A little over a year ago I asked about packaging Firefox extensions and
was told things would be better in Firefox 1.1. We now have 1.5. Are
we ready to start packaging Firefox extensions?
System Administrator 303-415-9701 x222
NWRA/CoRA Division FAX: 303-415-9702
3380 Mitchell Lane orion(a)cora.nwra.com
Boulder, CO 80301 http://www.cora.nwra.com
Just booted to kernel-xen-2.6.17-1.2617.2.1.fc6 and after a few minutes
I noticed my cursor in X would switch between normal, non-visible, and
graphically corrupt, then after about 15-20 minutes my X locked up
requiring a restart. I've booted to kernel-2.6.17-1.2617.2.1.fc6 and
have been running X for over three hours now without this problem.
Anybody else seen this ?
I'm latest rawhide :
Using an "Intel Corporation 945G/GZ Express Integrated Graphics
Controller (rev 02)".
I'm posting here first since I'm not sure against what to file this
bug. Maybe someone has the same issue...
Since this week-end at least (I hadn't rebooted my laptop in a few
days), after booting my laptop and logging in through gdm, I see the
GNOME slash screen, then I see lots of disk I/O going on for a few
seconds, but then once the splash screen went away but before anything
else appears on the screen (no wallpaper, no panel, only the solid blue
background)... complete freeze. Impossible to switch back to a console
or even ping the system.
The weird thing is that if I first open a failsafe session, disable
desktop effects with the "desktop-effects" tool, log out, log back into
GNOME and re-enable desktop effects once logged in, everything works
fine, wobbly and all.
Anyone else seeing this? Does this seem more of an X issue, a GNOME
issue or a compiz issue? :-/
My laptop has an ATI Mobility 9600 (r300) and I'm using the default
free radeon driver.
Clean custom Red Hat Linux rpm packages : http://freshrpms.net/
Fedora Core release 5.91 (FC6 Test2) - Linux kernel
2.6.17-1.2617.2.1.fc6 Load : 0.42 0.29 0.32
I have a problem with rebuilding the TinyERP package for Fedora Extras.
It needs to start a X server during the build process, so there is a
"BuildRequires" dependency on xorg-x11-server-xvfb. The X server needs a
default font "fixed", so there is a "BuildRequires" dependency on
xorg-x11-fonts-base. When you look into the fonts-base package than the
font "fixed" is defined in the /usr/share/X11/fonts/misc/fonts.alias
file. This was sufficient at least until the last build of this package
on July 24. In current Fedora Devel there is a problem because the font
"fixed" cannot be found. So what happened to the font "fixed"?
You can find the build logs at
TinyERP maintainer for Fedora Extras
The docs don't mention anything about it but I'm wondering if yum can
handle this sort of functionality :
rpm -Uvh --oldpackage http://kickstart/blah/name
Right now it seems yum can install a specific version of a package, but
it can't 'rollback' to a version if that RPM is already installed.
I'm guessing that's pretty easy for yum to do so maybe it already can
and I'm missing something?
With recent 3 or 4 kernels suspend to disk stopped working. When
suspending the process completes and the power button light blinks, This
is normal on my box when suspended. But now the botton blinks once and
the video comes back with the mouse and keyboard dead.
Support for the Mylex or LSILogic DAC960PG is broken on Fedora Core 5.
It works fine on Redhat 9 and FC4 but Anaconda just hangs on loading the DAC960 driver.
I am wondering what I can do to get find if this is a known issue and as to whether it will or has been fixed in FC6 ?
Should I post it as a bug on bugzilla.
Many thanks in advance,
(not much luck on fedora-extras-list, reposting here)
Short version: are 'undefined-non-weak-symbol' rpmlint warnings
acceptable when you have dynamic libraries circular dependencies within
a project ?
When a dynamic library links with symbols from another dynamic library,
you'll typically add the dependency on a LDADD line in Makefile.am, so
that the resulting .so file will show the other library explicitely in
its 'ldd' listing, and everyone is happy. Now the LDADD line is
optional: without it the dynamic library has undefined symbols that will
be resolved at run-time, albeit with a (small?) performance cost.
Now when the 2 dynamic libraries each link with each other symbols, you
cannot have a circular dependency in automake, so one of the 2 libraries
will have undefined symbols. And rpmlint will complain about that. Under
the circumstances, can the rpmlint warning be considered
false-positives, or is there a way to patch the automake files to
prevent the undefined symbols ?
For reference, this affects