Mozilla/5.0 (X11; Linux x86_64; rv:10.0.3) Gecko/20120329 Icedove/10.0.3
On 03.05.2012 20:26, Bruno Haible wrote:
> Vladimir wrote:
>>> # define W(flag, data) ((flag) ? SWAP (data) : (data))
>> That exactly proves my point: you get the penalty from the check
>> independently of whether you needed to swap in the first place. This may
>> make even no-swap version slower than always-swap.
> The difference is not measurable. It's a waste of time to discuss
> performance improvements or deteriorations of less than 0.1%.
That was exactly my point. Performance argument in this discussion is
mostly imaginary. The benefits of easier code and binary files
unification are real. If it was up to me I'd just make msgfmt output in
some fixed endianness, but making a fixed endianness (rather than host
one) a default would already be good.
>> Was the answer out of the list intentional?
> You sent me a private email; I replied in private.
Bad button, sorry.
> Bruno
>
>
--
Regards
Vladimir 'φ-coder/phcoder' Serbinenko
signature.asc Description: OpenPGP digital signature