usr/local/OSS_JDK/classpath/lib/classpath/libgtkpeer.so
        libgtk.so.1 (libc6) => /usr/lib/libgtk.so.1
        libgtk.so (libc6) => /usr/lib/libgtk.so
        libgtk-1.2.so.0 (libc6) => /usr/lib/libgtk-1.2.so.0
        libgnorbagtk.so.0 (libc6) => /usr/lib/libgnorbagtk.so.0
        libgnorbagtk.so (libc6) => /usr/lib/libgnorbagtk.so      
=======================================

There's a FrameTest.java file in the testsuite. Does this test work for you?

> > BTW: At this time, there's no configure switch to compile Classpath
> > for ORP, so you have to use the japhar or jdk includes for jni? Will
> > you add a with-orp switch at some point or is there a good reason
> > not to do so?
> 
> Going to try to take out the VM switches completely.

But you have to get the JNI stuff from somewhere and the Classpath tools dir is
of no use at all on my box.

Ciao,
Andreas


From a_rueckert@gmx.net Mon Oct 15 10:03:41 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15t8LF-0000Ue-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 10:03:41 -0400
Received: (qmail 9804 invoked by uid 0); 15 Oct 2001 14:03:38 -0000
Received: from krsn-d933648b.pool.mediaways.net (HELO krsn-d933648b) (217.51.100.139)
  by mail.gmx.net (mp009-rz3) with SMTP; 15 Oct 2001 14:03:38 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
Date: Mon, 15 Oct 2001 15:49:54 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <01101417020801.01309@krsn-d9336412> <01101510104100.00934@hun.bre.de> <m38zedur7f.fsf@lyta.haphazard.org>
In-Reply-To: <m38zedur7f.fsf@lyta.haphazard.org>
MIME-Version: 1.0
Message-Id: <01101515575100.01591@hun.bre.de>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mon, 15 Okt 2001 Brian Jones wrote:

--<snip>--
 
> No, there is a library that should be created during compilation
> called libgtkpeer.so.  There is no System.loadLibrary("gtkpeer") in
> that class.  Your error could also be because LD_LIBRARY_PATH
> (Solaris/Linux) did not point to a directory with libgtkpeer.so in it.

The only System.loadLibrary("gtkpeer") call I could find, is in
gnu/java/awt/peer/gtk/GtkToolkit.java
I also checked the gnu/java/awt/peer/gtk/TestAWT.java
file and it sets the toolkit with
=======================================
  public static void main(String args[])
  {
    if (args.length==0)
      {
        Properties prop = System.getProperties ();
        prop.put ("awt.toolkit", "gnu.java.awt.peer.gtk.GtkToolkit");
      }
    MainWindow f = new MainWindow();
    f.show();
  }             
=======================================
This app fails with the same problems on my box.

Ciao,
Andreas


From jsturm@one-point.com Mon Oct 15 10:21:15 2001
Received: from cc730106-b.taylor1.mi.home.com ([24.13.63.84])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15t8cF-0001l3-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 10:21:15 -0400
Received: from localhost (jsturm@localhost)
	by cc730106-b.taylor1.mi.home.com (8.9.3/8.9.3) with ESMTP id KAA06323
	for <classpath@gnu.org>; Mon, 15 Oct 2001 10:41:51 -0400
X-Authentication-Warning: mars.deadcafe.org: jsturm owned process doing -bs
Date: Mon, 15 Oct 2001 10:41:51 -0400 (EDT)
From: Jeff Sturm <jsturm@one-point.com>
X-Sender: jsturm@mars.deadcafe.org
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
In-Reply-To: <01101512565101.00769@krsn-d93364a2>
Message-ID: <Pine.LNX.4.10.10110151035560.6319-100000@mars.deadcafe.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>


On Mon, 15 Oct 2001, Andreas Rueckert wrote:
> > Your error could also be because LD_LIBRARY_PATH
> > (Solaris/Linux) did not point to a directory with libgtkpeer.so in it.
> 
> Don't think so...

Did you try:

 ldd /usr/local/OSS_JDK/classpath/lib/classpath/libgtkpeer.so

Note that loadLibrary fails if a DSO is not found _or_ any of its
DT_NEEDED libs could not be loaded.  The error message may be misleading
in the latter case.

Jeff



From a_rueckert@gmx.net Mon Oct 15 10:51:55 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15t95v-0003Fu-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 10:51:55 -0400
Received: (qmail 24428 invoked by uid 0); 15 Oct 2001 14:51:51 -0000
Received: from krsn-d93364b5.pool.mediaways.net (HELO krsn-d93364b5) (217.51.100.181)
  by mail.gmx.net (mp006-rz3) with SMTP; 15 Oct 2001 14:51:51 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
Date: Mon, 15 Oct 2001 16:49:45 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <Pine.LNX.4.10.10110151035560.6319-100000@mars.deadcafe.org>
In-Reply-To: <Pine.LNX.4.10.10110151035560.6319-100000@mars.deadcafe.org>
MIME-Version: 1.0
Message-Id: <01101516510405.01591@krsn-d93364b5>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mon, 15 Okt 2001 Jeff Sturm wrote:

--<snip>--

> Did you try:
> 
>  ldd /usr/local/OSS_JDK/classpath/lib/classpath/libgtkpeer.so
> 
> Note that loadLibrary fails if a DSO is not found _or_ any of its
> DT_NEEDED libs could not be loaded.  The error message may be misleading
> in the latter case.

It seems everything is there. But shouldn't the compile of the Gtk peer lib
fail anyway, if one of the libs is missing?

Ciao,
Andreas


From jsturm@one-point.com Mon Oct 15 14:05:45 2001
Received: from cc730106-b.taylor1.mi.home.com ([24.13.63.84])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tC7V-0006k5-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 14:05:45 -0400
Received: from localhost (jsturm@localhost)
	by cc730106-b.taylor1.mi.home.com (8.9.3/8.9.3) with ESMTP id OAA06625
	for <classpath@gnu.org>; Mon, 15 Oct 2001 14:26:54 -0400
X-Authentication-Warning: mars.deadcafe.org: jsturm owned process doing -bs
Date: Mon, 15 Oct 2001 14:26:54 -0400 (EDT)
From: Jeff Sturm <jsturm@one-point.com>
X-Sender: jsturm@mars.deadcafe.org
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
In-Reply-To: <01101516510405.01591@krsn-d93364b5>
Message-ID: <Pine.LNX.4.10.10110151421560.6623-100000@mars.deadcafe.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>


On Mon, 15 Oct 2001, Andreas Rueckert wrote:
> It seems everything is there. But shouldn't the compile of the Gtk peer lib
> fail anyway, if one of the libs is missing?

Not necessarily.  Any paths seen by ld (e.g. with -L) will not be
automatically searched by the loader (assuming an ELF target).

So each lib directory must be one of the "system" lib dirs or mentioned in
LD_LIBRARY_PATH, unless you also link with -rpath.

Jeff



From chris.gray@acunia.com Mon Oct 15 03:55:40 2001
Received: from nat.be.acunia.com ([194.7.211.211] helo=pigeon.smove.be)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15t2b4-0000q3-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 03:55:39 -0400
Received: from localhost.localdomain (IDENT:nobody@loki.leuven.intern.acunia.com [192.168.8.254])
	by pigeon.smove.be (8.9.3/8.9.3) with ESMTP id JAA15150
	for <classpath@gnu.org>; Mon, 15 Oct 2001 09:55:39 +0200
Received: from acunia.com (IDENT:gray@sophocles [127.0.0.1])
	by localhost.localdomain (8.9.3/8.9.3) with ESMTP id JAA06376;
	Mon, 15 Oct 2001 09:55:35 +0200
Message-ID: <3BCA9676.6310815F@acunia.com>
Date: Mon, 15 Oct 2001 09:55:34 +0200
From: Chris Gray <chris.gray@acunia.com>
Organization: ACUNIA NV
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: tromey@redhat.com
CC: Mark Wielaard <mark@klomp.org>, classpath@gnu.org
Subject: Re: Partial Double/Float merge with libgcj
References: <20011012165931.A26185@klomp.org> <878zeg22m9.fsf@creche.redhat.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey wrote:

> >>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:
>
> Mark> Maybe someone who knows a bit about float/doubles in java can
> Mark> comment on these differences.
>
> I don't know if I qualify, but ...
>

I too am not worthy ...

>
> Mark>    public static boolean isNaN (double v)
> Mark>    {
> Mark> -    return (doubleToLongBits (v) == 0x7ff8000000000000L);
> Mark> +    long bits = doubleToLongBits (v);
> Mark> +    long e = bits & 0x7ff0000000000000L;
> Mark> +    long f = bits & 0x000fffffffffffffL;
> Mark> +
> Mark> +    return e == 0x7ff0000000000000L && f != 0L;
> Mark>    }
>
> I think this is a remnant of when libgcj's doubleToLongBits would
> return the actual raw value instead of turning all NaNs into a single
> value.  Both implementations are correct, but the Classpath
> implementation is more desirable.
>

An interesting twist on this: jikes seems to have its own idea of the correct
bit-pattern for NaN (0xffc00000 instead of 0x7fc0000).  This shouldn't matter
if all NaN's are canonicalized to the same value.

Chris Gray
VM Architect, ACUNIA




From tromey@creche.cygnus.com Mon Oct 15 17:03:59 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tEty-0001TK-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 17:03:58 -0400
Received: from creche.cygnus.com (ta0203.peakpeak.com [204.144.244.203])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id PAA05347;
	Mon, 15 Oct 2001 15:03:28 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id PAA14107;
	Mon, 15 Oct 2001 15:19:21 -0600
To: Chris Gray <chris.gray@acunia.com>
Cc: Mark Wielaard <mark@klomp.org>, classpath@gnu.org
Subject: Re: Partial Double/Float merge with libgcj
References: <20011012165931.A26185@klomp.org> <878zeg22m9.fsf@creche.redhat.com> <3BCA9676.6310815F@acunia.com>
X-Zippy:  My mind is making ashtrays in Dayton....
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 15 Oct 2001 15:19:21 -0600
In-Reply-To: Chris Gray's message of "Mon, 15 Oct 2001 09:55:34 +0200"
Message-ID: <87het0vbxi.fsf@creche.redhat.com>
Lines: 19
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Chris" == Chris Gray <chris.gray@acunia.com> writes:

Chris> An interesting twist on this: jikes seems to have its own idea
Chris> of the correct bit-pattern for NaN (0xffc00000 instead of
Chris> 0x7fc0000).  This shouldn't matter if all NaN's are
Chris> canonicalized to the same value.

My understanding is that floatToIntBits is required to return
0x7fc00000 for any NaN, and that Float.NaN must be the same as
Float.intBitsToFloat(0x7fc00000).

This doesn't necessarily imply that there is a Jikes bug though.
It depends on exactly when Jikes uses that particular value.
Float.floatToRawIntBits can return other NaN values, and presumably
the VM could use them as well.  For instace libgcj uses whatever
the underlying hardware decides, and only does the canonicalization
when the real bit pattern might be seen by Java.

Tom


From ebb9@email.byu.edu Mon Oct 15 17:54:43 2001
Received: from femail26.sdc1.sfba.home.com ([24.254.60.16])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tFh5-0003t5-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 17:54:43 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail26.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011015215442.UCFA2746.femail26.sdc1.sfba.home.com@email.byu.edu>
          for <classpath@gnu.org>; Mon, 15 Oct 2001 14:54:42 -0700
Message-ID: <3BCB5A64.69839F1D@email.byu.edu>
Date: Mon, 15 Oct 2001 15:51:32 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>
Subject: Re: Partial Double/Float merge with libgcj
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Shoot, I meant to hit reply-to-all.

Eric Blake wrote:
> 
> Chris Gray wrote:
> >
> >
> > An interesting twist on this: jikes seems to have its own idea of the correct
> > bit-pattern for NaN (0xffc00000 instead of 0x7fc0000).  This shouldn't matter
> > if all NaN's are canonicalized to the same value.
> 
> Which version of jikes? My guess is version 1.13 or sooner, as floating
> point emulation mode was made default in later versions. When doing
> floating point emulation jikes should always spit out 0x7fc00000 for
> Float.NaN, but with native math, it uses whatever the CPU thinks is NaN
> (which for x86 architecture is 0xffc00000).  But you are right, as long
> as Float.floatToLongBits canonicalizes NaN, there should be no problem
> with the current algorithm.
> 
> However, I argue that this is a faster implementation (it avoids a
> native call):
> 
> public static boolean isNaN(double v)
> {
>   return v != v;
> }
> 
> This works since NaN != NaN is the only reflexive inequality comparison
> which returns true.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From tromey@creche.cygnus.com Mon Oct 15 18:04:21 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tFqP-0004Le-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 18:04:21 -0400
Received: from creche.cygnus.com (ta0203.peakpeak.com [204.144.244.203])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id QAA17906;
	Mon, 15 Oct 2001 16:04:17 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id QAA04998;
	Mon, 15 Oct 2001 16:20:12 -0600
To: Eric Blake <ebb9@email.byu.edu>
Cc: Classpath list <classpath@gnu.org>
Subject: Re: Partial Double/Float merge with libgcj
References: <3BCB5A64.69839F1D@email.byu.edu>
X-Zippy:  Are BOTH T.V.S on??
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 15 Oct 2001 16:20:11 -0600
In-Reply-To: Eric Blake's message of "Mon, 15 Oct 2001 15:51:32 -0600"
Message-ID: <877ktwv944.fsf@creche.redhat.com>
Lines: 13
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Eric" == Eric Blake <ebb9@email.byu.edu> writes:

Eric> However, I argue that this is a faster implementation (it avoids a
Eric> native call):
Eric> 
Eric> public static boolean isNaN(double v)
Eric> {
Eric> return v != v;
Eric> }

This seems reasonable to me too.

Tom


From toraneko@kun.ne.jp Mon Oct 15 23:34:25 2001
Received: from ms1.nttdata.co.jp ([163.135.193.232])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tKzo-00042c-00
	for <classpath@gnu.org>; Mon, 15 Oct 2001 23:34:25 -0400
Received: from mail1.nttdata.co.jp ([163.135.10.21])
	by ms1.nttdata.co.jp (8.9.3/3.7W-NTTDATA-TOP-03/21/01) with ESMTP id MAA01829
	for <classpath@gnu.org>; Tue, 16 Oct 2001 12:34:15 +0900 (JST)
Received: from geb.rd.nttdata.co.jp (localhost [127.0.0.1])
	by mail1.nttdata.co.jp (8.9.1/3.7W-NTTDATA-TOP-10/15/01) with ESMTP id MAA29882
	for <classpath@gnu.org>; Tue, 16 Oct 2001 12:32:28 +0900 (JST)
Received: from onuris.rd.nttdata.co.jp (onuris.rd.nttdata.co.jp [10.8.144.22])
	by geb.rd.nttdata.co.jp (8.9.3/3.7W/R8V8) with ESMTP id MAA18593
	for <classpath@gnu.org>; Tue, 16 Oct 2001 12:34:20 +0900 (JST)
Received: from arpa.rd.nttdata.co.jp (thunderlord.rd.nttdata.co.jp [10.8.151.217])
	by onuris.rd.nttdata.co.jp (8.10.1/3.7W/R8V8) with ESMTP id f9G3YYK14000
	for <classpath@gnu.org>; Tue, 16 Oct 2001 12:34:34 +0900 (JST)
Date: Tue, 16 Oct 2001 12:34:17 +0900
Message-ID: <878zecmf5y.wl@arpa.rd.nttdata.co.jp>
From: Takashi Okamoto <toraneko@kun.ne.jp>
To: Classpath Project <classpath@gnu.org>
Subject: Re: FYI: RMIC
In-Reply-To: <87r8s8pkkr.fsf@creche.redhat.com>
References: <87r8s8pkkr.fsf@creche.redhat.com>
User-Agent: Wanderlust/2.7.1 (Too Funky) SEMI/1.14.3 (Ushinoya) FLIM/1.14.2 (Yagi-Nishiguchi) APEL/10.3 Emacs/20.7 (i386-debian-linux-gnu) MULE/4.1 (AOI)
MIME-Version: 1.0 (generated by SEMI 1.14.3 - "Ushinoya")
Content-Type: text/plain; charset=US-ASCII
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

At 12 Oct 2001 16:20:04 -0600,
Tom Tromey wrote:
> I just checked in some code that lets the rmic program actually invoke
> the Java compiler.  It is based on the interface already exposed in
> RMIC.java; I didn't read the kaffe implementation to write this one.
> I also changed it a little bit.

Someone may want to use java compiler for other purpose. Then the
compiler interface should be moved to gnu.tools.javac or like so.


And someone may want following method for Compiler.java:

  /**
   * compile some sources at once.
   */
  public abstract void compile (String sources[]) throws Exception;

  /**
   * compile but output messages to specified stream. 
   */
  public abstract void compile (String sources[], OutputStream out) throws Exception;

  /**
   * set classpath
   */
  public abstract void setClasspath(String path) throws Exception;

regards.
----
Takashi Okamoto







From a_rueckert@gmx.net Tue Oct 16 03:51:45 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tP0q-0008Ki-00
	for <classpath@gnu.org>; Tue, 16 Oct 2001 03:51:44 -0400
Received: (qmail 9642 invoked by uid 0); 16 Oct 2001 07:51:42 -0000
Received: from krsn-d9336383.pool.mediaways.net (HELO krsn-d9336383) (217.51.99.131)
  by mail.gmx.net (mp001-rz3) with SMTP; 16 Oct 2001 07:51:42 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
Date: Tue, 16 Oct 2001 09:48:45 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <Pine.LNX.4.10.10110151421560.6623-100000@mars.deadcafe.org>
In-Reply-To: <Pine.LNX.4.10.10110151421560.6623-100000@mars.deadcafe.org>
MIME-Version: 1.0
Message-Id: <01101609515702.00769@krsn-d9336383>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mon, 15 Okt 2001 Jeff Sturm wrote:
> On Mon, 15 Oct 2001, Andreas Rueckert wrote:
> > It seems everything is there. But shouldn't the compile of the Gtk peer lib
> > fail anyway, if one of the libs is missing?
> 
> Not necessarily.  Any paths seen by ld (e.g. with -L) will not be
> automatically searched by the loader (assuming an ELF target).
> 
> So each lib directory must be one of the "system" lib dirs or mentioned in
> LD_LIBRARY_PATH, unless you also link with -rpath.

All these libs are in /etc/ld.so.conf , so it should work. Is there any way to
get additional info what lib is missing?

I honestly think that Classpath is ridiculously complex to install. I guess
you're doing a great job, but as it is, noone will notice. Have you ever
thought of creating RPMs for popular Linux distros, or so?

Ciao,
Andreas


From bryce@waitaki.otago.ac.nz Tue Oct 16 20:57:39 2001
Received: from waitaki.otago.ac.nz ([139.80.75.140])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tf1e-0007lb-00
	for <classpath@gnu.org>; Tue, 16 Oct 2001 20:57:38 -0400
Received: from waitaki.otago.ac.nz (IDENT:bryce@reason.otago.ac.nz [139.80.119.17])
	by waitaki.otago.ac.nz (8.11.2/8.11.2) with ESMTP id f9H0xkW27163;
	Wed, 17 Oct 2001 13:59:46 +1300
Message-ID: <3BCCD801.4000200@waitaki.otago.ac.nz>
Date: Wed, 17 Oct 2001 13:59:45 +1300
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.5) Gecko/20011012
X-Accept-Language: en-us
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>, a_rueckert@gmx.net
Subject: Re: Missing AWT peer
References: <Pine.LNX.4.10.10110151421560.6623-100000@mars.deadcafe.org> <01101609515702.00769@krsn-d9336383>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Andreas Rueckert wrote:

>>Not necessarily.  Any paths seen by ld (e.g. with -L) will not be
>>automatically searched by the loader (assuming an ELF target).
>>
>>So each lib directory must be one of the "system" lib dirs or mentioned in
>>LD_LIBRARY_PATH, unless you also link with -rpath.
>>
>
>All these libs are in /etc/ld.so.conf , so it should work. 
>

Nope. The path to the libraries must be in LD_LIBRARY_PATH. I believe 
this is a limitation of dlopen or maybe libtool's ltdl.

>Is there any way to
>get additional info what lib is missing?
>

Classpath's reference runtime implementation should throw the missing 
library name with the UnsatisfiedLinkError. For some reason your stack 
trace didnt mention it. Which VM is this?

>I honestly think that Classpath is ridiculously complex to install. I guess
>you're doing a great job, but as it is, noone will notice. Have you ever
>thought of creating RPMs for popular Linux distros, or so?
>

Here is the build procedure I use to get a classpath suitable for use 
with ORP:

1. create gnu/classpath/Configuration.java by hand, from gnu/classpath/Configuration.in
2. mkdir build
3. javac -d build -classpath .:build `find java gnu vm -name "*.java"`
4. cp gnu/java/locale/*.uni build/gnu/java/locale/
5. cp gnu/java/locale/*.properties build/gnu/java/locale/


I admit that this doesn't help if you want the native libraries, 
however. The lack of an easy-to-use configure/build is a known problem, 
and we'd love it if someone were to help fix it.

regards

Bryce.




From cbj@gnu.org Tue Oct 16 23:27:00 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15thMC-0004Y8-00
	for <classpath@gnu.org>; Tue, 16 Oct 2001 23:27:00 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Tue, 16 Oct 2001 23:26:59 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9H3QRV06634;
	Tue, 16 Oct 2001 23:26:27 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: build ideas
From: Brian Jones <cbj@gnu.org>
Date: 16 Oct 2001 23:26:27 -0400
Message-ID: <m3heszt09o.fsf@lyta.haphazard.org>
Lines: 20
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I'm seeking suggestions or ideas from individuals about how they would
like to be able to build Classpath.  I guess two cases exist, building
from CVS and building from a release.  I'll start the discussion with my
suggestions but I'm hoping for others and some concensus at the end.

Like other GNU projects, I would like to continue to use the
automake/autoconf system for driving the compilation of Classpath.
For releases I'd like to continue to distribute compiled .class files
and generated JNI headers.  I want CVS builds to generate JNI headers
and .class files relying only on free software.  There exists a need
within our community to generate .class files without compiling native
libraries and I'd like that to be easier... it seems to me that to be
successful this must be faster than 'jikes -d /tmp `find java gnu
vm/reference -regex ".*\.java$"` or some working equivalent.

Additions?  Contrary notions?  Devil's advocate?

Brian
-- 
Brian Jones <cbj@gnu.org>


From ebb9@email.byu.edu Tue Oct 16 23:52:50 2001
Received: from femail4.sdc1.sfba.home.com ([24.0.95.84])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15thlB-0005QB-00; Tue, 16 Oct 2001 23:52:50 -0400
Received: from email.byu.edu ([65.12.64.100]) by femail4.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011017035239.YFPB15859.femail4.sdc1.sfba.home.com@email.byu.edu>;
          Tue, 16 Oct 2001 20:52:39 -0700
Message-ID: <3BCD0061.93A4DFCA@email.byu.edu>
Date: Tue, 16 Oct 2001 21:52:01 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Brian Jones <cbj@gnu.org>, Classpath list <classpath@gnu.org>
Subject: Re: build ideas
References: <m3heszt09o.fsf@lyta.haphazard.org>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I'm all for a way to compile the .class files without the shared
libraries.  I develop on a cygwin platform, which does not have the best
support for free VMs or for shared libraries, so I tend to just compile
the java files.

By the way, I think developers should have the autotools installed on
their machine, so I see no reason for CVS to keep track of generated
files like configure and Makefile.in, only the originals like
Makefile.am.  The generated files should be part of the distribution,
though, since end users should not be required to have all the developer
tools just to use the final product.  Fortunately, automake does this
correctly.

Who knows, improving the build framework might inspire someone to
improve the dependency tracking that jikes does, to make that part of
compilation even more efficient and compatible with automake.

Brian Jones wrote:
> 
> Like other GNU projects, I would like to continue to use the
> automake/autoconf system for driving the compilation of Classpath.
> For releases I'd like to continue to distribute compiled .class files
> and generated JNI headers.  I want CVS builds to generate JNI headers
> and .class files relying only on free software.  There exists a need
> within our community to generate .class files without compiling native
> libraries and I'd like that to be easier... it seems to me that to be
> successful this must be faster than 'jikes -d /tmp `find java gnu
> vm/reference -regex ".*\.java$"` or some working equivalent.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From reali@inf.ethz.ch Wed Oct 17 04:49:08 2001
Received: from core.inf.ethz.ch ([129.132.178.196])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tmNv-0007XU-00; Wed, 17 Oct 2001 04:49:07 -0400
Received: from lillian-ics.inf.ethz.ch (root@lillian-ics.inf.ethz.ch [129.132.134.2])
	by core.inf.ethz.ch (8.9.3/8.9.3) with ESMTP id KAA23099;
	Wed, 17 Oct 2001 10:49:00 +0200 (MET DST)
Received: from x.oberon.ethz.ch (pookie.inf.ethz.ch [129.132.134.15])
	by lillian-ics.inf.ethz.ch (8.9.3/8.9.3) with SMTP id KAA26256;
	Wed, 17 Oct 2001 10:49:03 +0200 (MET DST)
Date: Wed, 17 Oct 2001 10:49:03 +0200 (MET DST)
Message-Id: <200110170849.KAA26256@lillian-ics.inf.ethz.ch>
From: reali@inf.ethz.ch
X-Mailer: Oberon Mail (ejz) on Aos 03.10.2001
To: classpath@gnu.org
CC: cbj@gnu.org
Cc: reali@inf.ethz.ch
Subject: Re: build ideas
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

> Additions?  Contrary notions?  Devil's advocate?

For my own JVM running on the ETH-Oberon System (thus no C compiler
available and no UNIX-like platform) I need only the .class files
and the data files.

I also need to compile a few own classes like VMSystem.
One option is to have a glibj.zip with the whole classpath,
recompile only my own classes and add them to glibj.zip.
Personally I rather prefer to build everything directly
with my own classes.


A side note: I tried to build classpath last week. Apparently
config.in is in the wrong directory (classpath root instead of
include).

-Patrik
--
Patrik Reali, reali@inf.ethz.ch
http://www.inf.ethz.ch/personal/reali/
http://www.oberon.ethz.ch/native/



From reali@inf.ethz.ch Wed Oct 17 05:12:33 2001
Received: from core.inf.ethz.ch ([129.132.178.196])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tmka-0000Jf-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 05:12:33 -0400
Received: from lillian-ics.inf.ethz.ch (root@lillian-ics.inf.ethz.ch [129.132.134.2])
	by core.inf.ethz.ch (8.9.3/8.9.3) with ESMTP id LAA24454;
	Wed, 17 Oct 2001 11:12:02 +0200 (MET DST)
Received: from x.oberon.ethz.ch (pookie.inf.ethz.ch [129.132.134.15])
	by lillian-ics.inf.ethz.ch (8.9.3/8.9.3) with SMTP id LAA27083;
	Wed, 17 Oct 2001 11:12:04 +0200 (MET DST)
Date: Wed, 17 Oct 2001 11:12:04 +0200 (MET DST)
Message-Id: <200110170912.LAA27083@lillian-ics.inf.ethz.ch>
From: reali@inf.ethz.ch
X-Mailer: Oberon Mail (ejz) on Aos 03.10.2001
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
To: classpath@gnu.org
Cc: reali@inf.ethz.ch
Subject: Error in FileInputStream?
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

One of the FileInputStream constructors is
implemented as follows:

public
FileInputStream(FileDescriptor fd) throws SecurityException
{
  // Hmmm, no other exception but this one to throw, but if the descriptor
  // isn't valid, we surely don't have "permission" to read from it.
  if (!fd.valid())
    throw new SecurityException("Invalid FileDescriptor");

  SecurityManager sm = System.getSecurityManager();
  if (sm != null)
    {
//      try
//        {
          //sm.checkRead(fd);
          throw new SecurityException("ffo");
//        }
//      catch (AccessControlException e)
//        {
//          throw new SecurityException(e.getMessage());
//        }
    }

  native_fd = fd.getNativeFD();
}


When the JVM boots, System.in is allocated using a file descriptor.
If the security manager is already installed, then an exception
is thrown.

Is there a reason, for assuming that the security manager is not
available at this point or is this only some forgotten code that,
by chance, nobody else but me run into?

-Patrik
--
Patrik Reali, reali@inf.ethz.ch
http://www.inf.ethz.ch/personal/reali/
http://www.oberon.ethz.ch/native/



From a_rueckert@gmx.net Wed Oct 17 06:40:59 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15to8A-0008F1-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 06:40:58 -0400
Received: (qmail 21451 invoked by uid 0); 17 Oct 2001 10:40:22 -0000
Received: from krsn-d9336175.pool.mediaways.net (HELO krsn-d9336175) (217.51.97.117)
  by mail.gmx.net (mp009-rz3) with SMTP; 17 Oct 2001 10:40:22 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: GNU Classpath <classpath@gnu.org>
To: GNU Classpath <classpath@gnu.org>
Subject: Re: build ideas
Date: Wed, 17 Oct 2001 11:05:30 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <m3heszt09o.fsf@lyta.haphazard.org>
In-Reply-To: <m3heszt09o.fsf@lyta.haphazard.org>
MIME-Version: 1.0
Message-Id: <01101711114308.00771@hun.bre.de>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mit, 17 Okt 2001 Brian Jones wrote:
> I'm seeking suggestions or ideas from individuals about how they would
> like to be able to build Classpath.  I guess two cases exist, building
> >from CVS and building from a release.  I'll start the discussion with my
> suggestions but I'm hoping for others and some concensus at the end.
> 
> Like other GNU projects, I would like to continue to use the
> automake/autoconf system for driving the compilation of Classpath.

I'm not sure if I understand all your arguments for this, but I can tell you
that moving from make to Ant was a huge step forward for the ArgoUML project (to
which I'm contributing a bit). I don't think anyone wants to go back to the old
make tools.

> For releases I'd like to continue to distribute compiled .class files
> and generated JNI headers.

If you need libs like GTK, I'd rather like to see a distribution with automatic
dependency checking, like a rpm file (I'm using RedHat Linux) or so.

> I want CVS builds to generate JNI headers
> and .class files relying only on free software.  There exists a need
> within our community to generate .class files without compiling native
> libraries and I'd like that to be easier... it seems to me that to be
> successful this must be faster than 'jikes -d /tmp `find java gnu
> vm/reference -regex ".*\.java$"` or some working equivalent.

I guess jakarta-ant is the answer to this problem...
http://jakarta.apache.org/ant/
I thought someone is already working on a build.xml file?

Ciao,
Andreas


From nferrier@tapsellferrier.co.uk Wed Oct 17 08:41:50 2001
Received: from tf1.tapsellferrier.co.uk ([195.7.242.226])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tq17-0004Lc-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 08:41:49 -0400
Received: from nferrier by tf1.tapsellferrier.co.uk with local (Exim 3.22 #3 (Debian))
	id 15tqAs-0004br-00; Wed, 17 Oct 2001 13:51:54 +0100
From: Nic Ferrier <nferrier@tf1.tapsellferrier.co.uk>
To: classpath@gnu.org
Subject: Re: build ideas
Message-Id: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
Date: Wed, 17 Oct 2001 13:51:54 +0100
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Brian said:
> I want CVS builds to generate JNI headers
> and .class files relying only on free software.  There exists a need
> within our community to generate .class files without compiling native
> libraries and I'd like that to be easier... it seems to me that to be
> successful this must be faster than 'jikes -d /tmp `find java gnu
> vm/reference -regex ".*\.java$"` or some working equivalent.

Andreas said:

>I guess jakarta-ant is the answer to this problem...
>  http://jakarta.apache.org/ant/
>I thought someone is already working on a build.xml file?


Personally, I don't like Ant. It's not *free* software either.

There's not much you can do with Ant that can't be done with Make
(given a little thought).

Ant is also very Java centric, Classpath has a lot of C (and C++ if we
include the GCJ stuff). Make is perfect for those languages.

Brian, would you like to check out the Paperclips and Classpathx make
files? They use the the @filelist convieniance to facilitate better
make control over java projects.


Nic Ferrier


From abies@pg.gda.pl Wed Oct 17 08:57:24 2001
Received: from mail.engelsinfo.de ([212.6.209.241])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tqGB-0005HJ-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 08:57:23 -0400
Received: from pg.gda.pl (dvalin [212.6.209.232])
	by mail.engelsinfo.de (8.9.3+Sun/8.9.1) with ESMTP id OAA09407
	for <classpath@gnu.org>; Wed, 17 Oct 2001 14:48:01 +0200 (MET DST)
Message-ID: <3BCD8026.3070009@pg.gda.pl>
Date: Wed, 17 Oct 2001 14:57:10 +0200
From: Artur Biesiadowski <abies@pg.gda.pl>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.5) Gecko/20011012
X-Accept-Language: en-us
MIME-Version: 1.0
To: classpath@gnu.org
Subject: Re: build ideas
References: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Nic Ferrier wrote:


> Ant is also very Java centric, Classpath has a lot of C (and C++ if we
> include the GCJ stuff). Make is perfect for those languages.

Yes, but on the other  hand 'make' is very C-centric (meaning that it 
sees dependencies as tree, not as graph). Fitting it to java is like 
putting a nail with screwdriver (?or was it other way around? - these 
idioms...)

For me main benefit of 'make' over 'ant', is that with 'make' and jikes 
you don't need a java set up to compile code.

Artur



From cbj@gnu.org Wed Oct 17 09:10:34 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tqSw-00061E-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 09:10:34 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Wed, 17 Oct 2001 09:10:33 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9HD9v407683;
	Wed, 17 Oct 2001 09:09:57 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: reali@inf.ethz.ch
Cc: classpath@gnu.org
Subject: Re: build ideas
References: <200110170849.KAA26256@lillian-ics.inf.ethz.ch>
From: Brian Jones <cbj@gnu.org>
Date: 17 Oct 2001 09:09:56 -0400
In-Reply-To: reali@inf.ethz.ch's message of "Wed, 17 Oct 2001 10:49:03 +0200 (MET DST)"
Message-ID: <m31yk2tntn.fsf@lyta.haphazard.org>
Lines: 12
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

reali@inf.ethz.ch writes:

> A side note: I tried to build classpath last week. Apparently
> config.in is in the wrong directory (classpath root instead of
> include).

In CVS it is in include/, so I'm not sure how it ended up in root for
you.

Brian
-- 
Brian Jones <cbj@gnu.org>


From etienne.gagnon@uqam.ca Wed Oct 17 09:12:45 2001
Received: from anis.telecom.uqam.ca ([132.208.250.6] helo=uqam.ca)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tqV3-00064E-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 09:12:45 -0400
Received: from er.uqam.ca (nobel.si.uqam.ca [132.208.219.1])
	by uqam.ca (8.11.6/8.11.0) with ESMTP id f9HD3wF22581;
	Wed, 17 Oct 2001 09:03:58 -0400 (EDT)
Received: from perseval (perseval.info.uqam.ca [132.208.137.96])
	by er.uqam.ca (8.10.0/8.10.0) with ESMTP id f9HD3vv03572;
	Wed, 17 Oct 2001 09:03:58 -0400 (EDT)
Received: from egagnon by perseval with local (Exim 3.32 #1 (Debian))
	id 15tqMW-0001Ap-00; Wed, 17 Oct 2001 09:03:56 -0400
Date: Wed, 17 Oct 2001 09:03:56 -0400
To: Nic Ferrier <nferrier@tf1.tapsellferrier.co.uk>
Cc: classpath@gnu.org
Subject: Re: build ideas
Message-ID: <20011017090356.A4494@perseval.info.uqam.ca>
References: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
User-Agent: Mutt/1.3.23i
From: "Etienne M. Gagnon" <etienne.gagnon@uqam.ca>
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I remember doing the following suggestion before on this list; you can also
find an implementation of my idea in Sablepath 
(see http://sourceforge.net/project/showfiles.php?group_id=5523&release_id=52287 )

My suggestion is to use the GNU auto* tools for the native code, and use Ant
for the Java code.

See comment below.
 
On Wed, Oct 17, 2001 at 01:51:54PM +0100, Nic Ferrier wrote:
> Personally, I don't like Ant. It's not *free* software either.

I think you are mistaken here.  Ant is indeed free software (BSD like license
with adv. clause), but it is not Copyleft software (GNU GPL).

If you don't want to use non-GPL software, you'll get in trouble: the Jikes
compiler is not GPLed either, and it has a license which is incompatible
with the GPL. (See http://www.gnu.org/licenses/license-list.html for a
good explanation of all this).

So, I see no reason not to use Ant for the Java stuff.  Yes, it is a Java
centric tool, but this is what you want here.  One could say that the
GNU auto* tools are C/C++ and traditional *NIX languages centric; they are
pretty ill suited for Java (I don't intend to open a debate on this, I am
just stating my *personal* view).

So, please give a look at Sablepath, which is divided into 2 packages:
 sablepath-classes : the Java code, Ant based build
 sablepath-libs : the native code, GNU auto* based build

This would satisfy the various groups using Classpath. Groups that don't
need the native stuff would simply get the *-classes package, people
needing standard JNI native stuff would download the *-lib package,
you could even add different native implementations, like *-cni-lib, etc.

Etienne
-- 
Etienne M. Gagnon                    http://www.info.uqam.ca/~egagnon/
SableVM:                                       http://www.sablevm.org/
SableCC:                                       http://www.sablecc.org/


From cbj@gnu.org Wed Oct 17 09:13:10 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tqVS-00065k-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 09:13:10 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Wed, 17 Oct 2001 09:13:09 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9HDCWY07693;
	Wed, 17 Oct 2001 09:12:32 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Nic Ferrier <nferrier@tf1.tapsellferrier.co.uk>
Cc: classpath@gnu.org
Subject: Re: build ideas
References: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
From: Brian Jones <cbj@gnu.org>
Date: 17 Oct 2001 09:12:32 -0400
In-Reply-To: Nic Ferrier's message of "Wed, 17 Oct 2001 13:51:54 +0100"
Message-ID: <m3wv1us94v.fsf@lyta.haphazard.org>
Lines: 22
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Nic Ferrier <nferrier@tf1.tapsellferrier.co.uk> writes:

> Personally, I don't like Ant. It's not *free* software either.
> 
> There's not much you can do with Ant that can't be done with Make
> (given a little thought).

Last I checked this is quite true.
 
> Ant is also very Java centric, Classpath has a lot of C (and C++ if we
> include the GCJ stuff). Make is perfect for those languages.

I think Ant could only be useful for Java.  I would not want to use it
for making shared libraries.
 
> Brian, would you like to check out the Paperclips and Classpathx make
> files? They use the the @filelist convieniance to facilitate better
> make control over java projects.

I'll do that.
-- 
Brian Jones <cbj@gnu.org>


From arenn@enteract.com Wed Oct 17 09:13:32 2001
Received: from smtp-2.enteract.com ([207.229.143.4])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tqVn-00067F-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 09:13:31 -0400
Received: from shell-3.enteract.com (shell-3.enteract.com [207.229.143.42])
	by smtp-2.enteract.com (Postfix) with ESMTP
	id 400145FA3; Wed, 17 Oct 2001 08:13:31 -0500 (CDT)
Received: (from arenn@localhost)
	by shell-3.enteract.com (8.11.6/8.11.2) id f9HDDUu97617;
	Wed, 17 Oct 2001 08:13:30 -0500 (CDT)
	(envelope-from arenn)
Date: Wed, 17 Oct 2001 08:13:30 -0500
From: "Aaron M. Renn" <arenn@urbanophile.com>
To: reali@inf.ethz.ch
Cc: classpath@gnu.org
Subject: Re: Error in FileInputStream?
Message-ID: <20011017081330.A97548@urbanophile.com>
Mail-Followup-To: "Aaron M. Renn" <arenn@urbanophile.com>,
	reali@inf.ethz.ch, classpath@gnu.org
References: <200110170912.LAA27083@lillian-ics.inf.ethz.ch>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
In-Reply-To: <200110170912.LAA27083@lillian-ics.inf.ethz.ch>; from reali@inf.ethz.ch on Wed, Oct 17, 2001 at 11:12:04AM +0200
Organization: GNU's Not Unix!
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

reali@inf.ethz.ch (reali@inf.ethz.ch) wrote:
> When the JVM boots, System.in is allocated using a file descriptor.
> If the security manager is already installed, then an exception
> is thrown.
> 
> Is there a reason, for assuming that the security manager is not
> available at this point or is this only some forgotten code that,
> by chance, nobody else but me run into?

It looks like code that was commented out to get around another bug or
to work around the fact that something wasn't implemented yet.  Probably
should be cleaned up.  I'll try to do it in the next couple days.

-- 
Aaron M. Renn (arenn@urbanophile.com) http://www.urbanophile.com/arenn/

"We have a choice, either to change the way we live, which is unacceptable, 
or to change the way that they live, and we chose the latter."
  - Sec. of Defense Donald Rumsfeld, News briefing, 2001-09-18



From cbj@gnu.org Wed Oct 17 09:20:17 2001
Received: from mail7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tqcL-0006OI-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 09:20:17 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Wed, 17 Oct 2001 09:20:15 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9HDJcK07707;
	Wed, 17 Oct 2001 09:19:38 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Eric Blake <ebb9@email.byu.edu>
Cc: Classpath list <classpath@gnu.org>
Subject: Re: build ideas
References: <m3heszt09o.fsf@lyta.haphazard.org> <3BCD0061.93A4DFCA@email.byu.edu>
From: Brian Jones <cbj@gnu.org>
Date: 17 Oct 2001 09:19:38 -0400
In-Reply-To: Eric Blake's message of "Tue, 16 Oct 2001 21:52:01 -0600"
Message-ID: <m3r8s2s8t1.fsf@lyta.haphazard.org>
Lines: 34
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Eric Blake <ebb9@email.byu.edu> writes:

> I'm all for a way to compile the .class files without the shared
> libraries.  I develop on a cygwin platform, which does not have the best
> support for free VMs or for shared libraries, so I tend to just compile
> the java files.

I'd like to get native compilation working with cygwin and perhaps
eventually with Mingw.  I actually managed to write a working .dll
with Mingw a couple of weeks ago... and I don't consider myself a
Windows programmer.  The fact it just worked was amazing to me.

Gcj might work on cygwin.  Porting ORP would probably not be too hard
although it would help if they used autoconf with that one.

> By the way, I think developers should have the autotools installed on
> their machine, so I see no reason for CVS to keep track of generated
> files like configure and Makefile.in, only the originals like
> Makefile.am.  The generated files should be part of the distribution,
> though, since end users should not be required to have all the developer
> tools just to use the final product.  Fortunately, automake does this
> correctly.

I think we're doing this correctly today.

> Who knows, improving the build framework might inspire someone to
> improve the dependency tracking that jikes does, to make that part of
> compilation even more efficient and compatible with automake.

I tried working with jikes dependency tracking some time ago and I
think I pretty much gave up on it.  What is the current state of this?

-- 
Brian Jones <cbj@gnu.org>


From a_rueckert@gmx.net Wed Oct 17 09:28:20 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tqk7-0006f2-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 09:28:19 -0400
Received: (qmail 3506 invoked by uid 0); 17 Oct 2001 13:28:16 -0000
Received: from krsn-d9336160.pool.mediaways.net (HELO krsn-d9336160) (217.51.97.96)
  by mail.gmx.net (mp004-rz3) with SMTP; 17 Oct 2001 13:28:16 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: Missing AWT peer
Date: Wed, 17 Oct 2001 15:22:33 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <Pine.LNX.4.10.10110151421560.6623-100000@mars.deadcafe.org> <01101609515702.00769@krsn-d9336383> <3BCCD801.4000200@waitaki.otago.ac.nz>
In-Reply-To: <3BCCD801.4000200@waitaki.otago.ac.nz>
MIME-Version: 1.0
Message-Id: <0110171525490P.00771@hun.bre.de>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mit, 17 Okt 2001 Bryce McKinlay wrote:
> Andreas Rueckert wrote:
> 
> >>Not necessarily.  Any paths seen by ld (e.g. with -L) will not be
> >>automatically searched by the loader (assuming an ELF target).
> >>
> >>So each lib directory must be one of the "system" lib dirs or mentioned in
> >>LD_LIBRARY_PATH, unless you also link with -rpath.
> >>
> >
> >All these libs are in /etc/ld.so.conf , so it should work. 
> >
> 
> Nope. The path to the libraries must be in LD_LIBRARY_PATH. I believe 
> this is a limitation of dlopen or maybe libtool's ltdl.

Ok, I set the LD_LIBRARY_PATH var. Same result.

> >Is there any way to
> >get additional info what lib is missing?
> >
> 
> Classpath's reference runtime implementation should throw the missing 
> library name with the UnsatisfiedLinkError. For some reason your stack 
> trace didnt mention it. Which VM is this?

ORP 1.0.6

> >I honestly think that Classpath is ridiculously complex to install. I guess
> >you're doing a great job, but as it is, noone will notice. Have you ever
> >thought of creating RPMs for popular Linux distros, or so?
> >
> 
> Here is the build procedure I use to get a classpath suitable for use 
> with ORP:
> 
> 1. create gnu/classpath/Configuration.java by hand, from gnu/classpath/Configuration.in
> 2. mkdir build
> 3. javac -d build -classpath .:build `find java gnu vm -name "*.java"`
> 4. cp gnu/java/locale/*.uni build/gnu/java/locale/
> 5. cp gnu/java/locale/*.properties build/gnu/java/locale/
> 
> 
> I admit that this doesn't help if you want the native libraries, 
> however. The lack of an easy-to-use configure/build is a known problem, 
> and we'd love it if someone were to help fix it.

Console apps seem to work. At a HelloWorldApp worked as expected. It seems that
it's just a problem with the GTK lib.

Ciao,
Andreas


From ebb9@email.byu.edu Wed Oct 17 10:11:33 2001
Received: from femail15.sdc1.sfba.home.com ([24.0.95.142])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15trPx-0000L5-00; Wed, 17 Oct 2001 10:11:33 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail15.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011017141131.IVVA2467.femail15.sdc1.sfba.home.com@email.byu.edu>;
          Wed, 17 Oct 2001 07:11:31 -0700
Message-ID: <3BCD8B0F.97854522@email.byu.edu>
Date: Wed, 17 Oct 2001 07:43:43 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Brian Jones <cbj@gnu.org>
CC: Classpath list <classpath@gnu.org>
Subject: Re: build ideas
References: <m3heszt09o.fsf@lyta.haphazard.org>
	 <3BCD0061.93A4DFCA@email.byu.edu> <m3r8s2s8t1.fsf@lyta.haphazard.org>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Brian Jones wrote:
> 
> > Who knows, improving the build framework might inspire someone to
> > improve the dependency tracking that jikes does, to make that part of
> > compilation even more efficient and compatible with automake.
> 
> I tried working with jikes dependency tracking some time ago and I
> think I pretty much gave up on it.  What is the current state of this?

As far as I know, little work is currently being done on jikes'
dependency tracking (there are only 3 or 4 active developers at the
moment, and none of us have been tackling it).  There is one outstanding
patch that has yet to be applied to CVS which claims to allow
command-line specification of the target directory for the generated .u
files.  Yet there are also several standing bug reports on flaws in the
algorithm, such as questions over whether a .java file should be
dependent on inner classes.  Personally, I have never looked at that
part of the jikes code, so I have no idea what it would take to make +M
spit out the best possible dependence files.  However, I feel that I
could usefully spend some time on that project if it would help the
Classpath project.

I agree with the earlier sentiment that Ant has the drawback of
requiring a working JVM installed on the machine, while make does not. 
While Ant may be the de-facto build framework for Java projects, it
would be difficult to use when bootstrapping a free machine, which
probably would not come with a free VM binary already installed.  I also
agree that make CAN be made to do anything that Ant can, it is just a
tougher assignment to write the correct rules.  But with automake and
jikes dependency tracking (fixed, of course), I think this should be
feasable.  By the way, I don't have Ant installed on my machine; but
make comes standard with cygwin.

What about the idea of both a Makefile and build.xml approach to
building the .class files?  Then Ant users who just want classes.zip and
already have a VM don't have to worry about make, but developers that
care about the native code don't need to worry about Ant even for the
.class files.  The only drawback of this dual approach is that
developers must remember to list new .java files in two separate build
framework files, instead of one.
-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From etienne.gagnon@uqam.ca Wed Oct 17 10:37:19 2001
Received: from anis.telecom.uqam.ca ([132.208.250.6] helo=uqam.ca)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tros-0001WC-00; Wed, 17 Oct 2001 10:37:18 -0400
Received: from er.uqam.ca (nobel.si.uqam.ca [132.208.219.1])
	by uqam.ca (8.11.6/8.11.0) with ESMTP id f9HEYqF12093;
	Wed, 17 Oct 2001 10:34:52 -0400 (EDT)
Received: from perseval (perseval.info.uqam.ca [132.208.137.96])
	by er.uqam.ca (8.10.0/8.10.0) with ESMTP id f9HEYnv16225;
	Wed, 17 Oct 2001 10:34:51 -0400 (EDT)
Received: from egagnon by perseval with local (Exim 3.32 #1 (Debian))
	id 15trmS-0001Cp-00; Wed, 17 Oct 2001 10:34:48 -0400
Date: Wed, 17 Oct 2001 10:34:48 -0400
To: Eric Blake <ebb9@email.byu.edu>
Cc: Brian Jones <cbj@gnu.org>, Classpath list <classpath@gnu.org>
Subject: Re: build ideas
Message-ID: <20011017103448.A4634@perseval.info.uqam.ca>
References: <m3heszt09o.fsf@lyta.haphazard.org> <3BCD0061.93A4DFCA@email.byu.edu> <m3r8s2s8t1.fsf@lyta.haphazard.org> <3BCD8B0F.97854522@email.byu.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BCD8B0F.97854522@email.byu.edu>
User-Agent: Mutt/1.3.23i
From: "Etienne M. Gagnon" <etienne.gagnon@uqam.ca>
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Wed, Oct 17, 2001 at 07:43:43AM -0600, Eric Blake wrote:
> I agree with the earlier sentiment that Ant has the drawback of
> requiring a working JVM installed on the machine, while make does not. 

This is just a atandard bootstrapping problem. (If you look at autoconf
and M4, you'll find the same kind of bootstrapping problem).

This problem is easily resolved by having:
1- Jikes
2- using a simple shell script to compile all java classes (no dependency
   tracking, etc.) [something like: find . | grep \\.java$ | xargs jikes -d classes ]

Once a first set of class files are generated, a Free VM can then work and execute
Ant, allowing a much nicer build environment.

In fact, neither Make nor Ant should be *necessary* to compile the java sources.
One needs these tools only to be able to get partial recompilation of sources,
or use them to do additional things like installing the code, or building
source/binary packages.

Another example of bootstrapping can be seen in Ant itself.  Ant uses Ant to build,
but the source package also contains an initial bootstrapping shell script.

So, I do not *personally* think that the dependency of Ant on Java is a valid
argument against using it for the Java part of Classpath.

Etienne
-- 
Etienne M. Gagnon                    http://www.info.uqam.ca/~egagnon/
SableVM:                                       http://www.sablevm.org/
SableCC:                                       http://www.sablecc.org/


From a_rueckert@gmx.net Wed Oct 17 10:52:22 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ts3S-00022d-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 10:52:22 -0400
Received: (qmail 21760 invoked by uid 0); 17 Oct 2001 14:52:19 -0000
Received: from krsn-d9336381.pool.mediaways.net (HELO krsn-d9336381) (217.51.99.129)
  by mail.gmx.net (mp020-rz3) with SMTP; 17 Oct 2001 14:52:19 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: build ideas
Date: Wed, 17 Oct 2001 16:26:35 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: text/plain
References: <m3heszt09o.fsf@lyta.haphazard.org> <m3r8s2s8t1.fsf@lyta.haphazard.org> <3BCD8B0F.97854522@email.byu.edu>
In-Reply-To: <3BCD8B0F.97854522@email.byu.edu>
MIME-Version: 1.0
Message-Id: <0110171629030T.00771@krsn-d93363ca>
Content-Transfer-Encoding: 8bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

On Mit, 17 Okt 2001 Eric Blake wrote:

--<snip>--
 
> What about the idea of both a Makefile and build.xml approach to
> building the .class files?  Then Ant users who just want classes.zip and
> already have a VM don't have to worry about make, but developers that
> care about the native code don't need to worry about Ant even for the
> .class files.  The only drawback of this dual approach is that
> ddevelopers must remember to list new .java files in two separate build
> framework files, instead of one.

New .java files shouldn't be a problem with Ant. It should find them
automatically in most cases.

Ciao,
Andreas


From tromey@creche.cygnus.com Wed Oct 17 11:47:17 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tsub-0004ox-00; Wed, 17 Oct 2001 11:47:17 -0400
Received: from creche.cygnus.com (ta0201.peakpeak.com [204.144.244.201])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id JAA05792;
	Wed, 17 Oct 2001 09:47:08 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id KAA21035;
	Wed, 17 Oct 2001 10:03:32 -0600
To: Brian Jones <cbj@gnu.org>
Cc: Eric Blake <ebb9@email.byu.edu>, Classpath list <classpath@gnu.org>
Subject: Re: build ideas
References: <m3heszt09o.fsf@lyta.haphazard.org> <3BCD0061.93A4DFCA@email.byu.edu> <m3r8s2s8t1.fsf@lyta.haphazard.org>
X-Zippy:  The fact that 47 PEOPLE are yelling and sweat is cascading
 down my SPINAL COLUMN is fairly enjoyable!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 17 Oct 2001 10:03:31 -0600
In-Reply-To: Brian Jones's message of "17 Oct 2001 09:19:38 -0400"
Message-ID: <878zeamey4.fsf@creche.redhat.com>
Lines: 8
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> Gcj might work on cygwin.

Not yet, alas.  People talk about it sometimes, but the port is still
incomplete.

Tom


From tromey@creche.cygnus.com Wed Oct 17 14:01:27 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tv0R-0003Gh-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 14:01:27 -0400
Received: from creche.cygnus.com (ta0201.peakpeak.com [204.144.244.201])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id MAA02029;
	Wed, 17 Oct 2001 12:01:24 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id MAA24169;
	Wed, 17 Oct 2001 12:17:49 -0600
To: Classpath Project <classpath@gnu.org>
Cc: Anthony Green <green@redhat.com>
Subject: java.security.Security proposal
X-Zippy:  If I pull this SWITCH I'll be RITA HAYWORTH!!  Or a SCIENTOLOGIST!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 17 Oct 2001 12:17:49 -0600
Message-ID: <87bsj6ku5u.fsf@creche.redhat.com>
Lines: 19
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Anthony Green submitted a gcj patch that fixes a Security bug.  It
also changes Security.loadProviders to look for `gcj.security' and not
`classpath.security':

    http://gcc.gnu.org/ml/java-patches/2001-q3/msg00460.html

One problem with this is that it introduces a divergence between
Classpath and libgcj.  I prefer to avoid those.

What if we change this mechanism so that we can have a VM-dependent
`.security' file in addition to the `classpath.security' file?  The
idea would be to use the `java.vm.name' property to look up the
initial security file, and then fall back on classpath.security.

Comments?

Tom

PS I'm going to check in his bug fix shortly.


From tromey@creche.cygnus.com Wed Oct 17 15:10:16 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15tw52-0006v6-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 15:10:16 -0400
Received: from creche.cygnus.com (ta0201.peakpeak.com [204.144.244.201])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id NAA14559;
	Wed, 17 Oct 2001 13:10:07 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id NAA17190;
	Wed, 17 Oct 2001 13:26:33 -0600
To: Classpath Project <classpath@gnu.org>
Subject: java.security.Security question
X-Zippy:  I always have fun because I'm out of my mind!!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 17 Oct 2001 13:26:33 -0600
Message-ID: <877ktujceu.fsf@creche.redhat.com>
Lines: 12
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Right now java.security.Security looks for classpath.security in
<java.home>/lib/security/.  According to GNU standards this should be
<java.home>/share/security/ instead.

Using lib is fine if we're following the JDK convention and want to be
compatible with 3rd party packages that write to the .security file as
part of their installation process.  But in that case we ought to name
the file `java.security'.

I'd like to change this code to follow GNU standards.  Any objections?

Tom


From cbj@gnu.org Wed Oct 17 15:17:39 2001
Received: from mail7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15twCA-0007G3-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 15:17:39 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Wed, 17 Oct 2001 15:17:37 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9HJGuM08372;
	Wed, 17 Oct 2001 15:16:56 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: tromey@redhat.com
Cc: Classpath Project <classpath@gnu.org>, Anthony Green <green@redhat.com>
Subject: Re: java.security.Security proposal
References: <87bsj6ku5u.fsf@creche.redhat.com>
From: Brian Jones <cbj@gnu.org>
Date: 17 Oct 2001 15:16:56 -0400
In-Reply-To: Tom Tromey's message of "17 Oct 2001 12:17:49 -0600"
Message-ID: <m3lmiam5zr.fsf@lyta.haphazard.org>
Lines: 18
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey <tromey@redhat.com> writes:

> Anthony Green submitted a gcj patch that fixes a Security bug.  It
> also changes Security.loadProviders to look for `gcj.security' and not
> `classpath.security':
> 
>     http://gcc.gnu.org/ml/java-patches/2001-q3/msg00460.html
> 
> One problem with this is that it introduces a divergence between
> Classpath and libgcj.  I prefer to avoid those.

I apologize for not having read that link yet, but how does changing
the name to gcj.security have to do with fixing a bug?  Is this really
gcj specific?

Brian
-- 
Brian Jones <cbj@gnu.org>


From tromey@creche.cygnus.com Wed Oct 17 16:01:10 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15twsI-0000n4-00; Wed, 17 Oct 2001 16:01:10 -0400
Received: from creche.cygnus.com (ta0201.peakpeak.com [204.144.244.201])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id OAA24335;
	Wed, 17 Oct 2001 14:00:54 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id OAA23801;
	Wed, 17 Oct 2001 14:17:18 -0600
To: Brian Jones <cbj@gnu.org>
Cc: Classpath Project <classpath@gnu.org>, Anthony Green <green@REDHAT.COM>
Subject: Re: java.security.Security proposal
References: <87bsj6ku5u.fsf@creche.redhat.com> <m3lmiam5zr.fsf@lyta.haphazard.org>
X-Zippy:  Is there something I should be DOING with a GLAZED DONUT??
X-Attribution:  Tom
Reply-To: tromey@REDHAT.COM
From: Tom Tromey <tromey@REDHAT.COM>
Date: 17 Oct 2001 14:17:17 -0600
In-Reply-To: Brian Jones's message of "17 Oct 2001 15:16:56 -0400"
Message-ID: <87vghehvhu.fsf@creche.redhat.com>
Lines: 16
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> I apologize for not having read that link yet, but how does
Brian> changing the name to gcj.security have to do with fixing a bug?
Brian> Is this really gcj specific?

Sorry -- the patch had two parts.  One part was a bug fix; I checked
it in.

The other part changed the name of classpath.security.  I'd prefer to
have it search both <current-vm>.security and classpath.security.
That allows for VM-specific security providers.  As AG points out,
there might be a CNI-based shared library provider.  Putting that in
classpath.security might confuse other VMs using Classpath.

Tom


From cbj@gnu.org Wed Oct 17 17:42:50 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tySg-0006fd-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 17:42:50 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Wed, 17 Oct 2001 17:42:48 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9HLfvS08679;
	Wed, 17 Oct 2001 17:41:57 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: tromey@redhat.com
Cc: Classpath Project <classpath@gnu.org>
Subject: Re: java.security.Security question
References: <877ktujceu.fsf@creche.redhat.com>
From: Brian Jones <cbj@gnu.org>
Date: 17 Oct 2001 17:41:52 -0400
In-Reply-To: Tom Tromey's message of "17 Oct 2001 13:26:33 -0600"
Message-ID: <m3vgheq6zj.fsf@lyta.haphazard.org>
Lines: 16
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey <tromey@redhat.com> writes:

> Right now java.security.Security looks for classpath.security in
> <java.home>/lib/security/.  According to GNU standards this should be
> <java.home>/share/security/ instead.
> 
> Using lib is fine if we're following the JDK convention and want to be
> compatible with 3rd party packages that write to the .security file as
> part of their installation process.  But in that case we ought to name
> the file `java.security'.
> 
> I'd like to change this code to follow GNU standards.  Any objections?

Only what you mention above.
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Wed Oct 17 17:46:00 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15tyVk-0006n2-00
	for <classpath@gnu.org>; Wed, 17 Oct 2001 17:46:00 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Wed, 17 Oct 2001 17:45:53 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9HLj8e08687;
	Wed, 17 Oct 2001 17:45:08 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: "Anthony Green" <green@redhat.com>
Cc: <tromey@redhat.com>, "Classpath Project" <classpath@gnu.org>
Subject: Re: java.security.Security proposal
References: <87bsj6ku5u.fsf@creche.redhat.com> <m3lmiam5zr.fsf@lyta.haphazard.org> <015201c15742$60d340c0$5be6b4cd@cygnus.com>
From: Brian Jones <cbj@gnu.org>
Date: 17 Oct 2001 17:45:07 -0400
In-Reply-To: "Anthony Green"'s message of "Wed, 17 Oct 2001 12:31:54 -0700"
Message-ID: <m3r8s2q6u4.fsf@lyta.haphazard.org>
Lines: 18
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

"Anthony Green" <green@redhat.com> writes:

> Brian wrote:
> > I apologize for not having read that link yet, but how does changing
> > the name to gcj.security have to do with fixing a bug?  Is this really
> > gcj specific?
> 
> It's possible to have an installation with native shared library
> providers and no .jar implementation.  You really don't want to have
> a global classpath.security file in this case because gcj is the
> only implementation that will find and use those providers.

I'm not very familiar with Java's security framework here so I'll take
your word for it.

Brian
-- 
Brian Jones <cbj@gnu.org>


From ebb9@email.byu.edu Thu Oct 18 20:29:44 2001
Received: from femail43.sdc1.sfba.home.com ([24.254.60.37])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15uNXk-00058T-00
	for <classpath@gnu.org>; Thu, 18 Oct 2001 20:29:44 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail43.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011019002942.KFXL29972.femail43.sdc1.sfba.home.com@email.byu.edu>
          for <classpath@gnu.org>; Thu, 18 Oct 2001 17:29:42 -0700
Message-ID: <3BCF7392.2FE55628@email.byu.edu>
Date: Thu, 18 Oct 2001 18:28:02 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>
Subject: java.util
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I just committed a large change to some of the core classes of java.util
(Abstract*.java, Arrays, BasicMapEntry, and Collections).  A lot of it
was just adding javadoc, formatting to GNU standards, and replacing some
code with more efficient constructs, but I also fixed a few bugs.  In
particular, in Collections, I made all the nested classes compliant to
JDK serialization.  Let me know if I introduced any problems.

I have more changes pending on the implementation classes (such as
HashMap, ArrayList, Vector, ...), but want to make sure the abstract
framework is correct first.

I currently don't have write access to libgcj.  Should I apply for that,
so that I can do the work of keeping the two in synch, rather than
leaving that responsibility on others like Tom?

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From mark@klomp.org Fri Oct 19 13:56:35 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15udsp-0004Ro-00
	for <classpath@gnu.org>; Fri, 19 Oct 2001 13:56:35 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15udsg-0005FB-00; Fri, 19 Oct 2001 19:56:26 +0200
Date: Fri, 19 Oct 2001 19:56:26 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Slowly adding javax.naming (JNDI) classes from libgcj
Message-ID: <20011019195626.A20120@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.22i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

One of the biggest differences with the libgcj tree are the javax.naming
(JNDI) classes. See <http://gcc.gnu.org/java/libgcj-classpath-compare.html>.
But what is in the libgcj tree has no comments and is mainly interfaces and
stubs. Since I have no experience with this package it will take some time
to clean up and document these classes. I have committed what I have now.

Note that I also cleaned up our standard.omit file. Now that we have the
java.util.zip package all Jar related classes should compile out of the box.
Please yell if I broke your compile with my build file changes.

	* configure.in (AC_OUTPUT): removed gnu/classpath/tools/Makefile,
	added javax/naming/Makefile, javax/naming/directory/Makefile and
	javax/naming/spi/Makefile
	* gnu/java/text/Makefile.in: removed
	* javax/Makefile.am (SUBDIRS): add naming
	* javax/naming/.cvsignore: new file
	* javax/naming/BinaryRefAddr.java: ditto
	* javax/naming/InvalidNameException.java: ditto
	* javax/naming/Makefile.am: ditto
	* javax/naming/Name.java: ditto
	* javax/naming/NamingException.java: ditto
	* javax/naming/RefAddr.java: ditto
	* javax/naming/StringRefAddr.java: ditto
	* javax/naming/directory/.cvsignore: ditto
	* javax/naming/directory/Makefile.am: ditto
	* javax/naming/spi/.cvsignore: ditto
	* javax/naming/spi/Makefile.am: ditto
	* lib/.cvsignore: add javax
	* lib/gen-classlist.sh.in: ditto
	* lib/standard.omit: removed java/net, java/lang, java/text and
	java/util/jar classes, added javax/accessibility and javax/swing

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From bryce@waitaki.otago.ac.nz Fri Oct 19 19:52:20 2001
Received: from mta5-rme.xtra.co.nz ([210.86.15.138])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ujR6-0004gY-00; Fri, 19 Oct 2001 19:52:20 -0400
Received: from localhost ([210.55.152.38]) by mta5-rme.xtra.co.nz
          with ESMTP
          id <20011019235211.VBYD12812.mta5-rme.xtra.co.nz@localhost>;
          Sat, 20 Oct 2001 12:52:11 +1300
Date: Sat, 20 Oct 2001 12:52:45 +1300
Subject: Re: build ideas
Content-Type: text/plain; charset=US-ASCII; format=flowed
Mime-Version: 1.0 (Apple Message framework v472)
Cc: Eric Blake <ebb9@email.byu.edu>,
 Classpath list <classpath@gnu.org>
To: Brian Jones <cbj@gnu.org>
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
In-Reply-To: <m3r8s2s8t1.fsf@lyta.haphazard.org>
Message-Id: <64612DC1-C4EC-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
Content-Transfer-Encoding: 7bit
X-Mailer: Apple Mail (2.472)
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Thursday, October 18, 2001, at 02:19  AM, Brian Jones wrote:

> Eric Blake <ebb9@email.byu.edu> writes:
>
>> By the way, I think developers should have the autotools installed on
>> their machine, so I see no reason for CVS to keep track of generated
>> files like configure and Makefile.in, only the originals like
>> Makefile.am.  The generated files should be part of the distribution,
>> though, since end users should not be required to have all the 
>> developer
>> tools just to use the final product.  Fortunately, automake does this
>> correctly.

I don't agree. It makes things much easier for people to build if 
Makefile.in and configure are included in the tree. Its hard to remember 
to regenerate them in your local tree every time they might have 
changed, and its a pain to have to install automake and autoconf on 
every machine you might want to develop on. On the other hand, its 
almost no additional effort to check in updated copies of the files 
whenever actual changes to them are made.

regards

Bryce.




From bryce@waitaki.otago.ac.nz Fri Oct 19 19:53:32 2001
Received: from mta4-rme.xtra.co.nz ([210.86.15.132])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ujSG-0004l7-00
	for <classpath@gnu.org>; Fri, 19 Oct 2001 19:53:32 -0400
Received: from localhost ([210.55.152.38]) by mta4-rme.xtra.co.nz
          with ESMTP
          id <20011019235330.UPRI18935.mta4-rme.xtra.co.nz@localhost>;
          Sat, 20 Oct 2001 12:53:30 +1300
Date: Sat, 20 Oct 2001 12:54:04 +1300
Subject: Re: Error in FileInputStream?
Content-Type: text/plain; charset=US-ASCII; format=flowed
Mime-Version: 1.0 (Apple Message framework v472)
Cc: classpath@gnu.org
To: reali@inf.ethz.ch
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
In-Reply-To: <200110170912.LAA27083@lillian-ics.inf.ethz.ch>
Message-Id: <93A606A8-C4EC-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
Content-Transfer-Encoding: 7bit
X-Mailer: Apple Mail (2.472)
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Wednesday, October 17, 2001, at 10:12  PM, reali@inf.ethz.ch wrote:

> When the JVM boots, System.in is allocated using a file descriptor.
> If the security manager is already installed, then an exception
> is thrown.
>
> Is there a reason, for assuming that the security manager is not
> available at this point or is this only some forgotten code that,
> by chance, nobody else but me run into?

I agree that it looks wrong. My guess is that the SecurityManager wasn't 
working correctly at the time the FileInputStream was written so we 
ended up with some weird code.

In libgcj the implementation is simply:

   public FileInputStream(FileDescriptor fdObj)
   {
     SecurityManager s = System.getSecurityManager();
     if (s != null)
       s.checkRead(fdObj);
     fd = fdObj;
   }

Note that SecurityException is unchecked, so it does nto need to be 
declared in the throws clause.

regards

Bryce.




From bryce@waitaki.otago.ac.nz Fri Oct 19 19:55:32 2001
Received: from mta4-rme.xtra.co.nz ([210.86.15.132])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ujUC-0004t4-00
	for <classpath@gnu.org>; Fri, 19 Oct 2001 19:55:32 -0400
Received: from localhost ([210.55.152.38]) by mta4-rme.xtra.co.nz
          with ESMTP
          id <20011019235530.UQFU18935.mta4-rme.xtra.co.nz@localhost>
          for <classpath@gnu.org>; Sat, 20 Oct 2001 12:55:30 +1300
Date: Sat, 20 Oct 2001 12:56:04 +1300
Subject: Re: Missing AWT peer
Content-Type: text/plain; charset=US-ASCII; format=flowed
Mime-Version: 1.0 (Apple Message framework v472)
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
To: Classpath list <classpath@gnu.org>
Content-Transfer-Encoding: 7bit
In-Reply-To: <0110171525490P.00771@hun.bre.de>
Message-Id: <DB5B2F34-C4EC-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
X-Mailer: Apple Mail (2.472)
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Thursday, October 18, 2001, at 02:22  AM, Andreas Rueckert wrote:

> Console apps seem to work. At a HelloWorldApp worked as expected. It 
> seems that it's just a problem with the GTK lib.

It could be a problem with ORP and the native part of its 
System.loadLibrary() implementation. I've never tried to load any JNI 
libraries in ORP.

regards

Bryce.




From bryce@waitaki.otago.ac.nz Fri Oct 19 19:57:52 2001
Received: from mta2-rme.xtra.co.nz ([210.86.15.130])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ujWR-000515-00; Fri, 19 Oct 2001 19:57:52 -0400
Received: from localhost ([210.55.152.38]) by mta2-rme.xtra.co.nz
          with ESMTP
          id <20011019235749.VCJX12093.mta2-rme.xtra.co.nz@localhost>;
          Sat, 20 Oct 2001 12:57:49 +1300
Date: Sat, 20 Oct 2001 12:58:24 +1300
Subject: Re: build ideas
Content-Type: text/plain; charset=US-ASCII; format=flowed
Mime-Version: 1.0 (Apple Message framework v472)
Cc: Brian Jones <cbj@gnu.org>,
 Classpath list <classpath@gnu.org>
To: Eric Blake <ebb9@email.byu.edu>
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
In-Reply-To: <3BCD8B0F.97854522@email.byu.edu>
Message-Id: <2E4A955D-C4ED-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
Content-Transfer-Encoding: 7bit
X-Mailer: Apple Mail (2.472)
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Thursday, October 18, 2001, at 02:43  AM, Eric Blake wrote:

> What about the idea of both a Makefile and build.xml approach to
> building the .class files?

This sounds reasonable to me.

regards

Bryce.




From tromey@creche.cygnus.com Fri Oct 19 20:10:21 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ujiX-0005X4-00
	for <classpath@gnu.org>; Fri, 19 Oct 2001 20:10:21 -0400
Received: from creche.cygnus.com (ta0198.peakpeak.com [204.144.244.198])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id SAA24187;
	Fri, 19 Oct 2001 18:10:18 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id SAA21493;
	Fri, 19 Oct 2001 18:27:20 -0600
To: Eric Blake <ebb9@email.byu.edu>
Cc: Classpath list <classpath@gnu.org>
Subject: Re: java.util
References: <3BCF7392.2FE55628@email.byu.edu>
X-Zippy:  I'm wearing PAMPERS!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 19 Oct 2001 18:27:19 -0600
In-Reply-To: Eric Blake's message of "Thu, 18 Oct 2001 18:28:02 -0600"
Message-ID: <871yjzjguw.fsf@creche.redhat.com>
Lines: 19
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Eric" == Eric Blake <ebb9@email.byu.edu> writes:

Eric> I currently don't have write access to libgcj.  Should I apply
Eric> for that, so that I can do the work of keeping the two in synch,
Eric> rather than leaving that responsibility on others like Tom?

Please do!  Use this form:

    http://sources.redhat.com/cgi-bin/pdw/ps_form.cgi

Put my email address in the "approver" field.
It may take a few days to get an account right now.

Note that the process for getting libgcj patches in is a bit more
restrictive than the Classpath process.  You have to send the patch
(with ChangeLog and description) to java-patches@gcc.gnu.org and get
approval (at least initially; this changes with time).

Tom


From ebb9@email.byu.edu Fri Oct 19 20:36:50 2001
Received: from femail35.sdc1.sfba.home.com ([24.254.60.25])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15uk8A-0006XZ-00
	for <classpath@gnu.org>; Fri, 19 Oct 2001 20:36:50 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail35.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011020003648.ZHDV6181.femail35.sdc1.sfba.home.com@email.byu.edu>
          for <classpath@gnu.org>; Fri, 19 Oct 2001 17:36:48 -0700
Message-ID: <3BD0C60E.5D2C1554@email.byu.edu>
Date: Fri, 19 Oct 2001 18:32:14 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>
Subject: question about vm/reference files
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

What is the reason that some of the VM dependent core .java files are
not located directly in the top tree, while others are?  For example,
java/lang/Object.java is in the main tree, but java/lang/Throwable.java
is only in the vm/reference subtree.

I understand why things like VMClassLoader.java are not in the main
tree, since that is not a public API, even if it does compile to the
java.lang package.  But for compilation 3rd-party projects, jikes needs
a .jar file with all the public APIs, even .class files for things like
Object (where the VM probably never loads the .class file, because the
it just uses native code).  It is annoying to have to look in two places
to build this .jar file.

Would there be anything wrong with moving these .java files from
vm/reference/java/lang to the main tree at java/lang?
Class
Runtime
Thread
Throwable
reflect/Constructor
reflect/Field
reflect/Method

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From cbj@gnu.org Fri Oct 19 22:47:32 2001
Received: from fe8.southeast.rr.com ([24.93.67.55] helo=mail8.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15umAe-0004sG-00
	for <classpath@gnu.org>; Fri, 19 Oct 2001 22:47:32 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail8.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Fri, 19 Oct 2001 22:47:17 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9K2k5123754;
	Fri, 19 Oct 2001 22:46:05 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Eric Blake <ebb9@email.byu.edu>
Cc: Classpath list <classpath@gnu.org>
Subject: Re: question about vm/reference files
References: <3BD0C60E.5D2C1554@email.byu.edu>
From: Brian Jones <cbj@gnu.org>
Date: 19 Oct 2001 22:46:04 -0400
In-Reply-To: Eric Blake's message of "Fri, 19 Oct 2001 18:32:14 -0600"
Message-ID: <m3bsj3owpf.fsf@lyta.haphazard.org>
Lines: 33
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Eric Blake <ebb9@email.byu.edu> writes:

> What is the reason that some of the VM dependent core .java files are
> not located directly in the top tree, while others are?  For example,
> java/lang/Object.java is in the main tree, but java/lang/Throwable.java
> is only in the vm/reference subtree.
> 
> I understand why things like VMClassLoader.java are not in the main
> tree, since that is not a public API, even if it does compile to the
> java.lang package.  But for compilation 3rd-party projects, jikes needs
> a .jar file with all the public APIs, even .class files for things like
> Object (where the VM probably never loads the .class file, because the
> it just uses native code).  It is annoying to have to look in two places
> to build this .jar file.
> 
> Would there be anything wrong with moving these .java files from
> vm/reference/java/lang to the main tree at java/lang?
> Class
> Runtime
> Thread
> Throwable
> reflect/Constructor
> reflect/Field
> reflect/Method

The thought was that Classpath either would not provide the classes
that are VM dependent (leaving those to the VM) or that there would be
multiple directories (vm/gcj, vm/orp, vm/kaffe, etc.)  The ugly
complexity appears to be necessary.

Brian
-- 
Brian Jones <cbj@gnu.org>


From ebb9@email.byu.edu Fri Oct 19 23:04:01 2001
Received: from femail29.sdc1.sfba.home.com ([24.254.60.19])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15umQb-0005DK-00; Fri, 19 Oct 2001 23:04:01 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail29.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011020030358.XEN21044.femail29.sdc1.sfba.home.com@email.byu.edu>;
          Fri, 19 Oct 2001 20:03:58 -0700
Message-ID: <3BD0E872.6AB424BC@email.byu.edu>
Date: Fri, 19 Oct 2001 20:58:58 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Brian Jones <cbj@gnu.org>
CC: Classpath list <classpath@gnu.org>
Subject: Re: question about vm/reference files
References: <3BD0C60E.5D2C1554@email.byu.edu> <m3bsj3owpf.fsf@lyta.haphazard.org>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

> The thought was that Classpath either would not provide the classes
> that are VM dependent (leaving those to the VM) or that there would be
> multiple directories (vm/gcj, vm/orp, vm/kaffe, etc.)  The ugly
> complexity appears to be necessary.
> 
> Brian

That still doesn't explain why Object is in the top level when the other
core classes are not.  Even if these classes have no Java
implementation, it is still useful to have a .java file in Classpath for
making the .class file which jikes requires for compilation purposes.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From bryce@waitaki.otago.ac.nz Sat Oct 20 03:06:06 2001
Received: from mta4-rme.xtra.co.nz ([210.86.15.132])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15uqCr-0006Yn-00; Sat, 20 Oct 2001 03:06:05 -0400
Received: from localhost ([210.55.119.168]) by mta4-rme.xtra.co.nz
          with ESMTP
          id <20011020070559.ZETU18935.mta4-rme.xtra.co.nz@localhost>;
          Sat, 20 Oct 2001 20:05:59 +1300
Date: Sat, 20 Oct 2001 12:58:24 +1300
Subject: Re: build ideas
Content-Type: text/plain; charset=US-ASCII; format=flowed
Mime-Version: 1.0 (Apple Message framework v472)
Cc: Brian Jones <cbj@gnu.org>,
 Classpath list <classpath@gnu.org>
To: Eric Blake <ebb9@email.byu.edu>
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
In-Reply-To: <3BCD8B0F.97854522@email.byu.edu>
Message-Id: <2E4A955D-C4ED-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
Content-Transfer-Encoding: 7bit
X-Mailer: Apple Mail (2.472)
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Thursday, October 18, 2001, at 02:43  AM, Eric Blake wrote:

> What about the idea of both a Makefile and build.xml approach to
> building the .class files?

This sounds reasonable to me.

regards

Bryce.




From a_rueckert@gmx.net Sat Oct 20 07:00:41 2001
Received: from mail.gmx.net ([213.165.64.20])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15utrt-0002hu-00
	for <classpath@gnu.org>; Sat, 20 Oct 2001 07:00:41 -0400
Received: (qmail 17517 invoked by uid 0); 20 Oct 2001 11:00:37 -0000
Received: from b83d3.pppool.de (HELO hun.bre.de) (213.7.131.211)
  by mail.gmx.net (mp020-rz3) with SMTP; 20 Oct 2001 11:00:37 -0000
From: Andreas Rueckert <a_rueckert@gmx.net>
Reply-To: Classpath list <classpath@gnu.org>
To: Classpath list <classpath@gnu.org>
Subject: Re: build ideas
Date: Sat, 20 Oct 2001 12:50:04 +0200
X-Mailer: KMail [version 1.0.28]
Content-Type: Multipart/Mixed;
  boundary="Boundary-=_nWlrBbmQBhCDarzOwKkYHIDdqSCD"
References: <2E4A955D-C4ED-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
In-Reply-To: <2E4A955D-C4ED-11D5-887F-003065F97F7C@waitaki.otago.ac.nz>
MIME-Version: 1.0
Message-Id: <01102012524604.00767@hun.bre.de>
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

--Boundary-=_nWlrBbmQBhCDarzOwKkYHIDdqSCD
Content-Type: text/plain
Content-Transfer-Encoding: 8bit

Hi!

On Sam, 20 Okt 2001 Bryce McKinlay wrote:
> On Thursday, October 18, 2001, at 02:43  AM, Eric Blake wrote:
> 
> > What about the idea of both a Makefile and build.xml approach to
> > building the .class files?
> 
> This sounds reasonable to me.

I'm trying to build the classpath classes with a converted Argo build.xml file.
This is, how far I've come. I had lots of problems with some packages, that I
simply excluded in the compile task for now.

Ciao,
Andreas

--Boundary-=_nWlrBbmQBhCDarzOwKkYHIDdqSCD
Content-Type: application/x-zip;
  name="build.zip"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="build.zip"
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==

--Boundary-=_nWlrBbmQBhCDarzOwKkYHIDdqSCD--


From ebb9@email.byu.edu Sat Oct 20 21:15:53 2001
Received: from femail2.sdc1.sfba.home.com ([24.0.95.82])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15v7DU-0003zL-00
	for <classpath@gnu.org>; Sat, 20 Oct 2001 21:15:52 -0400
Received: from email.byu.edu ([65.12.64.100]) by femail2.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011021011548.KDXP5454.femail2.sdc1.sfba.home.com@email.byu.edu>
          for <classpath@gnu.org>; Sat, 20 Oct 2001 18:15:48 -0700
Message-ID: <3BD21F6B.DA34DDC9@email.byu.edu>
Date: Sat, 20 Oct 2001 19:05:47 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>
Subject: java.lang.reflect.Proxy
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I finally got all the bugs worked out enough to check in my first
version of java.lang.reflect.Proxy (I started this project several
months ago).  Feel free to provide suggestions on ways to improve it. 
In the process, I also did some cleanup in the java.lang.reflect
package.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From mark@klomp.org Sun Oct 21 14:17:50 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vNAS-0004r4-00; Sun, 21 Oct 2001 14:17:50 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15vNAF-0006Pq-00; Sun, 21 Oct 2001 20:17:35 +0200
Date: Sun, 21 Oct 2001 20:17:35 +0200
From: Mark Wielaard <mark@klomp.org>
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: build ideas
Message-ID: <20011021201735.A24503@klomp.org>
References: <m3heszt09o.fsf@lyta.haphazard.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <m3heszt09o.fsf@lyta.haphazard.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi Brian,

On Tue, Oct 16, 2001 at 11:26:27PM -0400, Brian Jones wrote:
> Like other GNU projects, I would like to continue to use the
> automake/autoconf system for driving the compilation of Classpath.
This seems like a good idea. Especially for the native C/C++ code.

> For releases I'd like to continue to distribute compiled .class files
> and generated JNI headers.
And the configure, Makefile, etc files that are generated from the CVS
source I presume. Is there a way to ship with the generic jni.h file
that Etienne Gagnon made? He released it in the public domain. It is a
real pain to have to have the jni.h file from a particular VM around to
compile the native libraries.

> I want CVS builds to generate JNI headers and .class files relying only
> on free software.
I would suggest to use the GNU Compiler Collection (GCC) tools by default.
It includes a C compiler (gcc), a java to bytecode generator (gcj -C)
and a JNI header generator (gcjh -jni). Although it would be nice to be
able to override these defaults in the configure script (jikes, kaffeh, etc).

> There exists a need
> within our community to generate .class files without compiling native
> libraries and I'd like that to be easier... it seems to me that to be
> successful this must be faster than 'jikes -d /tmp `find java gnu
> vm/reference -regex ".*\.java$"` or some working equivalent.
It would be nice to include a 'classes' file in the release that lists
all the java source files. Such a file can easily be used with any java
compiler that supports the "@file" construct (I think almost all do).
We do have such a file at the moment but it does not seem to be widely
known. Maybe this is just a documentation issue. But maybe it is not
very convenient that this files lives in the lib directory and contains
../java entries. Move it to the toplevel directory?

It would also be nice to have different classes @files for subprojects
such as jazzlib which contains only the java.util.zip classes or a
version that excludes all (LGPLed) AWT code, or standalone versions of
the RMI sources, SQL, etc.
But if such a subproject also includes some native files we also need
a easy way to define which native files belong to which subproject.
It seems libgcj tries to do something like this.
 
> Additions?  Contrary notions?  Devil's advocate?
In the (far, far) future it would be nice to split the libgcj library in
a real runtime libary and a classes library, then we could create the
(native) gcj classes library straight from the Classpath sources.

Cheers,

Mark

P.S. I just bought a paper version of "The Goat Book"
     <http://sources.redhat.com/autobook/>
     So please let me know if you need any help.
     (then I will have a real reason to actually read it :)
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sun Oct 21 15:07:32 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vNwa-0006SN-00
	for <classpath@gnu.org>; Sun, 21 Oct 2001 15:07:32 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15vNw7-0006fX-00; Sun, 21 Oct 2001 21:07:03 +0200
Date: Sun, 21 Oct 2001 21:07:03 +0200
From: Mark Wielaard <mark@klomp.org>
To: Nic Ferrier <nferrier@tf1.tapsellferrier.co.uk>
Cc: classpath@gnu.org
Subject: Re: build ideas
Message-ID: <20011021210703.B24503@klomp.org>
References: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <E15tqAs-0004br-00@tf1.tapsellferrier.co.uk>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Wed, Oct 17, 2001 at 01:51:54PM +0100, Nic Ferrier wrote:
> 
> Personally, I don't like Ant. It's not *free* software either.
It Free Software. It is distributed under the Apache License.
Which is a Free Software license. This does make it GPL incompatible
since it includes the trademark license clauses which is a added restriction
according to the GPL. But since ANT is a standalonce tool which you will
probably never embed into a GPLed program I don't think this is such a
problem. For more info see:
<http://www.gnu.org/licenses/license-list.html#GPLIncompatibleLicenses>

> There's not much you can do with Ant that can't be done with Make
> (given a little thought).
> 
> Ant is also very Java centric, Classpath has a lot of C (and C++ if we
> include the GCJ stuff). Make is perfect for those languages.
But this is a valid argument.
And even if we needed to only compile the java source code and generate
a zip/jar file Ant would be a bit overkill. In that case a classes @file,
a java compiler and a zip/jar tool are enough. Just put that into a simple
Makefile or shell script and you are done.

Cheers,

Mark

-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sun Oct 21 15:44:46 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vOWb-0007tx-00
	for <classpath@gnu.org>; Sun, 21 Oct 2001 15:44:46 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15vOWC-0006kU-00; Sun, 21 Oct 2001 21:44:20 +0200
Date: Sun, 21 Oct 2001 21:44:20 +0200
From: Mark Wielaard <mark@klomp.org>
To: Eric Blake <ebb9@email.byu.edu>
Cc: Classpath list <classpath@gnu.org>
Subject: Re: java.util
Message-ID: <20011021214420.C24503@klomp.org>
References: <3BCF7392.2FE55628@email.byu.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BCF7392.2FE55628@email.byu.edu>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Thu, Oct 18, 2001 at 06:28:02PM -0600, Eric Blake wrote:
> I just committed a large change to some of the core classes of java.util
> (Abstract*.java, Arrays, BasicMapEntry, and Collections).  A lot of it
> was just adding javadoc, formatting to GNU standards, and replacing some
> code with more efficient constructs, but I also fixed a few bugs.

Great! I really appreciate all the javadoc comments, this will allow us to
generate a reference manual from our sources. I also saw that you updated
a couple of classes to the 1.4 (beta) spec and included a @since 1.4 tag.
This will make it easy to track any changes in that spec.

The Java Collections Clearinghouse <http://javacollections.org/> contain
some Test Suites that you might want to run.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sun Oct 21 16:45:13 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vPT7-0001fe-00; Sun, 21 Oct 2001 16:45:13 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15vPSt-0006o6-00; Sun, 21 Oct 2001 22:44:59 +0200
Date: Sun, 21 Oct 2001 22:44:58 +0200
From: Mark Wielaard <mark@klomp.org>
To: Eric Blake <ebb9@email.byu.edu>
Cc: Brian Jones <cbj@gnu.org>, Classpath list <classpath@gnu.org>
Subject: Re: question about vm/reference files
Message-ID: <20011021224458.D24503@klomp.org>
References: <3BD0C60E.5D2C1554@email.byu.edu> <m3bsj3owpf.fsf@lyta.haphazard.org> <3BD0E872.6AB424BC@email.byu.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BD0E872.6AB424BC@email.byu.edu>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Fri, Oct 19, 2001 at 08:58:58PM -0600, Eric Blake wrote:
> > The thought was that Classpath either would not provide the classes
> > that are VM dependent (leaving those to the VM) or that there would be
> > multiple directories (vm/gcj, vm/orp, vm/kaffe, etc.)  The ugly
> > complexity appears to be necessary.
> 
> That still doesn't explain why Object is in the top level when the other
> core classes are not.  Even if these classes have no Java
> implementation, it is still useful to have a .java file in Classpath for
> making the .class file which jikes requires for compilation purposes.

The original idea was that some of these classes were to VM specific to
include a generic Classpath version. Some classes have a VMXXX counterpart
that contains the VM generic code but since they do not have to have direct
access to

Now that we have some VMs that actually use Classpath as their standard
class library we can reevaluate the way we split these classes. Which
classes are really so VM specific that they cannot be implemented with a
VMXXX bridge?

What we could do is build two zip files, one with the standard classes and
one with the generic implementation classes (glibj-runtime.zip?) that each
VM should implement in their own VM specific way. (So this second zip would
be very small and contain the 'default' - noop - implementation of the VM
specific classes.)
Then those two zip combined could be used as input for compilers like jikes.
And each VM would have the generic zip file plus an vmspecific zip file in
its bootclasspath.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sun Oct 21 17:06:01 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vPnF-0002bT-00
	for <classpath@gnu.org>; Sun, 21 Oct 2001 17:06:01 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15vPn2-0006pU-00; Sun, 21 Oct 2001 23:05:48 +0200
Date: Sun, 21 Oct 2001 23:05:48 +0200
From: Mark Wielaard <mark@klomp.org>
To: Eric Blake <ebb9@email.byu.edu>
Cc: Classpath list <classpath@gnu.org>
Subject: Re: java.lang.reflect.Proxy
Message-ID: <20011021230548.E24503@klomp.org>
References: <3BD21F6B.DA34DDC9@email.byu.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BD21F6B.DA34DDC9@email.byu.edu>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Sat, Oct 20, 2001 at 07:05:47PM -0600, Eric Blake wrote:
> I finally got all the bugs worked out enough to check in my first
> version of java.lang.reflect.Proxy (I started this project several
> months ago).  Feel free to provide suggestions on ways to improve it. 
Impressive. I will print out the source code and see if I understand it
and give some comments on it.

> In the process, I also did some cleanup in the java.lang.reflect
> package.
I saw that you also changed some exception things "in anticipation of 1.4
compatibility". I have been working on that, but it has been a while ago.
An old version can be found in the Classpath mail archives. I also have
some local libgcj hacks that almost work. But the Classpath/libgcj merge
work is currently eating up all my Classpath time so it might take a while
for that code to appear in Classpath in full working order. I will see if
I can make some more time to really finish it.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From ebb9@email.byu.edu Mon Oct 22 00:01:49 2001
Received: from femail38.sdc1.sfba.home.com ([24.254.60.32])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15vWHd-0007fZ-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 00:01:49 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail38.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011022040147.CRNR624.femail38.sdc1.sfba.home.com@email.byu.edu>
          for <classpath@gnu.org>; Sun, 21 Oct 2001 21:01:47 -0700
Message-ID: <3BD399A0.5A4710EA@email.byu.edu>
Date: Sun, 21 Oct 2001 21:59:28 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>
Subject: more java.util
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I made another commit to java.util, this time to the List related
files.  Again, my biggest change was standardizing the javadoc and
formatting, but I also patched a few bugs.

2001-10-21  Eric Blake  <ebb9@email.byu.edu>

	* java/util/AbstractList.java:
	(modCount): Make sure it is updated in all needed places.
	* java/util/ArrayList.java: Improve javadoc. Implements
	RandomAccess. Add serialVersionUID. Reorder methods.
	(modCount): Make sure it is updated in all needed places.
	(rangeExclusive, rangeInclusive): Add common methods for bounds
	check.
	(isEmpty): Add missing method.
	* java/util/Collections.java: (class SynchronizedList): Make
	package visible.
	* java/util/ConcurrentModificationException.java: Improve
	javadoc.
	* java/util/EmptyStackException.java: Improve javadoc.
	* java/util/LinkedList.java: Improve javadoc.
	(modCount): Make sure it is updated in all needed places.
	(rangeExclusive, rangeInclusive): Add common methods for bounds
	check.
	* java/util/NoSuchElementException.java: Improve javadoc.
	* java/util/Stack.java: Improve javadoc. Fix synchronization
	issues.
	(modCount): Make sure it is updated in all needed places.
	* java/util/Vector.java: Improve javadoc. Fix synchronization
	issues. Implements RandomAccess. Reorder methods.
	(modCount): Make sure it is updated in all needed places.
	(setSize): Fix according to specifications: this does not dictate
	the backing array size.
	(removeAll, retainAll): Faster implementations.

Still outstanding in my local tree are modifications to the Set/Map
classes.  In the meantime, feel free to point out any bugs in my recent
commits.

I've noticed two main styles in java.util .java layouts, which probably
extend to other files.  One style lists public methods in alphabetical
order, the other lists them in the same order as the JDK web pages.  Is
there a preference as to which is better?  Also, is there a preference
as to whether field declarations should come before or after methods? 
I've sort of leaning to this approach (in part because it is similar to
the JDK web page layout):

class Foo
{
  static fields (constants first)
  instance fields
  constructors
  abstract methods
  other public methods, in alphabetical order
  helper (package-)private methods
  nested classes
}

Any preferences one way or another? Or flames in my general direction
for messing things up? If one layout style is preferred as the standard,
it would be a nice thing to document on the Classpath website and the
HACKING instructions.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From ebb9@email.byu.edu Mon Oct 22 01:56:24 2001
Received: from femail37.sdc1.sfba.home.com ([24.254.60.31])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15vY4V-0002Z3-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 01:56:24 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail37.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011022055622.GGOJ14145.femail37.sdc1.sfba.home.com@email.byu.edu>;
          Sun, 21 Oct 2001 22:56:22 -0700
Message-ID: <3BD3B442.423BD1F8@email.byu.edu>
Date: Sun, 21 Oct 2001 23:53:06 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
CC: Classpath list <classpath@gnu.org>
Subject: Re: classpath ./ChangeLog java/util/AbstractList.ja...
References: <E15vW2R-0000H0-00@subversions.gnu.org> <3BD3A4A1.5090402@waitaki.otago.ac.nz>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Agreed, and I just committed a patch.

Thanks for the feedback - it's nice to know my work has an audience.

Bryce McKinlay wrote:
> 
> Hi Eric,
> 
> Thanks!!
> 
> >       (rangeExclusive, rangeInclusive): Add common methods for bounds
> >       check.
> >
> 
> These should really be called checkBoundsInclusive and
> checkBoundsExclusive for consistency with AbstractList (plus, it makes
> it clearer what they actually do when reading the code...)
> 

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From bryce@waitaki.otago.ac.nz Mon Oct 22 02:06:04 2001
Received: from waitaki.otago.ac.nz ([139.80.75.140])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vYDs-0002n3-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 02:06:04 -0400
Received: from waitaki.otago.ac.nz (IDENT:bryce@reason.otago.ac.nz [139.80.119.17])
	by waitaki.otago.ac.nz (8.11.2/8.11.2) with ESMTP id f9M68AW18775;
	Mon, 22 Oct 2001 19:08:10 +1300
Message-ID: <3BD3B7C3.3060002@waitaki.otago.ac.nz>
Date: Mon, 22 Oct 2001 19:08:03 +1300
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.5) Gecko/20011012
X-Accept-Language: en-us
MIME-Version: 1.0
To: Eric Blake <ebb9@email.byu.edu>
CC: Classpath list <classpath@gnu.org>
Subject: Re: classpath ./ChangeLog java/util/AbstractList.ja...
References: <E15vW2R-0000H0-00@subversions.gnu.org> <3BD3A4A1.5090402@waitaki.otago.ac.nz> <3BD3B442.423BD1F8@email.byu.edu>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Eric Blake wrote:

>Agreed, and I just committed a patch.
>
>Thanks for the feedback - it's nice to know my work has an audience.
>

You're welcome!

I also looked at your implementations of removeAll/retainAll in Vector 
-- nice. But don't you think its odd that there would be optimized 
implementations in Vector but not ArrayList? I wonder if we should put 
those in ArrayList as well. In the past I have avoided including any 
public methods that wern't part of the Javadoc spec, but that was more 
to help ensure that the implementation worked similarly to Sun's rather 
than for any real correctness reasons, since of course adding extra 
virtual methods should not effect binary compatibility in any way in 
Java. What do you think?

regards

Bryce.




From ebb9@email.byu.edu Mon Oct 22 02:44:26 2001
Received: from femail25.sdc1.sfba.home.com ([24.254.60.15])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15vYp0-00043N-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 02:44:26 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail25.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011022064425.EOOA4954.femail25.sdc1.sfba.home.com@email.byu.edu>;
          Sun, 21 Oct 2001 23:44:25 -0700
Message-ID: <3BD3BF71.A1E59567@email.byu.edu>
Date: Mon, 22 Oct 2001 00:40:49 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
CC: Classpath list <classpath@gnu.org>
Subject: Re: classpath ./ChangeLog java/util/AbstractList.ja...
References: <E15vW2R-0000H0-00@subversions.gnu.org>
	 <3BD3A4A1.5090402@waitaki.otago.ac.nz> <3BD3B442.423BD1F8@email.byu.edu> <3BD3B7C3.3060002@waitaki.otago.ac.nz>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Bryce McKinlay wrote:

> I also looked at your implementations of removeAll/retainAll in Vector
> -- nice. But don't you think its odd that there would be optimized
> implementations in Vector but not ArrayList? I wonder if we should put
> those in ArrayList as well. In the past I have avoided including any
> public methods that wern't part of the Javadoc spec, but that was more
> to help ensure that the implementation worked similarly to Sun's rather
> than for any real correctness reasons, since of course adding extra
> virtual methods should not effect binary compatibility in any way in
> Java. What do you think?

I dunno - I'm hesitant to add unspecified public methods to public
classes, even though as you point out they will not destroy binary
compatibility.

I think Sun didn't notice the possibility for my optimization.  I guess
Vector has removeAll when ArrayList doesn't simply because Vector needs
to be synchronized.  Which is odd, because Sun didn't allow for iterator
or listIterator in Vector, so I was unable to create iterators which
were properly synchronized.  Maybe we could use package-private hooks to
work around these spec shortcomings while still maintaining the public
API:

AbstractList.java:
public boolean removeAll(Collection c)
{
  return internalRemoveAll(c);
}
boolean internalRemoveAll(Collection c)
{
  // existing implementation
}
public Iterator iterator()
{
  return internalIterator();
}
Iterator internalIterator()
{
  // existing implementation
}

ArrayList.java:
boolean internalRemoveAll(Collection c)
{
  // optimized implementation
}

Vector.java:
Iterator internalIterator()
{
  // synchronized implementation
}

It was a little easier working with the nested classes in Collections. 
For example, I added methods like Collections.EmptySet.toArray, which
Sun did not implement (of course, I tested this using only
java.lang.reflect, and not by decompiling or looking at JDK source). 
But in those nested classes, I did not feel guilty adding public methods
because the entire class is private and undocumented except for its
serializable nature.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From cbj@gnu.org Mon Oct 22 12:56:54 2001
Received: from fe7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15viNi-0008VU-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 12:56:54 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Mon, 22 Oct 2001 12:56:52 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9MGt3I04035;
	Mon, 22 Oct 2001 12:55:03 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: build ideas
References: <m3heszt09o.fsf@lyta.haphazard.org> <20011021201735.A24503@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 22 Oct 2001 12:55:02 -0400
In-Reply-To: Mark Wielaard's message of "Sun, 21 Oct 2001 20:17:35 +0200"
Message-ID: <m34ror3995.fsf@lyta.haphazard.org>
Lines: 66
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> Hi Brian,
> 
> On Tue, Oct 16, 2001 at 11:26:27PM -0400, Brian Jones wrote:
> > Like other GNU projects, I would like to continue to use the
> > automake/autoconf system for driving the compilation of Classpath.
> This seems like a good idea. Especially for the native C/C++ code.
> 
> > For releases I'd like to continue to distribute compiled .class files
> > and generated JNI headers.
>
> And the configure, Makefile, etc files that are generated from the CVS
> source I presume. Is there a way to ship with the generic jni.h file
> that Etienne Gagnon made? He released it in the public domain. It is a
> real pain to have to have the jni.h file from a particular VM around to
> compile the native libraries.

Yes, I want to add jni.h with the machine dependent parts generated
via configure.  No idea about the cross-compile implications yet.  I'm
not happy with the gtk/glib stuff breaking cross-compile either.

> > I want CVS builds to generate JNI headers and .class files relying only
> > on free software.
>
> I would suggest to use the GNU Compiler Collection (GCC) tools by default.
> It includes a C compiler (gcc), a java to bytecode generator (gcj -C)
> and a JNI header generator (gcjh -jni). Although it would be nice to be
> able to override these defaults in the configure script (jikes, kaffeh, etc).

Something has to be sort of preferred, so that's fine with me.  I've
actually already got the class, jni, and cni stuff split in configure
but I spent too much time on home improvement over the weekend to get
something to the point I could check it in.
 
> > There exists a need
> > within our community to generate .class files without compiling native
> > libraries and I'd like that to be easier... it seems to me that to be
> > successful this must be faster than 'jikes -d /tmp `find java gnu
> > vm/reference -regex ".*\.java$"` or some working equivalent.
>
> It would be nice to include a 'classes' file in the release that lists
> all the java source files. Such a file can easily be used with any java
> compiler that supports the "@file" construct (I think almost all do).
> We do have such a file at the moment but it does not seem to be widely
> known. Maybe this is just a documentation issue. But maybe it is not
> very convenient that this files lives in the lib directory and contains
> ../java entries. Move it to the toplevel directory?

The documentation probably should mention things like this.  Is there
a way to exec a javac-like program within the lib directory and use
the @file syntax without including the relative path as in '../' ?
 
> It would also be nice to have different classes @files for subprojects
> such as jazzlib which contains only the java.util.zip classes or a
> version that excludes all (LGPLed) AWT code, or standalone versions of
> the RMI sources, SQL, etc.
> But if such a subproject also includes some native files we also need
> a easy way to define which native files belong to which subproject.
> It seems libgcj tries to do something like this.

If you have a specific need then it can be put into a list of things
to do.
  
-- 
Brian Jones <cbj@gnu.org>


From nferrier@tapsellferrier.co.uk Mon Oct 22 13:41:11 2001
Received: from tf1.tapsellferrier.co.uk ([195.7.242.226])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vj4Y-0002bY-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 13:41:10 -0400
Received: from nferrier by tf1.tapsellferrier.co.uk with local (Exim 3.22 #3 (Debian))
	id 15vjF6-0000VB-00; Mon, 22 Oct 2001 18:52:04 +0100
From: Nic Ferrier <nferrier@tf1.tapsellferrier.co.uk>
To: classpath@gnu.org
Subject: Re: build ideas
Message-Id: <E15vjF6-0000VB-00@tf1.tapsellferrier.co.uk>
Date: Mon, 22 Oct 2001 18:52:04 +0100
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark said:
> It would be nice to include a 'classes' file in the release that lists
> all the java source files. Such a file can easily be used with any java
> compiler that supports the "@file" construct (I think almost all do).
> We do have such a file at the moment but it does not seem to be widely
> known. Maybe this is just a documentation issue. But maybe it is not
> very convenient that this files lives in the lib directory and contains
> ../java entries. Move it to the toplevel directory?

Brian replied:

>The documentation probably should mention things like this.  Is there
>a way to exec a javac-like program within the lib directory and use
>the @file syntax without including the relative path as in '../' ?

Not without a script no.


Mark said:
> It would also be nice to have different classes @files for subprojects
> such as jazzlib which contains only the java.util.zip classes or a
> version that excludes all (LGPLed) AWT code, or standalone versions of
> the RMI sources, SQL, etc.
> But if such a subproject also includes some native files we also need
> a easy way to define which native files belong to which subproject.
> It seems libgcj tries to do something like this.

The ClasspathX/Paperclips make file system has tried to address this issue.

In ClasspathX makefiles there is a source file list which describes
all the source files for a particular target. These are automatically
turned into equivalent .class files.

Multiple targets can easily be maintained by simply using multiple
source lists.


Nic


From bryce@waitaki.otago.ac.nz Mon Oct 22 23:02:31 2001
Received: from waitaki.otago.ac.nz ([139.80.75.140])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15vrpm-0006wG-00
	for <classpath@gnu.org>; Mon, 22 Oct 2001 23:02:30 -0400
Received: from waitaki.otago.ac.nz (IDENT:bryce@reason.otago.ac.nz [139.80.119.17])
	by waitaki.otago.ac.nz (8.11.2/8.11.2) with ESMTP id f9N34aW22045;
	Tue, 23 Oct 2001 16:04:36 +1300
Message-ID: <3BD4DE3D.90908@waitaki.otago.ac.nz>
Date: Tue, 23 Oct 2001 16:04:29 +1300
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.5) Gecko/20011012
X-Accept-Language: en-us
MIME-Version: 1.0
To: Eric Blake <ebb9@email.byu.edu>
CC: Classpath list <classpath@gnu.org>
Subject: Re: classpath ./ChangeLog java/util/AbstractList.ja...
References: <E15vW2R-0000H0-00@subversions.gnu.org>	 <3BD3A4A1.5090402@waitaki.otago.ac.nz> <3BD3B442.423BD1F8@email.byu.edu> <3BD3B7C3.3060002@waitaki.otago.ac.nz> <3BD3BF71.A1E59567@email.byu.edu>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Eric Blake wrote:

>Bryce McKinlay wrote:
>
>>I also looked at your implementations of removeAll/retainAll in Vector
>>-- nice. But don't you think its odd that there would be optimized
>>implementations in Vector but not ArrayList? I wonder if we should put
>>those in ArrayList as well. In the past I have avoided including any
>>public methods that wern't part of the Javadoc spec, but that was more
>>to help ensure that the implementation worked similarly to Sun's rather
>>than for any real correctness reasons, since of course adding extra
>>virtual methods should not effect binary compatibility in any way in
>>Java. What do you think?
>>
>
>I dunno - I'm hesitant to add unspecified public methods to public
>classes, even though as you point out they will not destroy binary
>compatibility.
>
>I think Sun didn't notice the possibility for my optimization.  I guess
>Vector has removeAll when ArrayList doesn't simply because Vector needs
>to be synchronized.  Which is odd, because Sun didn't allow for iterator
>or listIterator in Vector, so I was unable to create iterators which
>were properly synchronized.  Maybe we could use package-private hooks to
>work around these spec shortcomings while still maintaining the public
>API:
>

There is no need to synchronize the Vector iterator. Its made pretty 
clear in the docs that iterators should never be shared among different 
threads (hence the ConcurrentModificationException), and of course any 
operations you actually perform on the iterator are still synchronized 
internally against the underlying Vector. Note also that there is no 
Collections.synchronizedIterator - it doesn't make sense because even if 
it were synchronized it still wouldn't be thread-safe.

>It was a little easier working with the nested classes in Collections. 
>For example, I added methods like Collections.EmptySet.toArray, which
>Sun did not implement 
>

Actually I think these particular "optimizations" are very dubious!! 
;-). We should definatly try to optimize when the default behaviour can 
take quadratic time etc, but in this case I think the methods are not 
performance critical and don't justify the extra code. Your comment says 
that they "provide a performance advantage by not allocating unnecessary 
iterators in AbstractList", but I don't see why an iterator ever needs 
to be allocated. Why not just have a single, static empty iterator like 
before?

regards

Bryce.




From green@redhat.com Wed Oct 17 15:26:32 2001
Received: from runyon.cygnus.com ([205.180.230.5] helo=cygnus.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15twKl-0007lM-00; Wed, 17 Oct 2001 15:26:31 -0400
Received: from locutus (locutus.cygnus.com [205.180.230.91])
	by runyon.cygnus.com (8.8.7-cygnus/8.8.7) with SMTP id MAA05251;
	Wed, 17 Oct 2001 12:26:25 -0700 (PDT)
Message-ID: <015201c15742$60d340c0$5be6b4cd@cygnus.com>
From: "Anthony Green" <green@redhat.com>
To: <tromey@redhat.com>, "Brian Jones" <cbj@gnu.org>
Cc: "Classpath Project" <classpath@gnu.org>
References: <87bsj6ku5u.fsf@creche.redhat.com> <m3lmiam5zr.fsf@lyta.haphazard.org>
Subject: Re: java.security.Security proposal
Date: Wed, 17 Oct 2001 12:31:54 -0700
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4522.1200
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Brian wrote:
> I apologize for not having read that link yet, but how does changing
> the name to gcj.security have to do with fixing a bug?  Is this really
> gcj specific?

It's possible to have an installation with native shared library providers and
no .jar implementation.  You really don't want to have a global
classpath.security file in this case because gcj is the only implementation
that will find and use those providers.

AG




From chris.gray@acunia.com Thu Oct 18 08:35:43 2001
Received: from nat.be.acunia.com ([194.7.211.211] helo=pigeon.smove.be)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15uCOh-0004Rt-00
	for <classpath@gnu.org>; Thu, 18 Oct 2001 08:35:43 -0400
Received: from localhost.localdomain (IDENT:nobody@loki.leuven.intern.acunia.com [192.168.8.254])
	by pigeon.smove.be (8.11.6/8.11.6) with ESMTP id f9ICZeC20812
	for <classpath@gnu.org>; Thu, 18 Oct 2001 14:35:40 +0200
Received: from acunia.com (IDENT:gray@sophocles [127.0.0.1])
	by localhost.localdomain (8.9.3/8.9.3) with ESMTP id OAA06252;
	Wed, 17 Oct 2001 14:38:34 +0200
Message-ID: <3BCD7BC9.936DAD55@acunia.com>
Date: Wed, 17 Oct 2001 14:38:33 +0200
From: Chris Gray <chris.gray@acunia.com>
Organization: ACUNIA NV
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: tromey@redhat.com
CC: Mark Wielaard <mark@klomp.org>, classpath@gnu.org
Subject: Re: Partial Double/Float merge with libgcj
References: <20011012165931.A26185@klomp.org> <878zeg22m9.fsf@creche.redhat.com> <3BCA9676.6310815F@acunia.com> <87het0vbxi.fsf@creche.redhat.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey wrote:

> >>>>> "Chris" == Chris Gray <chris.gray@acunia.com> writes:
>
> Chris> An interesting twist on this: jikes seems to have its own idea
> Chris> of the correct bit-pattern for NaN (0xffc00000 instead of
> Chris> 0x7fc0000).  This shouldn't matter if all NaN's are
> Chris> canonicalized to the same value.
>
> My understanding is that floatToIntBits is required to return
> 0x7fc00000 for any NaN, and that Float.NaN must be the same as
> Float.intBitsToFloat(0x7fc00000).
>
> This doesn't necessarily imply that there is a Jikes bug though.

No.  It just means that Float.floatToRawIntBits(Float.NaN) returns
0xffc00000 when you might naively have expected 0x7fc00000.

>
> It depends on exactly when Jikes uses that particular value.
> Float.floatToRawIntBits can return other NaN values, and presumably
> the VM could use them as well.  For instace libgcj uses whatever
> the underlying hardware decides, and only does the canonicalization
> when the real bit pattern might be seen by Java.
>
> Tom
>
> _______________________________________________
> Classpath mailing list
> Classpath@gnu.org
> http://mail.gnu.org/mailman/listinfo/classpath



From mark@klomp.org Tue Oct 23 17:40:07 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15w9HL-0002mA-00
	for <classpath@gnu.org>; Tue, 23 Oct 2001 17:40:07 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15w9H8-0001fr-00; Tue, 23 Oct 2001 23:39:54 +0200
Date: Tue, 23 Oct 2001 23:39:54 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: Status Update
Message-ID: <20011023233954.A6191@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi all,

Time to give a little status update. The past month there has been lots
of work on merging more of Classpath and libgcj. According to Tom Tromey
we are now 100 classes closer to a full merge! And according to the status
page we have about 90 classes to go. See:
<http://gcc.gnu.org/java/libgcj-classpath-compare.html>

I hoped to contribute a bit more to this effort so we would have most
of it done by the end of the year. But I don't think we can really finish
it in only two months because a large part of the merge work up till now
consisted of simple Exception and Interface merges. The remaining classes
have lots of native code and lots of underlying gnu.* classes that are not
shown in the comparison page. And I think there are a couple of more
important things to do for the project at this point then finishing the
merge.

A few recent events made me realize that we should really work more on
our packaging, build process, documentation and VM interface. A friend
of mine wanted to help hack on the Classpath source but had a very hard
time getting a recent VM/Classpath setup that worked (to be honest he
still doesn't have a working setup, but is determined to get one).
Debian (sid/unstable) currently comes with three (!) different versions
of Classpath, one for gcj, one for Kissme and one for Orp.
And a discussion with David Brownell and Nic Ferrier of the ClasspathX
project (on the GCJ java@gcc.gnu.org mailing list, yeah for integration!)
made me realize that we don't have an easy way in share standard libraries
with the other GNU projects (or even non-GNU projects).

I know some people on the list have said this before, but until now I
didn't realize what all the fuss was about. You may call me slow...

So the coming month I will spend some time on getting our documentation
more up to date. Try to get both the Kissme and Orp VM (maybe others?) to
work with the current Classpath from CVS. It seems that this will be easier
then finishing the libgcj port at this moment. And it will give me a better
insight in our VM interface. I will also work on my 1.4 Throwable support
since I think this will make our VM interface simpler.

If there is time (or just because it is fun) I will look if I can help with
the build infrastructure. It works for me, but lots of people on the list
say it is complex. I also like the makeJazzlibDist and mkcollections scripts
because they give us ways to release parts of Classpath. Maybe they can be
genralized to other packages?

Brian when do you expect to make a new Classpath release? And what do you
think should be in it?

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Tue Oct 23 22:51:41 2001
Received: from fe8.southeast.rr.com ([24.93.67.55] helo=mail8.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wE8r-0005PK-00
	for <classpath@gnu.org>; Tue, 23 Oct 2001 22:51:41 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail8.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Tue, 23 Oct 2001 22:51:37 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9O2nRs07356;
	Tue, 23 Oct 2001 22:49:27 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: Status Update
References: <20011023233954.A6191@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 23 Oct 2001 22:49:27 -0400
In-Reply-To: Mark Wielaard's message of "Tue, 23 Oct 2001 23:39:54 +0200"
Message-ID: <m3d73doiq0.fsf@lyta.haphazard.org>
Lines: 22
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> Brian when do you expect to make a new Classpath release? And what do you
> think should be in it?

I'd like to finish my updates to the build and let folks provide
feedback on that.  Then I'd like to consider going ahead with a
release to be followed up by a focus on just fixing broken native code
(any any necessary .java problems) followed by another release.

With regards to ORP I did a quick hack to get it to use our JNI
libraries and something like that will end up in a subsequent release
of ORP.  Unfortunately this did not make it into the most recent
release a few days ago.

I'd like more experience with gcj and other JVMs using Classpath as
well.  But I mostly find I just need a platform to write and test Java
and JNI native code with that just works.

Brian
-- 
Brian Jones <cbj@gnu.org>


From reali@inf.ethz.ch Wed Oct 24 03:35:21 2001
Received: from core.inf.ethz.ch ([129.132.178.196])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wIZM-0006ux-00
	for <classpath@gnu.org>; Wed, 24 Oct 2001 03:35:20 -0400
Received: from lillian-ics.inf.ethz.ch (root@lillian-ics.inf.ethz.ch [129.132.134.2])
	by core.inf.ethz.ch (8.9.3/8.9.3) with ESMTP id JAA26919;
	Wed, 24 Oct 2001 09:35:09 +0200 (MET DST)
Received: from x.oberon.ethz.ch (pookie.inf.ethz.ch [129.132.134.15])
	by lillian-ics.inf.ethz.ch (8.9.3/8.9.3) with SMTP id JAA25217;
	Wed, 24 Oct 2001 09:35:12 +0200 (MET DST)
Date: Wed, 24 Oct 2001 09:35:12 +0200 (MET DST)
Message-Id: <200110240735.JAA25217@lillian-ics.inf.ethz.ch>
From: reali@inf.ethz.ch
X-Mailer: Oberon Mail (ejz) on Aos 03.10.2001
To: classpath@gnu.org
Cc: reali@inf.ethz.ch
Subject: java/io/File, design question
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi!

Can somebody explain, why the native methods of
java/io/File are not static? The call would be slightly
faster (static call, and no self reference parameter
needed).

Thanks!

-Patrik
--
Patrik Reali, reali@inf.ethz.ch
http://www.inf.ethz.ch/personal/reali/
http://www.oberon.ethz.ch/native/



From mark@klomp.org Thu Oct 25 02:48:21 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15weJR-0007pJ-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 02:48:21 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15weJ7-0005jM-00; Thu, 25 Oct 2001 08:48:01 +0200
Date: Thu, 25 Oct 2001 08:48:01 +0200
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: [hindm@us.ibm.com: Re: [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 - questions about licensing.]
Message-ID: <20011025084801.A21952@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

Found on the debian-java mailinglist.

The Jikes VM is interesting because they say it is written mostly in java.
<http://www-124.ibm.com/developerworks/oss/jikesrvm/>
And they encourage to develop a Free Classlibrary for it.
(It would ofcourse be better if we could convince them to donate their
current class library to Classpath.)
I am currently investigating the KissMe VM which is already fairly
integrated with Classpath. If someone else wants to investigate how
hard it would be to support Jikes VM that would be really nice.

Cheers,

Mark
----- Forwarded message from Michael Hind <hindm@us.ibm.com> -----

Importance: Normal
Subject: Re: [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 - questions
 about licensing.
To: burton@openprivacy.org
Cc: "Sam Ruby" <rubys@us.ibm.com>, debian-java <debian-java@lists.debian.org>,
        debian-legal@lists.debian.org,
        jikesrvm-researchers@www-126.southbury.usf.ibm.com
From: "Michael Hind" <hindm@us.ibm.com>
Date: Wed, 24 Oct 2001 22:50:58 -0400


Kevin,

     Thanks for your interest and encouragement regarding Jikes RVM.  The
team is happy to provide a system that others will find useful.

You are correct that the license for the libraries is more restrictive than
the license governing the core RVM.  I agree this is not ideal. The reason
for this is addressed in the Q&A Overview (first link on left navigational
menu), question 12, including here for convenience.

   12. Why are different licenses used for the Jikes RVM and the libraries?

   The Jikes RVM is released under the CPL (Common Public License), which
has been approved
   by the OSI (Open Source Initiative) as a fully certified open source
license. The Jikes RVM was
   written as a test-bed for evaluating new virtual machine technologies.
Research in library issues was
   not a major goal. However, a VM is not useful without a minimal set of
libraries. The libraries
   available with the Jikes RVM were developed independently by developers
at Object Technology
   International, an IBM company. Although the library license is not an
open source license, it does
   facilitate open source development of the RVM. We encourage any users
who would like to
   develop open source libraries that work with Jikes RVM.

You wrote

> I would suggest that a sub-project be started to integrate
> the code from GNU Classpath which should be able to replace the JLibrary
code.

As the last sentence says, we would also encourage such an activity.

Mike

- - - - - - - - - - - - - - - -
Michael Hind
Manager, Dynamic Optimization Group, Jalape?o Project
IBM Watson Research Center
www.research.ibm.com/people/h/hind
hind@watson.ibm.com
914 784-7589, tieline: 863-7589


burton@openprivacy.org (Kevin A. Burton)@www-126.southbury.usf.ibm.com on
10/24/2001 10:15:45 PM

Please respond to jikesrvm-researchers@www-126.southbury.usf.ibm.com

Sent by:  jikesrvm-researchers-admin@www-126.southbury.usf.ibm.com


To:   jikesrvm-researchers@www-126.southbury.usf.ibm.com
cc:   Sam Ruby/Raleigh/IBM@IBMUS, debian-java
      <debian-java@lists.debian.org>, debian-legal@lists.debian.org
Subject:  [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 -
      questions about licensing.



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


Sorry for the x-post but I think this is relevant to everyone included.

Hello.

For starters I would like to congratulate IBM and the RVM team on your
recent
RVM 2.0 release (and published white-papers).  This is definitely a great
contribution to the Java and Open Source communities!

Unfortunately, I have issues with the current RVM licensing terms.

Specifically the licensing of the JLibrary package.

http://www-124.ibm.com/cgi-bin/jikesrvm/jlibraries?rev=R-2001-10-13-22-25-43&jrv=2.0.0


This is a very restrictive license and I don't understand why it was chosen
as
it seems to go against the Open Source nature of the RVM.

This prohibits innovative use of the RVM code.  Specifically integration
with
GNU Classpath, distribution with Debian GNU/Linux, etc.

Is there anyone at IBM we can talk with about changing the terms of the
JLibrary
package to an Open Source license?  If the problem is a reliance on
proprietary
code (AKA SUNs code), I would suggest that a sub-project be started to
integrate
the code from GNU Classpath which should be able to replace the JLibrary
code.

Again.  Congrats.  I am very excited about the RVM release as I think that
the
lack of a quality (stable) Open Source JVM is hurting both the Java and
Open
Source communities.

Thanks.

Kevin

- --

   Need a good Engineer?  Hire me!  ( Java | P2P | XML | Linux | Open
Source )

Kevin A. Burton ( burton@apache.org, burton@openprivacy.org,
burtonator@acm.org )
             Location - San Francisco, CA, Cell - 415.595.9965
        Jabber - burtonator@jabber.org,  Web - http://relativity.yi.org

The invariable mark of wisdom is to see the miraculous in the common.
    - Ralph Waldo Emerson (1803-1882)



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt

iD8DBQE713XAAwM6xb2dfE0RAlAfAKCFhNdFzf3NsW0jARDHat9/53iLgACfQaYW
L8pZxwjCIV53p/ZkYHVXYN8=
=CeMK
-----END PGP SIGNATURE-----
_______________________________________________
Jikesrvm-researchers mailing list
Jikesrvm-researchers@www-124.ibm.com
http://www-124.ibm.com/developerworks/oss/mailman/listinfo/jikesrvm-researchers




--
To UNSUBSCRIBE, email to debian-java-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


----- End forwarded message -----

-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From burton@openprivacy.org Thu Oct 25 03:34:30 2001
Received: from dsl081-061-015.sfo1.dsl.speakeasy.net ([64.81.61.15] helo=openprivacy.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wf25-0001e8-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 03:34:29 -0400
Received: from openprivacy.org (localhost [127.0.0.1])
	by openprivacy.org (8.12.1/8.12.1/Debian -2) with ESMTP id f9P7YSBO027340;
	Thu, 25 Oct 2001 00:34:28 -0700
Received: from localhost (localhost [[UNIX: localhost]])
	by openprivacy.org (8.12.1/8.12.1/Debian -2) id f9P7YSUf027334;
	Thu, 25 Oct 2001 00:34:28 -0700
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: [hindm@us.ibm.com: Re: [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 - questions about licensing.]
References: <20011025084801.A21952@klomp.org>
From: burton@openprivacy.org (Kevin A. Burton)
In-Reply-To: <20011025084801.A21952@klomp.org>
Message-ID: <87zo6gmawt.fsf@universe.yi.org>
User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/21.1
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: 25 Oct 2001 00:34:23 -0700
Lines: 46
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Mark Wielaard <mark@klomp.org> writes:

> Hi,
> 
> Found on the debian-java mailinglist.

I would have CC'd this mailing list but the CC list was already long.  

> The Jikes VM is interesting because they say it is written mostly in java.
> <http://www-124.ibm.com/developerworks/oss/jikesrvm/> And they encourage to
> develop a Free Classlibrary for it.  (It would ofcourse be better if we could
> convince them to donate their current class library to Classpath.)  I am
> currently investigating the KissMe VM which is already fairly integrated with
> Classpath. If someone else wants to investigate how hard it would be to
> support Jikes VM that would be really nice.
<snip>

I took a look at the directory listings in the jar.  It looks like they only
have the basics of a 1.1 style VM.  AKA only the basic stuff.

It would be at least nice to compare and contrast between the different
implementations to see if there is anythign we can bring over.  Of course if it
isn't Free Software there is nothing we can do.

Kevin

- -- 

   Need a good Engineer?  Hire me!  ( Java | P2P | XML | Linux | Open Source )

Kevin A. Burton ( burton@apache.org, burton@openprivacy.org, burtonator@acm.org )
             Location - San Francisco, CA, Cell - 415.595.9965
        Jabber - burtonator@jabber.org,  Web - http://relativity.yi.org

For great justice.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt

iD8DBQE718BCAwM6xb2dfE0RAufdAKCOlX/CviwiXAHTkKfu+NayyKqQ9QCgtKdO
X0HM8dj0rbvtqQgLSq4H7bU=
=P/Wj
-----END PGP SIGNATURE-----


From ebb9@email.byu.edu Thu Oct 25 03:37:18 2001
Received: from femail12.sdc1.sfba.home.com ([24.0.95.108])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15wf4o-0001p5-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 03:37:18 -0400
Received: from email.byu.edu ([65.12.64.100])
          by femail12.sdc1.sfba.home.com
          (InterMail vM.4.01.03.20 201-229-121-120-20010223) with ESMTP
          id <20011025073717.CFL1132.femail12.sdc1.sfba.home.com@email.byu.edu>
          for <classpath@gnu.org>; Thu, 25 Oct 2001 00:37:17 -0700
Message-ID: <3BD7C12E.D65BCE99@email.byu.edu>
Date: Thu, 25 Oct 2001 01:37:18 -0600
From: Eric Blake <ebb9@email.byu.edu>
Organization: BYU Student
X-Mailer: Mozilla 4.76 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Classpath list <classpath@gnu.org>
Subject: Yet more java.util
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tonight I committed my final installment of java.util Collections API
documentation.  I also fixed various bugs I found in the Set and Map
classes.  Once again, feel free to point out bugs or inconsistencies
that I should repair.  Now I can start focusing on porting all my recent
changes to libgcj.

2001-10-25  Eric Blake  <ebb9@email.byu.edu>

	* java/util/AbstractCollection.java (removeAllInternal),
	(retainAllInternal): Add hooks for use by ArrayList.
	* java/util/AbstractList.java: Minor code updates. Fix some
	scoping.
	* java/util/AbstractMap.java: ditto
	* java/util/ArrayList.java (readObject, writeObject): ditto
	(removeAllInternal, retainAllInternal): Optimize.
	* java/util/Arrays.java: ditto
	* java/util/Collections.java: ditto. Change order of parameters
	to equals(Object, Object) to match specs.
	* java/util/Dictionary.java: Improve javadoc.
	(Dictionary): Add explicit constructor.
	* java/util/HashMap.java: Improve javadoc. Rearrange methods to
	follow order in JDK. Cleanups related to recent code migration to
	AbstractMap. Fix some scoping.
	(entrySet): Cache the result.
	(modCount): Ensure that this is updated correctly.
	* java/util/HashSet.java: Improve javadoc. Fix some scoping.
	(init): Add hooks for LinkedHashSet.
	(map): Use "" instead of Boolean.TRUE in backing map. Use
	package-private API where possible for less overhead.
	(readObject, writeObject): Fix serialization.
	* java/util/Hashtable.java: Improve javadoc. Fix some scoping.
	(entrySet, keySet, values): Cache the result.
	(modCount): Ensure that this is updated correctly.
	(contains, remove): Fix NullPointer checking to match specs.
	(class Enumeration): Make more like HashIterator.
	* java/util/IdentityHashMap.java: Minor code updates.
	(modCount): Ensure that this is updated correctly.
	(readObject, writeObject): Fix serialization.
	* java/util/LinkedHashMap.java: Minor code updates. Cleanups
	related to recent code migration to AbstractMap.
	* java/util/LinkedHashSet.java: New file.
	* java/util/LinkedList.java:
	(readObject, writeObject): Fix serialization.
	* java/util/Makefile.am: List recently added files.
	* java/util/Stack.java: Minor code updates.
	* java/util/TreeMap.java: Improve javadoc. Overhaul the class to
	be more efficient. Fix some scoping. Rearrange the methods.
	(nil): Ensure that this can be thread-safe, and make it a static
	final. Initialize it to be more useful as a sentinal node.
	(Node): Specify color in constructor.
	(deleteFixup, insertFixup): Improve comments and algorithm.
	(fabricateTree): Redesign with less overhead.
	(lowestGreaterThan): Add parameter first to make SubMap easier.
	(removeNode): Patch hole where nil was being modified. Choose
	predecessor instead of successor so in-place swap works.
	(class VerifyResult, verifyTree, verifySub, verifyError): Remove
	this dead code after verifying the class works.
	(class SubMap): Rewrite several algorithms to avoid problems with
	comparing nil.
	* java/util/TreeSet.java: Improve javadoc. Fix some scoping.
	(clone): Fix ClassCastException when cloning subSet().
	(readObject, writeObject): Fix serialization.
	* java/util/WeakHashMap.java: Improve javadoc. Fix some scoping.
	(NULL_KEY): Make it compare as null, for ease elsewhere.
	(Class WeakEntry): Rename from Entry, to avoid shadowing
	Map.Entry. Add missing toString.
	(modCount): Ensure that this is updated correctly.
	(clear, containsValue, keySet, putAll, values, WeakHashMap(Map)):
	Add missing methods and constructor.

-- 
This signature intentionally left boring.

Eric Blake             ebb9@email.byu.edu
  BYU student, free software programmer


From cbj@gnu.org Thu Oct 25 18:32:19 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wt2w-000834-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 18:32:18 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Thu, 25 Oct 2001 18:32:17 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9PMWQc01425;
	Thu, 25 Oct 2001 18:32:26 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: burton@openprivacy.org (Kevin A. Burton)
Cc: Mark Wielaard <mark@klomp.org>, classpath@gnu.org
Subject: Re: [hindm@us.ibm.com: Re: [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 - questions about licensing.]
References: <20011025084801.A21952@klomp.org> <87zo6gmawt.fsf@universe.yi.org>
From: Brian Jones <cbj@gnu.org>
Date: 25 Oct 2001 18:32:26 -0400
In-Reply-To: burton@openprivacy.org's message of "25 Oct 2001 00:34:23 -0700"
Message-ID: <m36693mjut.fsf@lyta.haphazard.org>
Lines: 36
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

burton@openprivacy.org (Kevin A. Burton) writes:

> Mark Wielaard <mark@klomp.org> writes:
> 
> > Hi,
> > 
> > Found on the debian-java mailinglist.
> 
> I would have CC'd this mailing list but the CC list was already long.  
> 
> > The Jikes VM is interesting because they say it is written mostly in java.
> > <http://www-124.ibm.com/developerworks/oss/jikesrvm/> And they encourage to
> > develop a Free Classlibrary for it.  (It would ofcourse be better if we could
> > convince them to donate their current class library to Classpath.)  I am
> > currently investigating the KissMe VM which is already fairly integrated with
> > Classpath. If someone else wants to investigate how hard it would be to
> > support Jikes VM that would be really nice.
> <snip>
> 
> I took a look at the directory listings in the jar.  It looks like they only
> have the basics of a 1.1 style VM.  AKA only the basic stuff.
> 
> It would be at least nice to compare and contrast between the different
> implementations to see if there is anythign we can bring over.  Of course if it
> isn't Free Software there is nothing we can do.

Treat their class libraries as you would treat Sun source code.  For
the purposes of the Classpath project, that means don't refer to it.

The Jikes VM team is interested in finding a free software class
library that would support research modifications and distribution
just as their VM does.  I'm hopeful that they will join in our effort.

Brian
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Thu Oct 25 22:15:24 2001
Received: from fe8.southeast.rr.com ([24.93.67.55] helo=mail8.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wwWq-0006o1-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 22:15:24 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail8.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Thu, 25 Oct 2001 22:15:22 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9Q2FYL08553;
	Thu, 25 Oct 2001 22:15:34 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: recent rmi additions
From: Brian Jones <cbj@gnu.org>
Date: 25 Oct 2001 22:15:34 -0400
Message-ID: <m3adyf87uh.fsf@lyta.haphazard.org>
Lines: 6
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Is the $param_0 and similar stuff in the rmi code ok?  gcj -C doesn't
seem to like it very much.

Brian
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Thu Oct 25 23:59:25 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wy9V-0001NY-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 23:59:25 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Thu, 25 Oct 2001 23:59:06 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9Q3xHQ24448;
	Thu, 25 Oct 2001 23:59:17 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: build update
From: Brian Jones <cbj@gnu.org>
Date: 25 Oct 2001 23:59:17 -0400
Message-ID: <m36693831m.fsf@lyta.haphazard.org>
Lines: 12
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I've made progress with making the build much easier, at least with
configure anyway.  I've temporarily broken my own native libraries
build locally until I add in the jni.h stuff.  So I can't commit what
I have yet but I thought folks would like to know my progress.

Btw, I couldn't make gcj the default compiler because it fails to
compile things that jikes is compiling without problems.  However I am
using gcc 2.96 so that could also be the problem there.

Brian
-- 
Brian Jones <cbj@gnu.org>


From mark@klomp.org Fri Oct 26 07:19:28 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15x51L-0001yJ-00; Fri, 26 Oct 2001 07:19:28 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15x518-0003ud-00; Fri, 26 Oct 2001 13:19:14 +0200
Date: Fri, 26 Oct 2001 13:19:10 +0200
From: Mark Wielaard <mark@klomp.org>
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: build update
Message-ID: <20011026131909.A14928@klomp.org>
References: <m36693831m.fsf@lyta.haphazard.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <m36693831m.fsf@lyta.haphazard.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Thu, Oct 25, 2001 at 11:59:17PM -0400, Brian Jones wrote:
> I've made progress with making the build much easier, at least with
> configure anyway.  I've temporarily broken my own native libraries
> build locally until I add in the jni.h stuff.  So I can't commit what
> I have yet but I thought folks would like to know my progress.
Post the diff! I want to play with it :)
 
> Btw, I couldn't make gcj the default compiler because it fails to
> compile things that jikes is compiling without problems.  However I am
> using gcc 2.96 so that could also be the problem there.
Use gcc 3.0.x. I sometimes compile with gcj 3.0.x and 3.1 (from CVS) and it
compiles almost everything out of the box.

There are currently two construct in Classpath that cannot be compiled with
gcj 3.0.2 but they have easy workaround. I have attached the needed patch
and I will check it in. With this gcj 3.0.2 can compile classpath out of the
box with gcj -C -d lib @classes (but see below).

2001-10-26  Mark Wielaard  <mark@klomp.org>

        * java/lang/reflect/Proxy.java (getProxyClass): 
        Workaround for bug in gcj 
        * java/util/TreeMap.java (TreeIterator (int)): likewise

gcj 3.1 has some trouble with our Class.java file (and ClassLoader.java)
and has a new bug that makes LinkedHashMap uncompilable (bug #java/4693).
But I don't think we should worry about gcc 3.1 CVS that much.

But I have not been able to make gcj work from within the lib dir when the
@classes file contains ../java/etc entries. But from the toplevel dir with
a @classes file that contains "normal" java/etc entries it works OK. You do
have to make sure that the @classes list does not contain duplicates though.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Fri Oct 26 07:39:33 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15x5Kn-0002jB-00; Fri, 26 Oct 2001 07:39:33 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15x5Km-0003wb-00; Fri, 26 Oct 2001 13:39:32 +0200
Date: Fri, 26 Oct 2001 13:39:32 +0200
From: Mark Wielaard <mark@klomp.org>
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: build update
Message-ID: <20011026133932.B14928@klomp.org>
References: <m36693831m.fsf@lyta.haphazard.org> <20011026131909.A14928@klomp.org>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="0OAP2g/MAC+5xKAE"
Content-Disposition: inline
In-Reply-To: <20011026131909.A14928@klomp.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

--0OAP2g/MAC+5xKAE
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

Hi,

On Fri, Oct 26, 2001 at 01:19:10PM +0200, Mark Wielaard wrote:
> 
> There are currently two construct in Classpath that cannot be compiled with
> gcj 3.0.2 but they have easy workaround. I have attached the needed patch
> and I will check it in. With this gcj 3.0.2 can compile classpath out of the
> box with gcj -C -d lib @classes (but see below).
> 
> 2001-10-26  Mark Wielaard  <mark@klomp.org>
> 
>         * java/lang/reflect/Proxy.java (getProxyClass): 
>         Workaround for bug in gcj 
>         * java/util/TreeMap.java (TreeIterator (int)): likewise

Forgot to actually attach the patch. It has been checked in.

Cheers,

Mark

-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore

--0OAP2g/MAC+5xKAE
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="gcj-workaround.diff"

Index: java/lang/reflect/Proxy.java
===================================================================
RCS file: /cvs/classpath/java/lang/reflect/Proxy.java,v
retrieving revision 1.1
diff -u -r1.1 Proxy.java
--- java/lang/reflect/Proxy.java	2001/10/21 01:12:06	1.1
+++ java/lang/reflect/Proxy.java	2001/10/26 11:01:13
@@ -256,9 +256,19 @@
 	    ProxyData data = (Configuration.HAVE_NATIVE_GET_PROXY_DATA
 			      ? getProxyData0(loader, interfaces)
 			      : ProxyData.getProxyData(pt));
-	    clazz = (Configuration.HAVE_NATIVE_GENERATE_PROXY_CLASS
-		     ? generateProxyClass0(loader, data)
-		     : new ClassFactory(data).generate(loader));
+
+	    // FIXME workaround for bug in gcj 3.0.x
+	    // Not needed with the latest gcj from cvs
+	    //clazz = (Configuration.HAVE_NATIVE_GENERATE_PROXY_CLASS
+	    //	       ? generateProxyClass0(loader, data)
+	    //         : new ClassFactory(data).generate(loader));
+	    if (Configuration.HAVE_NATIVE_GENERATE_PROXY_CLASS)
+	      clazz = generateProxyClass0(loader, data);
+	    else
+	      {
+	        ClassFactory cf = new ClassFactory(data);
+		clazz = cf.generate(loader);
+	      }
 	  }
 
 	Object check = proxyClasses.put(pt, clazz);
Index: java/util/TreeMap.java
===================================================================
RCS file: /cvs/classpath/java/util/TreeMap.java,v
retrieving revision 1.14
diff -u -r1.14 TreeMap.java
--- java/util/TreeMap.java	2001/10/25 07:34:19	1.14
+++ java/util/TreeMap.java	2001/10/26 11:01:13
@@ -1391,7 +1391,11 @@
      */
     TreeIterator(int type)
     {
-      this(type, firstNode(), nil);
+      // FIXME gcj cannot handle this. Bug java/4695
+      // this(type, firstNode(), nil);
+      this.type = type;
+      this.next = firstNode();
+      this.max = nil;
     }
 
     /**

--0OAP2g/MAC+5xKAE--


From cbj@gnu.org Fri Oct 26 09:52:10 2001
Received: from fe7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15x7P7-0006wv-00
	for <classpath@gnu.org>; Fri, 26 Oct 2001 09:52:09 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Fri, 26 Oct 2001 09:52:08 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9QDqDp03255;
	Fri, 26 Oct 2001 09:52:13 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: build update
References: <m36693831m.fsf@lyta.haphazard.org> <20011026131909.A14928@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 26 Oct 2001 09:52:12 -0400
In-Reply-To: Mark Wielaard's message of "Fri, 26 Oct 2001 13:19:10 +0200"
Message-ID: <m31yjqmrub.fsf@lyta.haphazard.org>
Lines: 25
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="=-=-="
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

--=-=-=

Mark Wielaard <mark@klomp.org> writes:

> On Thu, Oct 25, 2001 at 11:59:17PM -0400, Brian Jones wrote:
> > I've made progress with making the build much easier, at least with
> > configure anyway.  I've temporarily broken my own native libraries
> > build locally until I add in the jni.h stuff.  So I can't commit what
> > I have yet but I thought folks would like to know my progress.
>
> Post the diff! I want to play with it :)
>  

Current diff attached.  By default will only compile Java source.  To
attempt JNI library compilation use --enable-jni (it will fail because
of no jni.h).  To use gcj for bytecode compilation, use --with-gcj.
Since jikes is (a) faster and (b) works for me, jikes is still the
default.  To change the default, modify lib/Makefile.am.

Let me remind everyone this is not finished, so please hold off on the
flame throwers until I'm finished.  Suggestions for improvement
welcome.  :)
-- 
Brian Jones <cbj@gnu.org>


--=-=-=
Content-Type: application/octet-stream
Content-Disposition: attachment; filename=cp.diff.tmp
Content-Description: making the build easier, one patch at a time

Index: acinclude.m4
===================================================================
RCS file: /cvsroot/classpath/classpath/acinclude.m4,v
retrieving revision 1.19
diff -u -r1.19 acinclude.m4
--- acinclude.m4	2000/12/15 04:58:55	1.19
+++ acinclude.m4	2001/10/26 13:42:53
@@ -1,11 +1,11 @@
 dnl
 dnl Added macros
 dnl JAPHAR_GREP_CFLAGS
-dnl CLASSPATH_CHECK_JAPHAR
-dnl CLASSPATH_CHECK_KAFFE
-dnl CLASSPATH_CHECK_THREADS
-dnl CLASSPATH_CHECK_GLIB
-dnl CLASSPATH_CHECK_GTK
+dnl CLASSPATH_WITH_JAPHAR
+dnl CLASSPATH_WITH_KAFFE
+dnl CLASSPATH_WITH_THREADS
+dnl CLASSPATH_WITH_GLIB
+dnl CLASSPATH_WITH_GTK
 dnl
 
 dnl JAPHAR_GREP_CFLAGS(flag, cmd_if_missing, cmd_if_present)
@@ -20,8 +20,8 @@
 esac
 ])
 
-dnl CLASSPATH_INTERNAL_CHECK_JAPHAR
-AC_DEFUN(CLASSPATH_INTERNAL_CHECK_JAPHAR,
+dnl CLASSPATH_CHECK_JAPHAR
+AC_DEFUN(CLASSPATH_CHECK_JAPHAR,
 [
   if test "$1" = ""; then
     AC_PATH_PROG(JAPHAR_CONFIG, japhar-config, "", $PATH:/usr/local/japhar/bin:/usr/japhar/bin)
@@ -77,8 +77,8 @@
   AC_SUBST(JAPHAR_CLASSLIB)
 ])
 
-dnl CLASSPATH_INTERNAL_CHECK_KAFFE
-AC_DEFUN(CLASSPATH_INTERNAL_CHECK_KAFFE,
+dnl CLASSPATH_CHECK_KAFFE
+AC_DEFUN(CLASSPATH_CHECK_KAFFE,
 [
   AC_PATH_PROG(KAFFE_CONFIG, kaffe-config, "", $PATH:/usr/local/kaffe/bin:/usr/kaffe/bin)
   if test "$KAFFE_CONFIG" = ""; then
@@ -161,16 +161,16 @@
   AC_SUBST(KAFFE_CLASSLIB)
 ])
 
-dnl CLASSPATH_CHECK_JAPHAR - checks for japhar
-AC_DEFUN(CLASSPATH_CHECK_JAPHAR,
+dnl CLASSPATH_WITH_JAPHAR - checks for japhar
+AC_DEFUN(CLASSPATH_WITH_JAPHAR,
 [
   AC_ARG_WITH(japhar, 
   [  --with-japhar		  configure GNU Classpath for Japhar [default=yes]],
   [
     if test ${withval} = "yes" || test ${withval} = ""; then
-      CLASSPATH_INTERNAL_CHECK_JAPHAR
+      CLASSPATH_CHECK_JAPHAR
     elif test ${withval} != "no" || test ${withval} != "false"; then
-      CLASSPATH_INTERNAL_CHECK_JAPHAR(${withval})
+      CLASSPATH_CHECK_JAPHAR(${withval})
     fi
   ],
   [ 
@@ -180,14 +180,14 @@
   ])
 ])
 
-dnl CLASSPATH_CHECK_KAFFE - checks for which java virtual machine to use
-AC_DEFUN(CLASSPATH_CHECK_KAFFE,
+dnl CLASSPATH_WITH_KAFFE - checks for which java virtual machine to use
+AC_DEFUN(CLASSPATH_WITH_KAFFE,
 [
   AC_ARG_WITH(kaffe, 
   [  --with-kaffe		  configure GNU Classpath for Kaffe [default=no]],
   [   
     if test ${withval} = "yes" || test ${withval} = ""; then
-      CLASSPATH_INTERNAL_CHECK_KAFFE
+      CLASSPATH_CHECK_KAFFE
     fi
   ],
   [ conditional_with_kaffe=false
@@ -253,59 +253,276 @@
     fi
   fi
 ])
+
 
-dnl CLASSPATH_INTERNAL_CHECK_JIKES
-AC_DEFUN(CLASSPATH_INTERNAL_CHECK_JIKES,
+AC_DEFUN(CLASSPATH_FIND_JAVAC,
 [
-  AC_PATH_PROG(JIKES, jikes, "", $PATH:/usr/local/bin)
-  if test "$JIKES" = ""; then
-    echo "configure: cannot find jikes: is jikes in your path?" 1>&2
-    exit 1
+  user_specified_javac=
+
+  CLASSPATH_WITH_GCJ
+  CLASSPATH_WITH_JIKES
+  CLASSPATH_WITH_KJC
+
+  case "$user_specified_javac" in
+    gcj) AM_CONDITIONAL(FOUND_GCJ, test x = x) ;;
+    jikes) AM_CONDITIONAL(FOUND_JIKES, test x = x) ;;
+    kjc) AM_CONDITIONAL(FOUND_KJC, test x = x) ;;
+  esac
+
+  if test "$user_specified_javac" = ""; then
+    AM_CONDITIONAL(FOUND_GCJ, test "$GCJ" != "")
+    AM_CONDITIONAL(FOUND_JIKES, test "$JIKES" != "")
+  fi
+
+  if test "$GCJ" = "" && test "$JIKES" = "" && test "$user_specified_javac" != "kjc"; then
+      echo "configure: cannot find javac, try --with-gcj, --with-jikes, or --with-kjc" 1>&2
+      exit 1    
   fi
 ])
 
 
-dnl CLASSPATH_CHECK_JIKES - checks for jikes
-AC_DEFUN(CLASSPATH_CHECK_JIKES,
+AC_DEFUN(CLASSPATH_WITH_GCJ,
 [
-  AC_ARG_WITH(jikes, 
-  [  --with-jikes		  compile classes with jikes [default=no]],
+  AC_ARG_WITH(gcj,
+  [  --with-gcj              bytecode compilation with gcj ],
   [
-    if test ${withval} != "no"; then
-      if test ${withval} = "" || test ${withval} = "yes"; then
-        CLASSPATH_INTERNAL_CHECK_JIKES
-      else 
-        JIKES=${withval}
-	AC_SUBST(JIKES)
+    if test ${withval} != "" && test ${withval} != "yes" && test ${withval} != "no"; then
+      CLASSPATH_CHECK_GCJ(${withval})
+    else
+      if test ${withval} != "no"; then
+        CLASSPATH_CHECK_GCJ
+      fi
+    fi
+    user_specified_javac=gcj
+    AM_CONDITIONAL(USER_SPECIFIED_GCJ, test "$GCJ" != "")
+  ],
+  [
+    CLASSPATH_CHECK_GCJ
+  ])
+  AC_SUBST(GCJ)
+])
+
+
+AC_DEFUN(CLASSPATH_CHECK_GCJ,
+[
+  if test "$1" != ""; then
+    if test -f "$1"; then
+      GCJ="$1"
+    else
+      AC_PATH_PROG(GCJ, "$1")
+    fi
+  else
+    AC_PATH_PROG(GCJ, "gcj")
+  fi  
+])
+
+
+AC_DEFUN(CLASSPATH_WITH_JIKES,
+[
+  AC_ARG_WITH(jikes,
+  [  --with-jikes		  bytecode compilation with jikes ],
+  [
+    if test ${withval} != "" && test ${withval} != "yes" && test ${withval} != "no"; then
+      CLASSPATH_CHECK_JIKES(${withval})
+    else
+      if test ${withval} != "no"; then
+        CLASSPATH_CHECK_JIKES
       fi
-      conditional_with_jikes=true
     fi
+    user_specified_javac=jikes
+    AM_CONDITIONAL(USER_SPECIFIED_JIKES, test "$JIKES" != "")
   ],
   [ 
-    conditional_with_jikes=false
+    CLASSPATH_CHECK_JIKES
   ])
+  AC_SUBST(JIKES)
+])
+
+
+AC_DEFUN(CLASSPATH_CHECK_JIKES,
+[
+  if test "$1" != ""; then
+    if test -f "$1"; then
+      JIKES="$1"
+    else
+      AC_PATH_PROG(JIKES, "$1")
+    fi
+  else
+    AC_PATH_PROG(JIKES, "jikes")
+  fi
 ])
 
 
-dnl CLASSPATH_CHECK_KJC - checks for kjc
-AC_DEFUN(CLASSPATH_CHECK_KJC,
+AC_DEFUN(CLASSPATH_WITH_KJC,
 [
   AC_ARG_WITH(kjc, 
-  [  --with-kjc=DIR	  compile classes with kjc [default=no]],
+  [  --with-kjc=<ksusu.jar>  bytecode compilation with kjc [default=no]],
   [
     if test ${withval} != "no"; then
       AC_MSG_CHECKING(for kjc)
       if test ${withval} = "" || test ${withval} = "yes"; then
-        AC_MSG_ERROR(specify a classpath to kjc)
+        AC_MSG_ERROR(specify the location of ksusu.jar or kjc CLASSPATH)
       fi
       KJC_CLASSPATH=${withval}
       AC_SUBST(KJC_CLASSPATH)
       conditional_with_kjc=true
       AC_MSG_RESULT(${withval})
     fi
+    user_specified_javac=kjc
+    AM_CONDITIONAL(USER_SPECIFIED_KJC, test x$conditional_with_kjc = xtrue)
   ],
   [ 
     conditional_with_kjc=false
   ])
+
+
+  if test x$conditional_with_kjc = xtrue && test "$USER_JABBA" = ""; then
+    CLASSPATH_FIND_JAVA
+    if test "$USER_JABBA" = ""; then
+      echo "configure: cannot find java, try --with-java" 1>&2
+      exit 1
+    fi
+  fi
+])
+
+
+AC_DEFUN(CLASSPATH_WITH_JAVA,
+[
+  AC_ARG_WITH(java,
+  [  --with-java		  specify path or name of a java-like program ],
+  [
+    if test ${withval} != "" && test ${withval} != "yes" && test ${withval} != "no"; then
+      CLASSPATH_CHECK_JAVA(${withval})
+    else
+      if test ${withval} != "no"; then
+        CLASSPATH_CHECK_JAVA
+      fi
+    fi
+    AM_CONDITIONAL(USER_SPECIFIED_JABBA, test "$USER_JABBA" != "")
+  ],
+  [ 
+    CLASSPATH_CHECK_JAVA
+  ])
+  AC_SUBST(USER_JABBA)
+])
+
+
+AC_DEFUN(CLASSPATH_CHECK_JAVA,
+[
+  if test "$1" != ""; then
+    if test -f "$1"; then
+      USER_JABBA="$1"
+    else
+      AC_PATH_PROG(USER_JABBA, "$1")
+    fi
+  else
+    AC_PATH_PROG(USER_JABBA, "java")
+  fi
+])
+
+
+AC_DEFUN(CLASSPATH_FIND_JAVA,
+[
+  dnl Place additional bytecode interpreter checks here
+
+  CLASSPATH_WITH_JAVA
+])
+
+
+AC_DEFUN(CLASSPATH_WITH_JAVAH,
+[
+  AC_ARG_WITH(javah,
+  [  --with-javah		  specify path or name of a javah-like program ],
+  [
+    if test ${withval} != "" && test ${withval} != "yes" && test ${withval} != "no"; then
+      CLASSPATH_CHECK_JAVAH(${withval})
+    else
+      CLASSPATH_CHECK_JAVAH
+    fi
+    AM_CONDITIONAL(USER_SPECIFIED_JAVAH, test "$USER_JAVAH" != "")
+  ],
+  [ 
+    CLASSPATH_CHECK_JAVAH
+  ])
+  AC_SUBST(USER_JAVAH)
+])
+
+dnl Checking for a javah like program 
+AC_DEFUN(CLASSPATH_CHECK_JAVAH,
+[
+  if test "$1" != ""; then
+    if test -f "$1"; then
+      USER_JAVAH="$1"
+    else
+      AC_PATH_PROG(USER_JAVAH, "$1")
+    fi
+  else
+    for javah_name in gcjh javah; do
+      AC_PATH_PROG(USER_JAVAH, "$javah_name")
+      if test "$USER_JAVAH" != ""; then
+        break
+      fi
+    done
+  fi
+  
+  if test "$USER_JAVAH" = ""; then
+    echo "configure: cannot find javah" 1>&2
+    exit 1
+  fi
 ])
 
+dnl CLASSPATH_WITH_CLASSLIB - checks for user specified classpath additions
+AC_DEFUN(CLASSPATH_WITH_CLASSLIB,
+[
+  AC_ARG_WITH(classpath,
+  [  --with-classpath        specify path to a classes.zip like file ],
+  [
+    if test ${withval} = "yes"; then
+      # set user classpath to CLASSPATH from env
+      AC_MSG_CHECKING(for classlib)
+      USER_CLASSLIB=${CLASSPATH}
+      AC_SUBST(USER_CLASSLIB)
+      AC_MSG_RESULT(${USER_CLASSLIB})
+      conditional_with_classlib=true      
+    elif test ${withval} != "" && test ${withval} != "no"; then
+      # set user classpath to specified value
+      AC_MSG_CHECKING(for classlib)
+      USER_CLASSLIB=${withval}
+      AC_SUBST(USER_CLASSLIB)
+      AC_MSG_RESULT(${withval})
+      conditional_with_classlib=true
+    fi
+  ],
+  [ conditional_with_classlib=false ])
+  AM_CONDITIONAL(USER_SPECIFIED_CLASSLIB, test x$conditional_with_classlib = xtrue)
+])
+
+
+dnl CLASSPATH_WITH_INCLUDEDIR - checks for user specified extra include directories
+AC_DEFUN(CLASSPATH_WITH_INCLUDEDIR,
+[
+  AC_ARG_WITH(includedir,
+  [  --with-includedir=DIR   specify path to an extra include dir ],
+  [
+    AC_MSG_CHECKING(for includedir)
+    if test ${withval} != "" && test ${withval} != "yes" && test ${withval} != "no"; then
+      if test -r ${withval}; then
+        if test "$EXTRA_INCLUDES" = ""; then
+          EXTRA_INCLUDES="-I${withval}"
+        else
+          EXTRA_INCLUDES="${EXTRA_INCLUDES} -I${withval}"
+        fi
+        AC_SUBST(EXTRA_INCLUDES)
+        AC_MSG_RESULT("added ${withval}")
+      else
+        AC_MSG_RESULT("${withval} does not exist")
+      fi
+    fi
+  ],
+  [
+    if test -z "$EXTRA_INCLUDES"; then
+      EXTRA_INCLUDES=""
+      AC_SUBST(EXTRA_INCLUDES)
+    fi
+  ])
+])
Index: configure.in
===================================================================
RCS file: /cvsroot/classpath/classpath/configure.in,v
retrieving revision 1.75
diff -u -r1.75 configure.in
--- configure.in	2001/10/19 17:49:52	1.75
+++ configure.in	2001/10/26 13:42:53
@@ -1,11 +1,14 @@
 dnl Process this file with autoconf to produce a configure script.
 
+dnl Turning off cache for debug reasons
+define([AC_CACHE_LOAD], )dnl
+define([AC_CACHE_SAVE], )dnl
 AC_INIT(java/lang/System.java)
 
 AC_CANONICAL_HOST
 
 PACKAGE="classpath"
-VERSION="0.02"
+VERSION="0.03"
 LIBVERSION="0:0:0"
 AC_SUBST(PACKAGE)
 AC_SUBST(VERSION)
@@ -16,198 +19,111 @@
 
 AC_PREFIX_DEFAULT(/usr/local/classpath)
 
-dnl Initialize libtool
-AM_DISABLE_STATIC
-AM_PROG_LIBTOOL
-
 dnl required if SUBDIRS is used
 AC_PROG_MAKE_SET
 
-dnl Checks for programs.
-dnl AC_PROG_CXX
-AC_PROG_AWK
-AC_PROG_CC
-AC_PROG_CPP
-AC_PROG_INSTALL
+AC_ARG_ENABLE(java,
+[  --enable-java           compile Java source [default=yes]],
+[case "${enableval}" in 
+  yes) COMPILE_JAVA=yes ;;
+  no) COMPILE_JAVA=no ;;
+  *) COMPILE_JAVA=yes ;;
+esac],
+[COMPILE_JAVA=yes])
+
+AC_ARG_ENABLE(jni,
+[  --enable-jni            compile JNI source [default=no]],
+[case "${enableval}" in 
+  yes) 
+    COMPILE_JNI=yes
+    COMPILE_JAVA=yes
+
+    ;;
+  no) COMPILE_JNI=no ;;
+  *) COMPILE_JNI=yes; COMPILE_JAVA=yes ;;
+esac],
+[COMPILE_JNI=no])
+
+AC_ARG_ENABLE(cni,
+[  --enable-cni            compile CNI source [default=no]],
+[case "${enableval}" in 
+  yes) COMPILE_CNI=yes; COMPILE_JAVA=yes ;;
+  no) COMPILE_CNI=no ;;
+  *) COMPILE_CNI=yes; COMPILE_JAVA=yes ;;
+esac],
+[COMPILE_CNI=no])
+
+AM_CONDITIONAL(CREATE_JNI_HEADERS, test x$COMPILE_JNI = xyes) 
+AM_CONDITIONAL(CREATE_JNI_LIBRARIES, test x$COMPILE_JNI = xyes)
+AM_CONDITIONAL(CREATE_CNI_LIBRARIES, test x$COMPILE_CNI = xyes)
+
 AC_PROG_LN_S
-AC_HEADER_STDC
-AC_CHECK_HEADERS(unistd.h sys/types.h sys/config.h inttypes.h stdint.h)
-AC_EGREP_HEADER(uint32_t, stdint.h, AC_DEFINE(HAVE_INT32_DEFINED))
-AC_EGREP_HEADER(uint32_t, inttypes.h, AC_DEFINE(HAVE_INT32_DEFINED))
-AC_EGREP_HEADER(u_int32_t, sys/types.h, AC_DEFINE(HAVE_BSD_INT32_DEFINED))
-AC_EGREP_HEADER(u_int32_t, sys/config.h, AC_DEFINE(HAVE_BSD_INT32_DEFINED))
-AC_C_CONST
-AC_CHECK_FUNCS(gethostname socket strerror fork pipe execve strtod atof)
-AC_HEADER_TIME
-AC_STRUCT_TM
-AC_STRUCT_TIMEZONE
-
-AC_C_BIGENDIAN
-
-dnl guile checks - shouldn't need now that mauve exists
-dnl GUILE_FLAGS
-
-dnl We want to avoid -O2 at the moment, gcc 2.96 from Red Hat 7
-dnl cannot compile some native functions (gcc errors) with this
-dnl flag.
-CFLAGS="-g"
-
-dnl We want warnings, lots of warnings :-)
-if eval "test x$GCC = xyes"; then
-  JAPHAR_GREP_CFLAGS(-Wall, [
-    CFLAGS="$CFLAGS -Wall"
-  ])
-fi
+AC_PROG_INSTALL
 
-dnl Check for cthreads, w32 threads, pthreads
-dnl CLASSPATH_CHECK_THREADS
+dnl Checks for programs.
+if test "${COMPILE_JNI}" = "yes"; then
+  dnl AC_PROG_CXX
+  dnl Initialize libtool
+  AM_DISABLE_STATIC
+  AM_PROG_LIBTOOL
+  dnl AC_PROG_AWK
+  AC_PROG_CC
+  AC_PROG_CPP
+  AC_HEADER_STDC
+  AC_CHECK_HEADERS(unistd.h sys/types.h sys/config.h inttypes.h stdint.h)
+  AC_EGREP_HEADER(uint32_t, stdint.h, AC_DEFINE(HAVE_INT32_DEFINED))
+  AC_EGREP_HEADER(uint32_t, inttypes.h, AC_DEFINE(HAVE_INT32_DEFINED))
+  AC_EGREP_HEADER(u_int32_t, sys/types.h, AC_DEFINE(HAVE_BSD_INT32_DEFINED))
+  AC_EGREP_HEADER(u_int32_t, sys/config.h, AC_DEFINE(HAVE_BSD_INT32_DEFINED))
+  AC_C_CONST
+  AC_CHECK_FUNCS(gethostname socket strerror fork pipe execve strtod atof)
+  AC_HEADER_TIME
+  AC_STRUCT_TM
+  AC_STRUCT_TIMEZONE
+  AC_C_BIGENDIAN
+
+  dnl We want to avoid -O2 at the moment, gcc 2.96 from Red Hat 7
+  dnl cannot compile some native functions (gcc errors) with this
+  dnl flag.
+  CFLAGS="-g" 
+
+  dnl We want warnings, lots of warnings :-)
+  if eval "test x$GCC = xyes"; then
+    JAPHAR_GREP_CFLAGS(-Wall, [ CFLAGS="$CFLAGS -Wall" ])
+  fi
 
-dnl Check for AWT related glib/gtk/libart_lgpl
-AM_PATH_GTK(1.2.4,,exit 1)
-AM_PATH_GLIB(1.2.4,,exit 1,gthread)
-dnl XXX Fix me when libart.m4 has the compile test fixed!
-enable_libarttest=no
-AM_PATH_LIBART(2.1.0,,exit 1)
-
-dnl -----------------------------------------------------------
-dnl Must do automatic checks before specific --with-* overrides
-dnl for vm programs, includes, libraries
-dnl -----------------------------------------------------------
-
-dnl --with-japhar
-CLASSPATH_CHECK_JAPHAR
-
-dnl --with-kaffe
-CLASSPATH_CHECK_KAFFE
+  dnl Check for AWT related glib/gtk/libart_lgpl
+  AM_PATH_GTK(1.2.4,,exit 1)
+  AM_PATH_GLIB(1.2.4,,exit 1,gthread)
 
-dnl -----------------------------------------------------------
+  dnl XXX Fix me when libart.m4 has the compile test fixed!
+  enable_libarttest=no
+  AM_PATH_LIBART(2.1.0,,exit 1)
 
+  CLASSPATH_WITH_JAVAH
 
-dnl -----------------------------------------------------------
-dnl Various overrides are allowed here to contradict defaults 
-dnl from automatic VM checks.
-dnl -----------------------------------------------------------
-
-AC_ARG_WITH(java,
-  [  --with-java		  specify path to a java-like program ],
-  [
-    if test ${withval} != "" || test ${withval} != "yes" || test ${withval} != "no"; then
-      # set javac to user input value
-      AC_MSG_CHECKING(for java)
-      USER_JABBA=${withval}
-      AC_SUBST(USER_JABBA)
-      AC_MSG_RESULT(${withval})
-      conditional_with_java=true
-    fi
-  ],
-  [ conditional_with_java=false])
-
-AM_CONDITIONAL(USER_SPECIFIED_JABBA, test x$conditional_with_java = xtrue)
-
-AC_ARG_WITH(javah,
-  [  --with-javah		  specify path to a javah-like program ],
-  [
-    if test ${withval} != "" || test ${withval} != "yes" || test ${withval} != "no"; then
-      # set javah to user input value
-      AC_MSG_CHECKING(for javah)
-      USER_JAVAH=${withval}
-      AC_SUBST(USER_JAVAH)
-      AC_MSG_RESULT(${withval})
-      conditional_with_javah=true
-    fi
-  ],
-  [ conditional_with_javah=false ])
-AM_CONDITIONAL(USER_SPECIFIED_JAVAH, test x$conditional_with_javah = xtrue)
-
-dnl define JIKES for Makefile.am if --with-jikes specified to configure
-CLASSPATH_CHECK_JIKES
-AM_CONDITIONAL(USER_SPECIFIED_JIKES, test x$conditional_with_jikes = xtrue)
-
-dnl define KJC_CLASSPATH for Makefile.am if --with-kjc specified to configure
-CLASSPATH_CHECK_KJC
-
-AC_ARG_WITH(classlib,
-  [  --with-classlib	  specify path to a classes.zip like file ],
-  [
-    if test ${withval} != "" || test ${withval} != "yes" || test ${withval} != "no"; then
-      # set class library to user input value
-      AM_CONDITIONAL(USER_SPECIFIED_CLASSLIB, test x = y)
-      AC_MSG_CHECKING(for classlib)
-      USER_CLASSLIB=${withval}
-      AC_SUBST(USER_CLASSLIB)
-      AC_MSG_RESULT(${withval})
-      conditional_with_classlib=true
-    fi
-  ],
-  [ conditional_with_classlib=false ])
-AM_CONDITIONAL(USER_SPECIFIED_CLASSLIB, test x$conditional_with_classlib = xtrue)
-
-AC_ARG_WITH(includedir,
-  [  --with-includedir=DIR   specify path to an extra include dir ],
-  [
-    AC_MSG_CHECKING(for includedir)
-    if test ${withval} != "" || test ${withval} != "yes" || test ${withval} != "no"; then
-      if test -r ${withval}; then
-        if test "$EXTRA_INCLUDES" = ""; then
-          EXTRA_INCLUDES="-I${withval}"
-        else
-          EXTRA_INCLUDES="${EXTRA_INCLUDES} -I${withval}"
-        fi
-        AC_SUBST(EXTRA_INCLUDES)
-        AC_MSG_RESULT("added ${withval}")
-      else
-        AC_MSG_RESULT("${withval} does not exist")
-      fi
-    fi
-  ],
-  [
-    if test -z "$EXTRA_INCLUDES"; then
-      EXTRA_INCLUDES=""
-      AC_SUBST(EXTRA_INCLUDES)
-    fi
-  ])
-
-dnl -----------------------------------------------------------
-
-dnl -----------------------------------------------------------
-dnl Initialize maintainer mode, not clear to me yet whether 
-dnl developer-mode is still needed but maintainer-mode definitely 
-dnl is.
-dnl -----------------------------------------------------------
-AM_MAINTAINER_MODE
+  CLASSPATH_WITH_INCLUDEDIR
 
-dnl This build mode indicates to configure it should compile java classes
-dnl but not rebuild the dependency files for doing so.  The default then is to
-dnl not build any Java classes (glibj.zip is distributed in the build using
-dnl the maintainer-mode)
-AC_ARG_ENABLE(developer-mode,
-[  --enable-developer-mode enable some of the rules in maintainer-mode
-                          (compilation of Java classes) [default=no]],
-[case "${enableval}" in 
-  yes) USE_DEVELOPER_MODE=yes ;;
-  no) USE_DEVELOPER_MODE=no ;;
-  *) AC_MSG_ERROR(bad value ${enableval} for --enable-developer-mode) ;;
-esac],
-[USE_DEVELOPER_MODE=no])
-if test x${USE_MAINTAINER_MODE} = "xyes"; then
-  USE_DEVELOPER_MODE=no
 fi
-AM_CONDITIONAL(DEVELOPER_MODE, test x$USE_DEVELOPER_MODE = xyes)
-dnl to mimic MAINT from AM_MAINTAINER_MODE
-DEVEL=$DEVELOPER_MODE_TRUE
-AC_SUBST(DEVEL)
 
-dnl -----------------------------------------------------------
+CLASSPATH_FIND_JAVAC
+
+CLASSPATH_WITH_CLASSLIB
 
 
 dnl -----------------------------------------------------------
+dnl Initialize maintainer mode
+dnl -----------------------------------------------------------
+AM_MAINTAINER_MODE
+
+dnl -----------------------------------------------------------
 dnl Enable debugging statements at compile time.  By default
 dnl these statements should be optimized out of the bytecode
 dnl produced by an optimizing Java compiler and not hinder 
-dnl performance.
+dnl performance because debugging is turns off by default.
 dnl -----------------------------------------------------------
 AC_ARG_ENABLE(debug,
-[  --enable-debug   enable runtime debugging code],
+[  --enable-debug          enable runtime debugging code],
 [case "${enableval}" in 
   yes) 
 	LIBDEBUG="true"
@@ -218,7 +134,6 @@
 esac],
 [LIBDEBUG="false"])
 AC_SUBST(LIBDEBUG)
-
 dnl -----------------------------------------------------------
 
 
@@ -229,64 +144,40 @@
 dnl them.
 dnl -----------------------------------------------------------
 AC_ARG_ENABLE(load-library,
-[  --enable-load-library   [default=yes]],
+[  --enable-load-library   enable to use JNI native methods [default=yes]
+                          (disabled automatically using --enable-cni)],
 [case "${enableval}" in 
   yes) INIT_LOAD_LIBRARY="true" ;;
   no) INIT_LOAD_LIBRARY="false" ;;
   *) AC_MSG_ERROR(bad value ${enableval} for --enable-load-library) ;;
 esac],
-[INIT_LOAD_LIBRARY="true"])
+[
+  if test "$COMPILE_CNI" = "yes"; then
+    INIT_LOAD_LIBRARY="false"
+  else
+    INIT_LOAD_LIBRARY="true"
+  fi
+])
 AC_SUBST(INIT_LOAD_LIBRARY)
-
 dnl -----------------------------------------------------------
 
 
 dnl -----------------------------------------------------------
-dnl This takes care of autodetecting our preferred VM when one 
-dnl isn't specified explicitly.
+dnl Small loop to add other things to EXTRA_INCLUDES from a VM
 dnl -----------------------------------------------------------
-dnl if no jvm yet defined, check for japhar
-dnl if test -z "$JVM" || test "$JVM" = "no"; then
-dnl     # need to have these defined if no vm selected, default to japhar
-dnl     if test -z "$JVM_REFERENCE"; then
-dnl X dnl conditional_with_classlib != xtrue || test x$conditional_with_javah !=  xtrue; then
-dnl X dnl      echo classlib $conditional_with_classlib
-dnl X dnl      echo javah $conditional_with_javah
-dnl       CLASSPATH_INTERNAL_CHECK_JAPHAR
-dnl       AC_MSG_CHECKING(for defined vm)
-dnl       if test -z "$JVM" || test "$JVM" = "no"; then
-dnl         AC_MSG_ERROR(no: try --with-japhar or --with-kaffe)
-dnl       else
-dnl         AC_MSG_RESULT(yes)
-dnl       fi
+dnl for i in $JAPHAR_CFLAGS $KAFFE_CFLAGS ; do
+dnl   add_to_extra_includes=1
+dnl   for j in $EXTRA_INCLUDES ; do
+dnl     if test $i = $j ; then
+dnl       add_to_extra_includes=0
 dnl     fi
-dnl fi
-
-dnl must check our conditionals from vm checks
-AM_CONDITIONAL(JAPHAR, test x$conditional_with_japhar = xtrue)
-AM_CONDITIONAL(KAFFE, test x$conditional_with_kaffe = xtrue)
-dnl can be defined by kaffe check, should be here should we autodetect
-dnl kaffe in the future
-AM_CONDITIONAL(USER_SPECIFIED_KJC, test x$conditional_with_kjc = xtrue)
-
-dnl -----------------------------------------------------------
-
-dnl -----------------------------------------------------------
-dnl Small loop to add other things to EXTRA_INCLUDES from a VM
+dnl   done
+dnl   if test $add_to_extra_includes -eq 1; then
+dnl     EXTRA_INCLUDES="$EXTRA_INCLUDES $i"
+dnl   fi
+dnl done
 dnl -----------------------------------------------------------
-for i in $JAPHAR_CFLAGS $KAFFE_CFLAGS ; do
-  add_to_extra_includes=1
-  for j in $EXTRA_INCLUDES ; do
-    if test $i = $j ; then
-      add_to_extra_includes=0
-    fi
-  done
-  if test $add_to_extra_includes -eq 1; then
-    EXTRA_INCLUDES="$EXTRA_INCLUDES $i"
-  fi
-done
 
-dnl -----------------------------------------------------------
 
 dnl Create a link to the VM specific files we're compiling with
 if test x"$JVM_REFERENCE" = x; then
Index: lib/Makefile.am
===================================================================
RCS file: /cvsroot/classpath/classpath/lib/Makefile.am,v
retrieving revision 1.36
diff -u -r1.36 Makefile.am
--- lib/Makefile.am	2001/06/25 04:43:56	1.36
+++ lib/Makefile.am	2001/10/26 13:42:55
@@ -15,18 +15,19 @@
 sinclude $(JAVA_DEPEND)
 sinclude $(HEADERS_DEPEND)
 
+KJC_CLASSPATH=@KJC_CLASSPATH@
 
+USER_CLASSLIB = @USER_CLASSLIB@
+
+GCJ = @GCJ@
+JIKES = @JIKES@
+USER_JABBA = @USER_JABBA@
+USER_JAVAH = @USER_JAVAH@
+
 #-----------------------------------------------------------------
 # defining a javac program written in java
 #-----------------------------------------------------------------
 KJC_MAIN = at.dms.kjc.Main
-SUN_JAVAC_MAIN = sun.tools.javac.Main
-
-if USER_SPECIFIED_KJC
-JAVAC_MAIN = $(KJC_MAIN)
-else
-JAVAC_MAIN = $(SUN_JAVAC_MAIN)
-endif
 #-----------------------------------------------------------------
 
 
@@ -34,15 +35,7 @@
 # defining how to designate classpath for a vm and a javac type 
 # program
 #-----------------------------------------------------------------
-if JAPHAR
-if USER_SPECIFIED_JABBA
-ARG_CLASSPATH_VM = -classpath
-else
-ARG_CLASSPATH_VM = --classpath
-endif
-else
 ARG_CLASSPATH_VM = -classpath
-endif
 
 if USER_SPECIFIED_KJC
 ARG_CLASSPATH_JAVAC = --classpath
@@ -53,73 +46,20 @@
 
 
 #-----------------------------------------------------------------
-# defining a java-like program
-#-----------------------------------------------------------------
-if USER_SPECIFIED_JABBA
-USER_JABBA = @USER_JABBA@
-else
-
-if JAPHAR
-USER_JABBA = @JAPHAR_JABBA@
-else
-
-if KAFFE
-USER_JABBA = @KAFFE_JABBA@
-else
-USER_JABBA = java
-endif
-
-endif
-
-endif
-#-----------------------------------------------------------------
-
-
-#-----------------------------------------------------------------
-# defining the classpath for the java-like program, and 
-# the javac-like program
-#-----------------------------------------------------------------
-if USER_SPECIFIED_CLASSLIB
-USER_CLASSLIB = @USER_CLASSLIB@
-else
-
-if JAPHAR
-USER_CLASSLIB = @JAPHAR_CLASSLIB@
-else
-
-if KAFFE
-USER_CLASSLIB = @KAFFE_CLASSLIB@
-else
-USER_CLASSLIB = /usr/local/java/lib/classes.zip
-endif
-
-endif
-
-endif
-
-#-----------------------------------------------------------------
-
-
-#-----------------------------------------------------------------
 # handling javac definition including programs like jikes here
 #-----------------------------------------------------------------
-if USER_SPECIFIED_KJC
-JAVAC = $(USER_JABBA) $(ARG_CLASSPATH_VM) @KJC_CLASSPATH@:$(USER_CLASSLIB) $(JAVAC_MAIN) $(ARG_CLASSPATH_JAVAC) .:$(USER_CLASSLIB) -d . @classes
+if FOUND_JIKES
+JAVAC = $(JIKES) -nowarn +F $(ARG_CLASSPATH_JAVAC) $(top_srcdir):$(top_builddir)/vm/current:.:$(USER_CLASSLIB) -d . @classes
 else
-
-if USER_SPECIFIED_JIKES
-
-if USER_SPECIFIED_CLASSLIB
-JAVAC = @JIKES@ -nowarn +F $(ARG_CLASSPATH_JAVAC) $(top_srcdir):$(top_builddir)/vm/current:.:@USER_CLASSLIB@ -d . @classes
+if FOUND_GCJ
+JAVAC = $(GCJ) -C $(ARG_CLASSPATH_JAVAC) $(top_srcdir):$(top_builddir)/vm/current:.:$(USER_CLASSLIB) -d . $(JAVA_SRCS)
 else
-JAVAC = @JIKES@ -nowarn +F $(ARG_CLASSPATH_JAVAC) $(top_srcdir):$(top_builddir)/vm/current:. -d . @classes
+if FOUND_KJC
+JAVAC = $(USER_JABBA) $(ARG_CLASSPATH_VM) $(KJC_CLASSPATH):$(USER_CLASSLIB) $(KJC_MAIN) $(ARG_CLASSPATH_JAVAC) .:$(USER_CLASSLIB) -d . @classes
+endif 
 endif
-
-else
-JAVAC = $(USER_JABBA) $(ARG_CLASSPATH_VM) $(USER_CLASSLIB) $(JAVAC_MAIN) $(ARG_CLASSPATH_JAVAC) .:$(USER_CLASSLIB) -d . $(JAVA_SRCS)
 endif
 
-endif
 #-----------------------------------------------------------------
 
 
@@ -127,57 +67,14 @@
 # defining how to designate classpath for a vm and a javac type 
 # program
 #-----------------------------------------------------------------
-if JAPHAR
-
-if USER_SPECIFIED_JAVAH
 ARG_JNI_JAVAH = -jni
-else
-ARG_JNI_JAVAH = --jni
-endif
-
-else
-ARG_JNI_JAVAH = -jni
-endif
 #-----------------------------------------------------------------
 
 
 #-----------------------------------------------------------------
-# defining a javah-like program
-#-----------------------------------------------------------------
-if USER_SPECIFIED_JAVAH
-USER_JAVAH = @USER_JAVAH@
-else
-
-if JAPHAR
-USER_JAVAH = @JAPHAR_JAVAH@
-else
-
-if KAFFE
-USER_JAVAH = @KAFFE_JAVAH@
-else
-USER_JAVAH = javah
-endif
-
-endif
-
-endif
-#-----------------------------------------------------------------
-
-
-#-----------------------------------------------------------------
 # defining classpath arg for javah-like program
 #-----------------------------------------------------------------
-if JAPHAR
-
-if USER_SPECIFIED_JAVAH
-ARG_CLASSPATH_JAVAH = -classpath
-else
-ARG_CLASSPATH_JAVAH = --classpath
-endif
-
-else
 ARG_CLASSPATH_JAVAH = -classpath
-endif
 #-----------------------------------------------------------------
 
 JAVAH = $(USER_JAVAH) $(ARG_JNI_JAVAH) $(ARG_CLASSPATH_JAVAH) .:$(USER_CLASSLIB)
@@ -188,7 +85,12 @@
 	if ! [ -e $(pkgdatadir) ]; then mkdir -p $(pkgdatadir); fi
 	unzip -d $(pkgdatadir) glibj.zip
 
+if CREATE_JNI_HEADERS
 all: $(HEADERS_DEPEND) glibj.zip $(CP_HEADERS)
+else
+all: glibj.zip
+endif
+
 #jar-classes
 
 # this makes headers always generate after making glibj.zip
@@ -201,10 +103,9 @@
 	if ! [ -e gnu/java/locale ]; then mkdir gnu/java/locale; fi
 	cp $(top_srcdir)/gnu/java/locale/*.properties gnu/java/locale/
 	cp $(top_srcdir)/gnu/java/locale/*.uni gnu/java/locale/
-	# do something with all the .u files from jikes
 	$(ZIP) -r -D glibj.zip gnu java javax > /dev/null
 
-classes: gen-classlist.sh standard.omit $(top_srcdir)/java $(top_srcdir)/javax $(top_srcdir)/gnu
+classes: gen-classlist.sh standard.omit
 	top_builddir=$(top_builddir) $(SHELL) ./gen-classlist.sh standard
 
 .PHONY: dep
@@ -218,7 +119,7 @@
 
 $(MAKEFILE_DEPEND): compile-classes
 	top_srcdir=@top_srcdir@ $(PERL) ./mkdep.pl -d
-	$(SHELL) -c find -regex ".*\.u$$" -exec rm \{\} \; > /dev/null
+#	$(SHELL) -c find -regex ".*\.u$$" -exec rm \{\} \; > /dev/null
 
 $(HEADERS_DEPEND): classes
 	if ! [ -e $(top_builddir)/include ]; then -mkdir $(top_builddir)/include; fi	
@@ -267,6 +168,7 @@
 MAINTAINERCLEANFILES = makefile.dep $(JAVA_DEPEND) $(HEADERS_DEPEND)
 
 maintainer-clean-local:
+	-rm -rf com
 	-rm -rf gnu
 	-rm -rf java
 	-rm -rf javax
Index: lib/gen-classlist.sh.in
===================================================================
RCS file: /cvsroot/classpath/classpath/lib/gen-classlist.sh.in,v
retrieving revision 1.7
diff -u -r1.7 gen-classlist.sh.in
--- lib/gen-classlist.sh.in	2001/10/19 17:49:52	1.7
+++ lib/gen-classlist.sh.in	2001/10/26 13:42:55
@@ -1,4 +1,4 @@
 #! @SH@
 # @configure_input@
-@FIND@ @top_srcdir@/java @top_srcdir@/javax @top_srcdir@/gnu ${top_builddir}/java ${top_builddir}/javax ${top_builddir}/gnu ${top_builddir}/vm/current -follow -type f -print | grep '\.java$' > ${top_builddir}/lib/classes
+@FIND@ @top_srcdir@/com @top_srcdir@/java @top_srcdir@/javax @top_srcdir@/gnu ${top_builddir}/java ${top_builddir}/javax ${top_builddir}/gnu ${top_builddir}/vm/current -follow -type f -print | grep '\.java$' > ${top_builddir}/lib/classes
 for filexp in `cat @top_srcdir@/lib/$1.omit` ; do { grep -v ${filexp} < ${top_builddir}/lib/classes > ${top_builddir}/lib/classes.new ; mv ${top_builddir}/lib/classes.new ${top_builddir}/lib/classes ; } ; done
Index: native/Makefile.am
===================================================================
RCS file: /cvsroot/classpath/classpath/native/Makefile.am,v
retrieving revision 1.17
diff -u -r1.17 Makefile.am
--- native/Makefile.am	2001/01/23 23:57:45	1.17
+++ native/Makefile.am	2001/10/26 13:42:55
@@ -1,5 +1,13 @@
 ## Input file for automake to generate the Makefile.in used by configure
 
+if CREATE_JNI_LIBRARIES
 SUBDIRS = jni lib vmi
-##java.lang java.net java.io java.lang.reflect java.util lib vmi gnu.java.awt.peer.gtk
+else
+if CREATE_CNI_LIBRARIES
+SUBDIRS = cni
+else
+SUBDIRS =
+endif
+endif
+
 
Index: native/jni/Makefile.am
===================================================================
RCS file: /cvsroot/classpath/classpath/native/jni/Makefile.am,v
retrieving revision 1.2
diff -u -r1.2 Makefile.am
--- native/jni/Makefile.am	2001/06/25 04:43:56	1.2
+++ native/jni/Makefile.am	2001/10/26 13:42:55
@@ -61,15 +61,5 @@
              -I$(top_srcdir)/native/lib/ \
 	     -I$(top_builddir)/include 
 
-if JAPHAR
-install-data-local : 
-	-cd $(pkglibdir) && ln -s libjavaio.so libjaphar_javaio.so
-	-cd $(pkglibdir) && ln -s libjavalang.so libjaphar_javalang.so
-	-cd $(pkglibdir) && ln -s libjavalangmath.so libjaphar_javalangmath.so
-	-cd $(pkglibdir) && ln -s libjavalangreflect.so libjaphar_javalangreflect.so
-	-cd $(pkglibdir) && ln -s libjavanet.so libjaphar_javanet.so
-	-cd $(pkglibdir) && ln -s libjavautil.so libjaphar_javautil.so
-endif
-
 clean-local:
 	-rm -rf .deps
Index: native/jni/awt/Makefile.am
===================================================================
RCS file: /cvsroot/classpath/classpath/native/jni/awt/Makefile.am,v
retrieving revision 1.2
diff -u -r1.2 Makefile.am
--- native/jni/awt/Makefile.am	2001/06/25 04:43:56	1.2
+++ native/jni/awt/Makefile.am	2001/10/26 13:42:55
@@ -41,11 +41,6 @@
 	    @LIBART_CFLAGS@ \
 	    -I$(top_builddir)/include
 
-if JAPHAR
-install-data-local : 
-	-cd $(pkglibdir) && ln -s libgtkpeer.so libjaphar_gtkpeer.so
-endif
-
 clean-local:
 	-rm -rf .deps
 
Index: native/lib/Makefile.am
===================================================================
RCS file: /cvsroot/classpath/classpath/native/lib/Makefile.am,v
retrieving revision 1.11
diff -u -r1.11 Makefile.am
--- native/lib/Makefile.am	2000/07/01 16:10:31	1.11
+++ native/lib/Makefile.am	2001/10/26 13:42:56
@@ -7,11 +7,6 @@
 INCLUDES += -I../lib -I../nsa -I../vmi -I. -DNO_VMI @EXTRA_INCLUDES@ \
 	    -I$(top_srcdir)/include
 
-if JAPHAR
-install-data-local : 
-	-cd $(pkglibdir) && ln -s libclasspath.so libjaphar_classpath.so
-endif
-
 clean-local:
 	-rm -rf .deps
 
Index: vm/Makefile.am
===================================================================
RCS file: /cvsroot/classpath/classpath/vm/Makefile.am,v
retrieving revision 1.3
diff -u -r1.3 Makefile.am
--- vm/Makefile.am	1999/12/30 15:36:08	1.3
+++ vm/Makefile.am	2001/10/26 13:42:56
@@ -1,15 +1,4 @@
 # used by automake to generate Makefile.in
 
-if JAPHAR
 
 SUBDIRS = reference
-
-else
-
-if KAFFE
-SUBDIRS = kaffe
-else
-SUBDIRS = reference
-endif
-
-endif

--=-=-=--


From tromey@creche.cygnus.com Fri Oct 26 11:34:42 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15x90M-0002Zj-00; Fri, 26 Oct 2001 11:34:42 -0400
Received: from creche.cygnus.com (ta0193.peakpeak.com [204.144.244.193] (may be forged))
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id JAA01786;
	Fri, 26 Oct 2001 09:34:38 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id JAA03598;
	Fri, 26 Oct 2001 09:53:29 -0600
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: recent rmi additions
References: <m3adyf87uh.fsf@lyta.haphazard.org>
X-Zippy:  My ELBOW is a remote FRENCH OUTPOST!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 26 Oct 2001 09:53:29 -0600
In-Reply-To: Brian Jones's message of "25 Oct 2001 22:15:34 -0400"
Message-ID: <87wv1i4cue.fsf@creche.redhat.com>
Lines: 12
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> Is the $param_0 and similar stuff in the rmi code ok?  gcj -C
Brian> doesn't seem to like it very much.

What exactly are you seeing?
As far as I know the RMI in the libgcj tree is the same as what is in
the Classpath tree, and gcj builds it there.
What version of gcj are you using?  Maybe there is a bug in some older
version.  I'm using 3.1 (the cvs trunk).

Tom


From sballard@netreach.com Fri Oct 26 12:51:17 2001
Received: from rawaki.netreach.net ([207.29.195.114])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xACT-0005m7-00
	for <classpath@gnu.org>; Fri, 26 Oct 2001 12:51:17 -0400
Received: from netreach.com (cc257928-a.neave1.pa.home.com [24.23.77.187]) by rawaki.netreach.net (n1/$Revision: 1.3 $) with ESMTP id MAA25958; Fri, 26 Oct 2001 12:51:11 -0400 (EDT)
Message-ID: <3BD99460.D56F53B5@netreach.com>
Date: Fri, 26 Oct 2001 12:50:40 -0400
From: Stuart Ballard <sballard@NetReach.Net>
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.4.7 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Eric Blake <ebb9@email.byu.edu>
CC: Bryce McKinlay <bryce@waitaki.otago.ac.nz>,
        Classpath list <classpath@gnu.org>
Subject: Re: classpath ./ChangeLog java/util/AbstractList.ja...
References: <E15vW2R-0000H0-00@subversions.gnu.org>
		 <3BD3A4A1.5090402@waitaki.otago.ac.nz> <3BD3B442.423BD1F8@email.byu.edu> <3BD3B7C3.3060002@waitaki.otago.ac.nz> <3BD3BF71.A1E59567@email.byu.edu>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Eric Blake wrote:
> 
> I dunno - I'm hesitant to add unspecified public methods to public
> classes, even though as you point out they will not destroy binary
> compatibility.

I've always been under the impression that it's absolutely fine to add
an overridden method for a method that's declared in a superclass; I
think of the fact that Sun does or doesn't do this in certain situations
as an implementation detail that's mistakenly published in the javadoc.
It's guaranteed to not break source or binary compatibility and it's
exactly what object orientation is all about.

Another Collections case that has a similar potential optimization is
LinkedList.subList(); the iterator() method on the sublist class would
normally be inherited from AbstractList, which calls
list.iterator(startIndex). On a linked list, iterator(n) is O(n). But if
the LinkedList sublist were to store the starting and ending Entry
objects, you can make this operation O(1).

Since creating an iterator is a pre-requisite for many many other
operations, I think making such an optimization for LinkedList sublists
is worth doing. Last I checked (which admittedly was at 1.2 release
time) Sun didn't do this. My old Classpath LinkedList implementation
did, but that was vaporized by the libgcj merge (and no great loss,
because in most other ways it sucked).

Stuart.


From chris.gray@acunia.com Thu Oct 25 03:30:38 2001
Received: from nat.be.acunia.com ([194.7.211.211] helo=pigeon.smove.be)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15weyL-0001YE-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 03:30:37 -0400
Received: from localhost.localdomain (IDENT:nobody@loki.leuven.intern.acunia.com [192.168.8.254])
	by pigeon.smove.be (8.11.6/8.11.6) with ESMTP id f9P7UV417328
	for <classpath@gnu.org>; Thu, 25 Oct 2001 09:30:31 +0200
Received: from acunia.com (IDENT:gray@sophocles [127.0.0.1])
	by localhost.localdomain (8.9.3/8.9.3) with ESMTP id JAA12121
	for <classpath@gnu.org>; Thu, 25 Oct 2001 09:30:35 +0200
Message-ID: <3BD7BF9A.3CBBC4E9@acunia.com>
Date: Thu, 25 Oct 2001 09:30:34 +0200
From: Chris Gray <chris.gray@acunia.com>
Organization: ACUNIA NV
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: classpath@gnu.org
Subject: Re: [hindm@us.ibm.com: Re: [Jikesrvm-researchers] Congratulations on 
 Jikes-RVM 2.0 - questions about licensing.]
References: <20011025084801.A21952@klomp.org>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard wrote:

> Hi,
>
> Found on the debian-java mailinglist.
>
> The Jikes VM is interesting because they say it is written mostly in java.
> <http://www-124.ibm.com/developerworks/oss/jikesrvm/>
> And they encourage to develop a Free Classlibrary for it.
> (It would ofcourse be better if we could convince them to donate their
> current class library to Classpath.)
> I am currently investigating the KissMe VM which is already fairly
> integrated with Classpath. If someone else wants to investigate how
> hard it would be to support Jikes VM that would be really nice.

I would also invite people to look at ACUNIA's Wonka VM,
<URL:http://wonka.acunia.com>.  Like IBM, we already developed our
own class libraries, but a Classpath/Wonka hybrid would appear to be
feasible from both a legal and a technical point of view.  Of more
practical value (probably) would be a gcj-Wonka synthesis.

Regards

Chris Gray
VM Architect, ACUNIA




From Yann.Secq@lifl.fr Thu Oct 25 04:07:21 2001
Received: from malonne.lifl.fr ([134.206.10.29])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15wfXs-0002jj-00
	for <classpath@gnu.org>; Thu, 25 Oct 2001 04:07:21 -0400
Received: from lifl.fr (akila.lifl.fr [134.206.10.4])
          by malonne.lifl.fr (8.9.3/jtpda-5.3.3) with ESMTP id KAA03895
          for <classpath@gnu.org>; Thu, 25 Oct 2001 10:07:08 +0200 (MET DST)
Message-ID: <3BD7C5FA.6060504@lifl.fr>
Date: Thu, 25 Oct 2001 09:57:46 +0200
From: Yann SECQ <Yann.Secq@lifl.fr>
Organization: SMAC
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.5) Gecko/20011023
X-Accept-Language: en, fr
MIME-Version: 1.0
To: classpath@gnu.org
Subject: Re: [hindm@us.ibm.com: Re: [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 - questions about licensing.]
References: <20011025084801.A21952@klomp.org> <87zo6gmawt.fsf@universe.yi.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Kevin A. Burton wrote:

> I took a look at the directory listings in the jar.  It looks like they only
> have the basics of a 1.1 style VM.  AKA only the basic stuff.
> It would be at least nice to compare and contrast between the different
> implementations to see if there is anythign we can bring over.  Of course if it
> isn't Free Software there is nothing we can do.

It's always the same question : what do you mean by free software ?
I think it is not copyleft, but it is an approved OpenSource license.

yann.

-- 
  / Yann SECQ            Equipe SMAC           secq@lifl.fr \
| Multi-Agent Systems Modeling & Agent Oriented Programming |
  \ http://www.lifl.fr/SMAC        http://www.lifl.fr/~secq /



From cbj@gnu.org Fri Oct 26 14:56:14 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xC9O-00026U-00
	for <classpath@gnu.org>; Fri, 26 Oct 2001 14:56:14 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Fri, 26 Oct 2001 14:56:14 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9QIuF603923;
	Fri, 26 Oct 2001 14:56:15 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: tromey@redhat.com
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: recent rmi additions
References: <m3adyf87uh.fsf@lyta.haphazard.org> <87wv1i4cue.fsf@creche.redhat.com>
From: Brian Jones <cbj@gnu.org>
Date: 26 Oct 2001 14:56:15 -0400
In-Reply-To: Tom Tromey's message of "26 Oct 2001 09:53:29 -0600"
Message-ID: <m3g0869qnk.fsf@lyta.haphazard.org>
Lines: 22
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey <tromey@redhat.com> writes:

> >>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:
> 
> Brian> Is the $param_0 and similar stuff in the rmi code ok?  gcj -C
> Brian> doesn't seem to like it very much.
> 
> What exactly are you seeing?

Complaints that they may or are not initialized.  Using gcj 2.96.  

> As far as I know the RMI in the libgcj tree is the same as what is in
> the Classpath tree, and gcj builds it there.
> What version of gcj are you using?  Maybe there is a bug in some older
> version.  I'm using 3.1 (the cvs trunk).

Of course.  So I'll leave gcj as an option but I can't see making it
the default at this time.

Brian
-- 
Brian Jones <cbj@gnu.org>


From tromey@creche.cygnus.com Fri Oct 26 17:22:10 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15xEQb-0007h1-00; Fri, 26 Oct 2001 17:22:10 -0400
Received: from creche.cygnus.com (ta0193.peakpeak.com [204.144.244.193] (may be forged))
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id PAA07713;
	Fri, 26 Oct 2001 15:21:43 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id PAA04677;
	Fri, 26 Oct 2001 15:40:34 -0600
To: Brian Jones <cbj@gnu.org>
Cc: Mark Wielaard <mark@klomp.org>, GNU Classpath <classpath@gnu.org>
Subject: Re: build update
References: <m36693831m.fsf@lyta.haphazard.org> <20011026131909.A14928@klomp.org> <m31yjqmrub.fsf@lyta.haphazard.org>
X-Zippy:  Did I do an INCORRECT THING??
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 26 Oct 2001 11:21:05 -0600
In-Reply-To: Brian Jones's message of "26 Oct 2001 09:52:12 -0400"
Message-ID: <87zo6e2u7y.fsf@creche.redhat.com>
Lines: 11
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> To use gcj for bytecode compilation, use --with-gcj.  Since
Brian> jikes is (a) faster and (b) works for me, jikes is still the
Brian> default.

FWIW, gcj 3.1 ought to be much faster here, maybe as fast as jikes.
Between 3.0 and 3.1, Per changed gcj so that it can compile multiple
.java files to .class files at once.

Tom


From jsturm@one-point.com Fri Oct 26 17:35:14 2001
Received: from cc730106-b.taylor1.mi.home.com ([24.13.63.84])
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xEdF-0008Sm-00; Fri, 26 Oct 2001 17:35:13 -0400
Received: from localhost (jsturm@localhost)
	by cc730106-b.taylor1.mi.home.com (8.9.3/8.9.3) with ESMTP id RAA10813;
	Fri, 26 Oct 2001 17:43:54 -0400
X-Authentication-Warning: mars.deadcafe.org: jsturm owned process doing -bs
Date: Fri, 26 Oct 2001 17:43:54 -0400 (EDT)
From: Jeff Sturm <jsturm@one-point.com>
X-Sender: jsturm@mars.deadcafe.org
To: Tom Tromey <tromey@redhat.com>
cc: Brian Jones <cbj@gnu.org>, Mark Wielaard <mark@klomp.org>,
        GNU Classpath <classpath@gnu.org>
Subject: Re: build update
In-Reply-To: <87zo6e2u7y.fsf@creche.redhat.com>
Message-ID: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>


On 26 Oct 2001, Tom Tromey wrote:
> Between 3.0 and 3.1, Per changed gcj so that it can compile multiple
> .java files to .class files at once.

3.0.2 can as well, and maybe earlier... I don't remember exactly when Per
added that feature.

Jeff



From tromey@creche.cygnus.com Fri Oct 26 17:38:40 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15xEga-00009W-00; Fri, 26 Oct 2001 17:38:40 -0400
Received: from creche.cygnus.com (ta0193.peakpeak.com [204.144.244.193] (may be forged))
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id PAA11074;
	Fri, 26 Oct 2001 15:38:36 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id PAA14347;
	Fri, 26 Oct 2001 15:57:31 -0600
To: Jeff Sturm <jsturm@one-point.com>
Cc: Brian Jones <cbj@gnu.org>, Mark Wielaard <mark@klomp.org>,
        GNU Classpath <classpath@gnu.org>
Subject: Re: build update
References: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org>
X-Zippy:  Please come home with me...  I have Tylenol!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 26 Oct 2001 15:57:30 -0600
In-Reply-To: Jeff Sturm's message of "Fri, 26 Oct 2001 17:43:54 -0400 (EDT)"
Message-ID: <87wv1irrn9.fsf@creche.redhat.com>
Lines: 13
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Jeff" == Jeff Sturm <jsturm@one-point.com> writes:

>> Between 3.0 and 3.1, Per changed gcj so that it can compile multiple
>> .java files to .class files at once.

Jeff> 3.0.2 can as well, and maybe earlier... I don't remember exactly
Jeff> when Per added that feature.

Thanks -- I looked in the ChangeLog and saw that some bug fixes went
in post-3.0, but I had forgotten that he had merged the work to the
3.0 branch.

Tom


From cbj@gnu.org Fri Oct 26 23:17:31 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xJyV-0002IO-00
	for <classpath@gnu.org>; Fri, 26 Oct 2001 23:17:31 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Fri, 26 Oct 2001 23:17:30 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9R3HQ916086;
	Fri, 26 Oct 2001 23:17:26 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Jeff Sturm <jsturm@one-point.com>
Cc: Tom Tromey <tromey@redhat.com>, Mark Wielaard <mark@klomp.org>,
   GNU Classpath <classpath@gnu.org>
Subject: Re: build update
References: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org>
From: Brian Jones <cbj@gnu.org>
Date: 26 Oct 2001 23:17:26 -0400
In-Reply-To: Jeff Sturm's message of "Fri, 26 Oct 2001 17:43:54 -0400 (EDT)"
Message-ID: <m3zo6d7ovt.fsf@lyta.haphazard.org>
Lines: 14
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Jeff Sturm <jsturm@one-point.com> writes:

> On 26 Oct 2001, Tom Tromey wrote:
> > Between 3.0 and 3.1, Per changed gcj so that it can compile multiple
> > .java files to .class files at once.
> 
> 3.0.2 can as well, and maybe earlier... I don't remember exactly when Per
> added that feature.

Thanks, I'll upgrade to the latest 3.0.x version soon.

Brian
-- 
Brian Jones <cbj@gnu.org>


From mark@klomp.org Sat Oct 27 09:09:55 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xTDm-0006Ld-00; Sat, 27 Oct 2001 09:09:54 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15xTDV-0004Uk-00; Sat, 27 Oct 2001 15:09:37 +0200
Date: Sat, 27 Oct 2001 15:09:37 +0200
From: Mark Wielaard <mark@klomp.org>
To: Brian Jones <cbj@gnu.org>
Cc: Jeff Sturm <jsturm@one-point.com>, Tom Tromey <tromey@redhat.com>,
	GNU Classpath <classpath@gnu.org>
Subject: Re: build update
Message-ID: <20011027150937.A5532@klomp.org>
References: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org> <m3zo6d7ovt.fsf@lyta.haphazard.org>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="ZGiS0Q5IWpPtfppv"
Content-Disposition: inline
In-Reply-To: <m3zo6d7ovt.fsf@lyta.haphazard.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

--ZGiS0Q5IWpPtfppv
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

Hi.

On Fri, Oct 26, 2001 at 11:17:26PM -0400, Brian Jones wrote:
> Jeff Sturm <jsturm@one-point.com> writes:
> 
> > On 26 Oct 2001, Tom Tromey wrote:
> > > Between 3.0 and 3.1, Per changed gcj so that it can compile multiple
> > > .java files to .class files at once.
> > 
> > 3.0.2 can as well, and maybe earlier... I don't remember exactly when Per
> > added that feature.
> 
> Thanks, I'll upgrade to the latest 3.0.x version soon.
You should. It is not as fast as jikes but even on my 300 Mhz machine it
takes less then a minute to compile Classpath completely to bytecode.
(jikes takes 

I finally found a way to get gcj to compile Classpath with our build
scripts. You have to make sure that no classes are defined twice on the
command lince. So I changed the gen-classlist script to only look in
the srcdir for java files and to get the gnu/classpath/Configuration.java
file explicitly from the builddir. (You have add it to the standard.omit
list in case you configure in the srcdir.) And gcj goes crazy when it finds
classes on its classpath that are also on its commandline (somehow it gets
confused because it cannot see that ../X.java is the same as ./../X.java).
But if you don't give it any classpath options at all then it compiles
just fine. (I also made sure that all needed files are actually compiled
by changing standard.omit to include some AWT files that have a different
definition in Claspath from the one in libgcj.jar. Already committed.)

Attached is a diff against your patch (and standard.omit against CVS).
Could you include these if you commit your build changes?

Everything was tested against gcc version 3.0.2 20010922 (Debian prerelease).

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore

--ZGiS0Q5IWpPtfppv
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="mybuildpatch.diff"

Index: lib/standard.omit
===================================================================
RCS file: /cvs/classpath/lib/standard.omit,v
retrieving revision 1.29
diff -u -r1.29 standard.omit
--- lib/standard.omit	2001/10/26 15:24:15	1.29
+++ lib/standard.omit	2001/10/27 12:44:13
@@ -1,3 +1,4 @@
+../gnu/classpath/Configuration.java$
 ../gnu/test/.*$
 ../gnu/java/awt/peer/gtk/TestAWT.java$
 ../gnu/java/awt/peer/gtk/Test.java$
--- lib/Makefile.am	Sat Oct 27 14:44:39 2001
+++ mark/lib/Makefile.am	Sat Oct 27 14:41:49 2001
@@ -52,7 +52,7 @@
 JAVAC = $(JIKES) -nowarn +F $(ARG_CLASSPATH_JAVAC) $(top_srcdir):$(top_builddir)/vm/current:.:$(USER_CLASSLIB) -d . @classes
 else
 if FOUND_GCJ
-JAVAC = $(GCJ) -C $(ARG_CLASSPATH_JAVAC) $(top_srcdir):$(top_builddir)/vm/current:.:$(USER_CLASSLIB) -d . $(JAVA_SRCS)
+JAVAC = $(GCJ) -C -d . @classes
 else
 if FOUND_KJC
 JAVAC = $(USER_JABBA) $(ARG_CLASSPATH_VM) $(KJC_CLASSPATH):$(USER_CLASSLIB) $(KJC_MAIN) $(ARG_CLASSPATH_JAVAC) .:$(USER_CLASSLIB) -d . @classes
--- lib/gen-classlist.sh.in	Sat Oct 27 14:44:39 2001
+++ mark/lib/gen-classlist.sh.in	Sat Oct 27 14:41:49 2001
@@ -1,4 +1,6 @@
 #! @SH@
 # @configure_input@
-@FIND@ @top_srcdir@/com @top_srcdir@/java @top_srcdir@/javax @top_srcdir@/gnu ${top_builddir}/java ${top_builddir}/javax ${top_builddir}/gnu ${top_builddir}/vm/current -follow -type f -print | grep '\.java$' > ${top_builddir}/lib/classes
+@FIND@ @srcdir@/../java @srcdir@/../javax @srcdir@/../gnu ${top_builddir}/vm/current -follow -type f -print | grep '\.java$' > ${top_builddir}/lib/classes
 for filexp in `cat @top_srcdir@/lib/$1.omit` ; do { grep -v ${filexp} < ${top_builddir}/lib/classes > ${top_builddir}/lib/classes.new ; mv ${top_builddir}/lib/classes.new ${top_builddir}/lib/classes ; } ; done
+# Explicitly add generated java file
+echo ${top_builddir}/gnu/classpath/Configuration.java >> ${top_builddir}/lib/classes

--ZGiS0Q5IWpPtfppv--


From cbj@gnu.org Sat Oct 27 09:23:59 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xTRP-0006e0-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 09:23:59 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sat, 27 Oct 2001 09:23:59 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9RDNnx17077;
	Sat, 27 Oct 2001 09:23:49 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: Jeff Sturm <jsturm@one-point.com>, Tom Tromey <tromey@redhat.com>,
   GNU Classpath <classpath@gnu.org>
Subject: Re: build update
References: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org> <m3zo6d7ovt.fsf@lyta.haphazard.org> <20011027150937.A5532@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 27 Oct 2001 09:23:49 -0400
In-Reply-To: Mark Wielaard's message of "Sat, 27 Oct 2001 15:09:37 +0200"
Message-ID: <m3vgh16wt6.fsf@lyta.haphazard.org>
Lines: 49
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> Hi.
> 
> On Fri, Oct 26, 2001 at 11:17:26PM -0400, Brian Jones wrote:
> > Jeff Sturm <jsturm@one-point.com> writes:
> > 
> > > On 26 Oct 2001, Tom Tromey wrote:
> > > > Between 3.0 and 3.1, Per changed gcj so that it can compile multiple
> > > > .java files to .class files at once.
> > > 
> > > 3.0.2 can as well, and maybe earlier... I don't remember exactly when Per
> > > added that feature.
> > 
> > Thanks, I'll upgrade to the latest 3.0.x version soon.
>
> You should. It is not as fast as jikes but even on my 300 Mhz machine it
> takes less then a minute to compile Classpath completely to bytecode.
> (jikes takes 
> 
> I finally found a way to get gcj to compile Classpath with our build
> scripts. You have to make sure that no classes are defined twice on the
> command lince. So I changed the gen-classlist script to only look in
> the srcdir for java files and to get the gnu/classpath/Configuration.java
> file explicitly from the builddir. (You have add it to the standard.omit
> list in case you configure in the srcdir.) 

Well, there is the distinct possibility that we need to generate more
classes than just Configuration.java so I'd rather not single it out.

> And gcj goes crazy when it finds classes on its classpath that are
> also on its commandline (somehow it gets confused because it cannot
> see that ../X.java is the same as ./../X.java).  

Sounds like a bug.

> But if you don't give it any classpath options at all then it
> compiles just fine. (I also made sure that all needed files are
> actually compiled by changing standard.omit to include some AWT
> files that have a different definition in Claspath from the one in
> libgcj.jar. Already committed.)
> 
> Attached is a diff against your patch (and standard.omit against CVS).
> Could you include these if you commit your build changes?

I'll take a look.
 
-- 
Brian Jones <cbj@gnu.org>


From mark@klomp.org Sat Oct 27 09:44:27 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xTlD-0007F7-00; Sat, 27 Oct 2001 09:44:27 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15xTl9-0005EN-00; Sat, 27 Oct 2001 15:44:23 +0200
Date: Sat, 27 Oct 2001 15:44:23 +0200
From: Mark Wielaard <mark@klomp.org>
To: Brian Jones <cbj@gnu.org>
Cc: Jeff Sturm <jsturm@one-point.com>, Tom Tromey <tromey@redhat.com>,
	GNU Classpath <classpath@gnu.org>
Subject: Re: build update
Message-ID: <20011027154422.A17472@klomp.org>
References: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org> <m3zo6d7ovt.fsf@lyta.haphazard.org> <20011027150937.A5532@klomp.org> <m3vgh16wt6.fsf@lyta.haphazard.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <m3vgh16wt6.fsf@lyta.haphazard.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Sat, Oct 27, 2001 at 09:23:49AM -0400, Brian Jones wrote:
> Mark Wielaard <mark@klomp.org> writes:
> > I finally found a way to get gcj to compile Classpath with our build
> > scripts. You have to make sure that no classes are defined twice on the
> > command lince. So I changed the gen-classlist script to only look in
> > the srcdir for java files and to get the gnu/classpath/Configuration.java
> > file explicitly from the builddir. (You have add it to the standard.omit
> > list in case you configure in the srcdir.) 
> 
> Well, there is the distinct possibility that we need to generate more
> classes than just Configuration.java so I'd rather not single it out.

The problem is when srcdir == builddir and you try to get a list by
adding both all java files from the srcdir and all java files from the
builddir then the generated list contains duplicates. Only explicitly adding
generated java files from the builddir works around this. But if you want
you can also solve this by using e.g. sort -u.

> > And gcj goes crazy when it finds classes on its classpath that are
> > also on its commandline (somehow it gets confused because it cannot
> > see that ../X.java is the same as ./../X.java).  
> 
> Sounds like a bug.
It is, but I have not been able to get a small testcase.
Note to gcj hackers: To reproduce just run the following in the lib dir
from Classpath "gcj --classpath=.. -C -d . @classes". You will get errors
like:
../java/lang/ClassNotFoundException.java:48: Class `java.lang.ClassNotFoundException' already defined in ././../java/lang/ClassNotFoundException.java:48.
   public class ClassNotFoundException extends Exception
                ^

But making sure that a source file is not both on the commandline and
on the -classpath list works around it.

> > Attached is a diff against your patch (and standard.omit against CVS).
> > Could you include these if you commit your build changes?
> 
> I'll take a look.

Thanks, it would be nice to support gcj 3.0.x out of the box.

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Sat Oct 27 12:09:09 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xW1F-0003Ub-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 12:09:09 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sat, 27 Oct 2001 12:08:44 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9RG8WV17338;
	Sat, 27 Oct 2001 12:08:32 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: jni.h VM dependencies
From: Brian Jones <cbj@gnu.org>
Date: 27 Oct 2001 12:08:31 -0400
Message-ID: <m3ofmt6p6o.fsf@lyta.haphazard.org>
Lines: 42
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hello all,

I was wondering if any of you have great ideas for how to handle these
opaque structures defined in a typical jni.h.

typedef struct _jfieldID *jfieldID;
typedef struct _jmethodID *jmethodID;

I am currently thinking something like this in jni.h

#include "jni_vm.h"

typedef JFIELD_ID jfieldID;
typedef JMETHOD_ID jmethodID;

-- cut

And in jni_vm.h, place VM specific things like...

#if JVM == GCJ
#define JFIELD_ID void *
#define JMETHOD_ID void *
#endif

#if JVM == ORP
#define JFIELD_ID void *
#define JMETHOD_ID void *
#endif

#if JVM == JAPHAR
#define JFIELD_ID struct FieldStruct *
#define JMETHOD_ID struct MethodStruct *
#endif


Is there a way to avoid this nasty situation or a more elegant
solution?  Like in the Japhar case, now what happens while compiling
and it can't find FieldStruct?  Ugh.  

Brian
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Sat Oct 27 12:10:24 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xW2S-0003Xx-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 12:10:24 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sat, 27 Oct 2001 12:10:24 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9RGACH17348;
	Sat, 27 Oct 2001 12:10:12 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: GCJ jni.h fixes
From: Brian Jones <cbj@gnu.org>
Date: 27 Oct 2001 12:10:12 -0400
Message-ID: <m3k7xh6p3v.fsf@lyta.haphazard.org>
Lines: 9
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I know there are some gcj hackers on the list so I'm posting this
here.  I notice in gcj's jni.h the following problems in latest CVS.

DefineClass missing const char * argument
GetStringLength return type is jsize, not jint (although these are the same)
GetStringUTFChars return type is const jbyte *, not const char*

-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Sat Oct 27 13:00:31 2001
Received: from fe8.southeast.rr.com ([24.93.67.55] helo=mail8.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xWow-00054J-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 13:00:30 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by mail8.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sat, 27 Oct 2001 13:00:30 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9RH0I715900;
	Sat, 27 Oct 2001 13:00:18 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: Re: jni.h VM dependencies
References: <m3ofmt6p6o.fsf@lyta.haphazard.org>
From: Brian Jones <cbj@gnu.org>
Date: 27 Oct 2001 13:00:18 -0400
In-Reply-To: Brian Jones's message of "27 Oct 2001 12:08:31 -0400"
Message-ID: <m3g0856msd.fsf@lyta.haphazard.org>
Lines: 20
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Brian Jones <cbj@gnu.org> writes:

> Hello all,
> 
> I was wondering if any of you have great ideas for how to handle these
> opaque structures defined in a typical jni.h.
> 
> typedef struct _jfieldID *jfieldID;
> typedef struct _jmethodID *jmethodID;
> 

Let me restate, at the moment it is as above with the below just
before.

struct _jfieldID;
struct _jmethodID;

This is as defined in JNI specification.
-- 
Brian Jones <cbj@gnu.org>


From tromey@creche.cygnus.com Sat Oct 27 14:09:07 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15xXtL-0007Gt-00; Sat, 27 Oct 2001 14:09:07 -0400
Received: from creche.cygnus.com (ta0204.peakpeak.com [204.144.244.204])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id MAA05560;
	Sat, 27 Oct 2001 12:09:00 -0600
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id MAA31317;
	Sat, 27 Oct 2001 12:28:09 -0600
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: GCJ jni.h fixes
References: <m3k7xh6p3v.fsf@lyta.haphazard.org>
X-Zippy:  I'm ANN LANDERS!!  I can SHOPLIFT!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 27 Oct 2001 12:28:08 -0600
In-Reply-To: Brian Jones's message of "27 Oct 2001 12:10:12 -0400"
Message-ID: <87pu799buv.fsf@creche.redhat.com>
Lines: 27
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> DefineClass missing const char * argument

We have this:

  jclass   (*DefineClass)                  (JNIEnv *, jobject,
					    const jbyte *, jsize);

To me this looks the same as the JNI spec:

    http://java.sun.com/products/jdk/1.2/docs/guide/jni/spec/functions.doc.html#15986

Brian> GetStringLength return type is jsize, not jint (although these
Brian> are the same)

Thanks, I fixed this.

Brian> GetStringUTFChars return type is const jbyte *, not const char*

We comply with the JNI 1.2 spec:

    http://java.sun.com/products/jdk/1.2/docs/guide/jni/spec/functions.doc.html#17265

Where are you looking?

Tom


From mark@klomp.org Sat Oct 27 16:47:20 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xaMR-0003tW-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 16:47:20 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15xaMI-00070e-00; Sat, 27 Oct 2001 22:47:10 +0200
Date: Sat, 27 Oct 2001 22:47:10 +0200
From: Mark Wielaard <mark@klomp.org>
To: Chris Gray <chris.gray@acunia.com>
Cc: classpath@gnu.org
Subject: Wonka
Message-ID: <20011027224710.A26801@klomp.org>
References: <20011025084801.A21952@klomp.org> <3BD7BF9A.3CBBC4E9@acunia.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BD7BF9A.3CBBC4E9@acunia.com>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Thu, Oct 25, 2001 at 09:30:34AM +0200, Chris Gray wrote:
> 
> I would also invite people to look at ACUNIA's Wonka VM,
> <URL:http://wonka.acunia.com>.  Like IBM, we already developed our
> own class libraries, but a Classpath/Wonka hybrid would appear to be
> feasible from both a legal and a technical point of view.  Of more
> practical value (probably) would be a gcj-Wonka synthesis.

Nice. From you package report page I get the impression that you have
a lot of classes for 1.1 and 1.2. But it does not seem as complete as
Classpath at the moment. But you have a working AWT implementation!
And some javax.crypto stuff. Cool.

Was there a reason you could not use (or cooperate with) GNU Classpath
for your VM?

Two legal nitpicks. (Please ignore them if you don't like legal issues.)
Your license is not GPL compatible (which may or may not be a problem).
The GPL incompatability come from clause 4:
4. Products derived from this software may not be called "Wonka",
   nor may "Wonka" appear in their name, without prior written
   permission of ACUNIA NV.
This is the same as can be found in the Apache license which is also not
GPL compatible since it is an added restriction (although a minor one)
according to GPL clause 6. But since you actaully have a trademark on Wonka
this clause is not necessary. You could replace it with a fact (outside the
actual license text) such as "The following are trademarks of Acunia, XXX,
YYY, ZZZ. This license does not grand you any permission to use any
trademarks of Acunia when distributing derived works." That way it would
be GPL compatible (since then your claim is based on trademark law and not
on copyright laa)w. You might want to take a look at how AbiSource did a
similar thing with their trademark on AbiWord.

Your download does contain a COPYING file with the same clauses as menioned
on <http://wonka.acunia.com/licence.html>. But all source files and
documentation comes with a header that does not sound very open/free.
It would be nice to make it a bit more clear that the license agreement
mentioned in that text is acutally the top level COPYING file. Note for
example that all Classpath files explicitly state the license conditions.
This makes it much more clear what your rights are if you somehow only
received part of the code archive.

Groetjes,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From mark@klomp.org Sat Oct 27 16:59:18 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xaY1-0004kv-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 16:59:17 -0400
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15xaXY-00071U-00; Sat, 27 Oct 2001 22:58:48 +0200
Date: Sat, 27 Oct 2001 22:58:48 +0200
From: Mark Wielaard <mark@klomp.org>
To: Yann SECQ <Yann.Secq@lifl.fr>
Cc: classpath@gnu.org
Subject: Re: [hindm@us.ibm.com: Re: [Jikesrvm-researchers] Congratulations on Jikes-RVM 2.0 - questions about licensing.]
Message-ID: <20011027225848.B26801@klomp.org>
References: <20011025084801.A21952@klomp.org> <87zo6gmawt.fsf@universe.yi.org> <3BD7C5FA.6060504@lifl.fr>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BD7C5FA.6060504@lifl.fr>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Thu, Oct 25, 2001 at 09:57:46AM +0200, Yann SECQ wrote:
> Kevin A. Burton wrote:
> 
> >It would be at least nice to compare and contrast between the different
> >implementations to see if there is anythign we can bring over.  Of course 
> >if it isn't Free Software there is nothing we can do.
> 
> It's always the same question : what do you mean by free software ?
Read <http://www.gnu.org/philosophy/free-sw.html>

> I think it is not copyleft, but it is an approved OpenSource license.
The Jikes VM itself is (probably) Free Software since it is released under
the Common Public License which is derived from the IBM Public License which
is a Free Software license. But the class libraries are certainly not Free
Software since they are distributed under a separate license which is very
restrictive. See <http://www-124.ibm.com/cgi-bin/jikesrvm/jlibraries>.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Sat Oct 27 21:03:51 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xeMh-0004iQ-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 21:03:51 -0400
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sat, 27 Oct 2001 21:03:45 -0400
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9S13Sc17552;
	Sat, 27 Oct 2001 21:03:28 -0400
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: tromey@redhat.com
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: GCJ jni.h fixes
References: <m3k7xh6p3v.fsf@lyta.haphazard.org> <87pu799buv.fsf@creche.redhat.com>
From: Brian Jones <cbj@gnu.org>
Date: 27 Oct 2001 21:03:28 -0400
In-Reply-To: Tom Tromey's message of "27 Oct 2001 12:28:08 -0600"
Message-ID: <m37ktg7ezj.fsf@lyta.haphazard.org>
Lines: 30
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey <tromey@redhat.com> writes:

> >>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:
> 
> Brian> DefineClass missing const char * argument
> 
> We have this:
> 
>   jclass   (*DefineClass)                  (JNIEnv *, jobject,
> 					    const jbyte *, jsize);
> 
> To me this looks the same as the JNI spec:
> 
>     http://java.sun.com/products/jdk/1.2/docs/guide/jni/spec/functions.doc.html#15986
>
> Brian> GetStringUTFChars return type is const jbyte *, not const char*
> 
> We comply with the JNI 1.2 spec:
> 
>     http://java.sun.com/products/jdk/1.2/docs/guide/jni/spec/functions.doc.html#17265
> 
> Where are you looking?

I'm looking in "The Java Native Interface" ISBN 0-201-32577-2, in
which I have found one error with regards to function index so it is
possible these are errors as well, maybe not.

Brian
-- 
Brian Jones <cbj@gnu.org>


From bryce@waitaki.otago.ac.nz Sat Oct 27 21:30:39 2001
Received: from 210-55-119-123.adsl.xtra.co.nz ([210.55.119.123] helo=localhost.localdomain)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xemc-0005ZU-00
	for <classpath@gnu.org>; Sat, 27 Oct 2001 21:30:38 -0400
Received: from waitaki.otago.ac.nz ([192.168.2.2])
	by localhost.localdomain (8.11.2/8.11.2) with ESMTP id f9S1Beh10041;
	Sun, 28 Oct 2001 14:11:40 +1300
Message-ID: <3BDB51AD.7040701@waitaki.otago.ac.nz>
Date: Sun, 28 Oct 2001 13:30:37 +1300
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
User-Agent: Mozilla/5.0 (X11; U; Linux ppc; en-US; rv:0.9.5) Gecko/20011016
X-Accept-Language: en-us
MIME-Version: 1.0
To: mark@klomp.org, classpath@gnu.org
Subject: Re: classpath ./ChangeLog lib/standard.omit
References: <E15x8qF-0001K9-00@subversions.gnu.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard wrote:

>Log message:
>	* lib/standard.omit: DataFlavor.java and AWTEvent.java should and must
>	be compiled by default
>

AWTEvent depends on Event.java, which for some reason doesn't exist in 
Classpath. If you compile classpath with javac its okay because javac 
finds Event in its own classes, but it won't work if you try to 
bootstrap with jikes or gcj -C without having Sun's class libraries 
around. So, java.awt should probibly be omitted by default until that 
gets sorted.

regards

Bryce.




From bryce@waitaki.otago.ac.nz Sat Oct 27 22:04:55 2001
Received: from 210-55-119-123.adsl.xtra.co.nz ([210.55.119.123] helo=localhost.localdomain)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xfJn-0006oV-00; Sat, 27 Oct 2001 22:04:55 -0400
Received: from waitaki.otago.ac.nz ([192.168.2.2])
	by localhost.localdomain (8.11.2/8.11.2) with ESMTP id f9S1jvh10055;
	Sun, 28 Oct 2001 14:45:57 +1300
Message-ID: <3BDB6948.6030002@waitaki.otago.ac.nz>
Date: Sun, 28 Oct 2001 15:11:20 +1300
From: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
User-Agent: Mozilla/5.0 (X11; U; Linux ppc; en-US; rv:0.9.5) Gecko/20011016
X-Accept-Language: en-us
MIME-Version: 1.0
To: Mark Wielaard <mark@klomp.org>
CC: Brian Jones <cbj@gnu.org>, Jeff Sturm <jsturm@one-point.com>,
   Tom Tromey <tromey@redhat.com>, GNU Classpath <classpath@gnu.org>
Subject: Re: build update
References: <Pine.LNX.4.10.10110261740500.10811-100000@mars.deadcafe.org> <m3zo6d7ovt.fsf@lyta.haphazard.org> <20011027150937.A5532@klomp.org> <m3vgh16wt6.fsf@lyta.haphazard.org> <20011027154422.A17472@klomp.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard wrote:

>Note to gcj hackers: To reproduce just run the following in the lib dir
>from Classpath "gcj --classpath=.. -C -d . @classes". You will get errors
>like:
>../java/lang/ClassNotFoundException.java:48: Class `java.lang.ClassNotFoundException' already defined in ././../java/lang/ClassNotFoundException.java:48.
>   public class ClassNotFoundException extends Exception
>                ^
>

This is almost the same as the bug which prevents libgcj from building 
on case-insensitive filesystems (Darwin and Cygwin). The problem here is 
that GCJ thinks they are different files due to the different path 
prefix at the start of the filename. We need to figure out a good way to 
determine if two files are really the same file, even if they are loaded 
via different paths and/or names...

regards

Bryce.




From mark@klomp.org Sun Oct 28 04:18:13 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xm57-0004yz-00
	for <classpath@gnu.org>; Sun, 28 Oct 2001 04:18:13 -0500
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15xm51-0002Ov-00; Sun, 28 Oct 2001 10:18:07 +0100
Date: Sun, 28 Oct 2001 10:18:07 +0100
From: Mark Wielaard <mark@klomp.org>
To: Bryce McKinlay <bryce@waitaki.otago.ac.nz>
Cc: classpath@gnu.org
Subject: Re: classpath ./ChangeLog lib/standard.omit
Message-ID: <20011028101807.A9185@klomp.org>
References: <E15x8qF-0001K9-00@subversions.gnu.org> <3BDB51AD.7040701@waitaki.otago.ac.nz>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BDB51AD.7040701@waitaki.otago.ac.nz>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Sun, Oct 28, 2001 at 01:30:37PM +1300, Bryce McKinlay wrote:
> AWTEvent depends on Event.java, which for some reason doesn't exist in 
> Classpath. If you compile classpath with javac its okay because javac 
> finds Event in its own classes, but it won't work if you try to 
> bootstrap with jikes or gcj -C without having Sun's class libraries 
> around. So, java.awt should probibly be omitted by default until that 
> gets sorted.

Theoretically you are right. But John Leuner recently removed all Event
support from Classpath. While not the most elegant solution this should
make Classpath awt compile out of the box. And it does for me.

How did you configure? Maybe your checkout is not completely up to date?
As a workaround you can use libgcj.jar or Klasses.zip (from kaffe) in
your classpath, but that should not be neccessary.

I would like to see the standard.omit file as small as possible.
Ideally for me all java files in the cvs tree should be compilable
with gcj, jikes or kjc without any additional (boot)classpaths set.
And we would only select the class files that we don't want to
ship by default.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Sun Oct 28 10:33:47 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xrwZ-0003JP-00
	for <classpath@gnu.org>; Sun, 28 Oct 2001 10:33:47 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sun, 28 Oct 2001 10:33:45 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9SFXJZ13657;
	Sun, 28 Oct 2001 10:33:19 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: Bryce McKinlay <bryce@waitaki.otago.ac.nz>, classpath@gnu.org
Subject: Re: classpath ./ChangeLog lib/standard.omit
References: <E15x8qF-0001K9-00@subversions.gnu.org> <3BDB51AD.7040701@waitaki.otago.ac.nz> <20011028101807.A9185@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 28 Oct 2001 10:33:19 -0500
In-Reply-To: Mark Wielaard's message of "Sun, 28 Oct 2001 10:18:07 +0100"
Message-ID: <m3u1wj6aps.fsf@lyta.haphazard.org>
Lines: 15
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> I would like to see the standard.omit file as small as possible.
> Ideally for me all java files in the cvs tree should be compilable
> with gcj, jikes or kjc without any additional (boot)classpaths set.
> And we would only select the class files that we don't want to
> ship by default.

Just a note on this, the latest beta kjc from kopisusu did not compile
Classpath for me.  The one from Kaffe may work however since these are
different forks of the compiler.

Brian
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Sun Oct 28 12:13:15 2001
Received: from fe7.southeast.rr.com ([24.93.67.54] helo=mail7.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xtUp-0008Ll-00
	for <classpath@gnu.org>; Sun, 28 Oct 2001 12:13:15 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by mail7.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sun, 28 Oct 2001 12:12:59 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9SHCSY13945;
	Sun, 28 Oct 2001 12:12:28 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: java.math.*
From: Brian Jones <cbj@gnu.org>
Date: 28 Oct 2001 12:12:28 -0500
Message-ID: <m3lmhv664j.fsf@lyta.haphazard.org>
Lines: 8
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I'd like to adopt the gcj java.math.BigInteger/BigDecimal classes and
the related gnu.gcj.math.MPN.  This latter class is from Kawa so I'd
like to make sure we can use this.  Are there any problems or
objections?

Brian
-- 
Brian Jones <cbj@gnu.org>


From per@bothner.com Sun Oct 28 13:59:27 2001
Received: from adsl-216-102-199-253.dsl.snfc21.pacbell.net ([216.102.199.253] helo=bothner.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15xv9b-0004I8-00; Sun, 28 Oct 2001 13:59:27 -0500
Received: from bothner.com (eureka.bothner.com [192.168.1.9])
	by bothner.com (8.11.0/8.11.0) with ESMTP id f9SJFgH05297;
	Sun, 28 Oct 2001 11:15:42 -0800
Message-ID: <3BDC63EC.80503@bothner.com>
Date: Sun, 28 Oct 2001 12:00:44 -0800
From: Per Bothner <per@bothner.com>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.2.1) Gecko/20010901
X-Accept-Language: en-us
MIME-Version: 1.0
To: Brian Jones <cbj@gnu.org>
CC: GNU Classpath <classpath@gnu.org>
Subject: Re: java.math.*
References: <m3lmhv664j.fsf@lyta.haphazard.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Brian Jones wrote:

>I'd like to adopt the gcj java.math.BigInteger/BigDecimal classes and
>the related gnu.gcj.math.MPN.  This latter class is from Kawa so I'd
>like to make sure we can use this.  Are there any problems or
>objections?
>
You have my blessing (as author of Kawa and the MPN class).

It would be nice if someone cleaned up the serialization. Right now
BigInteger has a bunch of extra fields which are only used
temporarily during serialization, to match Sun's format.  It would be
great if someone who knows the serialization format (or wants to
learn) could write the code to generate the correct serilaization
direcectly without using temporary fields.  (Even better would be to
avoid generating temporary objects when serializing, but I'll settle
for getting rid of the extra fields.)
    --Per Bothner




From cbj@gnu.org Sun Oct 28 23:09:30 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15y3ju-0005wm-00
	for <classpath@gnu.org>; Sun, 28 Oct 2001 23:09:30 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Sun, 28 Oct 2001 23:08:29 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9T47ue14825;
	Sun, 28 Oct 2001 23:07:56 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Per Bothner <per@bothner.com>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: java.math.*
References: <m3lmhv664j.fsf@lyta.haphazard.org> <3BDC63EC.80503@bothner.com>
From: Brian Jones <cbj@gnu.org>
Date: 28 Oct 2001 23:07:55 -0500
In-Reply-To: Per Bothner's message of "Sun, 28 Oct 2001 12:00:44 -0800"
Message-ID: <m3pu77qeas.fsf@lyta.haphazard.org>
Lines: 21
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Per Bothner <per@bothner.com> writes:

> Brian Jones wrote:
> 
> >I'd like to adopt the gcj java.math.BigInteger/BigDecimal classes and
> >the related gnu.gcj.math.MPN.  This latter class is from Kawa so I'd
> >like to make sure we can use this.  Are there any problems or
> >objections?
> >
> You have my blessing (as author of Kawa and the MPN class).

Thank you.
 
> It would be nice if someone cleaned up the serialization. Right now
> BigInteger has a bunch of extra fields which are only used
> temporarily during serialization, to match Sun's format.  

I've added this as a task to the Savannah project page for java.math.

-- 
Brian Jones <cbj@gnu.org>


From chris.gray@acunia.com Mon Oct 29 05:13:51 2001
Received: from nat.be.acunia.com ([194.7.211.211] helo=pigeon.smove.be)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15y9QU-0004yx-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 05:13:50 -0500
Received: from localhost.localdomain (IDENT:nobody@loki.leuven.intern.acunia.com [192.168.8.254])
	by pigeon.smove.be (8.11.6/8.11.6) with ESMTP id f9TADb404460
	for <classpath@gnu.org>; Mon, 29 Oct 2001 11:13:37 +0100
Received: from acunia.com (IDENT:gray@sophocles [127.0.0.1])
	by localhost.localdomain (8.9.3/8.9.3) with ESMTP id LAA12695;
	Sun, 28 Oct 2001 11:13:17 +0100
Message-ID: <3BDBDA3D.76EACA6@acunia.com>
Date: Sun, 28 Oct 2001 11:13:17 +0100
From: Chris Gray <chris.gray@acunia.com>
Organization: ACUNIA NV
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Mark Wielaard <mark@klomp.org>
CC: classpath@gnu.org
Subject: Re: Wonka
References: <20011025084801.A21952@klomp.org> <3BD7BF9A.3CBBC4E9@acunia.com> <20011027224710.A26801@klomp.org>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard wrote:

> Hi,
>
> On Thu, Oct 25, 2001 at 09:30:34AM +0200, Chris Gray wrote:
> >
> > I would also invite people to look at ACUNIA's Wonka VM,
> > <URL:http://wonka.acunia.com>.  Like IBM, we already developed our
> > own class libraries, but a Classpath/Wonka hybrid would appear to be
> > feasible from both a legal and a technical point of view.  Of more
> > practical value (probably) would be a gcj-Wonka synthesis.
>
> Nice. From you package report page I get the impression that you have
> a lot of classes for 1.1 and 1.2. But it does not seem as complete as
> Classpath at the moment. But you have a working AWT implementation!
> And some javax.crypto stuff. Cool.
>

Up to now the development was mainly driven by our own internal
projects, so our coverage is a bit skewed: we're probably ahead of
Classpath in some areas, but I'd agree that we're behind in others
(notable text handling).

>
> Was there a reason you could not use (or cooperate with) GNU Classpath
> for your VM?
>

When we began the project it was not known under what license the code
would one day be released, so we were wary of basing a large part of it
on a project which had its own licensing agenda, so to speak.  Although
using the Classpath stuff as-is didn't seem to present particular problems,
we weren't too sure about e.g. tweaking classes to take advantage of our
native code.  The decision to go open-source was only taken a few months
ago, and it could have gone the other way (i.e. making Wonka a closed-
source product).

>
> Two legal nitpicks. (Please ignore them if you don't like legal issues.)
> Your license is not GPL compatible (which may or may not be a problem).
> The GPL incompatability come from clause 4:
> 4. Products derived from this software may not be called "Wonka",
>    nor may "Wonka" appear in their name, without prior written
>    permission of ACUNIA NV.

I'm not in a position to ignore legal issues, even if I don't like them. :-0
The GNU licensing page mentions that the Apache licence is not GPL-
compatible, but annoyingly omits to say which clause is the offending
one.  We actually took one clause out, but it seems we picked the wrong
one from the GPL-compatibilty point of view. :(

>
> This is the same as can be found in the Apache license which is also not
> GPL compatible since it is an added restriction (although a minor one)
> according to GPL clause 6. But since you actaully have a trademark on Wonka
> this clause is not necessary. You could replace it with a fact (outside the
> actual license text) such as "The following are trademarks of Acunia, XXX,
> YYY, ZZZ. This license does not grand you any permission to use any
> trademarks of Acunia when distributing derived works." That way it would
> be GPL compatible (since then your claim is based on trademark law and not
> on copyright laa)w. You might want to take a look at how AbiSource did a
> similar thing with their trademark on AbiWord.
>

The clause is there to protect our `brand': if we can achieve the same result
without breaking GPL compatibility then this would be very nice.  I'll take it
up
with the powers that be.


>
> Your download does contain a COPYING file with the same clauses as menioned
> on <http://wonka.acunia.com/licence.html>. But all source files and
> documentation comes with a header that does not sound very open/free.
> It would be nice to make it a bit more clear that the license agreement
> mentioned in that text is acutally the top level COPYING file. Note for
> example that all Classpath files explicitly state the license conditions.
> This makes it much more clear what your rights are if you somehow only
> received part of the code archive.
>

Again, I'll pass this comment on.  The header is standard company-wide
boilerplate, with one clause already removed because it didn't apply to
open-source code: I'll try to negotiate some more customisation.

>
> Groetjes,
>

:)

Beste wensen,

Chris Gray
VM Architect, ACUNIA




From tromey@creche.cygnus.com Mon Oct 29 13:02:04 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15yGjb-0005pk-00; Mon, 29 Oct 2001 13:02:03 -0500
Received: from creche.cygnus.com (ta0192.peakpeak.com [204.144.244.192] (may be forged))
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id LAA23904;
	Mon, 29 Oct 2001 11:01:54 -0700
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id LAA15272;
	Mon, 29 Oct 2001 11:21:35 -0700
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: GCJ jni.h fixes
References: <m3k7xh6p3v.fsf@lyta.haphazard.org> <87pu799buv.fsf@creche.redhat.com> <m37ktg7ezj.fsf@lyta.haphazard.org>
X-Zippy:  I just remembered something about a TOAD!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 29 Oct 2001 11:21:34 -0700
In-Reply-To: Brian Jones's message of "27 Oct 2001 21:03:28 -0400"
Message-ID: <87d736cno1.fsf@creche.redhat.com>
Lines: 13
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

>> We comply with the JNI 1.2 spec:
>> http://java.sun.com/products/jdk/1.2/docs/guide/jni/spec/functions.doc.html#17265

Brian> I'm looking in "The Java Native Interface" ISBN 0-201-32577-2,
Brian> in which I have found one error with regards to function index
Brian> so it is possible these are errors as well, maybe not.

When was this published?
I'm wondering if the online docs are newer.

Tom


From tromey@creche.cygnus.com Mon Oct 29 13:10:26 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15yGrh-0006Cm-00; Mon, 29 Oct 2001 13:10:25 -0500
Received: from creche.cygnus.com (ta0192.peakpeak.com [204.144.244.192] (may be forged))
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id LAA25628;
	Mon, 29 Oct 2001 11:10:23 -0700
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id LAA15317;
	Mon, 29 Oct 2001 11:30:04 -0700
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: jni.h VM dependencies
References: <m3ofmt6p6o.fsf@lyta.haphazard.org>
X-Zippy:  YOW!!  What should the entire human race DO??  Consume a fifth
 of CHIVAS REGAL, ski NUDE down MT. EVEREST, and have a wild
 SEX WEEKEND!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 29 Oct 2001 11:30:04 -0700
In-Reply-To: Brian Jones's message of "27 Oct 2001 12:08:31 -0400"
Message-ID: <87adyacn9v.fsf@creche.redhat.com>
Lines: 20
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> I was wondering if any of you have great ideas for how to
Brian> handle these opaque structures defined in a typical jni.h.
Brian> typedef struct _jfieldID *jfieldID;
Brian> typedef struct _jmethodID *jmethodID;

Just define them as `void *'.
Or define them the way you have them now.

Brian> I am currently thinking something like this in jni.h
Brian> #include "jni_vm.h"
Brian> typedef JFIELD_ID jfieldID;
Brian> typedef JMETHOD_ID jmethodID;
Brian> [ ... ]

How will `JVM' get defined?
I think you can't require user code to set it.

Tom


From mark@klomp.org Mon Oct 29 16:48:17 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yKGX-0008GW-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 16:48:17 -0500
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15yKGU-0007rc-00; Mon, 29 Oct 2001 22:48:14 +0100
Date: Mon, 29 Oct 2001 22:48:14 +0100
From: Mark Wielaard <mark@klomp.org>
To: classpath@gnu.org
Subject: java.awt status LGPL -> GPL?
Message-ID: <20011029224814.A30080@klomp.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

I was going through our documentation pages in doc/www.gnu.org and found
the following in status.wml:

    The AWT classes we have are licensed under the LGPL although RMS
    has demanded this be changed to the GPL.  At the same time the GCJ
    team has written their own AWT that is available under a license
    more like the rest of GNU Classpath and this may be available in
    GCC 3.0.  For now these AWT issues remain unresolved and GNU
    Classpath does not have a working AWT.

I was not aware that the license of the awt classes should have been
changed from LGPL to GPL. I might have missed something but I cannot
remember that was ever an issue (during the time we discussed switching
from LGPL to GPL+exception and merging with libgcj).

What is the current status of this?

Note that GCC 3.0.x ships with a partial AWT implementation under
GPL+exception. And the just announced Wonka class libraries seem to come
with a complete AWT under a BSD+trademark style license.

Cheers,

Mark

P.S. I note that we now have another CVS tree on savannah for the web
documentation. Which tree should I use when updating documentation?

-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Mon Oct 29 17:16:07 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yKhT-0001am-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 17:16:07 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Mon, 29 Oct 2001 17:16:05 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9TMFMM19735;
	Mon, 29 Oct 2001 17:15:22 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: tromey@redhat.com
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: GCJ jni.h fixes
References: <m3k7xh6p3v.fsf@lyta.haphazard.org> <87pu799buv.fsf@creche.redhat.com> <m37ktg7ezj.fsf@lyta.haphazard.org> <87d736cno1.fsf@creche.redhat.com>
From: Brian Jones <cbj@gnu.org>
Date: 29 Oct 2001 17:15:22 -0500
In-Reply-To: Tom Tromey's message of "29 Oct 2001 11:21:34 -0700"
Message-ID: <m3itcyqeit.fsf@lyta.haphazard.org>
Lines: 19
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey <tromey@redhat.com> writes:

> >>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:
> 
> >> We comply with the JNI 1.2 spec:
> >> http://java.sun.com/products/jdk/1.2/docs/guide/jni/spec/functions.doc.html#17265
> 
> Brian> I'm looking in "The Java Native Interface" ISBN 0-201-32577-2,
> Brian> in which I have found one error with regards to function index
> Brian> so it is possible these are errors as well, maybe not.
> 
> When was this published?
> I'm wondering if the online docs are newer.

(c) 1999, covers Java 2 (1.2).

Brian
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Mon Oct 29 17:16:42 2001
Received: from fe8.southeast.rr.com ([24.93.67.55] helo=mail8.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yKi2-0001cG-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 17:16:42 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by mail8.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Mon, 29 Oct 2001 17:16:41 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9TMFw419739;
	Mon, 29 Oct 2001 17:15:58 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: tromey@redhat.com
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: jni.h VM dependencies
References: <m3ofmt6p6o.fsf@lyta.haphazard.org> <87adyacn9v.fsf@creche.redhat.com>
From: Brian Jones <cbj@gnu.org>
Date: 29 Oct 2001 17:15:58 -0500
In-Reply-To: Tom Tromey's message of "29 Oct 2001 11:30:04 -0700"
Message-ID: <m3elnmqeht.fsf@lyta.haphazard.org>
Lines: 16
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Tom Tromey <tromey@redhat.com> writes:

> >>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:
> 
> Brian> I was wondering if any of you have great ideas for how to
> Brian> handle these opaque structures defined in a typical jni.h.
> Brian> typedef struct _jfieldID *jfieldID;
> Brian> typedef struct _jmethodID *jmethodID;
> 
> Just define them as `void *'.
> Or define them the way you have them now.

Okay.

-- 
Brian Jones <cbj@gnu.org>


From rao@gnu.org Mon Oct 29 18:51:46 2001
Received: from kosh.nivin.org ([64.81.66.61] helo=localhost.localdomain)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yMC2-00066l-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 18:51:46 -0500
Received: (from rao@localhost)
	by localhost.localdomain (8.11.2/8.11.2) id f9TNpiI10804;
	Mon, 29 Oct 2001 15:51:44 -0800
X-Authentication-Warning: localhost.localdomain: rao set sender to rao@gnu.org using -f
To: classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org>
Mail-Copies-To: never
From: Paul Fisher <rao@gnu.org>
Date: Mon, 29 Oct 2001 15:51:44 -0800
In-Reply-To: <20011029224814.A30080@klomp.org> (Mark Wielaard's message of
 "Mon, 29 Oct 2001 22:48:14 +0100")
Message-ID: <m38zdugg33.fsf@zathras.nivin.org>
Lines: 19
User-Agent: Gnus/5.090004 (Oort Gnus v0.04) Emacs/21.1 (i686-pc-linux-gnu)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> I was not aware that the license of the awt classes should have been
> changed from LGPL to GPL.

The AWT classes are GPL'd.  New licensing headers need to be checked
into CVS.  If someone could go ahead and do this, that would be a
help.  I haven't had time to do so.

> Note that GCC 3.0.x ships with a partial AWT implementation under
> GPL+exception.

The AWT files that are part of the GCC release are GPL'd.  The
licensing information at the top of such files is incorrect.

> And the just announced Wonka class libraries seem to come
> with a complete AWT under a BSD+trademark style license.

I'll have to look at this.


From mark@klomp.org Mon Oct 29 19:41:33 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yMyD-0000HA-00; Mon, 29 Oct 2001 19:41:33 -0500
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15yMyA-00086t-00; Tue, 30 Oct 2001 01:41:30 +0100
Date: Tue, 30 Oct 2001 01:41:30 +0100
From: Mark Wielaard <mark@klomp.org>
To: Paul Fisher <rao@gnu.org>
Cc: classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
Message-ID: <20011030014130.A30957@klomp.org>
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <m38zdugg33.fsf@zathras.nivin.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Mon, Oct 29, 2001 at 03:51:44PM -0800, Paul Fisher wrote:
> Mark Wielaard <mark@klomp.org> writes:
> 
> > I was not aware that the license of the awt classes should have been
> > changed from LGPL to GPL.
> 
> The AWT classes are GPL'd.  New licensing headers need to be checked
> into CVS.  If someone could go ahead and do this, that would be a
> help.  I haven't had time to do so.

I am a bit annoyed by this. Why was this never clearly announced!

I must not be the only one that missed the fact that we suddenly
changed from LGPL to GPL. I actually searched through the archives
to make sure I didn't miss any discussion about this and I couldn't
find anything about it!

> > Note that GCC 3.0.x ships with a partial AWT implementation under
> > GPL+exception.
> 
> The AWT files that are part of the GCC release are GPL'd.  The
> licensing information at the top of such files is incorrect.
Are you sure the libgcj people know about this?!?!

I really like working on Classpath and libgcj. But the AWT situation is
not funny anymore. I was hoping for months that there would be some
clearification or consensus on this issue. I would have loves to work on
finishing the AWT but I have not looked at the code since I wanted to
wait till we had decided to go forward with the LGPLed Classpath or
the GPL+exception libgcj implementation. If anybody had clued me in that
both versions were actually under the same license I could have started
merging a long long time ago! (Although working on a GPLed implementation
now that we have a AWT implementation under a Apache style license
available, which Acunia seems to want to actually make GPL compatible,
does not make much sense to me anymore.)

Could you please explain how AWT is defined? I have done a little work
on java.awt.datatransfer, which has not much to do with the rest of AWT
except for its package name, would that now also fall under GPL? Does it
include packages not in the Kaffe implementation like java.awt.geom,
java.awt.color or java.awt.print? What about packages that depend on
AWT classes like java.applet or java.beans?

Thanks,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From rao@gnu.org Mon Oct 29 20:12:30 2001
Received: from kosh.nivin.org ([64.81.66.61] helo=localhost.localdomain)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yNSA-0001aR-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 20:12:30 -0500
Received: (from rao@localhost)
	by localhost.localdomain (8.11.2/8.11.2) id f9U1CSt11003;
	Mon, 29 Oct 2001 17:12:28 -0800
X-Authentication-Warning: localhost.localdomain: rao set sender to rao@gnu.org using -f
To: classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org>
	<m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org>
From: Paul Fisher <rao@gnu.org>
Date: Mon, 29 Oct 2001 17:12:28 -0800
In-Reply-To: <20011030014130.A30957@klomp.org> (Mark Wielaard's message of
 "Tue, 30 Oct 2001 01:41:30 +0100")
Message-ID: <m34roigccj.fsf@zathras.nivin.org>
Lines: 19
User-Agent: Gnus/5.090004 (Oort Gnus v0.04) Emacs/21.1 (i686-pc-linux-gnu)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> I am a bit annoyed by this. Why was this never clearly announced!

Things have been murky for some time on this issue.  Only recently did
we get clarification on the status of the AWT.

> Are you sure the libgcj people know about this?!?!

Yes.  They are aware.

> If anybody had clued me in that both versions were actually under
> the same license I could have started merging a long long time ago!

We didn't know a long time ago.  This is recent news.  We'd very much
like to merge the AWTs.

Sorry for not going into great detail, but I have a plane to catch
(literally).  I'll comment more when I have some free time.


From mark@klomp.org Mon Oct 29 20:53:17 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yO5c-0003Vx-00; Mon, 29 Oct 2001 20:53:16 -0500
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15yO5a-0008Dq-00; Tue, 30 Oct 2001 02:53:14 +0100
Date: Tue, 30 Oct 2001 02:53:13 +0100
From: Mark Wielaard <mark@klomp.org>
To: Paul Fisher <rao@gnu.org>
Cc: classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
Message-ID: <20011030025313.A31569@klomp.org>
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org> <m34roigccj.fsf@zathras.nivin.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <m34roigccj.fsf@zathras.nivin.org>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Dear Paul,

On Mon, Oct 29, 2001 at 05:12:28PM -0800, Paul Fisher wrote:
> 
> Sorry for not going into great detail, but I have a plane to catch
> (literally).  I'll comment more when I have some free time.

Please do when you have time.

I don't have a very good feeling about this and I hope you can take that
bad feeling away. I don't mind if this is actually a good decission for
the GNU project as a whole, but in that case I really have to see more
explanation of why this is a good move. I am very pessimistic about targeting
our GPLed library against a free beer library from Sun. I don't think this
will create more Free Software or cooperation between different free software
projects at the moment. There are a lot of Free JVMs that are currently
using or are trying to use a version of GNU Classpath as their standard
library implementation. I hope that switching to GPL for an (essential)
part of the library does not effect that.

Thanks,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From cbj@gnu.org Mon Oct 29 23:35:16 2001
Received: from fe5.southeast.rr.com ([24.93.67.52] helo=mail5.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yQcO-0000nB-00
	for <classpath@gnu.org>; Mon, 29 Oct 2001 23:35:16 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by mail5.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Mon, 29 Oct 2001 23:35:14 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9U4YS020222;
	Mon, 29 Oct 2001 23:34:28 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: Mark Wielaard <mark@klomp.org>
Cc: classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 29 Oct 2001 23:34:28 -0500
In-Reply-To: Mark Wielaard's message of "Mon, 29 Oct 2001 22:48:14 +0100"
Message-ID: <m3668xrbjf.fsf@lyta.haphazard.org>
Lines: 15
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> P.S. I note that we now have another CVS tree on savannah for the web
> documentation. Which tree should I use when updating documentation?

I updated the website to use WML instead of MetaHTML earlier in
October.  This is an easy set of tools to get today and they work well
for statically generating our pages in a manner that is similar to
before.  Please make appropriate changes to doc/www.gnu.org/* and run
'make' from that directory.  There is 'make publish' in there to make
it easier to update the separate CVS tree.

Brian
-- 
Brian Jones <cbj@gnu.org>


From cbj@gnu.org Tue Oct 30 00:18:01 2001
Received: from fe6.southeast.rr.com ([24.93.67.53] helo=Mail6.nc.rr.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yRHl-0002Fc-00
	for <classpath@gnu.org>; Tue, 30 Oct 2001 00:18:01 -0500
Received: from lyta.haphazard.org ([66.26.244.135]) by Mail6.nc.rr.com  with Microsoft SMTPSVC(5.5.1877.687.68);
	 Tue, 30 Oct 2001 00:18:00 -0500
Received: (from cbj@localhost)
	by lyta.haphazard.org (8.11.6/8.11.0) id f9U5HD720279;
	Tue, 30 Oct 2001 00:17:13 -0500
X-Authentication-Warning: lyta.haphazard.org: cbj set sender to cbj@gnu.org using -f
To: GNU Classpath <classpath@gnu.org>
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org> <m34roigccj.fsf@zathras.nivin.org> <20011030025313.A31569@klomp.org>
From: Brian Jones <cbj@gnu.org>
Date: 30 Oct 2001 00:17:13 -0500
In-Reply-To: Mark Wielaard's message of "Tue, 30 Oct 2001 02:53:13 +0100"
Message-ID: <m31yjlr9k6.fsf@lyta.haphazard.org>
Lines: 48
User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Mark Wielaard <mark@klomp.org> writes:

> Dear Paul,
> 
> On Mon, Oct 29, 2001 at 05:12:28PM -0800, Paul Fisher wrote:
> > 
> > Sorry for not going into great detail, but I have a plane to catch
> > (literally).  I'll comment more when I have some free time.
> 
> Please do when you have time.
> 
> I don't have a very good feeling about this and I hope you can take that
> bad feeling away. I don't mind if this is actually a good decission for
> the GNU project as a whole, but in that case I really have to see more
> explanation of why this is a good move. 

The decision was made by Richard Stallman.  Previous to that I tried
to press for changing the license to be more like the rest of GNU
Classpath.  This is what I've wanted and by popular vote is also what
many of the contributors wanted, but that view did not prevail.
Apparently things have settled a bit since I made that note concerning
the AWT and we can now update the web pages again.

Legally speaking, the number of contributors to our AWT in Classpath
is still quite low.  Aaron did most of the work with myself and
another contributing bits and pieces (ignoring the peers).  We have
the option of licensing and using our code in any way we'd like as
long as we supply the FSF with 30 days notice in writing (This is my
reading of the contributor agreement).  If someone would like to use
my code under the GPL + exception, then I am more than willing to
supply the FSF with a writing to that affect which should make it legal.

> I am very pessimistic about targeting our GPLed library against a
> free beer library from Sun. I don't think this will create more Free
> Software or cooperation between different free software projects at
> the moment. 

The GPL'd library will still support research interests but it will
probably not be of interest to certain types of users.  For one, can
you execute a proprietary Java GUI application using a GPL'd AWT?
Two, I would like to know how this plays with Red Hat's embedded
customers of GCJ for instance?  Is it the same there, that customers
buy a more appropriate license based on something similar to the
individual contributor agreement with the FSF?

Brian
-- 
Brian Jones <cbj@gnu.org>


From per@bothner.com Tue Oct 30 03:57:10 2001
Received: from adsl-216-102-199-253.dsl.snfc21.pacbell.net ([216.102.199.253] helo=bothner.com)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yUhp-0001wg-00; Tue, 30 Oct 2001 03:57:09 -0500
Received: from bothner.com (eureka.bothner.com [192.168.1.9])
	by bothner.com (8.11.0/8.11.0) with ESMTP id f9U8vDH10621;
	Tue, 30 Oct 2001 00:57:13 -0800
Message-ID: <3BDE6B68.7060903@bothner.com>
Date: Tue, 30 Oct 2001 00:57:12 -0800
From: Per Bothner <per@bothner.com>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.5) Gecko/20011012
X-Accept-Language: en-us
MIME-Version: 1.0
To: Brian Jones <cbj@gnu.org>
CC: GNU Classpath <classpath@gnu.org>
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org> <m34roigccj.fsf@zathras.nivin.org> <20011030025313.A31569@klomp.org> <m31yjlr9k6.fsf@lyta.haphazard.org>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I would not be too hasty about changing any files quite yet.
I (in my capacity as a member of the Gcc steering comittee)
am pushing for a better resolution of this issue.
    --Per




From tromey@creche.cygnus.com Tue Oct 30 10:13:30 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15yaa2-0005zJ-00
	for <classpath@gnu.org>; Tue, 30 Oct 2001 10:13:30 -0500
Received: from creche.cygnus.com (ta0198.peakpeak.com [204.144.244.198])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id IAA21332;
	Tue, 30 Oct 2001 08:13:24 -0700
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id IAA13315;
	Tue, 30 Oct 2001 08:33:20 -0700
To: classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org>
X-Zippy:  BARRY..  That was the most HEART-WARMING rendition of
 ``I DID IT MY WAY'' I've ever heard!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 30 Oct 2001 08:33:20 -0700
In-Reply-To: Paul Fisher's message of "Mon, 29 Oct 2001 15:51:44 -0800"
Message-ID: <87ofmp5ein.fsf@creche.redhat.com>
Lines: 16
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Paul" == Paul Fisher <rao@gnu.org> writes:

Mark> Note that GCC 3.0.x ships with a partial AWT implementation
Mark> under GPL+exception.

Paul> The AWT files that are part of the GCC release are GPL'd.  The
Paul> licensing information at the top of such files is incorrect.

This is RMS' claim, but that doesn't necessarily make it true.  It has
a sort of revisionist quality to it -- namely, nobody who worked on
the libgcj AWT implementation was aware of this at the time.  So even
if his position is legally correct, morally it can't be.  IMHO of
course.  Certainly if I had known he was planning to GPL this code, I
would never have assigned it to the FSF.

Tom


From tromey@creche.cygnus.com Tue Oct 30 10:16:20 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15yacm-0006EU-00; Tue, 30 Oct 2001 10:16:20 -0500
Received: from creche.cygnus.com (ta0198.peakpeak.com [204.144.244.198])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id IAA21790;
	Tue, 30 Oct 2001 08:16:17 -0700
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id IAA13331;
	Tue, 30 Oct 2001 08:36:13 -0700
To: Mark Wielaard <mark@klomp.org>
Cc: Paul Fisher <rao@gnu.org>, classpath@gnu.org
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org>
X-Zippy:  I'm also pre-POURED pre-MEDITATED and pre-RAPHAELITE!!
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 30 Oct 2001 08:36:13 -0700
In-Reply-To: Mark Wielaard's message of "Tue, 30 Oct 2001 01:41:30 +0100"
Message-ID: <87lmht5edu.fsf@creche.redhat.com>
Lines: 26
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:

Paul> The AWT files that are part of the GCC release are GPL'd.  The
Paul> licensing information at the top of such files is incorrect.

Mark> Are you sure the libgcj people know about this?!?!

This is a recent development.

Mark> If anybody had clued me in that both versions were actually
Mark> under the same license I could have started merging a long long
Mark> time ago! (Although working on a GPLed implementation now that
Mark> we have a AWT implementation under a Apache style license
Mark> available, which Acunia seems to want to actually make GPL
Mark> compatible, does not make much sense to me anymore.)

I think merging is still premature, since I dispute RMS'
interpretation of the libgcj AWT license.  I also dispute the
contention that a GPL AWT is usable by random code.  I don't see how
that can be.

Mark> Could you please explain how AWT is defined?

I've always assumed it means `java.awt.*', including sub-packages.

Tom


From tromey@creche.cygnus.com Tue Oct 30 13:52:48 2001
Received: from [207.174.178.17] (helo=gash2.peakpeak.com)
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15ye0G-0000wm-00; Tue, 30 Oct 2001 13:52:48 -0500
Received: from creche.cygnus.com (ta0198.peakpeak.com [204.144.244.198])
	by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id LAA00739;
	Tue, 30 Oct 2001 11:52:40 -0700
Received: (from tromey@localhost)
	by creche.cygnus.com (8.9.3/8.9.3) id MAA14561;
	Tue, 30 Oct 2001 12:12:39 -0700
To: Brian Jones <cbj@gnu.org>
Cc: GNU Classpath <classpath@gnu.org>
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org> <m34roigccj.fsf@zathras.nivin.org> <20011030025313.A31569@klomp.org> <m31yjlr9k6.fsf@lyta.haphazard.org>
X-Zippy:  --``I love KATRINKA because she drives a PONTIAC.  We're going away now.
 I fed the cat. - Zippy''
X-Attribution:  Tom
Reply-To: tromey@redhat.com
From: Tom Tromey <tromey@redhat.com>
Date: 30 Oct 2001 12:12:39 -0700
In-Reply-To: Brian Jones's message of "30 Oct 2001 00:17:13 -0500"
Message-ID: <878zdtt00o.fsf@creche.redhat.com>
Lines: 26
X-Mailer: Gnus v5.7/Emacs 20.5
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

>>>>> "Brian" == Brian Jones <cbj@gnu.org> writes:

Brian> For one, can you execute a proprietary Java GUI application
Brian> using a GPL'd AWT?

Richard and Tim Wilkinson say yes, but I don't understand the
reasoning.  If the answer is yes, then how is the GPL different from
the LGPL here?

Brian> Two, I would like to know how this plays with Red Hat's
Brian> embedded customers of GCJ for instance?  Is it the same there,
Brian> that customers buy a more appropriate license based on
Brian> something similar to the individual contributor agreement with
Brian> the FSF?

The AWT in libgcj is assigned to the FSF.  Red Hat wasn't really
involved in writing it.  The work I did on it was done on my own time
(but using Red Hat equipment -- so it is a bit muddy).

If Richard does intend to license the libgcj AWT under the GPL, then I
guess Red Hat will have to write a new AWT (or work on the Acunia one)
if and when it ever needs one for a customer.  Certainly the LGPL
would be unsuitable for Red Hat's embedded customers, and the GPL
would be unsuitable (imo) for anybody.

Tom


From chris.gray@acunia.com Wed Oct 31 07:16:05 2001
Received: from nat.be.acunia.com ([194.7.211.211] helo=pigeon.smove.be)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yuHr-0002hV-00; Wed, 31 Oct 2001 07:16:04 -0500
Received: from localhost.localdomain (IDENT:nobody@loki.leuven.intern.acunia.com [192.168.8.254])
	by pigeon.smove.be (8.11.6/8.11.6) with ESMTP id f9VCFg420743;
	Wed, 31 Oct 2001 13:15:42 +0100
Received: from acunia.com (IDENT:gray@sophocles [127.0.0.1])
	by localhost.localdomain (8.9.3/8.9.3) with ESMTP id NAA21486;
	Wed, 31 Oct 2001 13:19:09 +0100
Message-ID: <3BDFEC3C.90C6AD44@acunia.com>
Date: Wed, 31 Oct 2001 13:19:08 +0100
From: Chris Gray <chris.gray@acunia.com>
Organization: ACUNIA NV
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Mark Wielaard <mark@klomp.org>
CC: Paul Fisher <rao@gnu.org>, classpath@gnu.org, chris@kiffer.eunet.be
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

> (Although working on a GPLed implementation
> now that we have a AWT implementation under a Apache style license
> available, which Acunia seems to want to actually make GPL compatible,
> does not make much sense to me anymore.)

Let me modulate this a bit (particularly as I won't be around much in the
next
four days, and I'm nervous about where this discussion may go without any
real-world input).

Firstly, although the blurb says "Full 1.1.8 AWT", this is also release
0.7.2,
not 1.0.  So whilst all of 1.1.8 is all on our roadmap somewhere in the near

future, we're not there yet.  See
<URL:http://wonka.acunia.com/package-report/awt-package-report.html>
for the state of the art here and now.

Secondly, although I responded positively to Mark's suggestions regarding
GPL-compatibility, it is still possible that our management would decide
that
protecting our trademark is more important, and that it is essential to have
the
trademark clause -in- the licence, not next to it.  IANAL, TINLA, etc..

Of course we would love to have Mark working on our AWT 8->, the more
so as the main thing standing between us and the advertised full 1.1.8 is
the number of hours in the day.  And I'm sure that our Rudolph is already
much nearer to that target than any other code base you could use.  Just
don't everybody run away with the idea that the problem is already solved.

Best wishes

Chris Gray
VM Architect, ACUNIA




From chris.gray@acunia.com Wed Oct 31 11:39:14 2001
Received: from nat.be.acunia.com ([194.7.211.211] helo=pigeon.smove.be)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15yyOX-0005y0-00; Wed, 31 Oct 2001 11:39:14 -0500
Received: from localhost.localdomain (IDENT:nobody@loki.leuven.intern.acunia.com [192.168.8.254])
	by pigeon.smove.be (8.11.6/8.11.6) with ESMTP id f9VGcs426779;
	Wed, 31 Oct 2001 17:38:54 +0100
Received: from acunia.com (IDENT:gray@sophocles [127.0.0.1])
	by localhost.localdomain (8.9.3/8.9.3) with ESMTP id RAA22037;
	Wed, 31 Oct 2001 17:42:21 +0100
Message-ID: <3BE029ED.6D9024D8@acunia.com>
Date: Wed, 31 Oct 2001 17:42:21 +0100
From: Chris Gray <chris.gray@acunia.com>
Organization: ACUNIA NV
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Mark Wielaard <mark@klomp.org>, Paul Fisher <rao@gnu.org>,
   classpath@gnu.org, chris@kiffer.eunet.be
Subject: Re: java.awt status LGPL -> GPL?
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org> <3BDFEC3C.90C6AD44@acunia.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

(Following-up to my own post is a sign of what?)

Chris Gray wrote:

> Secondly, although I responded positively to Mark's suggestions regarding
> GPL-compatibility, it is still possible that our management would decide
> that
> protecting our trademark is more important, and that it is essential to have
> the
> trademark clause -in- the licence, not next to it.  IANAL, TINLA, etc..

O.K., I've hoofed it from desk to desk and Clause 4 is now gone.  The only
diff between our licence and revised BSD now is a request (-not- a requirement)

to submit any improvements or extensions and to grant ACUNIA NV the rights to
redistribute these changes.  So I guess we're now GPL compatible.

Regards

Chris Gray
VM Architect, ACUNIA




From mark@klomp.org Wed Oct 31 18:45:35 2001
Received: from node1c5f.a2000.nl ([62.108.28.95] helo=nescio.wildebeest.org)
	by fencepost.gnu.org with esmtp (Exim 3.22 #1 (Debian))
	id 15z539-0001nX-00; Wed, 31 Oct 2001 18:45:35 -0500
Received: from mark by nescio.wildebeest.org with local (Exim 3.32 #1 (Debian))
	id 15z52m-0002jt-00; Thu, 01 Nov 2001 00:45:12 +0100
Date: Thu, 1 Nov 2001 00:45:12 +0100
From: Mark Wielaard <mark@klomp.org>
To: Chris Gray <chris.gray@acunia.com>
Cc: Paul Fisher <rao@gnu.org>, classpath@gnu.org, chris@kiffer.eunet.be
Subject: Re: java.awt status LGPL -> GPL?
Message-ID: <20011101004512.A10456@klomp.org>
References: <20011029224814.A30080@klomp.org> <m38zdugg33.fsf@zathras.nivin.org> <20011030014130.A30957@klomp.org> <3BDFEC3C.90C6AD44@acunia.com> <3BE029ED.6D9024D8@acunia.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <3BE029ED.6D9024D8@acunia.com>
User-Agent: Mutt/1.3.23i
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

Hi,

On Wed, Oct 31, 2001 at 05:42:21PM +0100, Chris Gray wrote:
> (Following-up to my own post is a sign of what?)
Greatness :)

My reponse to your last email would have been: "Thank you very much about
being so clear and honest with us about what is going on. Don't worry that
I will actually start hacking on your code base until it is GPL compatible
since I don't really like hacking on code that cannot be used in GPLed
programs". But now that excuse is gone!

It would be great if our management (read RMS) acted so fast :)
I will still wait a bit on the outcome of the negotiation that Per Bothner
is doing. But you can be sure that I will seriously look into Rudolph now.

Thanks,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore


From eland_at_jleuner@yahoo.com Tue Oct 30 14:27:57 2001
Received: from web21002.mail.yahoo.com ([216.136.227.56])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15yeYH-0002UN-00
	for <classpath@gnu.org>; Tue, 30 Oct 2001 14:27:57 -0500
Message-ID: <20011030192756.77464.qmail@web21002.mail.yahoo.com>
Received: from [57.66.12.99] by web21002.mail.yahoo.com via HTTP; Tue, 30 Oct 2001 11:27:56 PST
Date: Tue, 30 Oct 2001 11:27:56 -0800 (PST)
From: John Leuner <eland_at_jleuner@yahoo.com>
Subject: Potential bug in Hashtable
To: classpath@gnu.org
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

I ran the latest classpath code against mauve, a
hashtable test which checks the size() after a clone()
failed:

I think copy.size should be set to 0 in the clone
method (before the putAll)?

With my JVM it prints out 28, the IBM JDK prints 14 as
expected by the mauve test. 


 public synchronized Object clone()
  {
    Hashtable copy = null;
    try
      {
        copy = (Hashtable) super.clone();
      }
    catch (CloneNotSupportedException x)
      {
        // This is impossible.
      }
    copy.buckets = new HashEntry[buckets.length];
    copy.putAll(this);
    // Clear the caches.
    copy.keys = null;
    copy.values = null;
    copy.entries = null;
    return copy;
  }



  harness.checkPoint("Checking clone()");
      Hashtable copy = (Hashtable) hash.clone();
      hash.clear();
      harness.check (hash.toString(), "{}");
      harness.check (hash.size(), 0);
      // harness.check (copy.toString());
      System.out.println("copy.size() is " +
copy.size());
      harness.check (copy.size(), 14);


John Leuner

__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com


From eland_at_jleuner@yahoo.com Tue Oct 30 14:34:11 2001
Received: from web21001.mail.yahoo.com ([216.136.227.55])
	by fencepost.gnu.org with smtp (Exim 3.22 #1 (Debian))
	id 15yeeJ-0002kF-00
	for <classpath@gnu.org>; Tue, 30 Oct 2001 14:34:11 -0500
Message-ID: <20011030193409.75703.qmail@web21001.mail.yahoo.com>
Received: from [57.66.12.99] by web21001.mail.yahoo.com via HTTP; Tue, 30 Oct 2001 11:34:09 PST
Date: Tue, 30 Oct 2001 11:34:09 -0800 (PST)
From: John Leuner <eland_at_jleuner@yahoo.com>
Subject: Re: classpath ./ChangeLog lib/standard.omit
To: classpath@gnu.org
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: classpath-admin@gnu.org
Errors-To: classpath-admin@gnu.org
X-BeenThere: classpath@gnu.org
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:classpath-request@gnu.org?subject=help>
List-Post: <mailto:classpath@gnu.org>
List-Subscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=subscribe>
List-Id: <classpath.gnu.org>
List-Unsubscribe: <http://mail.gnu.org/mailman/listinfo/classpath>,
	<mailto:classpath-request@gnu.org?subject=unsubscribe>
List-Archive: <http://mail.gnu.org/pipermail/classpath/>

On Sun, Oct 28, 2001 at 01:30:37PM +1300, Bryce
McKinlay wrote:   
> AWTEvent depends on Event.java, which for some
reason doesn't exist in 
> Classpath.

Not according to the files I have. AWTEvent depends on
java.util.EventObject.

I removed the AWT 1.0 stuff to get the AWT compiling
(which was far more important than the AWT 1.0
support). I have subsequently been able to test it
(with the GTK peers), unfortunately there are problems
with the peer handling. It shouldn't take *too* much
effort to fix.

John Leuner

__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com


