This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Convert libiberty to use ISO C prototype style 4/n
>>>>> "Marcin" == Marcin Dalecki <martin@dalecki.de> writes:
Marcin> zlib development is in limbo and has goals which are impeding
Marcin> what should be desirable here. In particular the whole
Marcin> platform compatibility shims in zlib are definitely not what
Marcin> would be the best for GCC.
Can you elaborate on this a bit? What in particular is undesirable
about zlib's direction (I don't really follow it). And why should we
care about how zlib handles platform compatibility? Is it notably bad
in some way, or just different?
Marcin> Sharing code for the sake of it isn't always the best
Marcin> strategy.
I think the bar for making local changes to zlib, other than, say,
build trivia, should be very high. My reason is that zlib is
peripheral to GCC development. The only user of zlib in the tree, as
far as I know, is java, and as far as I know none of us is interested
in maintaining a zlib fork.
FWIW, zlib has been very stable over the years. There was a security
advisory or two, but these were easily solved by a new import. I
don't recall that we ever had to actually fix a real bug, and
"ChangeLog.gcj" indicates that most of what we've done locally is
tweak the build system from time to time.
Tom