Gnus/5.110006 (No Gnus v0.6) Emacs/21.4 (gnu/linux)
Eric Blake <address@hidden> writes:
> However, since coreutils accepts hh as an extension, should it issue a
> warning if the string being printed does not fit in an unsigned char? Or
> should it silently wrap around, and print 0 instead of 256? Or are we
> okay with keeping status quo that the size specifiers must be valid
> according to printf(3) rules, but are otherwise ignored?
ksh and bash ignore size specifiers, so there's good precedent for the
current behavior. I'd leave it alone.