Re: [expert] How do I get ext3 journal available on initial mount of /?
On Fri, Sep 19, 2003 at 06:53:41PM -0400, Bill Mullen wrote: > Yes, that is precisely what initrd images are for; they contain whichever > module(s) will be required to mount the root filesystem. This is also why > you can't just use an initrd from a different kernel version, as modules > must be compiled for that exact kernel to be able to be loaded at all. Thanks. I have been wondering for years why we needed the initrd's. > Not to sound contentious, but the behavior you describe is hardly the sort > of thing that I would describe as booting "fine" ... :-/ You may have point :-) /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] How do I get ext3 journal available on initial mount of /?
la On Fri, Sep 19, 2003 at 04:17:50PM -0400, Peter Møller Neergaard wrote: > > You might want to give us a little more info about why you think > > that ext3 is not working correctly. > > Because there is a difference between bringing back / and the other > mount points: Tourned out that the journal was missing (it must have been deleted at one point): when the journal is missing - initrd.img complains about file system trouble - the kernel apparently just mounts the file system as ext2, but claims that it is ext. I recreated the journal and will now embrace myself for the next crash where it hopefully will recover smoothly. /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] How do I get ext3 journal available on initial mount of /?
Thanks for the responses. On Fri, Sep 19, 2003 at 02:05:36PM -0400, Bryan Phinney wrote: > AFAIK, support for Ext3 and ReiserFS, both journaling file systems are built > into the Mandrake kernel, either integrated or as modules that are available > at boot. It is clearly a module on my version of Mandrake 9.1: pan:~ > lsmod | grep ext3 ext3 64608 3 (autoclean) jbd48692 3 (autoclean) [ext3] If I mount a Reiser drive I also get a reiserfs module. Of course this does not preclude the modules from being available at boot time. When the modules resides in /lib/modules/... they cannot be loaded before I have mounted / and that cannot happen before the files system has been checked. Is that what the initrd images in /boot are for? If so this explains my trouble. Somehow the process of upgrading to kernel 2.4.21-0.25mdk went haywire. I ended up with a useless initrd image and decided that it was easiest just to ignore it since it boots fine without. > You might want to give us a little more info about why you think > that ext3 is not working correctly. Because there is a difference between bringing back / and the other mount points: When mounting / (my /dev/hda6) e2fsck shows a slowly progressing progress bar. On the other partitions (/dev/hda8, /dev/hda9, and /dev/hda10) it simply says ``Recovering journal. As you can see they are all ext3: pan:/boot > mount /dev/hda6 on / type ext3 (rw,noatime) ... /dev/hda8 on /usr/local type ext3 (rw,noatime) /dev/hda9 on /usr/src type ext3 (rw,noatime) /dev/hda10 on /home type ext3 (rw,noatime) /Peter PS. I essentially use the strategy suggested by James: I do not press Y for recovery, let it run e2fsck, reboot with ctrl-d, and then it boots correctly the second time around. -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] How do I get ext3 journal available on initial mount of /?
I am running Mandrake on laptop. Unfortunately my APM suspend is not flawless so occasionally my laptop powers off when I try to suspend. Obviously this results in unsynchronized file systems and a check at boot time. For that reason I have formatted all my partitions (including /) as ext3 to have a journal to recover from. This does however not work for / where es2fsck is scanning the whole disk looking for errors. I presume that it is because ext3 is not build into the Mandrake kernel and thus have to be loaded as a module. Consequently the file system will only be mounted as ext2 and an old type scan will take place. One possible solution would of course be to compile the kernel with ext3 build. I would however prefer a solution where I could just use Mandrake's kernel. Is this possible? Thanks /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Hard disk error: how to locate from sector #, how to proceed?
I have had my laptop for a year and a half and now the first hard disk error has appeared. My syslog gives me ``nice'' messages like the following (repeated): Aug 22 07:22:51 pan kernel: hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error } Aug 22 07:22:51 pan kernel: hda: read_intr: error=0x01 { AddrMarkNotFound }, LBAsect=23887713, sector=850440 I have gone through the syslog and the sectors that are reported bad are the following: LBAsect=11615376, sector=3679200 LBAsect=11615376, sector=3679203 LBAsect=23236658, sector=199376 LBAsect=23236658, sector=199385 LBAsect=23236670, sector=199397 LBAsect=23887713, sector=850440 LBAsect=23887721, sector=850448 LBAsect=23889817, sector=852544 which appears relatively confined. Using diskdrake it appears that the first LBA sector is on /dev/hda6 (it listed it's starting sector as 7,936,173 with /dev/hda7 starting at 18,217,773) and the rest are on /dev/hda10 (which starts at 23,037,273). My question now is how do I proceed? Should I backup the data, reformat the disks and then restore the backup. Thanks /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Another Windoze feature in 9.1: kernel freezes routinely
Thanks for suggestions for various stress tests---I will try them tonight. To answer James Sparenberg's question: I have tried both the 0.13mdk and the 0.18mdk. I usually use NeTraverse' win4lin enabled kernel, but have reverted to the standard Mandrake kernels to see if this is the problem. I do not use the SMP kernels. I would be slightly surprised if it was a hardware problem, given that I have run Mandrake 9.0 for 3--4 months and 8.1 for about a year without problems of this sort. For the same reason, X is not too likely to be the culprit as I did not upgrade my low-level video driver when upgrading to 9.1 (I'm using a locally compiled version of Nvidia's proprietary driver). I'll keep you posted on results. /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Another Windoze feature in 9.1: kernel freezes routinely
I have now been running Mandrake 9.1 with the 2.4.21-0.18mdk kernel for about 3 weeks. At this point it is starting to be annoying that this kernel locks up more often than even Micro$oft Windoze. The lock up will happen anything from 5 minutes to 10 hours of boot. It must be the kernel locking up since there is no response to the SysRq+Alt+... keys. This happens routinely, but irregular, so I have no idea how to track it. I tried maximizing the information to syslog by choosing *.* /var/log/syslog in /etc/syslog.conf. A typical entry looks like this: Jul 6 14:42:18 pan spamd[5914]: identified spam (9.3/5.0) for turtle:501 in 0.4 seconds, 3321 bytes. Jul 6 14:45:00 pan CROND[5928]: (turtle) CMD (/usr/sbin/anacron -t $HOME/bin/shell/cron/anacrontab) Jul 6 14:45:00 pan anacron[5929]: Anacron 2.3 started on 2003-07-06 Jul 6 14:45:00 pan anacron[5929]: Normal exit (0 jobs run) Jul 6 14:47:12 pan syslogd 1.4.1: restart. Jul 6 14:47:12 pan /etc/hotplug/net.agent: invoke ifplugd eth1 Jul 6 14:47:12 pan ifplugd[1526]: Using interface eth1/00:02:2D:40:D0:92 Jul 6 14:47:12 pan ifplugd[1526]: ETHTOOL_GLINK failed: Operation not supported which means that I have booted around 14:46:45. Thus there does not appear to be any programs running just before the lock up. So at this point I would like suggestions: - how can I get more debug information from the kernel - should I change to a different kernel, e.g., vanilla 2.4.21. Or should I consider one of the patched ones. Thanks /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Installing auto_fs on devfsd system causes pam_console of login to hang
I am running Mdk 9.1. and recently I tried to install auto_fs to have automounting of my cdrom and floppy. I used the following auto.master /vol/etc/auto.vol --timeout=3 and the following auto.vol to configure the /vol directory cdrom-fstype=iso9660,ro,uni_xlate=1,codepage=850,umask=0,nosuid,nodev :/dev/cdrom floppy -fstype=vfat,uni_xlate=1,sync,codepage=850,umask=0,nosuid,nodev :/dev/fd0 After installing the automounting, my login (both terminal and KDE) hanged when logging in as an ordinary user. Tracing the pam output it turned out that it was hanging when pam_console is changing the permissions of console devices. I am normally using devfsd (since that is standard). If I disabled it the login would go through. Also, if I run it in debug mode (where it according to the output does not do anything) it would also go through: it would get and successfully process the requests (but not do anything physically). So somehow it appears to be a compatibility issue between devfsd and auto_fs. Does anybody know what is causing this? Anybody who has a working installation with devfsd and auto_fs? Thanks /Peter In case you need this is my devfsd.conf # $Id: devfsd.conf,v 1.6 2002/11/11 05:06:34 turtle Exp $ # Sample /etc/devfsd.conf configuration file. # Richard Gooch <[EMAIL PROTECTED]>17-FEB-2002 # # Enable full compatibility mode for old device names. You may comment these # out if you don't use the old device names. Make sure you know what you're # doing! REGISTER.* MKOLDCOMPAT UNREGISTER .* RMOLDCOMPAT # You may comment out the above and uncomment the following if you've # configured your system to use the original "new" devfs names or the really # new names #REGISTER ^vc/MKOLDCOMPAT #UNREGISTER ^vc/RMOLDCOMPAT #REGISTER ^pty/ MKOLDCOMPAT #UNREGISTER ^pty/ RMOLDCOMPAT #REGISTER ^misc/ MKOLDCOMPAT #UNREGISTER ^misc/ RMOLDCOMPAT # You may comment these out if you don't use the original "new" names REGISTER.* MKNEWCOMPAT UNREGISTER .* RMNEWCOMPAT # Enable module autoloading. You may comment this out if you don't use # autoloading LOOKUP .* MODLOAD # Uncomment the following if you want to set the group to "tty" for the # pseudo-tty devices. This is necessary so that mesg(1) can later be used to # enable/disable talk requests and wall(1) messages. REGISTER^pty/s.*PERMISSIONS -1.tty 0600 REGISTER^pts/.* PERMISSIONS -1.tty 0600 # # Uncomment this if you want permissions to be saved and restored # Do not do this for pseudo-terminal devices REGISTER^pt[sy] IGNORE CREATE ^pt[sy] IGNORE CHANGE ^pt[sy] IGNORE DELETE ^pt[sy] IGNORE REGISTER.* COPY/lib/dev-state/$devname $devpath CREATE .* COPY$devpath /lib/dev-state/$devname CHANGE .* COPY$devpath /lib/dev-state/$devname DELETE .* CFUNCTION GLOBAL unlink /lib/dev-state/$devname RESTORE /lib/dev-state # # Uncomment this if you want the old /dev/cdrom symlink REGISTER^cdroms/cdrom0$ CFUNCTION GLOBAL mksymlink $devname cdrom UNREGISTER ^cdroms/cdrom0$ CFUNCTION GLOBAL unlink cdrom REGISTER^v4l/video0$CFUNCTION GLOBAL mksymlink v4l/video0 video UNREGISTER ^v4l/video0$CFUNCTION GLOBAL unlink video REGISTER^radio0$CFUNCTION GLOBAL mksymlink radio0 radio UNREGISTER ^radio0$CFUNCTION GLOBAL unlink radio # ALSA stuff #LOOKUP snd MODLOAD ACTION snd # Uncomment this to let PAM manage devfs REGISTER.* CFUNCTION /lib/security/pam_console_apply_devfsd.so pam_console_apply_single $devpath # Uncomment this to manage USB mouse #REGISTER ^input/mouse0$ CFUNCTION GLOBAL mksymlink $devname usbmouse #UNREGISTER ^input/mouse0$ CFUNCTION GLOBAL unlink usbmouse REGISTER^input/mice$CFUNCTION GLOBAL mksymlink $devname usbmouse UNREGISTER ^input/mice$CFUNCTION GLOBAL unlink usbmouse # If you have removable media and want to force media revalidation when looking # up new or old compatibility names, uncomment the following lines # SCSI NEWCOMPAT /dev/sd/* names LOOKUP ^(sd/c[0-9]+b[0-9]+t[0-9]+u[0-9]+)p[0-9]+$ EXECUTE /bin/dd if=$mntpnt/\1 of=/dev/null count=1 # SCSI OLDCOMPAT /dev/sd?? names LOOKUP ^(sd[a-z]+)[0-9]+$ EXECUTE /bin/dd if=$mntpnt/\1 of=/dev/null count=1 # IDE NEWCOMPAT /dev/ide/hd/* names LOOKUP ^(ide/hd/c[0-9]+b[0-9]+t[0-9]+u[0-9]+)p[0-9]+$ EXECUTE /bin/dd if=$mntpnt/\1 of=/dev/null count=1 # IDE OLDCOMPAT /dev/hd?? names LOOKUP ^(hd[a-z])[0-9]+$ EXECUTE /bin/dd if=$mntpnt/\1 of=/dev/null count=1
Why is KDE better? (Re: [expert] 9.1 / XFree86 4.3: choosing compose map)
Playing around a little more, I realized that it works under KDE, but not under WindowMaker and IceWM. That leaves me more puzzled---I thought that this configuration worked at the level of the X server and thus should be independent of the window manager. What is it that KDE does that other Window Managers fail to do? Thanks /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] 9.1 / XFree86 4.3: choosing compose map
I am having a US keyboard but would like my Multi_Key to choose ligatures from the Latin 1 Compose map. How do I accomplish this? I have tried including the following in /etc/sysconfig/i18n SYSFONTACM=iso15 LC_TELEPHONE=en_US.ISO-8859-1 LC_CTYPE=en_US.ISO-8859-1 LANGUAGE=en_US.ISO-8859-1 LC_MONETARY=en_US.ISO-8859-1 LC_ADDRESS=en_US.ISO-8859-1 LC_COLLATE=en_US.ISO-8859-1 LC_NAME=en_US.ISO-8859-1 LC_PAPER=en_US.ISO-8859-1 LC_NUMERIC=en_US.ISO-8859-1 SYSFONT=lat0-16 LC_TIME=en_US.ISO-8859-1 LC_MEASUREMENT=en_US.ISO-8859-1 LANG=en_US.ISO-8859-1 LC_MESSAGES=en_US.ISO-8859-1 LC_IDENTIFICATION=en_US.ISO-8859-1 I can see that I get the right settings after starting X, but I do not get any ligatures. I have also tried adding the following lines to /usr/X11R6/lib/X11/locale/compose.dir iso8859-1/Compose C iso8859-1/Compose en_US.ISO-8859-1 iso8859-1/Compose en_US to force a specific map. That has been in vain as well. Any help is appreciated. /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Wine and 9.1 (Re: [expert] crossover plugin?)
On Tue, Apr 01, 2003 at 02:05:59PM -0500, JOHAM,DAVID (HP-Boise,ex1) wrote: > Crossover (and wine in general) are currently broken on all distros that use > glibc 2.3.x with the native pthreads stuff. I just want to make sure that I get this right: It is currently not possible to run wine under 9.1 In particular, the RPM package of Codeweavers Wine available from MandrakeClub is useless?? Thanks /Peter -- http://www.linearity.org/turtle/contact.html ``If we're an arrogant nation, they'll resent us. If we're a humble nation but strong, they'll welcome us.''--George W. Bush Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Multi-profile Linux
On Wed, Mar 12, 2003 at 04:48:46PM -0800, James Sparenberg wrote: > How have any of you solved this? I agree with your experience: Mandrake does not provide nice support for roaming with your laptop (perhaps Starbucks is not widespread enough in France :-) ). What makes it even more depressing is that Linux pcmcia package actually comes pretty much with the tools you need through the command cardctl. However, Mandrake insists on having only one setting of the encryption key. So here is my solution: 1) I have installed the pcmcia-cs package 2) I have replaced /etc/sysconfig/network-scripts/ifup-wireless with the attached script which incorporates a few bits from pcmcia-cs package 3) I have changed /etc/pcmcia/wireless.opts to include settings for two schemes: a default unencrypted and my home zone. This is done with the following setting: # Lucent Wavelan IEEE (+ Orinoco, RoamAbout and ELSA) # Melco/Buffalo Networks WLI-PCM-L11 # Note : wvlan_cs driver only, and version 1.0.4+ for encryption support home,*,*,00:60:1D:*|home,*,*,00:02:2D:*) INFO="Wavelan IEEE example (Lucent default settings)" ESSID="any" MODE="Managed" KEY="------XX open" ;; *,*,*,00:60:1D:*|*,*,*,00:02:2D:*) INFO="Wavelan IEEE example (Lucent default settings)" ESSID="any" MODE="Managed" KEY="off" ;; After this, all I need to do to change between the two keys are cardctl scheme home or cardctl scheme default as root and then of course ifdown, ifup on the card. This is as seamless as I have been able to get it. I can however still envy Micro$oft user's the ability for the driver to recognize the name of the network and set the key accordingly. /Peter -- http://www.linearity.org/turtle/contact.html ``Power tends to corrupt, and absolute power corrupts absolutely. Great men are almost always bad men''---Lord Acton, 1887 #!/bin/bash # Network Interface Configuration System # Copyright (c) 1996-2002 Red Hat, Inc. all rights reserved. # # Based on PCMCIA wireless script by (David Hinds/Jean Tourrilhes) # # This software may be freely redistributed under the terms of the GNU # public license. # # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. # # Configure wireless network device options. See iwconfig(8) for more info. # Valid variables: #MODE: Ad-Hoc, Managed, etc. #ESSID: Name of the wireless network #NWID: Name of this machine on the network. Hostname is default #FREQ: Frequency to operate on. See CHANNEL #CHANNEL: Numbered frequency to operate on. See FREQ #SENS: Sensitivity threshold for packet rejection. #RATE: Transfer rate. Usually one of Auto, 11, 5, 2, or 1. #KEY: Encryption key for WEP. #RTS: Explicit RTS handshake. Usually not specified (auto) #FRAG: Fragmentation threshold to split packets. Usually not specified. #SPYIPS: List of IP addresses to "spy" on for link performance stats. #IWCONFIG: Extra parameters to pass directly to IWCONFIG #IWPRIV: Extra parameters to pass directly to IWPRIV # Only meant to be called from ifup. # Implement the scheme of cardctl taken from the pcmcia package # The following is copy-and-paste from pcmcia/shared and is used to obtain # various parameters for the card grep_stab () { # this should be cheaper than invoking "grep" local CLASS DEV while read SOCKET CLASS DRIVER INSTANCE DEV MAJOR MINOR ; do if [ "$1" = "$DEV" ] ; then return 0 ; fi done return 1 } get_info () { if [ -d /var/lib/pcmcia ] ; then STAB=/var/lib/pcmcia/stab else STAB=/var/run/stab fi if [ -f /var/lib/pcmcia/scheme ] ; then SCHEME=`cat /var/lib/pcmcia/scheme` elif [ -f /var/run/pcmcia-scheme ] ; then SCHEME=`cat /var/run/pcmcia-scheme` fi if [ -z "$SCHEME" ] ; then SCHEME="default" ; fi grep_stab $1 < $STAB || usage } WI_HWADDR=`/sbin/ifconfig $DEVICE | sed -n -e 's/.*addr \([^ ]*\) */\1/p'` get_info $DEVICE ADDRESS="$SCHEME,$SOCKET,$INSTANCE,$WI_HWADDR" if [ -r ./wireless.opts ] ; then . ./wireless.opts else . /etc/pcmcia/wireless.opts fi echo "$MODE # $KEY # $ESSID" # Mode need to be first : some settings apply only in a specific mode ! if [ -n "$MODE" ] ; then iwconfig $DEVICE mode $MODE fi # This is a bit hackish, but should do the job right... if [ -n "$ESSID" -o -n "$MODE" ] ; then NICKNAME=`/bin/hostname` iwconfig $DEVICE nick "$NICKNAME" >/dev/null 2>&1 fi # Regular stuff... if [ -n "$NWID" ] ; then iwconfig $DEVICE nwid $NWID fi if [ -n "$FREQ" -a "$MODE" != "Managed" ] ; then iwconfig $DEVICE freq $FREQ elif [ -n "$CHANNEL" -a "$MODE" != "Managed" ] ; then iwconfig $DEVICE channel $CHANNEL fi if [ -n "$SENS" ] ; then iwconfig $DEV
[expert] Suggestions on backing up to a hard disk
I have a laptop with a 20Gb hard disk. I have just acquired an 80Gb external ieee1394 harddisk that I will use for backup and storage of some rarely used files. I would appreciate some advises on putting the new space to use in the best way: - on backing up is it most advantageous to copy the whole disk image (dd ...), copy the file structure (cp -R ...), or do some synchronization (unison, rsync)? - what about special files (devices, etc) if I just do it / - are the any of the file system formats that are more suited for this use (than standard ext2)? Thanks Peter -- http://www.linearity.org/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Latest updates (mainly kde) killed mouse dragging after suspend-resume; anybody else?
I am using Mandrake 9.0 on a Dell Inspiron laptop. Yesterday I used MandrakeUpdate to update my system with the updates that has been posted the last couple of days. After the update I experienced the following problem: after suspending and then resuming my laptop, I would no longer be able to drag a window (i.e., click on the left button in the title bar and then move the window or using meta-left). It worked before suspending so by rebooting the machine I would again be able to drag windows (until the first suspend, of course). This happened under both KDE and WindowMaker. It is very likely that this is caused by the updates. This morning I downloaded the original RPMS (from the 9.0 distribution) and downgraded to those RPMS. Now dragging no longer disappears after disabling. The packages that I have reinstalled are the following: arts-1.0.3-7mdk.i586.rpm dhcpcd-1.3.22pl1-3mdk.i586.rpm ftp-client-krb5-1.2.5-1mdk.i586.rpm kdeaddons-3.0.3-6mdk.i586.rpm kdeadmin-3.0.3-10mdk.i586.rpm kdeartwork-3.0.3-6mdk.i586.rpm kdebase-3.0.3-68mdk.i586.rpm kdebase-devel-3.0.3-68mdk.i586.rpm kdebase-nsplugins-3.0.3-68mdk.i586.rpm kdeedu-3.0.3-7mdk.i586.rpm kdegames-3.0.3-7mdk.i586.rpm kdegraphics-3.0.3-11mdk.i586.rpm kdegraphics-devel-3.0.3-11mdk.i586.rpm kdelibs-3.0.3-30mdk.i586.rpm kdelibs-devel-3.0.3-30mdk.i586.rpm kdemultimedia-3.0.3-7mdk.i586.rpm kdemultimedia-aktion-3.0.3-7mdk.i586.rpm kdemultimedia-devel-3.0.3-7mdk.i586.rpm kdenetwork-3.0.3-15mdk.i586.rpm kdenetwork-devel-3.0.3-15mdk.i586.rpm kdepim-3.0.3-4mdk.i586.rpm kdesdk-3.0.3-5mdk.i586.rpm kdetoys-3.0.3-3mdk.i586.rpm kdeutils-3.0.3-6mdk.i586.rpm krb5-libs-1.2.5-1mdk.i586.rpm libarts-1.0.3-7mdk.i586.rpm libcups1-1.1.16-0.4mdk.i586.rpm libldap2-2.0.25-7mdk.i586.rpm licq-kde-1.2.0a-4mdk.i586.rpm telnet-client-krb5-1.2.5-1mdk.i586.rpm xpdf-1.01-4mdk.i586.rpm Has anybody else experienced a similar problem? Any suggestions to which package my be the offender? Thanks in advance Peter -- http://www.linearity.org/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] I'm sorry, but in KDE how do I remove the title bar for an application
I'm really sorry for asking such an apparently simple question which shouldn't really need an expert advise. However, I've been trying for an hour I just cannot figure out how to specify to KDE that I don't want the title bar for a specific application. I can't really believe that it should matter, but I do run Mandrake 9.0 and I have the following KDE packages installed: kdesdk-3.0.5a-1.1mdk kdevelop-2.1.3-6mdk kdelibs-devel-3.0.5a-1.1mdk kdenetwork-3.0.5a-1.1mdk kde-i18n-en_GB-3.0.3-2mdk kdeadmin-3.0.5a-1.1mdk kdemultimedia-3.0.5a-1.1mdk kdeutils-3.0.5a-1.1mdk kdebase-nsplugins-3.0.5a-1.1mdk kdepim-3.0.5a-1.1mdk kdegraphics-devel-3.0.5a-1.1mdk kdebase-3.0.5a-1.1mdk kdenetwork-devel-3.0.5a-1.1mdk kdegraphics-3.0.5a-1.1mdk kdebase-devel-3.0.5a-1.1mdk kdelibs-3.0.5a-1.1mdk kdetoys-3.0.5a-1.1mdk kdemultimedia-devel-3.0.5a-1.1mdk Truly, thanks in advance /Peter -- http://www.linearity.org/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Opera and Java: which RPMs?
I would like touse Opera as web browser. I am however having trouble getting the it to work smoothly with java. Whenever I try to start opera I get the following message: pan:~ > opera -debugjava java: There seems to be a preloaded version of Xt. Typical causes for this are the LD_PRELOAD environment variable or that Opera indirectly depends on libXt.so Java will probably crash Opera. How to modify setup (example) SUN Java installed at: /opt/java/jre1.3.1_04 export OPERA_JAVADIR=/opt/java/jre1.3.1_04/lib/i386/ export LD_PRELOAD=${OPERA_JAVADIR}/libawt.so export LD_LIBRARY_PATH=${OPERA_JAVADIR}:${OPERA_JAVADIR}/client/ IBM Java installed at: /opt/java/IBMJava2-131 export OPERA_JAVADIR=/opt/java/IBMJava2-131/bin export LD_PRELOAD=${OPERA_JAVADIR}/libawt.so export LD_LIBRARY_PATH=${OPERA_JAVADIR}:${OPERA_JAVADIR}/client/ java: Disabling java due to potential dangers. If you know what you are doing, you can set the environment variable OPERA_FORCE_JAVA_ENABLED to '1' to override this. I have tried the following: - follow the descriptions above. It still resulted in the error message. - I have tried kaffe (1.0.7-1mdk) and j2re (1.4.0_01-1mdk) as my java runtime environment. - I have tried opera downloaded from opera's home page (6.10-20021029.4-shared-qt) as well as the version from MandrakeClub (6.03-0.static.1mdk). The latter is staticly linked and does not issue the error above, instead the opera stalls when trying to run java. Does anybody have a opera and Java working together. If so, what are the correct combination of RPMs? Thanks in advance Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Mdk 9.0: Gnome Control Center and URL Handler; where, what RPM?
I know this a FAQ: I am using Evolution under KDE and what to change the default browser, i.e., the browser used to open an URL. I've found a number of answers to this on the net, and they all suggest that I change the ``Default URL handler'' using Gnome Control Center. However, when I run gnome-control-center, I cannot find any setting for Default URL handler. I thought that I might be missing a package so I installed all packages containing gnome and gtk. Still I do not have the URL handler option. So where is the setting, or what is the package that I need to be able to change the browser used by Evolution. Thanks in advance Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] MDK 9.0 introduced occasional crashes of the X; how to debug?
After upgrading from Mandrake 8.1 to Mandrake 9.0, I'm now regularly experiencing (like every 2--3 day) experiencing that my X server crashes. I have checked the logs in /var/log and the only two lines are a one from kdm saying that the session ended unexpectedly and one from pam saying that the session terminated. Unfortunately, a new X server is launched immediately so I cannot see what was written last in /var/log/XFree.0.log. What can I do to trace the cause better? About my system: I am running on a Dell Inspiron 4100 laptop. I have a Nvidia GeForce2Go card. I experience the crashes both with Nvidia's nvidia driver and with the open source nv driver. So it is not so likely that it is a problem with the driver. I am now running XFree86-4.2.1-3mdk. Before upgrading from 8.1 to 9.0, I ran XFree86-4.2.0-10mdk. Should I try to downgrade? Are there other suggestions? Thanks in advance Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Mdk 9.0: Gnome Control Center and URL Handler; where, what RPM?
I know this a FAQ: I am using Evolution under KDE and what to change the default browser, i.e., the browser used to open an URL. I've found a number of answers to this on the net, and they all suggest that I change the ``Default URL handler'' using Gnome Control Center. However, when I run gnome-control-center, I cannot find any setting for Default URL handler. I thought that I might be missing a package so I installed all packages containing gnome and gtk. Still I do not have the URL handler option. So where is the setting, or what is the package that I need to be able to change the browser used by Evolution. Thanks in advance Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] MDK 9.0 introduced occassional crashes of the X; how to debug?
After upgrading from Mandrake 8.1 to Mandrake 9.0, I'm now regularly experiencing (like every 2--3 day) experiencing that my X server crashes. I haved checked the logs in /var/log and the only two lines are a one from kdm saying that the session ended unexpectedly and one from pam saying that the session terminated. Unfortunately, a new X server is launched immediately so I cannot see what was written last in /var/log/XFree.0.log. What can I do to trace the cause better? About my system: I am running on a Dell Inspiron 4100 laptop. I have a Nvidia GeForce2Go card. I experience the crashes both with Nvidia's nvidia driver and with the open source nv driver. So it is not so likely that it is a problem with the driver. I am now running XFree86-4.2.1-3mdk. Before upgrading from 8.1 to 9.0, I ran XFree86-4.2.0-10mdk. Should I try to downgrade? Are there other suggestions? Thanks in advance Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] kbdrate broken in 9.0
This is not necessarily a MDK9.0 issue. With MDK8.1, I had to use input pipe from /dev/console in a script I run after resuming my laptop. In other words I have something like kbdrate -s -r 20 -d 250 < /dev/console /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] hotplugging, pcmcia: too many modules loaded, corrupt modules.pcmcia or corrupt card?
I have just got a PCMCIA ieee1394 card which I am trying to get to work with Mandrake 8.1 and a self-compiled kernel 2.4.19. That made me discover the hotplug functionality from the hotplug package. So far I have been naïve enough to believe that the cardbus hotplugging was controlled just by cardctl (from kernel-pcmcia-cs) as the latter handles my wireless card. Now I have two questions: 1) First, is correct that I should have both the hotplug and cardctl? How do they play together 2) Second, and more seriously: when I insert the firewire card my display gets corrupted. From the syslog it appears to be because the kernel tries to load some framebuffer modules. Here is the the relevant part of the syslog Sep 26 17:42:14 pan kernel: cs: cb_alloc(bus 7): vendor 0x104c, device 0x8019 Sep 26 17:42:14 pan kernel: PCI: Enabling device 07:00.0 ( -> 0002) Sep 26 17:42:14 pan /sbin/hotplug: arguments (pci) env (PWD=/etc/hotplug PCI_CLASS=C0010 HOSTNAME=pan.prescott ACTION=add DEBUG=yes MACHTYPE=i586-mandrake-linux-gnu OLDPWD=/ SHLVL=1 PCI_SUBSYS_ID=104C:8010 PCI_SLOT_NAME=07:00.0 SHELL=/bin/bash HOSTTYPE=i586 OSTYPE=linux-gnu HOME=/ PCI_ID=104C:8019 TERM=dumb PATH=/bin:/sbin:/usr/sbin:/usr/bin _=/usr/bin/env) Sep 26 17:42:14 pan /sbin/hotplug: invoke /etc/hotplug/pci.agent () Sep 26 17:42:14 pan /etc/hotplug/pci.agent: pcimodules is scanning more than ... Sep 26 17:42:14 pan /etc/hotplug/pci.agent: Modprobe and setup ohci1394 3c59x rivafb i810_audio i810_rng usb-uhci for PCI slot Sep 26 17:42:14 pan kernel: ohci1394: $Rev: 530 $ Ben Collins <[EMAIL PROTECTED]> Sep 26 17:42:14 pan kernel: PCI: Setting latency timer of device 07:00.0 to 64 Sep 26 17:42:14 pan kernel: ohci1394_0: OHCI-1394 1.0 (PCI): IRQ=[11] MMIO=[f500-f5000800] Max Packet=[2048] Sep 26 17:42:14 pan kernel: rivafb: RIVA MTRR set to ON Sep 26 17:42:14 pan kernel: Console: switching to colour frame buffer device 80x30 Sep 26 17:42:14 pan kernel: devfs_register(0): could not append to parent, err: -17 Sep 26 17:42:14 pan kernel: rivafb: PCI nVidia NV10 framebuffer ver 0.9.3 (GeForce2-Go, 32MB @ 0xE000) Sep 26 17:42:15 pan kernel: i810_rng: RNG not detected Sep 26 17:42:15 pan insmod: /lib/modules/2.4.19-pmn-020913/kernel/drivers/char/i810_rng.o: init_module: No such device Sep 26 17:42:15 pan insmod: Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters Sep 26 17:42:15 pan insmod: /lib/modules/2.4.19-pmn-020913/kernel/drivers/char/i810_rng.o: insmod /lib/modules/2.4.19-pmn-020913/kernel/drivers/char/i810_rng.o failed Sep 26 17:42:15 pan insmod: /lib/modules/2.4.19-pmn-020913/kernel/drivers/char/i810_rng.o: insmod i810_rng failed Sep 26 17:42:15 pan /etc/hotplug/pci.agent: ... can't load module i810_rng Sep 26 17:42:15 pan /etc/hotplug/pci.agent: ... blacklisted module: usb-uhci Sep 26 17:42:15 pan kernel: ieee1394: Host added: Node[00:1023] GUID[000156010276] [Linux OHCI-1394] Now, I can see that the /etc/hotplug/pci.agent tries to some clever tricks to figure out what modules to load. However, I can see that the cleverness does not work in my case since the script ends up deciding to many modules (only the ohci1394 should be needed). Is this due to a corrupt modules.pcimap (included below), or is it because of my card providing wrong information. Any help will be appreciated. Thanks Peter /lib/modules/linux-2.4.19-pmn020913/modules.pcimap: # pci module vendor device subvendor subdevice class class_mask driver_data i810_rng 0x8086 0x2418 0x 0x 0x 0x 0x i810_rng 0x8086 0x2428 0x 0x 0x 0x 0x i810_rng 0x8086 0x2448 0x 0x 0x 0x 0x i810_rng 0x8086 0x244e 0x 0x 0x 0x 0x i810_rng 0x8086 0x245e 0x 0x 0x 0x 0x ohci1394 0x 0x 0x 0x 0x000c0010 0x00ff 0x 3c59x0x10b7 0x5900 0x 0x 0x 0x 0x 3c59x0x10b7 0x5920 0x 0x 0x 0x 0x0001 3c59x0x10b7 0x5970 0x 0x 0x 0x 0x0002 3c59x0x10b7 0x5950 0x 0x 0x 0x 0x0003 3c59x0x10b7 0x5951 0x 0x 0x 0x 0x0004 3c59x0x10b7 0x5952 0x 0x 0x 0x 0x0005 3c59x0x10b7 0x9000 0x 0x 0x 0x 0x000
[expert] LPRng, HP Laserjet 4L: leaving printer with some data in queue
I am using the LPRng printing system under Mandrake 8.1. The reason for this choice is that I could not get CUPS to use an arbitrary command to print: I am running an laptop and want to choose my printer based on my the network (= current IP). I have just added an HP Laserjet 4L printer on my local port using printtool. It installed the printer and can print the test page and other stuff without problems. However, after finishing the print job there is still some data stored in the printer since printer leaves is data light on. The data does not effect the next print job. Does anybody have an idea about what configuration option can remove this small annoyance? Thanks Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] apm-scripts-0.1-3mdk: bug fix in locking X on suspend
The apm-scripts-0.1-3mdk package contains various scripts to be run on suspend. One if them is /etc/sysconfig/apm-scripts/suspend.d/5XFree which includes code to lock the X display on suspend. Unfortunately, the author is under believe that if one user is running KDE, they all are. So if there is a DCOP server running the script tries only DCOP to lock any display. That scheme fails for the obvious reason. The best solution I have found is simply to try using DCOP before using the other, more hacky, methods available. (Testing whether the process specified in .DCOPserver_hostname_display is running, fails because a DCOP server might be around after logout.) The following rewritten script does the locking correct #!/bin/sh # -*- Mode: shell-script -*- # $Id:$ [ -e /etc/sysconfig/apmd ] && . /etc/sysconfig/apmd # Don't try to restore the terminal or lock X if X isn't running... [ "x`pidof X`" != x ] && { [ -n "$CHANGEVT" -a "$CHANGEVT" != "0" ] && chvt 1 [ "x$LOCK_X" = "xyes" -o "x$LOCK_X" = "x1" ] && { # Lock X, based on patch from Hannu Martikka # <[EMAIL PROTECTED]> # PMN: changed to use who rather than w since that lists the display who |awk '{print $1, $2}'|grep -v '-'|grep : | \ sort -u | \ while read line; do Usern=`echo $line |awk '{print $1}'` XDisp=`echo $line |awk '{print $2}'` if [ "x$Usern" = "xroot" ]; then xlock -display $XDisp &>/dev/null & else # PMN: First try to see if we can get a dcopserver (KDE) to respond; # otherwise do something ugly su $Usern -c "env DISPLAY=$XDisp dcop kdesktop KScreensaverIface 'lock()' || xscreensaver-command -display $XDisp -lock || (xscreensaver -display $XDisp -no-splash &sleep 1; xscreensaver-command -display $XDisp -lock) || (xlock -display $XDisp &)" &>/dev/null fi done } } Best Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] XFree86, 4.2, lost Compose
I've recently updated to XFree86 v4.2 (XFree86-4.2.0-10mdk). Since then I have no longer to use my Multi_Key to Compose characters. I have traced the problem so far that it stems from the fact that X chooses the empty compose map in /usr/X11R6/lib/X11/locale/C. Instead, I would like to use for instance the compose map in /usr/X11R6/lib/X11/locale/iso8859-1. As I understand the X(7x) man page the compose map is chosen based on LC_ALL, LC_CTYPE, LANG (in that order). Since the /etc/X11/Xsession script (called by startx through xinit) calls /etc/profile.d/lang.sh, I have tried changing my settings in /etc/sysconfig/i18n to the following SYSFONT=lat1-10 LC_ALL=iso8859-1 LC_CTYPE=iso8859-1 LC_MONETARY=en_US LANGUAGE=en_US:en LC_TIME=en_US LC_NUMERIC=en_US LC_COLLATE=en_US LC_MESSAGES=en_US LANG=iso8859-1 SYSFONTACM=iso15 Still, I end up with the C compose map. Of course, I could fix this by copying the compose map. However, I prefer a solution through configuration. Does anybody know what I have to do? /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Re: cardctl scheme has no effect (and why, MandrakeSoft?)
On June 20, 2002, Jan Lentfer wrote: JL> There is response that says "Changing scheme from xyz to JL> zxy" but the configuration just isn't changed. Is there a JL> different way in MDK to manage different pcmcia network JL> setups. The cardctl scheme is so nice and easy why doesn't it JL> work with MDK? When it comes to PCMCIA network cards, Mandrake (and supposedly also Redhat---see the note ), uses the network settings specified in /etc/sysconfig/network and /etc/sysconfig/network-scripts/ifcfg-*. These script does not take into account the scheme set using cardctl. You can try to compare the configuration files in /etc/pcmcia with configuration files accompaning the pcmcia package (these can for instance be found in the pcmcia-cs-*/etc directory of the kernel sources, e.g., /usr/src/linux-2.4.8/pcmcia-cs-3.1.29). You will notice that everything except the network-script is as provided by the pcmcia-package. Thus schemes actually work out of the box for all other types of PCMCIA cards. As for the script for network cards, it has been replaced with a script that calls the relevant /etc/sysconfig/network-scripts/ifup script. And this script is actually run when you do cardctl scheme But, obviousely, you see no change since the settings are independent of schemes. The configuration of the wireless card is BTW an interesting piece a code under MDK. The card configuration from the pcmcia package is copied in twice! First it can found in /etc/sysconfig/network-scripts/ifup-wireless which is processed when /etc/sysconfig/network-scripts/ifup figures out that we have wireless card. And shortly after the same section appears with all the variables names prefixed by WIRELESS_. I wonder if someone at MandrakeSoft has a good explanation for that duplication (downwards comtability??). Anyways, we can use this to get schemes back, by adding the following to /etc/sysconfig/network-scripts/ifup-wireless: # Implement the scheme of cardctl taken from the pcmcia package # The following is copy-and-paste from a couple of places in # pcmcia/shared and is used to obtain the settings for the card # with the current scheme. grep_stab () { # this should be cheaper than invoking "grep" local CLASS DEV while read SOCKET CLASS DRIVER INSTANCE DEV MAJOR MINOR ; do if [ "$1" = "$DEV" ] ; then return 0 ; fi done return 1 } get_info () { if [ -d /var/lib/pcmcia ] ; then STAB=/var/lib/pcmcia/stab else STAB=/var/run/stab fi if [ -f /var/lib/pcmcia/scheme ] ; then SCHEME=`cat /var/lib/pcmcia/scheme` elif [ -f /var/run/pcmcia-scheme ] ; then SCHEME=`cat /var/run/pcmcia-scheme` fi if [ -z "$SCHEME" ] ; then SCHEME="default" ; fi grep_stab $1 < $STAB || usage } WI_HWADDR=`/sbin/ifconfig $DEVICE | sed -n -e 's/.*addr \([^ ]*\) */\1/p'` get_info $DEVICE ADDRESS="$SCHEME,$SOCKET,$INSTANCE,$WI_HWADDR" if [ -r ./wireless.opts ] ; then . ./wireless.opts else . /etc/pcmcia/wireless.opts fi Best Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] laptop issue, external keyboard/mouse kills mouse driver
I have a laptop from Dell (Inspiron 4100). It has a PS/2 port on the back to connect an external mouse or keyboard. I've been trying to do so, but every time it results in the mouse driver dying. More precisely: I have the laptop running and get mouse events on /dev/psaux from the built-in touch pad. As soon, as I connect either an external mouse or an external keyboard through the PS/2 port, I stop getting input on /dev/psaux. Even when removing the external device, the mouse does not come back. Does anybody have a clue to what is happening or an idea to work around it. Thanks /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] devfs, hdc=ide-scsi (Mdk 8.1) and the symbolic link /dev/cdrom
On February 25, 2002, Larry Sword wrote: LS> Delete the links in /etc/dev and redo the links pointing to the LS> correct /dev/cdroms/ PMN> I don't have a directory or a file named /etc/dev. Did you /dev? LS> No!, I don't, I did have a brain fart. OK. From what package should I get the /etc/dev directory that is missing on my machine? Or should I create it by hand, and if so what links should I put into it. LS> To remove a sylink: rm /dev/cdrom Does not work! As I wrote in my previous mail PMN> peter:/dev > ls -la cdrom ; sudo rm cdrom ; ls -la cdrom PMN> lr-xr-xr-x1 root root 16 Feb 25 11:46 cdrom -> ../cdroms/cdrom0 PMN> lr-xr-xr-x1 root root 16 Feb 25 11:48 cdrom -> ../cdroms/cdrom0 As you can see, I must have been successful on deleting the link (since the time stamp have changed), but it is recreated as soon as I have deleted. So until I figure out how to disable whatever process that recreates link ln, rm, etc will not help me. /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] devfs, hdc=ide-scsi (Mdk 8.1) and the symbolic link /dev/cdrom
On February 25, 2002, you wrote: LS> Delete the links in /etc/dev and redo the links pointing to the LS> correct /dev/cdroms/ I don't have a directory or a file named /etc/dev. Did you /dev? Anyways, it is impossible for me to delete the link: peter:/dev > ls -la cdrom ; sudo rm cdrom ; ls -la cdrom lr-xr-xr-x1 root root 16 Feb 25 11:46 cdrom -> ../cdroms/cdrom0 lr-xr-xr-x1 root root 16 Feb 25 11:48 cdrom -> ../cdroms/cdrom0 and trying -f to ln does not help either: peter:/dev > sudo ln -s -f cdroms/cdrom3 cdrom ln: creating symbolic link `cdrom' to `cdroms/cdrom3': File exists So my assumption was that devfsd really insists on having the link due to a configuration option. /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Xterm replacements with arbitrary keybinding configuration
Which of the xterm replacements allow one to configure the keybindings arbitrarily, e.g., binding Ctrl+space to insert selection? Thanks Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] devfs, hdc=ide-scsi (Mdk 8.1) and the symbolic link /dev/cdrom
I have recently upgraded from 7.2 to 8.1 (better later than never). This has caused me problem with the handling of my CDROM drive and CD-writer on my computer; I have both, the CDROM is /dev/hdc and the CD-writer /dev/hdd. I prefer to have both drives being emulated SCSI using ide-scsi. I therefore have hdc=ide-scsi hdd=ide-scsi among my load options to the kernel. This creates the SCSI devices just fine. However, the symbolic link /dev/cdrom gets wrong; it points to the non-existing /dev/cdroms/cdrom0 rather than /dev/cdroms/cdrom1 or /dev/cdroms/cdrom3. It appears that this link is maintained by devfsd. However, do I tell it to link it to the correct file? /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] bulding the KDE menu, update-menus?
On February 23, 2002, Pesarif wrote: P> IIRC, you are not supposed to use update-menus (don't ask me why). P> Just rm -rf ~/.kde/share/applnk outside of KDE for your user and it P> should be fine when you log into KDE. Thanks, that brought back my menus. Actually, I realized that it is also safe to run update-menus (I presume that is what Menudrake uses when saving the menus). The reason that it didn't work on my system was that I had added my own directory .menu-methods and forgot to put in a method for kde. Gee, do I feel like a dofus. /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] bulding the KDE menu, update-menus?
For some reason I suddenly lost my menustructure for KDE, i.e., the menu now only consist of soffice (which I run recently), Bookmarks, Recent Documents, Quick Browser, Run command ..., Configure, Lock Screen, and Logout. I have checked with menudrake and it still know the full structure. I have run update-menus (as user and as root) to regenerate the menus, but that have been in vain. So now my (probably really stupid) question is: what do I do to regenerate the full menu structure for KDE. Thanks Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Laptop/desktop synchronization
I want to set up synchronization of my laptop with my desktop as a means of backup (in case my laptop gets stolen, hit by a 18-wheeler, etc). Doing some search on the net the following options seemed to emerge, with different pros and cons: - rsync: old, well-tested, but with a clear client/server distinction, should be run explicitly (or through cron) - unison: consider the two copies equal, should be run explicitly - InterMezzo: automatic when connection is available, only whole file systems (at this point), experimental, clearly client/server What experiences does people have with any of these systems (or others that I did not find). Thanks Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] apm-scripts and X locking, bug in etc/sysconfig/apm-scripts/suspend.d/5Xfree86?
I have worked on getting the X display to lock automatically when I use APM on my laptop. I used the standard scripts provided by the package apm-scripts. I run KDE and the locking is thus governed by the following part of the shell-script: w |tail +3 |awk '{print $1, $3}'|grep -v '-'|grep : | \ sort -u | \ while read line; do Usern=`echo $line |awk '{print $1}'` XDisp=`echo $line |awk '{print $2}'` if [ "x`/sbin/pidof 'kdeinit: kdesktop'`" != "x" ]; then # We're running KDE - a dcop call is sufficient. su $Usern -c "DISPLAY=$XDisp dcop kdesktop KScreensaverIface 'lock()'" else When trying to run the script I encountered two problems: - the listing from w do not include my DISPLAY, cf the following example output 1:25am up 7:40, 4 users, load average: 0.58, 0.61, 0.61 USER TTY FROM LOGIN@ IDLE JCPU PCPU WHAT turtle pts/1-12:59am 26:11 0.00s 0.00s /bin/cat so the line was removed by grep -v '-' in the pipe. I solved this problem by using: who |awk '{print $1, $2}'|grep -v '-'|grep : | \ instead. - the call to su produced an error that DISPLAY could not be found as a command. I solved that changing the line to: su $Usern -c "env DISPLAY=$XDisp dcop kdesktop KScreensaverIface 'lock() '" & So now the scripts are working for me, but I wonder if I am the only one who have experienced these problems. /Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] fetchmail problem
On February 4, 2002, skippi wrote: S> Is this a deamon which is suppose to be running in the S> background? Sendmail is not a daemon, but a program that can be invoked everytime you send the mail. It will take care of queueing and relaying the mail appropriately. What is appropriate depends on your system, i.e., are on stand-alone machine you will typically relay all your mail to an SMTP-server, while you on a machine with a mail server will leave the message to the mail server. The beauty of sendmail is that all this is hidden behind the scenes on a well-configured system. I know Mario mentioned sendmail, but if you take a close look at the following snip from the fetchmail man page: As each message is retrieved fetchmail normally delivers it via SMTP to port 25 on the machine it is running on (localhost), just as though it were being passed in over a normal TCP/IP link. The mail will then be delivered locally via your system's MDA (Mail Delivery Agent, usually sendmail(8) but your system may use a different one such as smail, mmdf, exim, or qmail). All the delivery-control mechanisms (such as .forward files) normally available through your system MDA and local delivery agents will therefore work. If no port 25 listener is available, but your fetchmail compilation detected or was told about a reliable local MDA, it will use that MDA for local delivery instead. At build time, fetchmail normally looks for executable procmail(1) and sendmail(1) binaries. you'll notice that the default behavior depends on whether you a running SMTP server or not. You initial message could therefore be caused by an ill-configured SMTP server. I would therefore also check whether you machine is listening on port 25. Best Peter -- http://www.linearity.org/turtle/contact.html ``When you have had all the experiences, met all the famous people, made some money, toured the world and got all the acclaim you still think--is that it? Some might be satisfied--but I wasn't'' -- G. Harrison Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] Have anybody successfully installed the Palm Emulator under 7.2
I have been unsuccessful at compiling the Palm Emulator source from Palm under 7.2. I wondered if anybody has been successful in doing it and could give me hints on where I go wrong. I am currently trying to compile the source version 3.3, but I had similar problems when I tried to compile the version 3.1 source. In both cases configure complains that it cannot find FLTK. I have checked that 7.2 comes with FLTK v. 1.0.9 which are later than the 1.0.5 and 1.0.7 named in the Docs/_Building.txt accompanying the emulator source. The library is installed with libfltk.{a,so} in /usr/lib. In an attempt to solve problems, I have alsodownloaded and build FLTK 1.0.11 with and without shared libraries; in both cases with files installed in /usr/local/lib. I have /usr/X11R6/lib /usr/i486-linux-libc5/lib /usr/local/lib in my /etc/ld.so.conf and have run ldconfig after installing the libraries. Following the instructions in the source bundle, I run ./configure of BuildUnix with ./configure --with-fltk=/usr/local/lib When I do so it fails with the last two lines being checking for 'int fl_height (void)' in -lfltk... no configure: error: *** FLTK *must* be installed before running configure. *** I have looked into the configure script and noticed that the test used is #line 3465 "configure" #include "confdefs.h" #include #include int main() { fl_height() ; return 0; } I can successfully compile that file with the following command line gcc -L /usr/local/lib test2.c++ -lfltk -L/usr/X11R6/lib -lX11 -o test2 provided that only libfltk.a and *not* libfltk.so is installed in /usr/local/lib (I presume that the compiler otherwise picks up the share library over the statically linked library). I am stuck at this point, because I cannot figure what options configure use to compile the test file; it is thus not clear how to approach the problem any further. I hope that somebody have successfully installed the software and can tell me what to do next. Thanks in advance Best Peter -- http://www.linearity.org/turtle/contact.html Her er dit vern mot vold her er dit sverd: Troen på livet vårt, menneskets verd. -- Nordahl Grieg (This is your shield, this is your sword: the belief in life, the value of man) Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] crond
On June 22, 2001, David E Fox wrote: PMN> env -i HOME=$home LOGNAME=$logname PATH=$path SHELL=$shell PMN> /usr/bin/sh -x $* DEF> Or in bash, 'export ' :) No. I used the command env with the -i to make sure to clear the current environment. The whole point of the script is to give you the same environment as when the script is run by cron. Using export gives you the current shell environment plus what you have exported. DEF> Anyway, I think intermixing korn shell and bash (or whatever DEF> Solaris's 'sh' is, probably Bourne-derived) complicates things DEF> unnecessarily. Again, this script was written to exactly mimic the behavior of cron; on solaris the man page explicitly states that cron command is passed as an command argument to sh. I checked the linux man page the other night, and as I recall the default behavior there boiled down to sh again. DEF> AFAIK, it's just another process (at least true of 'cron') so it, DEF> like any other process, it has a completely independent DEF> environment. Such things like PATH can be set explicitly and will DEF> be true of only the current process. True, the only thing to be aware of is that cron's PATH is whatever (probably not that big) path is used at system boot when crond is started. (That is of course unless you use an explicit setting of the PATH in your crontab.) Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
Re: [expert] crond
R> When tasks are running from my crontab It send to my mail next R> message "/bin/bash: root: command not found", but these tasks R> started successfully. Do I understand correctly that you get the error from cron, but not when you try the same commands on the command line? If so, this is most typical because cron sets up a very limited environment for the program, e.g., a very basic path, while you usually has a lot of extra settings in the environment. I use the following little script to test any command I put into crontab: #!/bin/ksh # Set up the right environment to test a shell file for cron home=$HOME logname=`whoami` path=/usr/bin:/usr/sbin:. shell=/usr/bin/sh env -i HOME=$home LOGNAME=$logname PATH=$path SHELL=$shell /usr/bin/sh -x $* One note: I originally wrote if for Sun Solaris. I could not find anywhere in the linux man pages where it is stated what setting PATH will have inside a cron job. So the setting above might be wrong. Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
Re: [expert] crond
R> When tasks are running from my crontab It send to my mail next R> message "/bin/bash: root: command not found", but these tasks R> started successfully. Do I understand correctly that you get the error from cron, but not when you try the same commands on the command line? If so, this is most typical because cron sets up a very limited environment for the program, e.g., a very basic path, while you usually has a lot of extra settings in the environment. I use the following little script to test any command I put into crontab: #!/bin/ksh # Set up the right environment to test a shell file for cron home=$HOME logname=`whoami` path=/usr/bin:/usr/sbin:. shell=/usr/bin/sh env -i HOME=$home LOGNAME=$logname PATH=$path SHELL=$shell /usr/bin/sh -x $* One note: I originally wrote if for Sun Solaris. I could not find anywhere in the linux man pages where it is stated what setting PATH will have inside a cron job. So the setting above might be wrong. Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
Re: [expert] Playmidi reports `/dev/sequencer unknown' with Soundblaster PCI64V
On June 18, 2001, Laura Conrad wrote: LC> As I said, it's probably worth downloading alsaconf and letting it LC> figure this stuff out. For some reason, mandrake didn't include LC> it with the 8.0 alsa stuff. I have done so. The only difference between your setup and mine is the choice of sound card. LC> I'd be surprised, because I had this working on Mandrake 6.x. My LC> guess is that you aren't loading all the right driver modules. LC> Here's what I has on 6.x in /etc/conf.modules: Thanks, I tried downloading it. However, sfxload still gave me an error message that I do not have an AWE device. And then, finally, after trying an evaluation version of the commercial oss, I found the source of all my troubles: though the Ensoniq chip has MIDI support, my specific Card,the AudioPCI, does not support MIDI playback (wave tables) in the hardware. (I would never have guessed that from the manual that to has the usual Windows non-distinction between hardware and software.) So I followed the advice that Oscar wrote on June 18, 2001, : O> On the other hand, If you can't config midi device, you always can O> use timidity to play midi files (it uses midi to wave). It's really O> good!. Thanks! AT LAST IT WORKED! I really appreciate all your help and patience with trying to get the card to work correctly. Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
Re: [expert] Playmidi reports `/dev/sequencer unknown' with Soundblaster PCI64V
I really appreciate your patience with trying to help me with out with this. Unfortunately, I am still not able to get things to work correctly. (It's one of those moments where I consider shifting back to Windoze :-( ) On June 17, 2001, Laura Conrad wrote: LC> My SBLive card uses the Ensoniq 1370 hardware. And I use sfxload LC> to load the soundfonts on it. I remember the documentation not LC> making that very clear, but that is what you do. Interesting! When I using sfxload, I get an error message: peter:~ > sudo sfxload hit7.sf2 No AWE synth device is found My RPM containing sfxload awesfx-0.4.3a-11mdk. Could it make any difference that 8.0 uses awesfx-0.4.3a-12mdk.i586.rpm? LC> Specifically, the disk that came with my card had a directory with LC> soundfonts in it somewhere, and I copied them to my hard drive, LC> and now when I want to use MIDI I say: I searched my disk, and it did not contain any *.sf2 or *.sbk files. So the hit7.sf2 file above, was a random file I downloaded from the net. Could that make any difference? Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
Re: [expert] Playmidi reports `/dev/sequencer unknown' with Soundblaster PCI64V
Thanks for the replies I got from Oscar and Laura. At least, it seems that I am a little closer now. On June 12, 2001, Oscar wrote: O> Is your soundcard MIDI detected? I have a SB Awe 64 and harddrake O> detected the sound device but not the midi port. Maybe your card's O> midi device has not been detected by the install program. It is that this is what have happened to me too: harddrake only reports the sound card, not any midi devices. Are there other ways to test whether MIDI has been detected? O> Try running "sndconfig" as root from console (without X) Also, be O> sure to select the correct midi port in kmid (not the external MIDI O> out port) I tried running sndconfig, but that did not change anything. In particular, kmid could not find any midi devices. I then followed Laura's advice and installed ALSA. Now kmid lists an ``External MIDI 0 MIDI 0-0 - ALSA Device''. However, I can still not get it to produce any sound. On June 12, 2001, Laura Conrad wrote: LC> After I ran alsaconf, I wasn't getting the MIDI errors any more, LC> but I still wasn't hearing the MIDI files playing, so then I LC> remembered that you have to sfxload some soundfonts. I am that same point now. I have installed alsa (it took a while to get this to work as ALSA applies some patches that are already in my Mandrake 7.2 kernel 2.2.17 to all kernels prior to 2.2.18) and usual sound playback works. However, playmidi still does not produce any sound: - playmidi with no options, option -f, option -4, reports ``No playback device found.'' - playmidi with option -e or option -a hangs for a while and then exits. I suppose it is playing the file, but sends the output to a device I cannot hear. I tried following your advice and use sfxload. Unfortunately, that seems only to work with AWE devices, not with my Ensoniq 1370. So is there a special program to load sound fonts on that card? Or could it be something as stupidly simple as the midi playback having its own volume control (I tried unmuting maximizing the volume on all channels using alsamixer). Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
Re: [expert] Playmidi reports `/dev/sequencer unknown' with Soundblaster PCI64V
On June 10, 2001, Larry Sword wrote: LS> Try using a switch with 'playmidi' such as: LS> playmidi -e or playmidi -a ,your midi file> This does solve the problem. I have tried -a, -D, -e, and -4 and all give the same message as in my initial posting. Furthermore, adding up to -v does not give any debugging information to spread any light on the problem. Unfortunately, I know virtually nothing about midi, so I have not tried any of the other options listed by -l LS> Try playmidi -h for all the switch listings. LS> man playmidi for further information. Of course! Given that this is the expert mailing list, I take it for granted that people (myself included) have tried that plus searching google or similar before bothering to write a question. I also tried signing up for the playmidi mailing list, but the mail server responded user unknown; contacting the author has given no response so far. Interestingly enough in my Mandrake 7.2 distribution, the man page is not in sync with program (option -D is not described on the manual page) and -l lists options (-k and -w) which playmidi response to with unknown options. Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
[expert] Playmidi reports `/dev/sequencer unknown' with Soundblaster PCI64V
I am using Mandrake 7.2 and have a SoundBlaster audio PCI64V, which Mandrake's hardware detection picks up as a Ensonic ES1370 (AudioPCI). As far as I can tell from the documentation should support midi. Nevertheless, I get the following error report when I try to run playmidi peter:~ > playmidi sicklysweet1b.midsound.mid Playmidi 2.4 Copyright (C) 1994-1997 Nathan I. Laredo, AWE32 by Takashi Iwai This is free software with ABSOLUTELY NO WARRANTY. For details please see the file COPYING. open /dev/sequencer: No such device When I do ls -la /dev/sequencer I get crw-rw-rw-1 turtle audio 14, 1 Sep 27 2000 /dev/sequencer so I would appear that /dev/sequencer is an existing device. Does anybody know how to solve this problem. Thanks in advance Peter -- http://cs-people.bu.edu/turtle/contact.html ``Deserves death! I daresay he does. Many that live deserve death. And some that die deserve life. Can you give it to them? Then do not be too eager to deal out death in judgement.'' -- Tolkien
RE: [expert] grub: how to specify booting from CDrom?
> Once the BIOS has released the system to the OS, in this case Grub, > you can not bypass or override the boot paramiters set by the BIOS. Correct me if I am wrong, but I thought that booting from the viewpoint of the BIOS consists in reading the Master Boot Record. In my case the Master Boot Record contains grub that then picks (depending on my choice) picks a partition to continuing loading the system (Linux, Windoze, ...). So what you are saying is that it is not possible to ask grub to load the system from the CD ROM (hd2) instead of my usual hard disk (hd0). > In the boot settings most BIOS offer an option to choose the sequence > that will be used to boot your system. I admit that it is a bit paranoid, but I prefer not to have automatic boot from CD ROMs or floppies as a precaution against viruses. Best Peter -- http://cs-people.bu.edu/turtle/contact.html ``Those who forget the past are condemned to repeat it''---Salman Rushdie
[expert] grub: how to specify booting from CDrom?
I am using grub as boot loader. Rather than turning loading from a bootable CD rom on and off in the BIOS, I would prefer just adding an option on my boot menu. I tried various settings in my /boot/grub/menu.lst without success. So if it is possible I would appreciate if somebody could point me to the right solution. Just for reference, I have my CD rom at the primary device on my second IDE controllor (i.e., /dev/hdc). Thanks in advance Peter -- http://cs-people.bu.edu/turtle/contact.html ``Those who forget the past are condemned to repeat it''---Salman Rushdie
[expert] Is it necessary have cron let htdig update its database when not running a web server
After installing Mandrake 7.2, for a good while got the following message DB2 problem...: missing or empty key value specified from running my cron.daily job. I tracked it down to the htdig-dbgen script where htmerge have problems because there are no words for it to index (as I don't have any web pages on my linux box). I have looked at the rpm's and it appears htdig-dbgen is installed as part of the htdig rpm. The dependencies tell that htdig by kdebase and thus part of the standard installation (as I understand it). I presume KDE uses it to generate the indexes of it help pages. So it does not seem that I can remove the htdig package. Now MandrakeSoft did not consider it a support problem that I got errors on a fresh installation, so they referred me to this list. My immediate work around was to make a single web page with a couple of words in /var/www/html/index.html. In this way htmerge always has a couple of words to index and is all happy. However, as I do not run a web server, I thought that I might just simply remove htdig-delete from my cron.daily directory. Is there any reason why I should not do this? Thanks in advance Peter -- http://cs-people.bu.edu/turtle/contact.html ``Those who forget the past are condemned to repeat it''---Salman Rushdie