GNOME Bugzilla – Bug 97883
Crash on startup [gnome_window_manager_get_settings, settings==NULL]
Last modified: 2004-12-22 21:47:04 UTC
Package: control-center Severity: critical Version: GNOME2.1.2 2.1.1.1 Synopsis: Crash on startup Bugzilla-Product: control-center Bugzilla-Component: theme-selector BugBuddy-GnomeVersion: 2.0 (2.1.2) Description: Description of Problem: gnome-theme-properties crashes when started. Steps to reproduce the problem: 1. Start the program 2. 3. Actual Results: Crash Expected Results: Running program How often does this happen? Every time. Additional Information: Debugging Information: Backtrace was generated from '/gnome/head/INSTALL/bin/gnome-theme-properties' [New Thread 1024 (LWP 27399)] 0x40a65219 in __wait4 () at __wait4:-1 in __wait4
+ Trace 29954
Thread 1 (Thread 1024 (LWP 27399))
------- Bug moved to this database by unknown@bugzilla.gnome.org 2002-11-06 18:15 ------- Unknown version 1.1.x in product control-center. Setting version to the default, "unspecified". Reassigning to the default owner of the component, control-center-maint@bugzilla.gnome.org.
Unique 2.1.x trace.
Still open in gnome-theme-manager.
Yep, I am seeing this too.
*** Bug 97934 has been marked as a duplicate of this bug. ***
In bug 97934, the reporter says he sees this on a terminal: gnome-theme-manager:2795): capplet-common-WARNING **: Couldn't load window manager settings module `/usr/lib/window-manager-settings/libmetacity.so' (/usr/lib/window-manager-settings/libmetacity.so: cannot open shared object file: No such file or directory) (gnome-theme-manager:2795): GLib-GObject-WARNING **: invalid cast from (NULL) pointer to `GObject' (gnome-theme-manager:2795): GLib-GObject-WARNING **: invalid (NULL) pointer instance (gnome-theme-manager:2795): GLib-GObject-CRITICAL **: file gsignal.c: line 1605 (g_signal_connect_data): assertion `G_TYPE_CHECK_INSTANCE (instance)' failed (gnome-theme-manager:2795): capplet-common-WARNING **: Couldn't load window manager settings module `/usr/lib/window-manager-settings/libmetacity.so' (/usr/lib/window-manager-settings/libmetacity.so: cannot open shared object file: No such file or directory)
Moving this up to 'urgent' since it blocks testing, and this code needs testing :/
Program received signal SIGSEGV, Segmentation fault. 0x83e58955 in ?? () (gdb) bt
+ Trace 30228
FAM, if fam is started it crash., without FAM, it works without fam : FAMOpen failed, FAMErrno=0 FAMOpen failed, FAMErrno=0 FAMOpen failed, FAMErrno=0 FAMOpen failed, FAMErrno=0 FAMOpen failed, FAMErrno=0 so where is the problem ?
*** Bug 98370 has been marked as a duplicate of this bug. ***
I'm seeing this, and I don't have FAM installed. Same back trace.
Fixed for me in the latest from CVS 15 nov. Please test.
I just pulled CVS (16 Nov 19:00 GMT), I also had to install metacity CVS due to a depedence on libmetacity-private. I still get a crash on startup: Backtrace was generated from '/opt/gnome-2.1/bin/gnome-theme-manager' [New Thread 1024 (LWP 32042)] 0x40afd839 in __wait4 () at __wait4:-1 in __wait4
+ Trace 30534
Thread 1 (Thread 1024 (LWP 32042))
Loz: could you look at the top entry in capplets/theme-switcher/ChangeLog so we know what version is still crashing. Anoncvs mirrors can lag quite a bit.
*** Bug 98794 has been marked as a duplicate of this bug. ***
That backtrace can not possibly correspond to CVS. gnome_window_manager_get_settings (wm=0x80856e8, settings=0x0) at gnome-window-manager.c:136 That call is now generated as gnome_window_manager_get_settings (window_manager, &wm_settings); settings can not be null
Based on jody's comment, and confirmation from jrb on IRC, I'm closing this bug as indicative of an old build. Thunder, jrb indicates that this may indicate a broken install of the glade file from the packages. Can you take a look at that?
Doh. Forgot to actually close.
*** Bug 98972 has been marked as a duplicate of this bug. ***
We're still getting traces for this one, dated recently
Andrew: we are, but that's because the Ximian packages are broken, not because CVS is broken.
*** Bug 99154 has been marked as a duplicate of this bug. ***
*** Bug 99729 has been marked as a duplicate of this bug. ***
*** Bug 100128 has been marked as a duplicate of this bug. ***
*** Bug 100129 has been marked as a duplicate of this bug. ***
bug 100129 provides a trace where settings is not null - while some of these traces are likely broken packages like Luis mentioned, I'm not sure all of them are.
We're (still) getting dups from 2.1.3. ftp.gnome.org indicates that this is a later build than Nov 24.
*** Bug 100098 has been marked as a duplicate of this bug. ***
*** Bug 100249 has been marked as a duplicate of this bug. ***
*** Bug 100396 has been marked as a duplicate of this bug. ***
*** Bug 100632 has been marked as a duplicate of this bug. ***
If anyone marks a duplicate later than 2.1.3, please point it out in this bug because I think this may be fixed (that's not a coder's opinion, but I haven't seen dups from the metatheme-type capplet). I'm leaving this bug open for a bit longer before assuming it's fixed.
*** Bug 101382 has been marked as a duplicate of this bug. ***
*** Bug 101398 has been marked as a duplicate of this bug. ***
(the first two are bogus) davh@cine$ valgrind gnome-theme-manager ~ ==2235== valgrind-1.0.4, a memory error detector for x86 GNU/Linux. ==2235== Copyright (C) 2000-2002, and GNU GPL'd, by Julian Seward. ==2235== Estimated CPU clock rate is 2424 MHz ==2235== For more details, rerun with: -v ==2235== valgrind's libpthread.so: KLUDGED call to: pthread_getschedparam ==2235== Syscall param write(buf) contains uninitialised or unaddressable byte(s) ==2235== at 0x40CB8CB4: __libc_write (in /lib/libc-2.3.1.so) ==2235== by 0x4070F9AB: (within /usr/X11R6/lib/libICE.so.6.3) ==2235== by 0x40710673: _IceTransWrite (in /usr/X11R6/lib/libICE.so.6.3) ==2235== by 0x407081FC: _IceWrite (in /usr/X11R6/lib/libICE.so.6.3) ==2235== Address 0x428F5EAF is 3 bytes inside a block of size 1024 alloc'd ==2235== at 0x400456AB: malloc (/home/cyrille/Code/2002/packages/my-packages/valgrind-1.0.4/vg_clientfuncs.c:100) ==2235== by 0x40705410: IceOpenConnection (in /usr/X11R6/lib/libICE.so.6.3) ==2235== by 0x406FB301: SmcOpenConnection (in /usr/X11R6/lib/libSM.so.6.0) ==2235== by 0x402CCA19: gnome_client_connect (/home2/garnomeb/garnome-0.19.3/gnome/libgnomeui/work/libgnomeui-2.1.3/libgnomeui/gnome-client.c:1486) ==2235== ==2235== Syscall param writev(vector[...]) contains uninitialised or unaddressable byte(s) ==2235== at 0x40CBE6D8: __libc_writev (in /lib/libc-2.3.1.so) ==2235== by 0x406EF0DF: write_data (/home2/garnomeb/garnome-0.19.3/gnome/linc/work/linc-0.7.0/src/linc-connection.c:554) ==2235== by 0x406EF26F: linc_connection_writev (/home2/garnomeb/garnome-0.19.3/gnome/linc/work/linc-0.7.0/src/linc-connection.c:667) ==2235== by 0x406B72C6: giop_send_buffer_write (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/GIOP/giop-send-buffer.c:419) ==2235== Address 0x42920C4E is 10 bytes inside a block of size 2048 alloc'd ==2235== at 0x400456AB: malloc (/home/cyrille/Code/2002/packages/my-packages/valgrind-1.0.4/vg_clientfuncs.c:100) ==2235== by 0x40B9EEB7: g_malloc (/home2/garnomeb/garnome-0.19.3/gnome/glib/work/glib-2.1.3/glib/gmem.c:138) ==2235== by 0x406B706C: get_next_indirect (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/GIOP/giop-send-buffer.c:273) ==2235== by 0x406B70F6: giop_send_buffer_append_copy (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/GIOP/giop-send-buffer.c:295)==2235== ==2235== Conditional jump or move depends on uninitialised value(s) ==2235== at 0x404BC4C3: gconf_engine_connect (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:341) ==2235== by 0x404BC839: gconf_engine_get_default (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:461) ==2235== by 0x404C21A5: gconf_client_get_default (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf-client.c:429) ==2235== by 0x402BAC22: ??? (/home2/garnomeb/garnome-0.19.3/gnome/libgnomeui/work/libgnomeui-2.1.3/libgnomeui/gnome-ui-init.c:420) ==2235== ==2235== Use of uninitialised value of size 4 ==2235== at 0x40B912C6: g_hash_table_insert (/home2/garnomeb/garnome-0.19.3/gnome/glib/work/glib-2.1.3/glib/ghash.c:204) ==2235== by 0x404BC368: gconf_engine_set_database (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:286) ==2235== by 0x404BC51D: gconf_engine_connect (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:353) ==2235== by 0x404BC839: gconf_engine_get_default (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:461) ==2235== ==2235== Use of uninitialised value of size 4 ==2235== at 0x40B912FA: g_hash_table_insert (/home2/garnomeb/garnome-0.19.3/gnome/glib/work/glib-2.1.3/glib/ghash.c:297) ==2235== by 0x404BC368: gconf_engine_set_database (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:286) ==2235== by 0x404BC51D: gconf_engine_connect (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:353) ==2235== by 0x404BC839: gconf_engine_get_default (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:461) ==2235== ==2235== Use of uninitialised value of size 4 ==2235== at 0x40B91340: g_hash_table_insert (/home2/garnomeb/garnome-0.19.3/gnome/glib/work/glib-2.1.3/glib/ghash.c:315) ==2235== by 0x404BC368: gconf_engine_set_database (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:286) ==2235== by 0x404BC51D: gconf_engine_connect (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:353) ==2235== by 0x404BC839: gconf_engine_get_default (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:461) ==2235== ==2235== Conditional jump or move depends on uninitialised value(s) ==2235== at 0x404BC420: gconf_engine_connect (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:312) ==2235== by 0x404BC552: gconf_engine_get_database (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:361) ==2235== by 0x404BCCD1: gconf_engine_notify_add (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:653) ==2235== by 0x404C2553: gconf_client_add_dir (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf-client.c:565) ==2235== ==2235== Conditional jump or move depends on uninitialised value(s) ==2235== at 0x404BCCD7: gconf_engine_notify_add (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:655) ==2235== by 0x404C2553: gconf_client_add_dir (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf-client.c:565) ==2235== by 0x402BAC46: ??? (/home2/garnomeb/garnome-0.19.3/gnome/libgnomeui/work/libgnomeui-2.1.3/libgnomeui/gnome-ui-init.c:427) ==2235== by 0x402BAD96: ??? (/home2/garnomeb/garnome-0.19.3/gnome/libgnomeui/work/libgnomeui-2.1.3/libgnomeui/gnome-ui-init.c:470) ==2235== ==2235== Conditional jump or move depends on uninitialised value(s) ==2235== at 0x406BD262: ORBit_small_invoke_stub (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/orb-core/orbit-small.c:588) ==2235== by 0x406BD243: ORBit_small_invoke_stub_n (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/orb-core/orbit-small.c:571) ==2235== by 0x404C9D77: ConfigDatabase3_add_listener_with_properties (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/GConf-stubs.c:756) ==2235== by 0x404BCD22: gconf_engine_notify_add (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:663) ==2235== ==2235== Use of uninitialised value of size 4 ==2235== at 0x406BD2A3: ORBit_small_invoke_stub (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/orb-core/orbit-small.c:595) ==2235== by 0x406BD243: ORBit_small_invoke_stub_n (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/orb-core/orbit-small.c:571) ==2235== by 0x404C9D77: ConfigDatabase3_add_listener_with_properties (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/GConf-stubs.c:756) ==2235== by 0x404BCD22: gconf_engine_notify_add (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/gconf.c:663) ==2235== ==2235== Invalid read of size 4 ==2235== at 0x406CD9FB: ORBit_small_handle_request (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/poa/orbit-adaptor.c:107) ==2235== by 0x406BD305: ORBit_small_invoke_stub (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/orb-core/orbit-small.c:602) ==2235== by 0x406BD243: ORBit_small_invoke_stub_n (/home2/garnomeb/garnome-0.19.3/gnome/ORBit2/work/ORBit2-2.5.0/src/orb/orb-core/orbit-small.c:571) ==2235== by 0x404C9D77: ConfigDatabase3_add_listener_with_properties (/home2/garnomeb/garnome-0.19.3/gnome/GConf/work/GConf-1.2.1/gconf/GConf-stubs.c:756) ==2235== Address 0x21 is not stack'd, malloc'd or free'd ==2235== discard 120 (1518 -> 22045) translations in range 0x42139000 .. 0x42142FFF ==2235== discard syms in /lib/libnss_compat-2.3.1.so due to munmap() ==2235== discard 31 (398 -> 5546) translations in range 0x42167000 .. 0x42179FFF ==2235== discard syms in /lib/libnsl-2.3.1.so due to munmap() ==2235== discard 177 (2328 -> 34026) translations in range 0x402F5000 .. 0x402FDFFF ==2235== discard syms in /lib/libnss_files-2.3.1.so due to munmap() ==2235== ==2235== ERROR SUMMARY: 24 errors from 11 contexts (suppressed: 29 from 2) ==2235== malloc/free: in use at exit: 104850 bytes in 2052 blocks. ==2235== malloc/free: 5008 allocs, 2956 frees, 387494 bytes allocated. ==2235== For a detailed leak analysis, rerun with: --leak-check=yes ==2235== For counts of detected errors, rerun with: -v ==2239== discard 120 (1518 -> 22045) translations in range 0x44E96000 .. 0x44E9FFFF ==2239== discard syms in /lib/libnss_compat-2.3.1.so due to munmap() ==2239== discard 31 (398 -> 5546) translations in range 0x44EB2000 .. 0x44EC4FFF ==2239== discard syms in /lib/libnsl-2.3.1.so due to munmap() ==2239== ==2239== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 10 from 2) ==2239== malloc/free: in use at exit: 72500 bytes in 1005 blocks. ==2239== malloc/free: 1446 allocs, 441 frees, 272560 bytes allocated. ==2239== For a detailed leak analysis, rerun with: --leak-check=yes ==2239== For counts of detected errors, rerun with: -v
*** Bug 101950 has been marked as a duplicate of this bug. ***
For the benefit of the dup finder, here is a similar trace that we want to mark as duplicates of this bug: [New Thread 1024 (LWP 4843)] 0x4091f4d9 in wait4 () from /lib/libc.so.6
+ Trace 32039
Thread 1 (Thread 1024 (LWP 4843))
*** Bug 101785 has been marked as a duplicate of this bug. ***
*** Bug 102245 has been marked as a duplicate of this bug. ***
Duplicate from 2.1.5. This is NOT FIXED.
I'm pretty sure this is fixed in cvs.
We keep getting recent duplicates. We can't keep closing the bug and then getting duplicates and reopening it. Earlier the traces had settings=NULL - in bug 102245 the trace has settings not being NULL, but it still crashed. Maybe this is a very slightly different bug.
*** Bug 102327 has been marked as a duplicate of this bug. ***
*** Bug 102424 has been marked as a duplicate of this bug. ***
*** Bug 102516 has been marked as a duplicate of this bug. ***
I'm getting a broken pipe in 2.1.5 (theme-thumbnail.c:453 [generate_theme_thumbnail]: ----- (452) pixbuf = gdk_pixbuf_new (GDK_COLORSPACE_RGB, TRUE, 8, ICON_SIZE_WIDTH, ICON_SIZE_HEIGHT); (453) write_bytes = write (pipe_to_factory_fd[1], meta_theme_info->gtk_theme_name, strlen (meta_theme_info->gtk_theme_name) + 1); ----- # ps -A | grep gnome-theme-man 18765 pts/3 00:00:00 gnome-theme-man 18766 pts/3 00:00:00 gnome-theme-man <defunct> The child (PID 187666) was created in theme-thumbnail.c:522 [theme_thumbnail_factory_init]. It doesn't wait around for the parent to start writing, it just exits. Is this the right behavior? Stack trace: Program received signal SIGPIPE, Broken pipe. 0x40ce1584 in __libc_write () at __libc_write:-1 -1 __libc_write: No such file or directory. in __libc_write (gdb) backtrace
+ Trace 32311
OK, I'm confused... was gnome-theme-properties recently renamed to gnome-theme-manager? See bug #101567.
Yes. A while ago, actually.
*** Bug 103106 has been marked as a duplicate of this bug. ***
I'm still getting this with 2.2RC1 (control-center 2.1.6) but the segfault is a little bit different from the ones I see here. I'm not sure if it's the same bug or not but I figured I'd post the segfault here just to be safe. Backtrace was generated from '/opt/garnome/bin/gnome-theme-manager' 0x40b6e089 in wait4 () from /lib/libc.so.6
+ Trace 32675
I patched a pile of these just before 2.1.6 and just patched the one mentioned in the last backtrace. Give cvs a try.
*** Bug 103453 has been marked as a duplicate of this bug. ***
*** Bug 103531 has been marked as a duplicate of this bug. ***
Bug 103531 has the "settings == NULL" bug and is current CVS.
Yet another of these damn things patched. Folks there is way way to much cut-n-pasting going on in there. The same idiom has come up in 5 places now.
Speaking as someone who's been affected by this bug for, well, gnome-theme-manager's entire existence, this is now fixed for me.
*** Bug 103727 has been marked as a duplicate of this bug. ***
*** Bug 103890 has been marked as a duplicate of this bug. ***
I've just built control-center-2.0.0 from ftp.gnome.org. Well, here gnome-theme-manager doesn't crash, it just hangs infinitely while scanning /usr/share/themes (as strace suggests). --snip-- write(17, "\0\0\0d", 4) = 4 write(17, "M221 1000 1000 /usr/share/themes"..., 100) = 100 lstat64("/usr/share/themes/CleanIce", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 lstat64("/usr/share/themes/CleanIce/index.theme", 0xbffff398) = -1 ENOENT (No such file or directory) getgroups32(0x20, 0x80c1770) = 15 geteuid32() = 1000 write(17, "\0\0\0Z", 4) = 4 write(17, "M222 1000 1000 /usr/share/themes"..., 90) = 90 lstat64("/usr/share/themes/CleanIce", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 lstat64("/usr/share/themes/CleanIce/gtk-2.0/gtkrc", 0xbffff3a8) = -1 ENOENT (No such file or directory) getgroups32(0x20, 0x80c1910) = 15 geteuid32() = 1000 write(17, "\0\0\0b", 4) = ? ERESTARTSYS (To be restarted) --
Well, I mean to say "control-center-2.2.0" When I disable fam (by stopping inetd), gnome-theme-manager complains loudly, but works! I'm using fam/libfam 2.6.8, if that means anything. I still hesitate upgrading to fam 2.6.9, as it conflicts with many other installed C++ programs (ongoing transition to GCC 3.2 in Debian)
j.rohr: you're probably seeing bug 103797
Andrew: You're right, thanks. Johannes
I have the same problem but the stop of inetd has no impact on the fact that gnome-control-center crashes when I launch it and a window tells me that gnome-theme manager has crashed All icons in nautilus as well as on desktop are like a blank sheet of paper icon (files and directories)
Created attachment 13944 [details] [review] control-center startup crash fix
Created attachment 13950 [details] [review] updated patch to fix theme-manager crash + added creation of default_icon_theme key if it didnot exist
Maintainers, is this patch relevant to CVS? I'm reopening for now just to keep the patch visible. If it's fixed in CVS, please close it/needinfo it again.
*** Bug 104914 has been marked as a duplicate of this bug. ***
*** Bug 104916 has been marked as a duplicate of this bug. ***
*** Bug 104877 has been marked as a duplicate of this bug. ***
How about applying patch to CVS. Becasue i saw alot of complains in debian gtk-gnome about that crash and the fact that all icons in nautilus after upgrade becoming exactly the same. p.s. Donot know to whom send a patch
Is this patch updated to reflect the recent renaming of the Default theme?
Where can i find inforamtion about the changing of the name of Default theme ?
Ah ok itis changing of the name for Default gtk and Metacity theme. I think the best solution is just add link ln -s Simple Default and ln -s gnome Default for icon-theme during installation of packages. I just checked neither gnome-themes create such link nor gnome-icon-theme. So this packages just should be fixed. Otherwise in theme manager all 'Default' should be replaced wirh 'Simple' and 'gnome' at least in 15 20 places.
Urm, no. The renaming happened to remove the word "Default" from the UI, due to reconfigurable defaults. The correct thing to do is to remove any occurences of the string "Default" from the code.
Sorry for me itis not clear - so instead of "Default" what should be used ? Just emty string ""? or "Simple", in last case name "Simple" can be changed later also and we need again to change code of theme-manager ?
Hmmm. Confused, I don't know.
What i care about is when user migrate to gnome 2.2 - he install all packages restart X and voila he has system with reasonable choosen themes. Here there are some problems. I observer at least on 2 from my 3 PCs - key "/gnome/desktop/icon_theme" is not created automatically by Nautilus so what we have is just the same icon for all object. (Strangely at one of my PC nautilus create that key automatically). So what i expect from theme manager - it should start and create that key and choose a proper value for that (and probably gtk_theme) key. So question is what should we written ? I prefer to have those links called "Default". And theme packages choose link to which one will be default.
Created attachment 14099 [details] [review] crash fix for theme-switcher updated to latest modifications in gnome-themes and gnome-icon-theme
*** Bug 105361 has been marked as a duplicate of this bug. ***
Maintainers, please could you review the patches in this bug?
*** Bug 106462 has been marked as a duplicate of this bug. ***
This should really be looked at if it's still relevant.
Hmm. I noticed who had filed this bug now :) I haven't seen this in *ages*. Anyone else seeing it still?
I reckon this is fixed, but there are patches so I don't want to close it.
But if it's fixed I guess the patches aren't really relevant? There were some stability improvements in CVS the other day too. Can people please retry with the latest from CVS or released tarball? (2.2.x)
A variant of this patch was applied later.