2008/8/27 Duncan <[EMAIL PROTECTED]> > > Greg Lee <[EMAIL PROTECTED]> posted > [EMAIL PROTECTED], excerpted below, on Tue, 26 Aug 2008 16:31:45 > +0000: > > > You mean you have a pre-compiled version of Pan > > Rather unlikely for someone running Gentoo. More likely, he compiled it > himself (as Gentooers generally do), but then upgraded something, the > environment moved out from under the pan package he'd built, and he > hasn't figured out what moved where and how to correct it yet. > > But that does bring up another way to fix it. Simply recompile it, and > it should be able to find what it built against. If it can't (which from > what he said seems to be the case), it's not a pan problem directly, but > one of pan's dependencies, perhaps one of the gtk+ libs, or pcre, or... > When he loads pan, it tries to load the lib, which tries to load a > libstdc++ it can't find, because it's the lib that had stuff move out > from under it, not pan. Of course, rebuilding that lib should solve the > problem, but finding the lib to rebuild is the difficult bit.
yesterday evening i was too tired to start cleaning the system. now i've found out what the problem was: i have 2 versions of gcc installed, 4.2.4 and 4.3.1-r something. usually i don't have 2 gcc versions since it's not a very good idea, but this time i use 2 of them since some packages fail with gcc 4.3.1. so when a package fails for unknown reasons, before i submit a bug report i check out if it builds with gcc 4.2.4. I use the amd64 system with multilib enabled, since i use some firefox-bin, openoffice-bin, mplayer-bin and mplayerplug-in compiled with ABI=x86. i've seen something strange, like the old gcc version still being there after removing them, but i've never given this too much importance since everything was ok. i've never had eselect-gcc installed on this system, but always used gcc-config, never removed stuff manually but with portage or paludis and never cleaned cruft files on the system. now what i've discovered thanks to pan is that the libstdc++.so.6 was pointing to the 4.2.4 version instead of the 4.3.1 while the inuse compiler was the 4.3.1. i don't know when this problem occured (i've last switched to 4.3.1 about one month ago and since then i've compiled and recompiled a lot of packages). i don't know how this mispointing could have impacted on the system, but it might have been a big bad stuff. this week end i really need to lauch the fincruft script, move out of the way all the unneeded files and then revdep. after that i have to make sure that the compiler is set to the 4.3.1 and recompile world to be sure that everything is ok and then rerun the cruft script. at that point everything should be ok (or at least i hope so).... for the moment i've temporarly set the symlink to the gcc 4.3.1 and pan is working well. well, i hope that next week i don't need to have too much support.... -- dott. ing. beso _______________________________________________ Pan-users mailing list Pan-users@nongnu.org http://lists.nongnu.org/mailman/listinfo/pan-users
