Should udev have created /dev/modem?
by Willem Riede
I plugged in my U.S.Robotics 56000 Voice USB Modem. When I booted fc5-devel,
"/dev/ttyACM0" was created for it. But no symbolic link to /dev/modem.
Should that have happened? If that is up to me to configure, what is the right
way to do so on fc5?
Thanks, Willem Riede.
18 years, 3 months
vmware player direct3d
by Peter Quiring
I've recently downloaded vmware player from vmware.com and it works with
Fedora Core 5 test1.
Using some sample images from their website (Browser-Appliance) I
managed to edit the vmx file and make it boot from a cdrom and installed
Win2000Pro (had to change the scsi type from lsilogic to buslogic). It
works! 100% free! I no longer have to boot to Windoze for "some" apps.
Then I learned that they have Direct3D support as well (many claim that
even WC3 works) but it doesn't work for me with FC5.
vmware says it can't create a 3d object and disables 3d support. I've
downloaded BillardGL from sourceforge compiled and run it without any
problems so I know my GL system is intact. The only issue I've noticed
is that GL apps don't seem to restore my resolution after exiting
(that's another unrelated issue). But vmware 1.0.1 doesn't see that GL
is available. I've also posted this at vmware's discussion forums.
Of course Wine is a better option, and once working again with X11R7 I
will give it a try again, but vmware has it's advantages too.
Any ideas?
Peter
18 years, 3 months
texdoc and TEXDOCVIEW_pdf
by Michael A. Peters
texdoc is a shell script from tetex.
By default for viewing pdf files that are part of the tex documentation,
it wants to use xpdf.
xpdf is no longer installed by default, so a user is required to either
install it - or set the TEXDOCVIEW_pdf environmental variable themselves
to something else (like evince) - otherwise they get an xpdf command not
found error.
This would be nice to resolve for FC5 if it can be properly done.
Would it be better to install something in /etc/profile.d/ that sets
TEXDOCVIEW_pdf to something like gnome-open?
That should still allow the a user to put in their own TEXDOCVIEW_pdf if
they wanted to, but would use their default pdf reader if they didn't
set anything themselves.
This may need to be done for more than just the TEXDOCVIEW_pdf
environmental variable, but I think something needs to be done to avoid
texdoc complaining about a lack of xpdf when a perfectly good PDF viewer
(such as evince or Acrobat) is installed. It also seems like it should
use the users preferred application for viewing PDF files.
Only potential issue I can think of would be KDE users etc. who don't
have gnome-open installed. A standard wrapper that uses the appropriate
viewer for the desktop being used would be better, if such a thing
exists (since tetex is not a gnome app).
18 years, 3 months
re: kudzu illegal instruction at startup
by Gianluca Cecchi
>You may want to look at the bug I posted on this:
>
>Bugzilla - 176490
Strange enough, yesterday before posting I searched through bugzilla with
kudzu as the key,
but fewer results appeared than today, and nothing in particular for the seg
fault thing. Thanks.
I updated this bug number with my debug information.
Gianluca
18 years, 3 months
re: kudzu illegal instruction at startup
by RaXeT
On Thu, Dec 29, 2005 at 07:45:03AM +0100, Gianluca Cecchi wrote:
> hello,
> when kudzu starts it gives
>
> Checking for hardware changes/etc/rc.d/init.d/kudzu: line 23: 3937 Illegal
> instruction /sbin/kudzu $KUDZU_ARGS
> [FAILED]
You may want to look at the bug I posted on this:
Bugzilla - 176490
18 years, 3 months
kudzu illegal instruction at startup
by Gianluca Cecchi
hello,
when kudzu starts it gives
Checking for hardware changes/etc/rc.d/init.d/kudzu: line 23: 3937 Illegal
instruction /sbin/kudzu $KUDZU_ARGS
[FAILED]
in "start" function there is
/sbin/kudzu $KUDZU_ARGS -z || /sbin/kudzu $KUDZU_ARGS
instead of typical
/usr/sbin/kudzu $KUDZU_ARGS -t 30
if [ "$RETVAL" -eq 5 ]; then
echo -n $"Hardware configuration timed out."
echo -n $"Run '/usr/sbin/kudzu' from the command line to
re-detect."
initlog -n kudzu -s $"Hardware configuration timed out."
initlog -n kudzu -s $"Run '/usr/sbin/kudzu' from the command
line to re-detect."
fi
Any reason for that?
Gianluca
18 years, 3 months
Can't upgrade to Rawhide
by Igor Jagec
I'm trying to upgrade to Rawhide from fully updated FC4 on my test
partition for quite some time, but I never manage to succeed it. Yum
constantly give me some errors, such as:
...
--> Running transaction check
--> Processing Conflict: initscripts conflicts kernel < 2.6.12
--> Processing Conflict: kudzu conflicts kernel < 2.6.13
--> Processing Dependency: python = 2.4.1 for package: smart
--> Restarting Dependency Resolution with new changes.
--> Populating transaction set with selected packages. Please wait.
--> Running transaction check
--> Processing Conflict: initscripts conflicts kernel < 2.6.12
--> Processing Conflict: kudzu conflicts kernel < 2.6.13
--> Processing Dependency: python = 2.4.1 for package: smart
--> Restarting Dependency Resolution with new changes.
--> Populating transaction set with selected packages. Please wait.
--> Running transaction check
Error: Missing Dependency: python = 2.4.1 is needed by package smart
Error: Unable to satisfy dependencies
Error: Package initscripts needs kernel < 2.6.12, this is not available.
Error: Package kudzu needs kernel < 2.6.13, this is not available.
Error: Package smart needs python = 2.4.1, this is not available.
------------------------------------------------------------------------
Is there any way to fix that manually? Is it possible to access to
Rawhide via SmartPM? If so, what channel would that be?
Thanks, cheers!
--
Igor Jagec
18 years, 3 months
Re: kudzu illegal instruction at startup
by Gianluca Cecchi
>What CPU are you running on ? Did you move disks between two
>different machines by any chance ?
The machine is a dual athlon Mp.
I previously installed fc5 devel and further updates on a vmware workstation
virtual machine on this host.
Then I dumped all files on a preformatted partition of this host.
Then with chroot, grub setup, dev files creation, mkinitrd re-run and other
conf file modifications I am able to boot natively in fc5.
I expected kudzu to help doing this reconfiguration....
>Do you get the same thing when you run it by hand from the command line ?
>If so, running it under gdb (with the kudzu-debuginfo package installed)
>should give you a backtrace of where the problem occured.
Yes it gives the same error.
I'm going to install kudzu-debuginfo and let you know...
Thankls,
Gianluca
18 years, 3 months
rawhide report: 20051229 changes
by Build System
Updated Packages:
GFS-kernel-2.6.14.1-20051219.162641.FC5.6
-----------------------------------------
* Wed Dec 28 2005 Jesse Keating <jkeating(a)redhat.com>
- rebuilt
cman-kernel-2.6.14.1-20051219.162641.FC5.5
------------------------------------------
* Wed Dec 28 2005 Jesse Keating <jkeating(a)redhat.com>
- rebuilt
dlm-kernel-2.6.14.1-20051219.162641.FC5.5
-----------------------------------------
* Wed Dec 28 2005 Jesse Keating <jkeating(a)redhat.com>
- rebuilt
gnbd-kernel-2.6.14.0-20051108.134753.FC5.12.28
----------------------------------------------
* Wed Dec 28 2005 Jesse Keating <jkeating(a)redhat.com>
- rebuilt
kernel-2.6.14-1.1796_FC5
------------------------
* Wed Dec 28 2005 Dave Jones <davej(a)redhat.com>
- 2.6.15-rc7-git2
- Reduce stack usage in block layer.
- Limit number of serial ports registered at boot time.
- Fix spinlock corruption in cfq with ub.
selinux-policy-2.1.6-18
-----------------------
* Wed Dec 28 2005 Dan Walsh <dwalsh(a)redhat.com> 2.1.6-18
- Fix /dev/ub[a-z] file context
Broken deps for i386
----------------------------------------------------------
jacorb - 2.2-3jpp_3fc.i386 requires libgcj.so.6
jonas - 4.3.3-1jpp_15fc.i386 requires libgcj.so.6
jonas-examples - 4.3.3-1jpp_15fc.i386 requires libgcj.so.6
Broken deps for ia64
----------------------------------------------------------
jacorb - 2.2-3jpp_3fc.ia64 requires libgcj.so.6()(64bit)
rgmanager - 1.9.31-3.ia64 requires ccs
Broken deps for ppc
----------------------------------------------------------
cman - 1.0.3-5.FC5.ppc requires cman-kernel-modules >= 0:2.6.11
dlm - 1.0.0-7.FC5.ppc requires dlm-kernel-modules >= 0:2.6.11
gnbd - 1.0.1-2.ppc requires gnbd-kernel-modules >= 0:2.6.11
jacorb - 2.2-3jpp_3fc.ppc requires libgcj.so.6
jonas - 4.3.3-1jpp_15fc.ppc requires libgcj.so.6
jonas-examples - 4.3.3-1jpp_15fc.ppc requires libgcj.so.6
Broken deps for ppc64
----------------------------------------------------------
cman - 1.0.3-5.FC5.ppc64 requires cman-kernel-modules >= 0:2.6.11
dlm - 1.0.0-7.FC5.ppc64 requires dlm-kernel-modules >= 0:2.6.11
emacs - 21.4-5.ppc64 requires fonts-xorg-75dpi
gnbd - 1.0.1-2.ppc64 requires gnbd-kernel-modules >= 0:2.6.11
jacorb - 2.2-3jpp_3fc.ppc64 requires libgcj.so.6()(64bit)
Broken deps for s390
----------------------------------------------------------
jacorb - 2.2-3jpp_3fc.s390 requires libgcj.so.6
jonas - 4.3.3-1jpp_15fc.s390 requires libgcj.so.6
jonas-examples - 4.3.3-1jpp_15fc.s390 requires libgcj.so.6
systemtap - 0.5.2-2.s390 requires kernel >= 0:2.6.9-11
systemtap - 0.5.2-2.s390 requires kernel-devel
Broken deps for s390x
----------------------------------------------------------
jacorb - 2.2-3jpp_3fc.s390x requires libgcj.so.6()(64bit)
libvte-java - 0.11.11-7.s390x requires libgtkjava-2.8.so()(64bit)
libvte-java - 0.11.11-7.s390x requires libgtkjni-2.8.so()(64bit)
systemtap - 0.5.2-2.s390x requires kernel >= 0:2.6.9-11
systemtap - 0.5.2-2.s390x requires kernel-devel
Broken deps for x86_64
----------------------------------------------------------
jacorb - 2.2-3jpp_3fc.x86_64 requires libgcj.so.6()(64bit)
jonas - 4.3.3-1jpp_15fc.x86_64 requires libgcj.so.6()(64bit)
jonas-examples - 4.3.3-1jpp_15fc.x86_64 requires libgcj.so.6()(64bit)
18 years, 3 months
fsck after crash not active in fc5 dev?
by Gianluca Cecchi
Hello,
I noticed after a crash (due to loss of electricity) that file systems was
not asked to be fscked
after reboot. The file /.autofsck is created at boot, as before, but
comparing rh el 4 with fc5 it seemes that inside rc.sysinit the piece below
is missing, so that
the system does not prompt if one wants to force fsck and one has to
manually create /forcefsck and reboot in this case
(or probably is there a boot flag to tell to force fsck? I don't know this).
if [ "$PROMPT" != "no" ]; then
if [ "$AUTOFSCK_DEF_CHECK" = "yes" ]; then
if /sbin/getkey -c $AUTOFSCK_TIMEOUT -m $"Press N
within %d seconds to not force file system integrity check..." n ; then
AUTOFSCK_OPT=
fi
else
if /sbin/getkey -c $AUTOFSCK_TIMEOUT -m $"Press Y
within %d seconds to force file system integrity check..." y ; then
AUTOFSCK_OPT=-f
fi
fi
echo
else
# PROMPT not allowed
if [ "$AUTOFSCK_DEF_CHECK" = "yes" ]; then
echo $"Forcing file system integrity check due to
default setting"
else
echo $"Not forcing file system integrity check due
to default setting"
fi
fi
I'm testing fedora right now and don't know previous versions, so this could
also be standard policy for fedora. Sorry in this case.
Is this ok, due to devel os or did it disappear uncorrectly?
I would like to have this chance also for devel/test systems.
Just my opinion.
Thanks for reading.
Gianluca
18 years, 3 months