SL Baur writes:
Marc Aurel <4-tea-2(a)bong.saar.de> writes in
xemacs-beta(a)xemacs.org:
>>>>>> In article <m2pvfa2a4m.fsf(a)bong.bongnet>, Marc Aurel
>>>>>> <4-tea-2(a)bong.saar.de> writes:
>>> I also updated to gcc 2.8 and friends.
>> I'm too scared from my past experience to try that again. I had to turn
>> off optimization when I used gcc-2.8.1, because even a humble '-O'
would
>> lead to a 'FATAL ERROR 6' on my AMD K5.
> No more crashes since I got rid of RedHat's gcc/egcs combination and
> replaced it with the latest stable egcs release (1.0.3).
I just updated egcs and glib from the RedHat site (or Sunsite UNC's
mirror) and will let you know what happens shortly (after I download
and build).
Does anyone have other similar experiences with RH 5.1? Should we add
this to the PROBLEMS file?
I built one with egcs and one with gcc 2.8 -- didn't realize the
egcs was so different. Had no problems with either one, though the
gcc 2.8 one seemed marginally faster.
I haven't built XEmacs on RH 5.1 yet, but I've built and run InfoDock
and it is O.K. (with gcc). Um, why didn't Red Hat include libc5
libraries for legacy binaries? That seems _really_ silly to me.
A check into Glint says libc 5.3.12-27 is in there.
--
Rebecca Ore