The gaim launcher will not start gaim if gaim is set to auto-login to an AIM account. If it is not set to auto-login, gaim will die as soon as you connect. The same problem occurs if you launch it from the Alt-F2 run dialog. The problem does not occur if run from a terminal OR if 'Run in Terminal' is checked in the launcher properties. In the forums, people have attributed it to gaim needing a stdout. Interestingly enough, running 'gaim > /dev/null' works only in the Alt-F2 dialog, but when the same command is put in the launcher, it crashes. This happens with many 0.6x gaim series, and gnome 2.2 and 2.4. Gaim works as expected from a KDE launcher. More info here: http://forums.gentoo.org/viewtopic.php?t=73106&highlight=gaim+stdout http://forums.gentoo.org/viewtopic.php?t=86908&highlight=gaim+launcher http://forums.gentoo.org/viewtopic.php?t=61842&highlight=gaim+launcher http://forums.gentoo.org/viewtopic.php?t=49815 Reproducible: Always Steps to Reproduce: 1. Start gaim from a gnome launcher 2. Connect to an AIM account 3. Actual Results: gaim crashes and disappears Expected Results: gaim should have connected as normal
gnome team doesn't handle gaim.
Hi, I got the same problem here, but found a workaround on the gentoo forum : try to compile gaim with low CFLAGS. My current CFLAGS are with -Os, and i got gaim to run and launch from gnome launcher with -O2.
Changing my CFLAGS from -Os to -O2 fixed the problem for me. Perhaps filter the -Os flag in the ebuild?
whoa, thats wierd. switching to -O2 did fix it for me too... perhaps some filtering is in order. the weird thing is that 'gaim > /dev/null' works (not as a launcher though) and just plain gaim works from a kde launcher... what a weird side affect of -Os
I can confirm this bug. Connecting Gaim to an ICQ account will infallibly cause the program to segfault and coredump. This holds for starts from terminal and fluxbox menu. Other accounts work fine. Emerge info> Portage 2.0.49-r15 (default-x86-1.4, gcc-3.2.3, glibc-2.3.1-r4, 2.4.22) ================================================================= System uname: 2.4.22 i686 AMD Athlon(tm) XP 2400+ Gentoo Base System version 1.4.2.8 distcc 2.9 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [enabled] ccache version 2.2 [enabled] ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CFLAGS="-mcpu=pentium -O2" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /var/qmail/control /usr/share/config /usr/kde/2/share/config /usr/kde/3/share/config /usr/X11R6/lib/X11/xkb" CONFIG_PROTECT_MASK="/etc/gconf /etc/env.d" CXXFLAGS="-O2 -mcpu=i686 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs sandbox userpriv usersandbox ccache distcc" GENTOO_MIRRORS="ftp://trumpetti.atm.tut.fi/gentoo/" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage" USE="x86 oss apm avi crypt encode gif jpeg libg++ libwww mad mikmod mpeg ncurses nls pdflib png quicktime spell truetype xml2 xmms xv zlib gdbm berkdb slang readline svga tcltk java sdl gpm tcpd pam ssl perl python esd imlib oggvorbis gtk qt opengl cdr X dga Xaw3d -arts -cups -foomaticdb -bonobo -guile -gnome -kde -gtk2 -motif" Th funny thing is that gaim worked flawlesly untill i had to do a hard reboot. Reemergin cleanly has not resolved it.
Created attachment 20437 [details] ebuild with replace -Os with -O2 Update ebuild to replace -Os with -O2? ('Fixed' ebuild attached, so at least gaim does not appear broken on gentoo systems)
I'll try to test this out and get it in. I don't use gnome but I don't see any real harm in setting -O2.
Daniel Armyr: I don't think your bug is the same thing. You seem to have an ICQ-specific bug and to me that sounds like an UPSTREAM issue with the gaim team.
OK I've committed the replace-flags to gaim-0.74-r1. Please test it out and let me know if it works for you and I'll close this bug.
This was working fine in -r1. However it appears to be broken in -r2 and we're having the same problem!
Hmm the replace-flags command is still in place in -r2 so I can't say what is causing the problem this time.
I honestly don't understand how changing optimizations affects the ability for this to work with gaim-launcher. Nor do I understand why it wouldn't work from 0.74-r1 to 0.74-r2 since nothing involving that fix in -r1 was touched.
The problem now (gaim-0.75-r5) is that the cflags (-Os -> -O2) are, in fact, not replaced. Moving the replace-flags section of the ebuild to the top of the src_compile section replaces the flags correctly. I can't find the core file so I don't know what the real problem is with starting gaim from a launcher as opposed to starting it from a terminal. All I can find is in -gnomerc-errors that gaim has segfaulted and attempted to dump a core file.
Alright I've moved it to the top of src_compile(). Can anyone confirm if gaim-0.75-r5 (please rsync first) is working for this issue yet?
It works for me, but I reckon you already knew that :-)
Alright then I'll again close this bug. Re-open if you need to.
Created attachment 24142 [details] Backtrace from gaim core dump after trying to login to yahoo OK, I managed to find the setting for core dumps, so here is a backtrace for gaim compiled with -Os.
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=2c297a5d6acec05aaa94de3215be947d12ef86a5 commit 2c297a5d6acec05aaa94de3215be947d12ef86a5 Author: Cristian Othón Martínez Vera <cfuga@cfuga.mx> AuthorDate: 2025-03-26 16:55:54 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2025-04-10 09:32:00 +0000 x11-plugins/pidgin-encryption: drop strip/replace compiler flags by default +20 years ago there's was a bug with ```net-im/gaim```, probably caused by gcc-2.95 and ```-Os``` optimization. The following developers and maintainers just copied those lines when bumping versions, and nobody tried to remove them with new Gaim/Pidgin/GCC versions. Also ```x11-plugins/pidgin-encryption``` has the same code; I guess the Gentoo maintainer/developer was extra cautious. Bug: https://bugs.gentoo.org/29394 Signed-off-by: Cristian Othón Martínez Vera <cfuga@cfuga.mx> Closes: https://github.com/gentoo/gentoo/pull/41309 Signed-off-by: Sam James <sam@gentoo.org> .../pidgin-encryption-3.1-r4.ebuild | 38 ++++++++++++++++++++++ 1 file changed, 38 insertions(+) https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=8d9891cab30df71423ccb306dde5eacacf146b56 commit 8d9891cab30df71423ccb306dde5eacacf146b56 Author: Cristian Othón Martínez Vera <cfuga@cfuga.mx> AuthorDate: 2025-03-26 16:42:29 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2025-04-10 09:32:00 +0000 net-im/pidgin: drop strip/replace compiler flags by default, enable py3.13 +20 years ago there was a bug with ```net-im/gaim```, probably caused by gcc-2.95 and ```-Os``` optimization. The following developers and maintainers just copied those lines when bumping versions, and nobody tried to remove them with new Gaim/Pidgin/GCC versions. Bug: https://bugs.gentoo.org/29394 Closes: https://bugs.gentoo.org/944837 Signed-off-by: Cristian Othón Martínez Vera <cfuga@cfuga.mx> Signed-off-by: Sam James <sam@gentoo.org> net-im/pidgin/{pidgin-2.14.14.ebuild => pidgin-2.14.14-r1.ebuild} | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)
