dsh8290@??? said: > Do you really want to give newbies a CVS snapshot of a compiler?
What an absurd question. Do you genuinely believe that Red Hat are shipping
just a CVS snapshot without any other fixes, or do you have other motives for
saying such strange things?¹
The current version appears to be gcc-2.96-110, released with Red Hat Linux
7.3. Aside from the CVS snapshot tarball, it contains 389 individual patch
files -- feeding them through diffstat gives us the following output:
615 files changed, 35145 insertions(+), 9955 deletions(-), 371 modifications(!)
I believe that where applicable, those patches were also fed back into the
master GCC tree, hence helping to get the 3.0 release out the door.
2.96-RH is actively supported and stable. It's _far_ from being just 'a CVS
snapshot'.
> Do you know why RH 7.x includes 'kgcc'?
Because the kernel was packed with bugs which were brought to light by more
aggressively optimising compilers. Now those bugs are mostly found and
fixed, Red Hat no longer compiles the kernels with an older compiler.
> This is totally contrary to my experiences. g++ 2.95 can fully handle
> the complex templates used by the SigC type-safe signal library
> whereas Sun's CC barfed trying to buid it. g++ 2.96 also thought that
> the words 'and' , 'or' , and 'not' were reserved words and not legal
> structure member identifiers (as used by the cctype library). cctype
> is NOT buggy in that respect. The C++ optimizer would just give up
> and die while trying to compile Mahogany. This is all on top of a
> buggy libc that defined LONG_BIG as 64 on my ia32 system (the python
> developers were quick enough to catch that and #error with a decent
> message). Need I try and recall other problems it caused?
Yes please. Although it would be somewhat off-topic for this list, so
private email would perhaps be more appropriate. Do not forget to provide
the Bugzilla bug numbers.
> While I agree that gcc/g++ 2.95 isn't perfect and fully standards
> compliant, it is *stable* and *supported* by the GCC development team.
And Red Hat's 2.96 is stable and supported by Red Hat's compiler team -
many of whom used to work for a small company by the name of Cygnus -- you
may have heard of them, especially if you've ever looked at GCC Changelogs.
> A CVS snapshot is neither; not to mention binary incompatible with all
> stable and supported gcc releases.
This is normal -- just as 2.95 is binary incompatible with all other stable
and supported gcc releases. Hopefully, this breakage of binary compatibility
with _every_ version should be fixed now in 3.1 and beyond². There's nothing
special about 2.96 in this respect.
--
dwmw2
¹ I work for Red Hat, btw. Although I'd probably have tried to inject some
sense into this discussion even if I didn't.
² I believe this was planned for 3.0, but it got broken again for IA64
before 3.1.
