eric wroteBTW. parabola seem to build the new version with gcc 6
The key word here is
seem[ed], It is not known whether they compiled it with gcc 6, 5 or 4 but that also makes me eager to contact the last maintainer to find out how he compiled it and or with what version of gcc and why he orphaned the package.
eric wrote the last glibc break the compilation for the old package too written error
Do you see the theme here? It isn't glibc, all of the bug reports I've posted lead to either gcc or gcc-libs *not* glibc
Like this one for instance:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69581
taco@ zen ~ % pacman -Qo /usr/lib/libstdc++.so
/usr/lib/libstdc++.so is owned by gcc-libs 6.1.1-1
eric wroteMaybe the issue come from obarun system.... i doubt
As you said 'obarun isn't arch', However the problem
IS coming from arch and fedora as they made the choice of pushing the updates straight from git, VERY bad idea. Their decision has made new bugs even MORE prominent and just look at whats happening now! ARGH. I wanted to use arch NOT fedora, fedora is known to do this shit.
eric wroteTaking a git release, grrrr, i don't like this ...
eric wrotefor me it is not gcc6 that i hate, but this kind of package
It would have been fine if it was just icecat, but no, It is also IceApe, WxPython, cataclysm-dda and even more. Hell anything that relies on stdlib.h is absolutely fucked. and all I get from researching it is "Try GCC 4.9" or the classic "It will be fixed in the NEXT version". If you disagree, Prove me wrong. The only way I can fathom why everything isn't becoming utterly broken is that the devs precompile the packages with a different version of gcc.
eric wroteCompilation take almost the same time that a kernel
That's because it takes more time for gcc6 to shit out more warnings and errors, than it does actually compiling it. It *didn't* have this problem on gcc 5.
eric wrotei will try to compile the gcc package with the last commit
I commend you for your efforts/work and believe me when you read this, I'm
not mad at you. I'm simply venting/mad at poor execution from arch that
affects obarun, because let me repeat- gentoo doesn't even
list gcc 6 as a testing package, they haven't even made it a priority to test. That's really REALLY bad. Void for instance has compiled the latest icecat, why? because they have
gcc 4.9
To seal the fate of my trust for arch, This the answer a Arch Trusted User gave when asked if it is possible to use Gcc 4.9, The OP had the same exact F!@ # $ing issue.
Xyne (0) wrote
A rolling release distro is not suitable for your needs. You should consider using something else for work. If you have a work computer then remove Arch and put whatever you actually need on it. If you don't have a work computer or need to keep Arch on the work computer, dual boot or use system virtualization. Keep a fully updated Arch host system and a nested build system.
https://bbs.archlinux.org/viewtopic.php?pid=1598675# p1598675
Therefore according to Xyne, Arch is the least applicable option for production. Which is the complete opposite of what I thought it was at first.
Would arch apply fedoras patch [1] for gcc6 so it doesn't suck ass? Allan says "No." [1]
wdirksen wrote
Allan,
Will you be updating the gcc related Arch packages with this or other solution?
Allan wrote
Why would I? gcc is not broken...
Yes, the SAME Allan that made propaganda for systemd, the same person that thought keysigning wasn't an issue back in 2011 and now he is using the "But it works for me(TM)" bullshit excuse. Honestly, fuck him.
http://allanmcrae.com/2012/08/are-we-removing-what-defines-arch-linux/
https://igurublog.wordpress.com/2011/03/24/lwn-picks-up-on-package-signing/
When I wrote in the other thread and I quote
bit wrotePretty much debating if I should just replace obarun with void unless gcc fixes their shit real fast, which they most likely won't.
I wasn't even remotely kidding. Arch did this. Not you.
[0]
http://pkgs.fedoraproject.org/cgit/rpms/flann.git/tree/flann-1.8.4-gcc6.patch
[1]
https://bbs.archlinux.org/viewtopic.php?pid=1626161# p1626161