You are not logged in.
That's not a problem here though since totem isn't required by any installed apps.
it's required by totem-plugin.
as for me so far, no big crash. I just had some wierd stuff happening in my session right after updating, e.g gdm crashed on logout, and then soe .ICEauth files were left, and I couldn't run the session, it was complaining about some non-writable .ICEauth file, which was plain bullshit. cleaned up, and ready set go, I was fine.
what's more, gnome was consistently complaining about at-spi not being installed on session startup, I silenced it by installing at-spi.
edit:
'ln -s /usr/lib/libtotem-plparser.so /usr/lib/libtotem-plparser.so.1' makes brasero start again.
what's more, this release is supposed to obsolete pam_keyring isn't it?
Last edited by lloeki (2007-09-03 20:58:36)
To know recursion, you must first know recursion.
Offline
That's not a problem here though since totem isn't required by any installed apps.
it's required by totem-plugin.
Right, the "here" meant "on my computer"
what's more, gnome was consistently complaining about at-spi not being installed on session startup, I silenced it by installing at-spi.
Oh, I forgot to forgot that issue.
I only reported the first issue with at-spi that even when installed, gnome still didn't detect it.
JGC fixed that quickly, but I forgot that at-spi wasn't originally installed, and gnome indeed still complains about at-spi not being there on a fresh gnome install.
So the question I wanted to ask is : why is at-spi not part of the gnome group ? Or, why doesn't a gnome package (the one that complains about at-spi not being installed) depend on at-spi ?
pacman roulette : pacman -S $(pacman -Slq | LANG=C sort -R | head -n $((RANDOM % 10)))
Offline
at-spi is required when you use accessibility. To use accessibility, you need tools like gok or orca. Both have gail and at-spi as dependencies.
Offline
The issue if I recall with at-spi, is that gnome 2.20 enables accessibility by default, and might change your setting to enable it. If you start gnome, and uncheck accessibility, it should hopefully not bug you about it.
Offline
Accessibility is enabled by default in development releases, that's why it's enabled here.
Offline
hi! I have a problem with metacity! it doesn't start
(~) walala $ metacity --replace
Advertencia del gestor de ventanas: Los arreglos para aplicaciones rotas se han deshabilitado. Algunas aplicaciones podrían no comportarse correctamente.
depth: 24
Advertencia del gestor de ventanas: Log level 16: Using INDIRECT rendering because LIBCM_DIRECT is not defined
Advertencia del gestor de ventanas: Log level 16: Using non-TFP rendering because LIBCM_TFP is not defined
metacity: symbol lookup error: /usr/lib/libcm.so.7: undefined symbol: glXGetProcAddress
Offline
libcm has been compiled against mesa and libgl-dri from testing. You need these packages from testing. In case you use nvidia while getting this error, you'll have to recompile libcm from source or recompile metacity without compositor, as nvidia doesn't provide the required symbols in their libGL implementation.
The sources are at http://www.archlinux.org/~jgc/gnome/desktop/metacity and http://www.archlinux.org/~jgc/hnome/external/libcm/
Offline
Hmm, it seems this symbol is available since mesa 6.5.3 or so, which is in extra. No need for the testing version in that case. The missing symbols are part of the GLX 1.4 API, which is not a final API yet. Since it's not final, ATI and Nvidia don't implement it. I'll add some code to libcm to fix this issue, found something on http://www.nabble.com/Re%3A-kwin_compos … l#a7188856 that could be useful.
In this case, to get the current version of libcm working on nvidia or ATI, there's a change needed on line 43 of src/wspixmap.c
Change:
gpointer f = glXGetProcAddress ((const guchar *)name);
into:
gpointer f = glXGetProcAddressARB ((const guchar *)name);
Offline
we're on a bumpy road
$ gnome-system-log
**clickety clickety**
(gnome-system-log:7080): Gtk-CRITICAL **: _gtk_tree_view_find_path: assertion `node != NULL' failed
Segmentation fault
Last edited by lloeki (2007-09-05 19:51:26)
To know recursion, you must first know recursion.
Offline
about gnome keyring pam integration, I found details here http://live.gnome.org/GnomeKeyring/Pam
it works like a charm, and has many more features than teh oldy pam_keyring, so that one from AUR is definitely obsoleted!
as for the current 'bugs', currently pam_gnome_keyring.so lies in /usr/lib... in the gnome-keyring PKGBUILD it should be moved to install into /lib/security, and upon install some comments should be added about the lines to add in the various pam.d files, and a link to the aforementioned page as for the behavior of the 'login' keyring, and other various details...
Last edited by lloeki (2007-09-05 21:34:10)
To know recursion, you must first know recursion.
Offline
Thanks for pointing out about the pam module, couldn't get it to work here, looks like it was just in the wrong directory afterall.
About the logviewer crash: that's a GTK bug. I'm waiting for a new GTK release that will fix the logviewer crash and the recent files monitoring.
Offline
The panel applets don't work for me.
This is in my xsession-errors:
** (gnome-panel:29329): WARNING **: panel-applet-frame.c:1278: failed to load applet OAFIID:GNOME_WindowMenuApplet:
Failed to resolve, or extend '!prefs_key=/apps/panel/applets/window_menu_screen0/prefs;background=none:;orient=down;size=x-small;locked_down=false
...for every applet I added to the panel.
When I start /usr/lib/gnome-applets/<applet> manually and add it to the panel while it's running, it works.
When I then press ctrl+c in the terminal running the applet it quits and gnome-panel tells me that
the applet quit unexpectly and asks me whether i want it reloaded. Answering yes leads to the
same error dialog I always have to click away at the start of session asking me whether to keep the applet in the panel or to remove it.
I tried deleting all gnome and gconf settings. Everything was reset to its defaults but the applets still don't work.
The rest of your gnome repo works fine for me. Thanks for updating!
Last edited by tradiaz (2007-09-06 12:24:04)
Offline
It's working now. No idea why....
Offline
some new findings. wanted to try some applet stuff today:
$ gdesklets
Starting gdesklets-daemon...
Connecting to daemon [ ### ]
(process:7488): GLib-GObject-WARNING (recursed) **: invalid unclassed pointer in cast to `GtkObject'
aborting...
Connecting to daemon [ ### ]
and never connects.
To know recursion, you must first know recursion.
Offline
After experiencing my applet problem some more times now, i found the problem:
When logging out from the session bonobo-activation-server doesn't quit. When starting a new session
there are 2 bono-activation-servers. Killing the old one makes applets work.
Offline
JGC, if you're building new nautilus packages, could you add the patches that make nautilus use xds if they don't include them in the new nautilus themselves?
bug report: http://bugzilla.gnome.org/show_bug.cgi?id=171655
Offline
JGC, if you're building new nautilus packages, could you add the patches that make nautilus use xds if they don't include them in the new nautilus themselves?
bug report: http://bugzilla.gnome.org/show_bug.cgi?id=171655
Same request here. This should solve the Drag and Drop from file-roller to nautilus.
Offline
Sure, I'll add it, it's quite hilarious that the GNOME archive manager can't interact with the GNOME filemanager actually.
Offline
Hi, since last update of gnome to 2.19, the fonts in the menu of thunderbird and firefox are quite small than the rest of the system... the are some way to solve this issue?
tks
Offline
c0r0nel, check for DPI settings discreptancies in:
- Gnome's preferences -> appearance -> fonts -> details
- firefox
- X
for X, take a peek at
xdpyinfo | grep resolution
I had this problem under KDE, because QT uses the DPI info provided by X (which in turn uses EDID information provided by nvidia driver) which is ~108 DPI on my lappy, but GTK uses 96 by default, which can be modified, and firefox uses GTK value in some places and its own damn settings in some others. I never managed to get them all to the real value, so I just force set everyone to 96 (well I make nvidia ignore EDID, flat out lie and report 96 to everyone), and screw WYSIWYG.
in your case maybe GTK uses X DPI info now, and firefox makes it 96 regardless. who knows.
Last edited by lloeki (2007-09-13 09:47:12)
To know recursion, you must first know recursion.
Offline
Thanks lloeki, the value was 84. I increase it to 96. Now the fonts in firefox and thunderbird have the same size that the rest of gnome but more bigger. Then i decrease the font size in preferences and all look very nice again.
sorry for my english.
Offline
JGC, care to upload your pkgbuilds? I'd really like to patch a few with ubuntu's patches, and I have no idea other than version changes from your pkgbuilds to those in extra.
Offline
JGC, care to upload your pkgbuilds? I'd really like to patch a few with ubuntu's patches, and I have no idea other than version changes from your pkgbuilds to those in extra.
Any chances you missed that post : http://bbs.archlinux.org/viewtopic.php? … 41#p277041 ?
pacman roulette : pacman -S $(pacman -Slq | LANG=C sort -R | head -n $((RANDOM % 10)))
Offline
Yes I did, thanks shining.
Offline
Today I see that hal stopped auto mounting drives. Anyone have the same problem?
in /dev/null no one can hear your scream...
Turkish Archlinux & KDEmod Mirror Administrator
http://server.elsistech.com/archlinux/ ~ http://server.elsistech.com/archlinux/kdemod/
Offline