Re: Cannot execute /bin/bash: Permission denied
David Jarvie wrote: >On Sunday 19 March 2006 18:47, Joey Hess wrote: > > >>David Jarvie wrote: >> >> >>>Permissions on root directories: all have as a minimum, 755. /tmp/ >>>and /var/tmp have 777. >>> >>> >>Have you checked the permissions of / ? Having it not world readable can >>definitly cause this problem. >> >> > >Ah! So obvious when you think of it - yes, that was the cause. But why the >permissions should have changed, I have no idea. > >Thank you for the help. > > Interesting. Wasn't reading that strace carefully enough ;-). All of those successful reads take place before the setuid32() and clone(). Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Cannot execute /bin/bash: Permission denied
Joey Hess wrote: >David Jarvie wrote: > > >>Permissions on root directories: all have as a minimum, 755. /tmp/ >>and /var/tmp have 777. >> >> > >Have you checked the permissions of / ? Having it not world readable can >definitly cause this problem. > > That certainly would be problematic, but I don't think that's the case here. Presumably David would have noticed the inability to read anything after becoming a regular user. Also, bash is able to read various files according to the strace. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Cannot execute /bin/bash: Permission denied
David Jarvie wrote: >My system (debian testing) has suddenly stopped letting ordinary users log >on. Root can still function normally, but any other user receives the above >message whenever they try to log on. I have done some investigation into >the problem, and can rule out the following causes: > >bash corruption: I have reinstalled bash, libc6 and libncurses5. Plus, the >same problem occurs with /bin/csh. > >/etc/passwd and /etc/group look fine. > >Permissions on root directories: all have as a minimum, 755. /tmp/ >and /var/tmp have 777. > >Permissions on /home directories: all are owned by the respective user, and >have 755 permissions. > >Disk space: plenty of space on all volumes. > >Permissions on libraries used by /bin/bash: I did a 'ldd /bin/bash': >linux-gate.so.1 => (0xe000) >libncurses.so.5 => /lib/libncurses.so.5 (0xb7f43000) >libdl.so.2 => /lib/tls/i686/cmov/libdl.so.2 (0xb7f3f000) >libc.so.6 => /lib/tls/i686/cmov/libc.so.6 (0xb7e09000) >/lib/ld-linux.so.2 (0xb7f9c000) >lrwxrwxrwx 1 root root 14 Feb 18 10:35 /lib/tls/i686/cmov/libdl.so.2 -> >libdl-2.3.5.so >-rw-r--r-- 1 root root 9592 Feb 6 23:39 /lib/tls/i686/cmov/libdl-2.3.5.so >lrwxrwxrwx 1 root root 17 Mar 19 17:46 /lib/libncurses.so.5 -> >libncurses.so.5.5 >-rw-r--r-- 1 root root 268396 Oct 14 22:21 /lib/libncurses.so.5.5 >lrwxrwxrwx 1 root root 13 Feb 18 10:35 /lib/tls/i686/cmov/libc.so.6 -> >libc-2.3.5.so >-rwxr-xr-x 1 root root 1262704 Feb 6 >23:39 /lib/tls/i686/cmov/libc-2.3.5.so >lrwxrwxrwx 1 root root 11 Mar 19 17:44 /lib/ld-linux.so.2 -> ld-2.3.5.so >-rwxr-xr-x 1 root root 88168 Feb 6 23:38 /lib/ld-2.3.5.so > > >Other libraries in /lib: not sure what their permissions should be. > >The output of 'strace su david -c /bin/bash' is attached. Perhaps somebody >who understands it can comment. > >Any ideas for fixing this would be extremely welcome. > > > What are the permissions on /dev/log and /etc/localtime? These are the two files that the strace reveals bash as being denied access to. Here are those files on my system: $ ls -lLh /etc/localtime /dev/log srw-rw-rw- 1 root root0 2006-03-16 23:01 /dev/log -rw-r--r-- 1 root root 1.3K 2006-03-01 21:16 /etc/localtime Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Can't use dpkg (and thus apt-get et al.)
[EMAIL PROTECTED] wrote: My suggestion, rather than messing around with editing and such, would be to do: dpkg --clear-avail : to clear the available list, then do: apt-get update : to repopulate it. Later, Seeker Ah, good advice. I was thinking of /var/lib/dpkg/status for some reason, which of course can't be regenerated so easily. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Can't use dpkg (and thus apt-get et al.)
David R. Litwin wrote: An update: # dpkg-reconfigure debconf dpkg-query: parse error, in file `/var/lib/dpkg/available' near line 359695 package `dcgui': field name ` /usr/sbin/dpkg-reconfigure: debconf is not installed But, when I apt-get install debconf debconf is already the newest version. What do you make of this? -- —A watched bread-crumb never boils. —My hover-craft is full of eels. —[...]and that's the he and the she of it. fsck? Your /var/lib/dpkg/available is corrupt. You might have to remove the offending lines. Just remember to save a copy somewhere before making changes. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: virtual package? aptitude says "not a real package".
[EMAIL PROTECTED] wrote: Kumar Appaiah wrote: On my system, apt-cache show emacs-dl-wn gives me black, install says that package is unavailable. Thanks for your response! I'm still puzzled. What's this "ghost" package, then? The "mplayer" package seems to be another example. If they are unavailable, then why do they show up when you type $ aptitude search mplayer for example. What does this mean? It means that there is no such package but that something depends or conflicts with the package. These packages are just referenced by others even though they arent in the archive. This happens for a variety of reasons. [EMAIL PROTECTED]:~$ apt-cache showpkg mplayer Package: mplayer Versions: Reverse Depends: xmms-xmmplayer,mplayer mozilla-mplayer,mplayer 1.0-pre5 Dependencies: Provides: Reverse Provides: -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: converting CHM files
Dave Thayer wrote: On Thu, Sep 22, 2005 at 06:47:42AM -0500, Robert D. Crawford wrote: The chmlib-tools package contains some low-level type programs which can be used to unpack chm and manipulate chm files: test_chmLib - extracts a file out of a chm file extract_chmLib - extracts a chm file into a directory enumdir_chmLib - lists the contents of chm file(s) enum_chmLib - lists the contents of a chm file chm_http - simplistic web server to allow browsing of chm files without extracting them These tools seem to be fairly crude, but you may be able to do somethng useful with them. dt aptitude install xchm HTH. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Hard links to directories
Hubert Chan wrote: On Sun, 4 Sep 2005 15:40:58 +0100 (BST), Max <[EMAIL PROTECTED]> said: Dear All, Is there a way of switching on permission to have hard links to directories? Is there a rationale for disallowing this? The Linux kernel VFS does not allow directory hard links for technical reasons. (i.e. they don't know how to implement it without potentially breaking things or killing performance.) If I remember correctly, this *used* to be possible. A couple of major kernel versions ago I recall being able to create hard links to directories on my ext2 filesystem. Doing so did issue a warning, but other than that I saw no problems. It seems they realized it is unwise to allow this and disabled it altogether; It may have never worked flawlessly to begin with. Does anybody know exactly when this happened and for what reason? I can see why its not necessary or efficient to implement properly, but I am curious about the details of this change. Thanks, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install problem: Kernel panic: No init found. Try passing init= option to kernel.
brodul wrote: First I start linux26. But there was an error . . . <0>Kernel panic Attempted to kill init! Then I just press Enter. And there is an error to: . . . Freeing unused kernel memory: 96 k freed Kernel panic: No init found. Try passing init= option to kernel. So you are just booting the CD.. that is odd. Have you tried with a different CD? Which version of Debian are you trying to install? Where did you get the CD, did you download it from debian.org? If so, which image did you download? -- Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: changelog names are not displayed properly
kamaraju kusumanchi wrote:
Jan C. Nordholz wrote:
On Tue, Aug 30, 2005 at 03:13:13PM -0400, kamaraju kusumanchi wrote:
Using Debian Sid, I have apt-listchanges package installed which
displays the changelog before installing the packages. However, in
this changelog, The names are not displayed properly. For example
the upgrade xpad package shows
xpad (2.9-2) unstable; urgency=low
* Fix broken entries in changelog
* Bump standards version, no changes
* Fix clean target to clean po/
-- Sren Boll Overgaard <[EMAIL PROTECTED]> Sun, 28 Aug 2005
10:30:11 +02
00
The author name is displayed as Sren .
How can I avoid this error?
Hi,
debian package changelogs are encoded in Unicode, where every character
with an ASCII value > 128 is represented as two bytes. To properly read
such files, invoke a terminal emulator in Unicode mode - e.g. xterm
has a
wrapper script "uxterm" which does just that, provided you have
configured
at least one UTF locale ('locale -a' shows all configured locales, and
'dpkg-reconfigure locales' will let you re-choose).
1) which terminal emulator does apt-listchanges use by default? man
apt-listchanges says that it uses $PAGER by default. But I do not have
$PAGER set, so I wonder which terminal emulator it is using.
2) How do I know whether my locales are UTF compatible?
I have
kusumanchi% locale -a
C
POSIX
Are they UTF compatible? or do I need to install new locales?
You have only one locale (with two names), and that is the only one
which cant be removed. You should install the locales package and
generate some locales. The encoding (ie UTF-8) follows the locale name
in the list. Currently I have the en_US and en_US.UTF-8 locales generated.
HTH,
Michael Spang
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install problem: Kernel panic: No init found. Try passing init= option to kernel.
brodul wrote: I am installing DEBIAN, I press Enter to boot, but then: . . . Freeing unused kernel memory: 96 k freed Kernel panic: No init found. Try passing init= option to kernel. I am a n00b please explain clear. Sorry for my EN. First off, "n00b" tends to a derogatory term. "newbie" is more applicable here. The error means you have no init binary on your root filesystem. The installation must have failed somehow for this to have happened. Can you provide more information about what you've done during your installation attempt? -- Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: DVD burning with 2.6.13
Dirk wrote: I thought that (since 2.6.8) it is a common problem that ppl can't burn stuff with cdrecord anymore... I was told to try/use dvdrecord instead of cdrecord... cdrecord works fine for burning cds, and I'm running 2.6.13-rc6. Its just a matter of setting the device correctly. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: What is the point of sudo?
Hendrik Boom wrote: On Sun, Aug 28, 2005 at 09:37:52PM -0400, Roberto C. Sanchez wrote: On Sun, Aug 28, 2005 at 05:20:37PM -0500, Hugo Vanwoerkom wrote: It's also easier to do root-type things while in X without having to log into X as root or muck about with the Xauthority files. Which running gdm I don't even know how to do: "sysadmin is not allowed to login here" or some such thing. And for good reason. Root should never be logged into a GUI. There is no good reason for this. There are reasons to do system administration from a GUI, especially if decent GUI applications exist for common tasks. The trouble with doing so, I have heard, is there are features in X that give a program that is talking to a window on an X screen to place input in another window ... this apparently was originally a feature whereby neat GUI things could easily be implemented, but it the other window happens to be a root login, it's a security nightmare. I believe window managers may use this stuff. By the way, I don't know about now, but 15 years ago, the window manager didn't even have to be running on the same magine as either your screen or any of your applications. -- hendrik Well, there are reasons to run specific administration tools within X, but the previous posts were referring to running all of X as root. This is frowned upon because X is huge and potentially buggy and because there is absolutely no benefit to running it as root. Whether X apps can hijack a root xterm I don't know, but I doubt this is the issue in question because superuser terminal sessions from within X are common while logged in as a regular user. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: I messed up bootmisc.sh - now can't log in!
[EMAIL PROTECTED] wrote: Michael Spang <[EMAIL PROTECTED]> wrote: Alan Ianson wrote: On Sat August 27 2005 06:38 pm, [EMAIL PROTECTED] wrote: Alan Ianson <[EMAIL PROTECTED]> wrote: On Sat August 27 2005 05:28 pm, [EMAIL PROTECTED] wrote: I've tried booting off a cd-rom and mounting the disk, but I can't get the disk to mount for some reason. I'm willing to try pretty much anything at this point. My default grub menu list has a "recovery" option. If that option isn't there for some reason edit your grub boot command line and try adding "single" to the end. I've never tried it manually before but I have used the recovery mode a few times and I think that's the only difference. Ah, but there's the problem. I can't log in to edit anything. Is there a way to make a menu come up? (Am I missing something, besides a system prompt of course ;-) At the grub menu, before the kernel boots. Hit a key to stop the default from booting and I believe if you press "e" you can edit the boot command. Just add "single" to the end and then boot that way. That will boot in recovery mode I think. I hope mysql isn't loading in that mode ;). If that doesn't work you may need to boot with a knopix or ubuntu disk and edit whatever needs editing that way. Ha. If its running from the startup runlevel then you could always add init=/bin/bash to your command line. But why would you have MySQL in /etc/rcS.d/? Also, you don't need to type out 'single', just S will work fine. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] It is running from the startup run level, but how do I add init=/bin/bash if I can't get a system prompt from the system? To take another angle on this, I have booted the system under Knoppix, but when I try to mount the disk, it says "/dev/sda1 NOT A BLOCK DEVICE". Is there any way to get the system to cough up what it thinks the device name of the disk is? Dave Williams __ Switch to Netscape Internet Service. As low as $9.95 a month -- Sign up today at http://isp.netscape.com/register Netscape. Just the Net You Need. New! Netscape Toolbar for Internet Explorer Search from anywhere on the Web and block those annoying pop-ups. Download now at http://channels.netscape.com/ns/search/install.jsp Are you using LILO or GRUB? As another post said, with GRUB you will have a menu on boot with a list of options, pressing 'e' on the desired option will allow you to add to the command line. With LILO you have to press something else. If all else fails you can use Knoppix to remove the offending startup script. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: What is the point of sudo?
Ian wrote: I know it provides a fake root environment for work, but why would you want that? No, sudo allows root privileges on a per-user, per-command, and per-host basis. Package: sudo Description: Provide limited super user privileges to specific users Sudo is a program designed to allow a sysadmin to give limited root privileges to users and log root activity. The basic philosophy is to give as few privileges as possible but still allow people to get their work done. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: I messed up bootmisc.sh - now can't log in!
Alan Ianson wrote: On Sat August 27 2005 06:38 pm, [EMAIL PROTECTED] wrote: Alan Ianson <[EMAIL PROTECTED]> wrote: On Sat August 27 2005 05:28 pm, [EMAIL PROTECTED] wrote: I've tried booting off a cd-rom and mounting the disk, but I can't get the disk to mount for some reason. I'm willing to try pretty much anything at this point. My default grub menu list has a "recovery" option. If that option isn't there for some reason edit your grub boot command line and try adding "single" to the end. I've never tried it manually before but I have used the recovery mode a few times and I think that's the only difference. Ah, but there's the problem. I can't log in to edit anything. Is there a way to make a menu come up? (Am I missing something, besides a system prompt of course ;-) At the grub menu, before the kernel boots. Hit a key to stop the default from booting and I believe if you press "e" you can edit the boot command. Just add "single" to the end and then boot that way. That will boot in recovery mode I think. I hope mysql isn't loading in that mode ;). If that doesn't work you may need to boot with a knopix or ubuntu disk and edit whatever needs editing that way. Ha. If its running from the startup runlevel then you could always add init=/bin/bash to your command line. But why would you have MySQL in /etc/rcS.d/? Also, you don't need to type out 'single', just S will work fine. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Diablo II in Wine for multiple users
Hendrik Boom wrote: My oldest son is running Diablo II and World of Warcraft on Wine on sarge. But my youngest son wants to play too. So my oldest son had a permissions question. How should he set permissions on everything involved so that they can use the same copy of these programs? (Not simultaneously of course). The trouble seems to be that these programs maintin history and save files and such, presumably created with the user ID of the first user to create them, in directories that are owned by someone, and are thereafter unavailable for writing later... And he doesn't really know where these games store all these data either. And he doesn't want his little brother to accidentally clobber his own saved games and such. If only all this stuff would be kept in ~/.diabloii or some such, but that doesn't seem to be the way the games were written. -- hendrik All D2 savegames are stored in $D2DIR/save. If this is properly backed up upon each use then there shouldn't be any problems. Look around in any hidden directories that belong to wine for the Windows tree. The game should be in "$WINROOT/Program Files/Diablo II". You could probably move the Windows tree into /usr/local/games, using setgid and some simple scripts to manage the savegames. Finally, if the savegames in question are on the multiplayer service (Battle.net), then they are stored on that service protected by a username and password. Each son could have his own account. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: blind spot for java
Graham Smith wrote: Hmmm I wonder if this is has something to do with the upgrades going on in unstable at the minute. Shouldn't this line /var/chroot/sid-ia32/lib from ld.so.conf have pulled in the required ld-linux.so.2 link? Graham As I understand it the location of the dynamic linker is hard coded in the binary, so I doubt it. $ readelf -l /bin/bash | grep ld-linux [Requesting program interpreter: /lib/ld-linux.so.2] Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: update-alternatives and window-manager and display-manager
Paolo Pantaleo wrote: 2005/8/26, Marcel Gschwandl <[EMAIL PROTECTED]>: On Mon, 2005-08-22 at 11:50 -0700, Marc Wilson wrote: I think that * update-alternatives is a good facility It isn't, really. IMHO it is. The alternatives system would be even better if it wasn't half broken. I mean, why does it leave all kinds of obsolete, broken symlinks around? There have also been times when the alternatives system was out of sync with reality thinking that a symlink pointed to one choice when it in fact pointed to another. Upon reselecting the correct choice nothing was fixed because it does nothing if you don't "change" your choice. I ended up replacing the symlink manually. Even if these issues are the fault of broken maintainer scripts or something similar, the system should have more tolerance when things go wrong. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: upgrade from kernel 2.4 to 2.6
Tong wrote: Hi, I want to upgrade my kernel from 2.4 to 2.6. Yeah, I know the simple answer is install the 2.6 kernel and boot into it. But I'm thinking, it might be more than that. For me particularly, I use SCSI emulation for my CDs in kernel 2.4. (I know I shouldn't but the ATAPI interface didn't work out of the box, and I didn't know how to fix it, so I had to revert to SCSI emulation) If I upgrade my kernel to 2.6, I think I have to tweak those CD modules manually -- to use the ATAPI interface only. What is the right setting? I've mess my module setting so much that I am afraid next change would be the last hay to break my camel. Thanks for you help. Another thing about kernel. Say I know a very cool kernel that has all nice features built within, will just grab it and boot into it works in Debian? I presume that lots of dependencies would be broken, wouldn't it? What is the right way? Thanks tong PS. The "very cool kernel" is precompiled Linux kernel 2.6.12.2 with unionfs and squashfs modules included, and with alsa sound drivers and madwifi drivers. http://www.linux-live.org/ SCSI emulation is nonexistent and unnecessary in 2.6. You can write CDs with cdrecord without trouble. You can also save a few command line options by tweaking /etc/default/cdrecord. My /etc/default/cdrecord follows; with it I need not specify anything about my writer. #ident @(#)cdrecord.dfl 1.4 02/07/07 Copyr 1998 J. Schilling # # This file is /etc/default/cdrecord # It contains defaults that are used if no command line option # or environment is present. # # The default device, if not specified elswhere # CDR_DEVICE=dvdrw # # The default FIFO size if, not specified elswhere # CDR_FIFOSIZE=4m dvdrom= ATA:1,1,0 -1 -1 "" dvdrw= ATA:1,0,0 -1 -1 burnfree HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Using RPM on Debian
Redefined Horizons wrote: I've got some software that I'd like to install on my Debian Sarge box. I can't get it in source, so I can't compile it, and it is only made available in binary form as RMP. What is the best/easiest way to convert an RPM to a Deb? Can I install RMP on my Debian machine, or is there a conversion utility? Thanks, Scott Huey The 'alien' package can convert rpms to debs for you. See the manpage for more info. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: adding modules to kernel
Adam Hardy wrote: > Is there some short-cut? Run xconfig again, select the option as a module, and recompile. If the tree hasn't been cleaned since you last built, it should only compile that one module and not the entire kernel. When this is done you can either reinstall the kernel or just copy the module in question into /lib/modules/. HTH, Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: /dev broken on upgrade to 2.6.9
Matthew Kay wrote: Hi all! Have got a nasty error on upgrading my kernel, using make-kpkg, to 2.6.9. I didn't change very much, just wanted to take advantage of the updated ACPI code. Specifically, the problem is that /dev is completely broken. It shows up but I can't cd into it (Not a Directory). I seem to be using udev; I couldn't honestly say whether I was using it before, but it is installed (0.046 I think). The errors only start coming up after INIT; it says 'Mounting a tmpfs over /dev...done and Creating initial device nodes...done' but warns that '.udevdb exists on existing /dev'. The errors are all along the lines of not being able to find nodes in /dev, and it asks to go into maintenance mode upon not being able to find /dev/hda3, my root. Funnily enough, it does mount this and I can remount it read-write in the usual way: mount -t ext2 -o remount,rw /dev/hda3 / My first instinct was to rollback to 2.6.8.1 which worked for me. However, since I can't stat /dev/hda, I can't update lilo! I've downloaded gnoppix to help, but doe anyone have any ideas as to how I can get my system working again?? Please ask if I've left out crucial stuff. Many thanks, Matt (on-list) Although I have no idea what is causing your problem, it may be helpful to know that udev will bind /.dev to your old /dev, so you can access any needed nodes there. Of course you could always stop udev with '/etc/init.d/udev stop' and that should unmount the tmpfs and may temporarily resolve the problem, allowing you to investigate further. udev needs hotplug: you did include that in your new kernel, right? Just a thought.. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Can't eject CD
Gerard Ceraso wrote: Steven Jones wrote: do a fuser on the cdrive look for the process that "holds" it and stop/kill it. man fuser for details regards Steven aka thing -Original Message- From: John Fleming [mailto:[EMAIL PROTECTED] Sent: Tuesday, 7 December 2004 11:28 a.m. To: [EMAIL PROTECTED] Subject: Can't eject CD Pse help newbie - My CD/RW lite is on, and I can't eject the CD. I really don't want to reboot just for this. My device is ATAPI:0,0,0. I've tried cdrecord -eject to no avail. Any help please? Debian Sarge Make sure that you are not in the cdrom directory. CD-ROM drives can be a terrible pain. I find the 'eject' command almost never fails. However it is likely this does the same thing as 'cdrecord -eject.' You could of course use a paperclip, though it might not be a good idea while the drive is powered. Are any processes locking the drive? Is it mounted? 'lsof /cdrom' may be of some help in finding out why the drive is locked. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: AGP
Hugo Vanwoerkom wrote: Paul Johnson wrote: On Tuesday 07 December 2004 8:30 am, Giorgio Raccanelli wrote: Hello everyone, I'm in trouble in the installation of the x-server. I have an ATI Radeon 9600, so I installed the driver radeon as suggested in the HOWTO. When I run dpkg-reconfigure --priority=low xserver-xfree86 I'm asked of the PCI bus ID. Now my videocard is on AGP, so what should I do? Leave it alone, I've yet to come across a video card that setting the PCI ID actually helped. Except if you run Ruby multi-seat Linux and use IsolateDevice= ;-) H Heh, true. It is important to note that many of the questions are "optional." It might be helpful if this was made more clear to users. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gdm and reboot
Sean O'Dubhghaill wrote: On Wednesday 01 Dec 2004 8:02 pm, Ralph Katz wrote: On 12/01/2004 02:10 PM, Frederico Rodrigues Abraham wrote: Isn't there an option on GDM to enable normal users to shutdown and reboot the machine? It always requires the root password to do it... -- Fred Yes. This is from my /etc/gdm/gdm.conf (using Sarge): And you can also change this from the gdmsetup tool by turning off secure actions menu in the Security tab # The Actions menu (formerly system menu) is shown in the greeter, this is the # menu that contains reboot, shutdown, suspend, config and chooser. None of # these is available if this is off. They can be turned off individually # however SystemMenu=true # The Actions in the Actions menu require the root password SecureSystemMenu=false Regards, Ralph I have often wondered why exactly it defaults to requiring a password. Requiring a user who has physical access to a computer root privileges to shut it down seems fundamentally flawed to me--they could easily shut it down by removing power. However this is potentially damaging, so it seems logical to allow them to initiate a proper shutdown. Is removing this limitation a security issue for remote users? Surely disabling 'secure actions' won't allow a remote user to shutdown via gdm.. right? Anyhow, just thought I'd throw this out there to see if anyone has a good explanation. It always seemed to be a completely unncessesary and potentially frustrating default. I find it especially strange since by default any user can shut down once logged into gnome via gdm, but they become stranded once back on the welcome screen. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: combining mp3s
Patrick Wiseman wrote: On Fri, 26 Nov 2004 16:02:29 -0500, Rick Pasotto <[EMAIL PROTECTED]> wrote: I have serveral mp3s that really should be a single file. What program should I use to combine them? I've had success with a simple 'cat file1.mp3 file2.mp3 > file3.mp3', but I'm sure there's something wrong with that! Patrick Well.. considering many players can play uncompressed RAR files with a bunch of mp3s in them, it's not surprising that this works. Though certainly the result doesn't follow the mp3 standard; it won't have correct headers at the beginning and will have extra headers in the middle of the file. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problem of compiling kernel
Lian Liming wrote: Perhaps you need to compile an initrd and you forgot to give the --initrd option to make-kpkg... This is something confusing me so long. Is it neccessary to make the initrd file? I just remember that i have read quite a lot articles about compiling kernel on Linux, a lot of them do not tell that you must make the initrd file. I don't check the "RAM disk support" when "make meuconfig". Then should i really need to add the step of making initrd file? Any suggestions are appreciated! Make sure to select your root disk driver and root filesystem as built-ins and you won't need to worry about an initrd. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: partition magic?
Hugo Vanwoerkom wrote: [EMAIL PROTECTED] wrote: Hi! belahcene <[EMAIL PROTECTED]> wrote on 25/11/2004 10:57:25 AM: > Hi every body, > I have a machine with only ONE partion where win$ is installed, > I want the equivalent of magicPartition ( something like diskdrake of > mandrake on debian), either on windows or on linux ( to run it from > knoppix) to create a new partition without loosing the content ( in > general after using fdisk to resize, the window$ doesn't run correctly > ) of window$. You can use qtparted (or qpartet, I'm not sure about the name) It is included in knoppix at least since 3.4 Presume that's just a frontend to parted. But parted always gives me messages about "a strange layout" although it is a brandnew partition created with cfdisk! I used to keep XP around for partition magic until the time it messed up my linux partition so bad I had to trash it. Now I use resize2fs, which does the job. H. I don't know about any 'strange layout' error, but 'couldn't align the partitions properly' is a problem I have had nightmares with. d-i twice broke windows on me (windows didn't like partition table after parted messed it up, it converted the whole thing to standard CHS or something similar; note that this was with the same d-i netinst cd on differnt computers), one requiring a full reinstall because I couldn't deal with the problem and the other I managed to convert the table back to something sane again. The whole fiasco was caused by Linux not using LBA for my disk drives. Linux doesn't detect LBA at all--I have to set it manually whenever I do partitioning or bad things happen. This is doable by typing 'echo bios_head:255 > /proc/ide/hdx/settings' before using parted/sfdisk/etc. I also have a hdx=c,h,s in my grub's menu.lst to bypass Linux's nondetection of LBA addressing. Linux is a hardy system, as it can use CHS on big drives without issues, however the fact that it refuses to detect when a partition table is aligned to LBA causes big problems for _other_ systems. I've lost massive amounts of [unimportant/backed-up] data and time due to disk addressing problems and programs which cry about partitions that aren't aligned to cylinder boundaries (this is terribly uncessary, why do they still demand this as in medieval times?). The state of things these days is terrible--too many hacks (ezbios/drive..*shudders*) and programs which just can't cope. (there was a time when I could boot into windows and linux but Partition Magic wouldn't dare touch my partitions which it detected as bogus, they really need a 'realign partitions with current geometry' option. Plus depending on who you ask (the OS, the bios in different interrupts) you will always get different geometry. Linux CAN cope, but it can also break things for other OSes. Be forewarned.. partitions need to be managed with care! Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: My CDROM/CDRW not fully working in Debian
Tong wrote: Hi, I've searched on the net, and it seems that it's only me having this problem -- my cdrom/cdrm is not fully working: I've noticed long time before that my eject command is not working properly: $ eject eject: unable to eject, last error: Inappropriate ioctl for device The result is same for both normal user and root. Today, I've found out that my "cdrecord -msinfo" is not working. Strangely, the "cdrecord -toc" command is working fine: $ cdrecord -msinfo cdrecord: Drive needs to reload the media to return to proper status. cdrecord: Input/output error. read track info: scsi sendcmd: no error CDB: 52 01 00 00 00 FF 00 00 1C 00 status: 0x2 (CHECK CONDITION) Sense Bytes: 70 00 05 00 00 00 00 0A 00 00 00 00 30 06 00 00 Sense Key: 0x5 Illegal Request, Segment 0 Sense Code: 0x30 Qual 0x06 (cannot format medium - incompatible medium) Fru 0x0 Sense flags: Blk 0 (not valid) cmd finished after 0.000s timeout 240s cdrecord: Cannot get next writable address for 'invisible' track. cdrecord: This means that we are checking recorded media. cdrecord: This media cannot be written in streaming mode anymore. cdrecord: If you like to write to 'preformatted' RW media, try to blank the media first. cdrecord: Cannot read first writable address $ cdrecord -toc Cdrecord-Clone 2.01a34 (i686-pc-linux-gnu) Copyright (C) 1995-2004 Jörg Schilling NOTE: this version of cdrecord is an inofficial (modified) release of cdrecord and thus may have bugs that are not present in the original version. Please send bug reports and support requests to <[EMAIL PROTECTED]>. The original author should not be bothered with problems of this version. scsidev: '0,0,0' scsibus: 0 target: 0 lun: 0 Linux sg driver version: 3.1.25 Using libscg version 'schily-0.8'. Device type: Removable CD-ROM Version: 0 Response Format: 2 Capabilities : Vendor_info: 'SONY' Identifikation : 'CD-RW CRX195E1 ' Revision : 'ZYS5' Device seems to be: Generic mmc CD-RW. Using generic SCSI-3/mmc CD-R/CD-RW driver (mmc_cdr). Driver flags : MMC-3 SWABAUDIO BURNFREE FORCESPEED Supported modes: TAO PACKET SAO SAO/R96P SAO/R96R RAW/R16 RAW/R96P RAW/R96R first: 1 last 1 track: 1 lba: 0 (0) 00:02:00 adr: 1 control: 4 mode: 2 track:lout lba:357394 ( 1429576) 79:27:19 adr: 1 control: 4 mode: -1 I can bear with the fact that eject not working, but not the "cdrecord -msinfo" -- I can't write multi-session to my CDR any more. I know it is a weired problem. Please spend some time and think it over for me, because I have no clue what could be wrong. Things has always been working out of box for me for other distros. FYI, I installed the SCSI simulation after noticing that "cdrecord -scanbus" not working after freshly install Debian, but I've forgot how I did it, so that I can't get back and try ATAPI... Thanks a lot for your help. tong PS, I'm using Debian testing, 2.4.25-1-386. My kernel drivers: $ lsmod | grep -Ei 'scsi|ide|cd' ide-scsi8464 1 scsi_mod 85312 3 [sg sd_mod ide-scsi] ide-cd 27936 0 cdrom 25056 0 [ide-cd] ide-detect 288 0 (autoclean) (unused) ide-disk 12512 12 (autoclean) ide-core 94108 12 (autoclean) [ide-scsi ide-cd ide-detect sis5513 ide-disk] I would say try kernel 2.6.x's new ATA writing support. It works flawless for me. ide-scsi is broken in 2.6, but it has become unnecessary anyway. If and when you get 2.6 up and running a howto is in /usr/share/doc/cdrecord/README.ATAPI.setup under kernel 2.6.x. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How do you make mkinitrd work!?
Benedict Verheyen wrote: Hugo Vanwoerkom wrote: Good point. So what's the reason that debian kernel images come with an initrd? An example I can think of is vesafb: in order for that to be used it must be either builtin or in initrd. Right? H. The kernels Debian provides need to run on a broad number of machines hence the use of an initrd which loads the modules needed for a particular machine. Benedict Exactly. Vesafb is not really the best example--it's there, but the initrd's job is to get your root filesystem mounted. I believe you can still use vesafb as a module without it being in the initrd, it just gets loaded later in the boot process. If we take a look at a stock debian kernel's initrd, we'd see modules for ide disk drives, raid arrays, scsi disk drives, tons of filesystems, and a bunch of miscellaneous drivers. Without an initrd to give the kernel access to these modules, they would all have to be built in leaving all users with a gigantic kernel with many unneeded built-in features. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How do you make mkinitrd work!?
Shawn McCuan wrote: This is what i get - but - ive already used make for the modules - - metion:/home/metiosarius/src/linux-2.6.9# mkinitrd -o /boot/initrd-2.6.9.img 2.6.9 /usr/sbin/mkinitrd: /lib/modules/2.6.9: Not a directory /usr/sbin/mkinitrd: MODULES needs to be set to none? metion:/home/metiosarius/src/linux-2.6.9# [snip] mkinitrd needs access to your modules. If there are no modules in /lib/modules/ you need to install some there to make an initrd. You're sure you typed 'make modules_install' in your kernel source directory? Are you sure you selected any modules? If you built your own kernel, then chances are you don't need an initrd. If your kernel can mount root without needing modules to access your hard disk or root filesystem, then theres no reason to create or use an initrd; it will likely have no modules in it even if you do (who would build their own kernel but neglect to select their hard disk drivers and root filesystem as built-ins?). If you're sure you need it and have successfully installed some modules, try 'ls /lib/modules' and using the correct directory from the list as your version. If you manage to boot the kernel and for some reason still want an initrd, 'uname -r' will tell you the kernel's version. For example, my kernel has an EXTRAVERSION of -tyr... # uname -r 2.6.9-tyr # mkinitrd -o /boot/initrd.img-2.6.9-tyr 2.6.9-tyr # ls /boot/initrd.img-2.6.9-tyr /boot/initrd.img-2.6.9-tyr However, if you have no modules, you can't make one (it wouldn't very useful even if you could... =) Finally, the 'standard' debian naming for the initrd is initrd.img- not initrd-.img. I didn't realize this during my previous post. I say 'standard' because thats simply what update-grub looks for, the name is mostly irrelevant. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How do you make mkinitrd work!?
Shawn McCuan wrote: Ive downloaded, configured, compiled and installed kernel 2.6.9 BUT- . the only thing I have left to do is "mkinitrd" Im not sure what commands I should be putting in... im trying: metion:/home/metiosarius/src/linux-2.6.9# mkinitrd /boot/initrd-2.6.9.img 2.6.9 $Id: mkinitrd,v 1.201 2004/05/16 22:00:48 herbert Exp $ Usage: /usr/sbin/mkinitrd [OPTION]... <-o outfile> [version] [snip] mkinitrd -o /boot/initrd-2.6.9.img 2.6.9 Didn't you see the -o in the usage: line? Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt pinning
ocl wrote: Hi, I am trying to get a certain package (dbmail) pinned to a certain version (v2*). This package is not part of debian distribution --yet, but an older version (v1*) does exist for it in the debian repositories. This is the stuff that I added to /etc/apt/sources.list deb http://debian.nfgd.net/debian woody/ deb-src http://debian.nfgd.net/debian woody/ deb http://debian.nfgd.net/debian sid/ deb-src http://debian.nfgd.net/debian sid/ deb http://debian.nfgd.net/debian stable/ deb-src http://debian.nfgd.net/debian stable/ deb http://debian.nfgd.net/debian unstable/ deb-src http://debian.nfgd.net/debian unstable/ deb http://debian.nfgd.net/debian experimental/ deb-src http://debian.nfgd.net/debian experimental/ Please notice the 'slash' at the end. It is necessary for this particular repository. When I add the following lines Package: dbmail Pin: release version 2* Pin-Priority: 1500 to the /etc/apt/prefrences file, I would have expected that 'apt-get install dbmail-pqsql' (i.e PostgreSQL version of DbMail) would install v2.x, but it does not, it keeps getting v1.x I did try it with 'apt-get --reinstall install dbmail-pqsql' but the result is the same. Same result with Synaptic too. Could someone help, please. Cheers, Ray Perhaps 'release' should be removed since you're not specifying a release? I'm not sure on how to pin by version, but a pin by release looks something like this: Package: apt Pin: release a=unstable Pin-Priority:1000 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: dselect problems
Giorgio Raccanelli wrote: Hello, I'm in trouble using dselect. I've just installed Debian Woody 3.0r3 i386 and I'm trying to add some package to my system. Using dselect I make the selection and then I hit Return. I'm warned about the dependencies. I press space and I see the dependency list. From what I understood from the dselect manual, if I'm happy with the selection made by the system I can simply hit Return and I should be fine. But, in my case, this is not happening. When I press Return I'm told about the dependencies again. Therefore, after seeing the dependency list again I press Return for accepting. And again, I'm warned of dependencies and so on! So far, I've been able to install only mc, but to do it I had to press Q instead of Return (which mean I forced the system to accept my selection): this shouldn't be the solution. Do you have any hint? Thank you Giorgio Raccanelli dselect will not allow you to continue if you ask it to do something impossible; there's likely a conflict which it prompts you to resolve. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: dpkg --root functionality with apt-get?
Bruce Edge wrote: Hi, I need to make apt-get operate on a debootstrap built tree. I don't want to chroot into it because I need to mount --bind my local debian mirror then. Is there some -o option::whatever I can use to tell it to use this built tree as the root? Lastly, is there a comprehensive list of all the supported -o options ? Thanks, Bruce The apt.conf(5) manpage has info on the the options which can be used with -o. You need to convince dpkg that your root is elsewhere, and to have apt use the other tree's status so that it knows whats installed. This should do it: apt-get install (packages) -o Dpkg::options::=--root=/path/to/root -o dir::state::status=/path/to/root/var/lib/dpkg/status Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: devfs vs. udev
Sean O'Dubhghaill wrote: [snip] - Does hotplug inform udev when a new device connects to the computer? Not sure, but new devices plugged in will cause new dev entries to be made. I think hotplug will just load the relevant modules and udev will sort out devices on its own Package: udev Description: /dev/ management daemon udev is a program which dynamically creates and removes device nodes from /dev/. It responds to /sbin/hotplug device events and requires a 2.6 kernel. Yep... udev is /dev intelligence plugged into hotplug. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: have a question about /proc
Chris Lale wrote: > This reminds me. I cannot view /proc in Nautilus - even as root. Is there a way to do it? Chris. Works for me.. does nothing appear or do you get an error? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Kernel 2.4
Andreas Janssen wrote: Almost all other mainstream distibutions (Mandrake, Slackware, Fedora, SuSE) install 2.6 by default. Knoppix comes with kernel 2.6 (although you may have to activate it using a boot option). Debian Sarge comes with kernel 2.6 (although I think right now the default installation kernel is 2.4). The only mainstream distribution I know that does not have kernel 2.6 is Debian Woody, the current stable release. It is nearly 2 1/2 years old. best regards Andreas Janssen Interesting. I was aware of the option in both Knoppix and d-i etc to use 2.6. Will Sarge default to 2.6? Considering it hasn't been released yet it most certainly is not old. Maybe I had the wrong impression, however I always see posts where people are using 2.4, and some people I know go as far as to consider 2.4.27 the latest 'stable.' Many programs and [actively developed] drivers (cdrecord comes to mind) have only experimental 2.6 support -- this makes me think many developers haven't made the move. After all, how can you work on a driver while using a kernel which it doesn't properly support? Of course they could be working on that experimental support, but overall I had the general impression that they were clinging to 2.4; 2.6 is quite mature and the general slowness of the move to 2.6 probably gave me the wrong impression. Anyway, thanks for your input. It's good to know I'm not an oddball for using the latest kernel, for a while I was sure there must have been something wrong with 2.6 which kept people from upgrading. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: have a question about /proc
[EMAIL PROTECTED] wrote: I was browsing through my file system one day on my linux partition and noticed somthing that i never did before in the /proc directory. firstly what is this directory I now that it has somthing to do with the procedure file system what ever that is I dont know, secondly why does it duplicate my root directory and then create nested copies of it. they seem to go on endlessly what perpose does that serve just curiouse. Mabey its just my computer but I have tried differant linux distrobutions and the all seem to do the same thing. Oh and for the record I can not delete them any at all very strange.? You can't delete anything in /proc because theres nothing to delete.. the information is retrieved from memory and formatted when requested by the kernel. If you examine it closely, you will see that only a few files actually report a size. 'proc' stands for 'process,' not procedure. This filesystem is important, as many commands need the information in provides. The proc(5) manpage explains what many of the files are. There are lots of symlinks and yes, you can follow them to get ridiculously long paths for any of the files on your system like /proc/self/root/proc/self/root/... Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: install package - dependies
Hugo Vanwoerkom wrote: Mariusz Antonik wrote: Hi. I want try to build a kernel 2.6.9 To do this I nead install qt develop package. Why not just get the Debian 2.6.9 kernel-source package or get the Linux 2.6.9 kernel, get you favorite .config, use make menuconfig to tune it, run make-kpkg --revision 1 kernel_image or such, dpkg -i the .deb it made and you got your 2.6.9 kernel. Isn't the installation of libqt3-dev a separate issue? You need qt for xconfig, so it is not a separate issue [for some]. Why not try apt-get update, see if that corrects the missing dependency issue. Then install qt and you're set. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: change conffiles
Richard Kemp wrote: I know that it's not recommended. But that would be very practical to be able to maintain its own conffiles easily. Maybe so. I just read the dpkg-divert manpage and it won't divert conffiles, either. So I don't think this is possible. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: change conffiles
Richard Kemp wrote: How to install my own conffiles like /etc/bash.bashrc or /etc/screenrc with a debian package without conflicts with the real package ? This is strongly not recommended (by some Debian doc..somewhere..), though it might be possible to divert the originals. Packages aren't supposed to mess with others' conffiles, even if they are custom built with that purpose in mind. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Kernel 2.4
Although this questionis not specific to Debian, it is relevant and I figure someone here has the answer. Why is the Linux community so opposed to moving to Kernel 2.6? Is 2.4 really that much more stable? Familiar? Upgrading to much trouble? Too many things changed too much? I think not--the open source community is a very dynamic one, things always change and are upgraded without a second thought. So why are so many afraid of moving defaults to the latest stable? 2.4 seems to be the default everywhere..(well to be honest, I don't know the default for most distributions, but Debian and Knoppix are still with 2.4) yet 2.6 is not exactly new. The only issue I had was the transition away from ide-scsi for IDE CD burning. Even for desktops which would not suffer at all upgrading, 2.4 seems to remain the standard. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: playing mkv files.
Alex Polite wrote: I'm trying to play som mkv (matroska) video files. So far I've tried mplayer, xine, vlc and gstreamer, all from unstable. None of them are able to play the files. Xine complains about a missing demuxer. I know that these programs support matroska so I guess the problem is the debian packs aren't compiled with matroska support. Are there any unofficial packs for unstable with matroska support, or do I have to compile it myself? alex There is no Debian package for mplayer, which plays mkv files (and just about anything else imaginable) extremely well. I am curious as to what you tried if you didn't build it from source. You can get the source from their webpage (google it) and install it to /usr/local. I don't know about any of the other players, but mplayer does work. The only matroska packages in debian unstable are libmatroska-dev and mkvtoolnix. mkvtoolnix can mux/demux/remux matroska files, but cannot play them. libmatroska-dev is development stuff for libmatroska, which apparently isn't a debian package. (does anybody know why that is?) As for unofficial packages, go to www.apt-get.org. There are mplayer packages there, though I prefer building from cvs source just for that warm fuzzy feeling of having the latest snapshot of a great program. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt-get Failure
Brian Nelson wrote: On Wed, Nov 17, 2004 at 09:13:12AM +, michael wrote: Michael Spang wrote: Sed's postinst contains this line, which fails: sudo install-info --quiet --section "General commands" "General commands" /usr/share/info/sed.info The bug is apparently resolved so a fix should be on the way. so what do I do meantime? I'm in the middle of switching from stable to unstable when it hit the `sed` issue & aborted the dist-upgrade. I guess i have to wait for the bug-fix to get into unstable and then try again? (BTW followed the link in another email but it didn't really help me) Either grab the new version from incoming.debian.org that presumably fixes that bug, or grab an older version from snapshot.debian.net. Or just temporarily comment out the line in ths postinst to pacify dpkg, as follows. Not installing the info shouldn't have serious consequences. The newer version seems to have appeared in my mirror's package list within the last hour, however. /var/lib/dpkg/info/sed.postinst: #! /bin/sh -e pkg=sed if [ ! "$1" = "configure" ]; then exit 0 fi #install-info --quiet --section "General commands" "General commands" \ #/usr/share/info/sed.info BTW, is there an easy way to make reply in Thunderbird send to the mailing list list instead of the author of the message? I find it a pain to change the address each time. Michael Spang -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt-get Failure
Sed's postinst contains this line, which fails: sudo install-info --quiet --section "General commands" "General commands" /usr/share/info/sed.info The bug is apparently resolved so a fix should be on the way. Robert Tilley wrote: The sed package causes apt-get upgrade to stop. [EMAIL PROTECTED]:~# apt-get -f upgrade Reading Package Lists... Done Building Dependency Tree... Done The following packages have been kept back: libqt3-compat-headers libqt3-headers libqt3-mt-dev libqt3c102-mt qt3-dev-tools The following packages will be upgraded: console-data inkscape kernel-image-2.6.9-1-686 kvim-perl libexif10 mozilla-stumbleupon python-gnome2 python2.3-gnome2 qt3-assistant qt3-doc qt3-qtconfig valgrind vim vim-common vim-doc 15 upgraded, 0 newly installed, 0 to remove and 5 not upgraded. 1 not fully installed or removed. Need to get 0B/35.8MB of archives. After unpacking 1061kB disk space will be freed. Do you want to continue? [Y/n] Reading changelogs... Preconfiguring packages ... Setting up sed (4.1.2-2) ... No `START-INFO-DIR-ENTRY' and no `This file documents'. install-info(/usr/share/info/sed.info): unable to determine description for `dir' entry - giving up dpkg: error processing sed (--configure): subprocess post-installation script returned error exit status 1 Errors were encountered while processing: sed E: Sub-process /usr/bin/dpkg returned an error code (1) Can someone with more knowledge than myself provide some insight into this problem? Bob -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
