April 2010 - openSUSE Users - openSUSE Mailing Lists

[opensuse] Linux-compatible GPS anyone ?
by Philippe Andersson 14 Apr '10

14 Apr '10

Hello List, I will have to replace my car's GPS device soon. Could anyone recommend a brand/model that is linux-friendly (and current) ? By that I mean: - maps can be updated from linux (or without a computer) - POIs can be uploaded and downloaded from linux Thanks in advance for sharing your experiences. Ph. A. -- *Philippe Andersson* Unix System Administrator IBA Particle Therapy | Tel: +32-10-475.983 Fax: +32-10-487.707 eMail: pan(a)iba-group.com <http://www.iba-worldwide.com> -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

4 6

[opensuse] Need Help building png2qrgb -- qt-mt not found -- but I'm looking at it??
by David C. Rankin 14 Apr '10

14 Apr '10

Guys, I'm stuck with another lack of knowledge on how adjust the ./configure options to get png2qrgb to build. The configure error complaints that it cannot find the library qt-mt, but I have it installed. I'm using as specific as possible a configure cli to make sure the libraries are found, but it gives the same error as running just './configure': ./configure --prefix=/opt/kde3/ \ --with-qt-dir=/usr/lib/qt3/ \ --with-qt-libraries=/usr/lib/qt3/lib/ <snip> checking for libz... -lz checking for libpng... -lpng -lz -lm checking for libjpeg6b... no checking for libjpeg... -ljpeg checking for Qt... configure: error: Qt (>= Qt 3.1 (20021021)) (library qt-mt) not found. Please check your installation! For more details about this problem, look at the end of config.log. Make sure that you have compiled Qt with thread support! So, 'locate qt-mt': 10:48 alchemy:~/dt/kde3/apps/bld/png2qrgb-0.3> locate qt-mt /usr/lib/libqt-mt.so.3 /usr/lib/qt3/lib/libqt-mt.so.3 /usr/lib/qt3/lib/libqt-mt.so.3.3 /usr/lib/qt3/lib/libqt-mt.so.3.3.8 /usr/lib/qt3/lib64/libqt-mt.la /usr/lib/qt3/lib64/libqt-mt.prl /usr/lib/qt3/lib64/libqt-mt.so /usr/lib/qt3/lib64/libqt-mt.so.3 /usr/lib/qt3/lib64/libqt-mt.so.3.3 /usr/lib/qt3/lib64/libqt-mt.so.3.3.8 /usr/lib64/libqt-mt.so.3 /usr/lib64/pkgconfig/qt-mt.pc Err -- I know this is simple, but it doesn't take much to stump me :-( Anybody see a solution? -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

3 2

[opensuse] Problem Starting NMBD Service on Samba 3.5 : symbol lookup error: /usr/sbin/nmbd: undefined symbol: _talloc_free
by Masim "Vavai" Sugianto 14 Apr '10

14 Apr '10

Having update Samba 3.2 on openSUSE 11.1 into Samba 3.5 on OBS. Update has completed sucessfully but nmbd service could not start : # service nmb restart Shutting down Samba NMB daemon Warning: daemon not running. done Starting Samba NMB daemon /usr/sbin/nmbd: symbol lookup error: /usr/sbin/nmbd: undefined symbol: _talloc_free startproc: exit status of parent of /usr/sbin/nmbd: 127 Anyone know how to solve this problem ? -- Best Regards, Masim "Vavai" Sugianto /************************************************************/ Blog (ID) : http://www.vavai.com Blog (EN) : http://www.vavai.net /************************************************************/ -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

2 1

[opensuse] Test version of sysvinit 2.88dsf
by Dr. Werner Fink 14 Apr '10

14 Apr '10

For the lion-heartedout there ... you'll find the new version 2.88dsf of sysvinit at http://download.opensuse.org/repositories/Base:/System:/Test/openSUSE_Facto… from the Changelog of the tar ball: sysvinit (2.88dsf) world; urgency=low [ Petter Reinholdtsen ] * Mention new home on Savannah in README. * Revert change from Fedora/RedHat where the now obsolete command INIT_CMD_CHANGECONS was introduced. Based on feedback and patch from Bill Nottingham. * Adjust makefile to make sure the install directories are created before files are copied into them. * Simplify build rules, based on patch from Mike Frysinger and Gentoo. * Fix minor bug in optimizing of argument parsing. Based on report from jakemus on freshmeat. * Add casts to get rid of compiler warning about signed/unsigned issues. * Change tty handling in init to make sure the UTF-8 flag is not cleared on boot. Patch from Samuel Thibault. * Add Makefile in toplevel directory. * Print usage information when shutdown is used by non-root user. Patch from Mike Frysinger and Gentoo. * Sync shutdown manual page and usage information. Patch from Mike Frysinger and Gentoo. * Fix race condition in utmp writing. Patch from Gil Kloepfer via Mike Frysinger and Gentoo. * Rewrite findtty() in bootlogd to recursively search /dev/ for the correct device, to handle terminal devices for example in /dev/pty/. Patch from Debian. * Make sure bootlogd findpty() returns an error value when it fails to find a usable pty. Patch from Rob Leslie via Debian. * Make sure bootlogd fflush() every line, even if asked not to flush to disk using fdatasync(). Patch from Scott Gifford via Debian. * Add compatibility code to handle old path "/etc/powerstatus" for a while. * Incude definition for MNT_DETACH which is missing in older GNU libc headers. * Do not strip binaries before installing them, to make it easier to get binaries with debug information installed. [ Werner Fink ] * Add the comment from Andrea Arcangeli about the correct place of setting the default childhandler within spawn(). * Make sure that newline is printed out for last(1) even if an utmp record entry is truncated. * Check if utmp not only exists but also is writable and delay writing out of the utmp runlevel record if utmp is not writable. * Be able to find libcrypt also on 64 bit based architectures. * Add option -w to the last command to display the full user and domain names in the output. Patch from Petr Lautrbach. * Add a manual page for utmpdump as this tool is sometimes very useful even if not intended for normal use. * Use paths.h macros for wall * Change path "/etc/powerstatus" to "/var/run/powerstatus" * Detected also removable block devices at halt/reboot to be able to flush data and send them the ATA standby command. This should avoid data loss on USB sticks and other removable block devices. * Flush block devices on halt/reboot if not done by the kernel. * Set SHELL to /bin/sh in the environmant of shutdown. * Retry to write out shutdown messages if interrupted. * pidof/killall5 - make omit pid list a dynamic one. * pidof - provide '-n' to skip stat(2) syscall on network based FS. * init - avoid compiler warnings * init - initialize console by using the macros from ttydefaults.h * init - add the possiblity to ignore further interrupts from keyboard * init - add the possiblity to set sane terminal line settings * sulogin - add the possibility to reset the terminal io * Fix some minor problems * init - enable is_selinux_enabled() to detect selinuxfs * Add fix for Debian bug #536574 -- Can be enabled by -DACCTON_OFF * Add helper program fstab-decode to make it easier to handle /etc/mtab content. Patch by Miloslav Trmac and Fedora. * Add fix for Debian bug #335023 - Make sure TERM is set on FreeBSD. * Add fix for Debian bug #374038 - Make it clear that shutdown -c can only cancel a waiting shutdown, not an active one. * Add note to pidof manual page about the use of readlink(2). Patch by Bill Nottingham and Fedora. * Add PAM patch contrib/notify-pam-dead.patch based on Debian bug #68621, which will add PAM support for programs spawned by init on the console like sulogin. Based on patch by Topi Miettinen. This patch is not applied by default yet while we review its usefullness. It is only helpful for session handling, as sulogin do not use and will not use a PAM conv() function. The current sulogin is able to handle DES as well as MD5, SHA, and Blowfish encrypted passwords due using getpwnam(3). * Move utmp/wtmp before the execvp() in spawn() to be sure to use the correct pid even on a controlling tty * Remaining problem is that the pid of the second fork() for getting a controlling tty isn't that reported by spawn() * Re-enable writting utmp/wtmp for boot scripts * Extend sulogin to support additional encryption algorithms * Re-enable maintenance message of sulogin * Enable the sulogin fallback password check to handle MD5, SHA, and Blowfish encrypted passwords in case of getpwnam(3) fails. * sulogin picking the SELinux context was broken. Patch by Daniel Walsh -- Petter Reinholdtsen <pere(a)hungry.com> Sun Apr 11 11:28:55 CEST 2010 sysvinit (2.87dsf) world; urgency=low * Fix typos and do minor updates in the manual pages. * Correct section of mountpoint(1). * Document -e and -t options for telinit in init(8). * Update address of FSF in the COPYRIGHT file. * Document in halt(8) that -n might not disable all syncing. Patch by Bill Nottingham and Fedora * Adjust output from "last -x". In reboot lines, print endpoint of uptime too. In shutdown lines print downtimes rather than the time between downs. Fix typo in string compare in last.c. Patch by Thomas Hood. * Improve handling of IPv6 addresses in last. Patch from Fedora. * Document last options in usage information, previously only mentioned in the manual page. * Add new option -F to last, to output full date string instead of the short form provided by default. Patch from Olaf Dabrunz and SuSe. * Adjust build rules to make sure the installed binaries are stripped. * Increase the compiler warning level when building. * Fix utmp/wtmp updating on 64-bit platforms. Patch by Bill Nottingham and Fedora. * Avoid unchecked return value from malloc() in utmpdump. Patch from Christian 'Dr. Disk' Hechelmann and Fedora. * Make sure to use execle and no execl when passing environment to the new process. Patch from RedHat. * Correct init to make sure the waiting status is preserved across re-exec. Patch from RedHat. * Correct init to avoid race condition when starting programs during boot. Patch from SuSe. * Allow 'telinit u' in runlevels 0 and 6. Patch from Thomas Hood. * Change install rules to make pidof an absolute symlink. Patch from Thomas Hood. * Improve error message from init if fork() fail. Patch found in Suse. * Add support for SE Linux capability handling. Patch from Manoj Srivastava, adjusted to avoid aborting if SE policy was loaded in the initrd with patch from Bill Nottingham and Fedora. * Add -c option to pidof for only matching processes with the same process root. Ignore -c when not running as root. Patch from Thomas Woerner and Fedora. * Adjust init to terminate argv0 with one 0 rather than two so that process name can be one character longer. Patch by Kir Kolyshkin. * Make sure bootlogd exit with non-error exit code when forking of the child successfully. * Add bootlogd option -s to make it possible to control the use of fdatasync(). Patch from Thomas Hood. * Add bootlogd option -c to tell it to create the log file if it does not exist. Patch from Thomas Hood. * Let bootlogd also look at ttyB* devices to work on HPPA. Patch from Thomas Hood. * Change init to use setenv() instead of putenv, make sure the PATH value is usable on re-exec. Patch from Thomas Hood. * Add usleep in killall5 after killing processes, to force the kernel to reschedule. Patch from SuSe. * Modify pidof to not print empty line if no pid was found. * Modify init and sulogin to fix emergency mode's tty, making sure ^C and ^Z work when booting with 'emergency' kernel option. Patch from Samuel Thibault. * Modify init to allow some time for failed opens to resolve themselves. Patch from Bill Nottingham and Fedora. * Modify init to shut down IDE, SCSI and SATA disks properly. Patches from Sebastian Reichelt, Werner Fink and SuSe. * Modify wall to use UT_LINESIZE from <utmp.h> instead of hardcoded string lengths. Patch from SuSe. * Change wall to make halt include hostname in output. * Change killall to avoid killing init by mistake. Patch from SuSe. * Change killall5 to use the exit value to report if it found any processes to kill. Patch from Debian. * Add option -o opmitpid to killall5, to make it possible to skip some pids during shutdown. Based on patch from Colin Watson and Ubuntu. * Add references between killall5 and pidof manual pages. Patch from Debian. * Modify killall to work better with user space file system, by changing cwd to /proc when stopping and killing processes, and avoiding stat() when the value isn't used. Also, lock process pages in memory to avoid paging when user processes are stopped. Patch from Debian and Goswin von Brederlow with changes by Kel Modderman. * Change shutdown to only accept flags -H and -P with the -h flag, and document this requirement in the manual page. * Change reboot/halt to work properly when used as a login shell. Patch by Dale R. Worley and Fedora. * Let sulogin fall back to the staticly linked /bin/sash if both roots shell and /bin/sh fail to execute. -- Petter Reinholdtsen <pere(a)hungry.com> Sun, 12 Jul 2009 19:58:10 +0200 the new home of the sysvinit project is now at http://savannah.nongnu.org/projects/sysvinit/ Werner -- "Having a smoking section in a restaurant is like having a peeing section in a swimming pool." -- Edward Burr -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

1 0

[opensuse] howto: zypper lock
by Patrick Shanahan 14 Apr '10

14 Apr '10

How do I apply a lock in zypper to only use packman for a particular package/group of packages, libxine1-* ?? zypper al -r Packman -t package libxine1 does not work ??? tks, -- Patrick Shanahan Plainfield, Indiana, USA HOG # US1244711 http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://counter.li.org -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

4 3

[opensuse] Xorg +KDE4.4.2 +FF ==> HANG
by Anton Aylward 14 Apr '10

14 Apr '10

I like KDE4. OK, so it has had bumps; yes 4.{0.1.2} were 'pre-release', but how else can you get feedback and bug reports. I can live with that. I did. But now I have a problem and lock-up, but I'm not convinced its the fault of KDE4, though perhaps KDE4 has uncovered it. This weekend I fixed up a few erroneous repository URLs (thanks Patrick) and finally got updates to Xorg. In fact I got a number of set of updates to Xorg over the period Friday though Sunday. Perhaps that was significant. My problem though that - and still - is that if I run KDE4 it freezes. It locks up. I can SSH to the machine and see that xorg is eating 100% of one of the CPUs - the thread seems to be locked-in. The load factor steadily grows. Oddly the RSS for xorg is small, only about 8% of 2G. I've established this is specific to KDE. I can run icewm and E16 and E17 all day with no problem. I've also wolf-fenced down things with KDE. It seems running Firefox triggers the lock-up. Not a URL, but running firefox. Even the "firefox crashed ..." startup dialogue will freeze the machine once I switch to it. So, how do I know the fault is in Xorg and KDE/F is merely triggering it? I can run in failsafe mode and that awful xorg.conf.install and it doesn't lock up. But the display & fonts are lousy and it doesn't recognise my USB mouse, only the touchpad. If I zap the xorg.conf altogether I get a similar result; it doesn't lock up but things like the right-click don't work properly since there is nothing in the xorg.conf to tell it about mouse. Or for that matter the graphics card's capabilities. And when I look at the /var/log/Xorg.0.log I see that in the absence of an xorg.conf it somehow load up many video drivers instead of the single one that was in the old xorg.conf. I now see [ 1378.656] (II) LoadModule: "fglrx" [ 1378.659] (II) LoadModule: "radeonhd" [ 1378.746] (II) LoadModule: "ati" [ 1378.748] (II) LoadModule: "radeon" [ 1378.749] (II) LoadModule: "fbdev" [ 1378.760] (II) LoadModule: "vesa" [ 1378.805] (II) LoadModule: "fbdevhw" [ 1378.859] (II) LoadModule: "vgahw" [ 1378.890] (II) LoadModule: "fb" [ 1378.893] (II) LoadModule: "ramdac" [ 1378.893] (II) LoadModule: "exa" What gives? Is it just probing? Probably: [ 1378.659] (II) UnloadModule: "fglrx" [ 1378.894] (II) UnloadModule: "radeonhd" [ 1378.894] (II) UnloadModule: "fbdev" [ 1378.894] (II) UnloadModule: "fbdevhw" [ 1378.894] (II) UnloadModule: "vesa" but that still means ati AND radeon AND vgahw AND fb were loaded. I don't understand this. Are they all necessary where only radeon was needed before? I'd really like to reconstruct an xorg.conf that bring back the capability of my USB mouse and the Radeon card. But my attempts so far have all ended in the hang. Actually this mail was interrupted by a hang. I tried alt-tab to another window and the panel came up and things froze. So perhaps the problem with the interaction between KDE4.4 and Xorg goes deeper than it appears. Someone is going to ask why I don't report this as a bug on bugzilla. Well, while I can reproduce it I can't say a lot about the what or why. Certainly not enough to pinpoint the fault to being in KDE as opposed to KDE just triggering a fault in Xorg. And while it seems to be FF that is related, it might just be the interaction with Gnome aspects of FF. Perhaps we need a FF based on Qt :-) But that doesn't explain the problem with compositing ... Or perhaps it does since icewm doesn't do that ... If you have any ideas about how I could isolate the cause of the problem, please let me know. -- Men fear thought as they fear nothing else on earth, more than ruin, more even than death. . . . Thought is subversive and revolutionary, destructive and terrible, thought is merciless to privilege, established institutions, and comfortable habit. Thought looks into the pit of hell and is not afraid. Thought is great and swift and free, the light of the world, and the chief glory of man. -- Bertrand Russell -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

1 0

[opensuse] Automount ext4 from 11.1 install?
by Oddball 13 Apr '10

13 Apr '10

Oddball schreef: > > # modprobe -v ext4dev > > # mount /dev/sd** /mnt (*Name*number) > These two have to be used on every boot if one wants to take a look at an ext4 partition, from my 11.1 install. Only root can mount such a partition, and using only the second command does not work over here. Also, i have to (re)start the filemanager to get it visual, but when mounted my user kan use the partition, and all filemanagers can be used. Not a big deal, but: Does anyone know a simple way to get this automated? > > -- Enjoy your time around, Oddball, aka M9. OS: Linux 2.6.27.19-3.2-default x86_64 Huidige gebruiker: oddball@AMD64x2-sfn1 Systeem: openSUSE 11.1 (x86_64) KDE: 4.2.1 (KDE 4.2.1) "release 103" -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

3 5

[opensuse] Re: AutoFS somewhat complex...
by Joachim Schrod 13 Apr '10

13 Apr '10

Dave Howorth wrote: > Joachim Schrod wrote: > >> Well, I would say that unavailability of the files beneith one's >> home directory usually makes one's system very unusable as well. >> What do you do after login? Play xpat2 (not possible, has been >> removed from openSUSE... ;-))? >> >> I would never trade the convenience of an NFS-based home against >> the minimal chance that I want to log in and the NFS server ain't >> there in this moment... Then I could not work anyhow, the server >> problem has to be fixed first. > > What you do after login is try to diagnose the problem with the server! > It's more difficult to diagnose it if you can't get a terminal running :) If you want to support that use case, you can make *one* user with a local home case in a different place than /home. Since all your files won't be available anyhow, this ain't different than logging in. (The first ls -l in the home directory with sub-dirs mounted per autofs on would try to mount the unavailable NFS server and would run into long timeouts anyhow.) > PPS e.g. Running different firefox environments is more complicated with > an NFS home as opposed to a local home with NFS-mounted subdirectories. I don't understand this use case. What are different firefox environments? -- Different configurations for the same firefox version? That's what firefox's profiles are for, aren't they? -- Installation of different firefox versions? That's what different user ids are for; one uses another GUI session on another console, or uses VirtualBox/VMware. (Or uses ssh w/ X forwarding, if one knows how to start firefox properly.) Why do you need local homes here? It leads to different home directories on different systems and thus different firefox configuration dirs. If you want that you can simply use a 2nd uid, no need to have a whole new host with the added complexity of NFS-mounted subdirs for that. Joachim -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Joachim Schrod Email: jschrod(a)acm.org Roedermark, Germany -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

1 0

[opensuse] Re: AutoFS somewhat complex...
by Joachim Schrod 13 Apr '10

13 Apr '10

Dave Howorth wrote: > Verner Kjærsgaard wrote: >> a) I wanted to use remote/nfs'd storage for /home > > This is a bad idea nowadays when users have graphical logins. Any > problem with NFS means their system becomes unusable. Better to have a > local home and NFS mount the subdirectories (or more likely, NFS-mount > one directory and symlink the various subdirectories to it). Well, I would say that unavailability of the files beneith one's home directory usually makes one's system very unusable as well. What do you do after login? Play xpat2 (not possible, has been removed from openSUSE... ;-))? I would never trade the convenience of an NFS-based home against the minimal chance that I want to log in and the NFS server ain't there in this moment... Then I could not work anyhow, the server problem has to be fixed first. Oh, and concerning the OP: I don't know what his problem is concerning firefox and it's SQLite database on NFS file systems. We have this here and have never encountered any problems or errors. Locking over NFS works; we have 2010 and this ain't NFSv2-over-UDP any more. Just my 0.02 EUR, Joachim -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Joachim Schrod Email: jschrod(a)acm.org Roedermark, Germany -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

4 4

[opensuse] Re: zypper: really no check for expiration of gpg keys?
by Joachim Schrod 13 Apr '10

13 Apr '10

Michael Schroeder wrote: > On Tue, Apr 13, 2010 at 12:16:29AM +0200, Joachim Schrod wrote: >> Meta data in OBS repo-md repositories (i.e., repomd. is usually >> signed with gpg. It seems that zypper does not check expiration of >> used gpg keys. (zypper 1.0.13 on openSUSE 11.1, in case that matters.) >> >> As an example: >> http://download.opensuse.org/repositories/Apache:/MirrorBrain/Apache_openSU… >> has a key that expired at April 1, 2010; i.e., 12 days ago. (The >> key has ID 0xBD6D129A and fingerprint EDDD C98D 96A0 F889 9AB0 7C78 >> 9584 A164 BD6D 129A.) >> >> I would have expected a warning or an error when this repository is >> refreshed, but nothing as such happens. > > Same as with rpm ;-) Good point; but actually I find check of repository meta-data signatures even more important than rpm signatures. RPMs may be validly older, since the software may have not changed for a long time -- I do not expect that to happen for repository meta-data. Joachim -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Joachim Schrod Email: jschrod(a)acm.org Roedermark, Germany -- To unsubscribe, e-mail: opensuse+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse+help(a)opensuse.org

1 0