Re: VirtualBox 4 fails building kernel module (the same with nvidia module)
MRH misiek_s...@o2.pl writes: On 23/01/11 04:53, Cameron Hutchison wrote: and I had to revert two commits to fix it: 7b8ea53d7f1865cd8f05dfb8f706a4ff5a72abcf (makefile: not need to regenerate kernel.release file when make kernelrelease) 01ab17887f4cdcb8bb5a5d1bc3b160d186e6e99b (Makefile: make kernelrelease should show the correct full kernel version) I'm afraid it's a bit beyond my skills at the moment - I guess it's about getting source tree by svn and reverting the changes applied by getting previous version of a file? You need to have the git kernel tree, and you use the git revert command. The second commit to revert (01ab1788) has a conflict when reverting, but that's pretty easy to resolve (if you know what you're doing :-)) Thanks anyway - I hope it will start working for me at some point :) I haven't had time to look into what the correct fix would be, but I may have some time in the next few weeks. I'll put it on my todo list. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/3d34.4d3f5ba2.3d...@getafix.xdna.net
Re: VirtualBox 4 fails building kernel module (the same with nvidia module)
tv.deb...@googlemail.com tv.deb...@googlemail.com writes: The problem with extraversion and 2.6.37 is known, it's been reported here if I remember, look for a message starting with kernel-package: 2.6.37 in the archives. But since it seems to affect only proprietary software I guess they'll have to adapt. Since I was the one to report that, I should correct some things I got wrong. The problem also exists in 2.6.36 - that's when it was introduced, and I had to revert two commits to fix it: 7b8ea53d7f1865cd8f05dfb8f706a4ff5a72abcf (makefile: not need to regenerate kernel.release file when make kernelrelease) 01ab17887f4cdcb8bb5a5d1bc3b160d186e6e99b (Makefile: make kernelrelease should show the correct full kernel version) I don't know if this is related to the OP's issues - I don't use nvidia drivers. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1f88.4d3bb44a.f1...@getafix.xdna.net
kernel-package: 2.6.37 kernelrelease target no longer works
I've built a new 2.6.37 kernel with kernel-package where I use an --append-to-version option to add a hostname a build number. Due to a change in the kernel Makefile, make kernelrelease no longer gets the kernel release from include/config/kernel.release , but instead uses scripts/setlocalversion . This causes problems because the extra stuff added to EXTRAVERSION by --append-to-version is no longer there when you run make kernelrelease, so I get a bare kernel version instead of my version. The Linux Makefile change was made in commit 7b8ea53d. I don't know whether this should be considered a bug in kernel-package or the kernel. The kernel is no longer preserving the value of EXTRAVERSION passed on the make command line when make is subsequently invoked, whereas it previously did. But then, the kernel people don't seem to care much about stable interfaces, so perhaps kernel-package should adapt somehow. What are people's thoughts? For now, I'm going to revert 7b8ea53d in my tree so I can get VirtualBox installing again (it compares $(uname -r) to $(make kernelrelease)), but this should probably be fixed somewhere. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/f0a.4d34d5fb.9...@getafix.xdna.net
Re: permissions all zero when using 'cp'
Martin Lorenz mar...@lorenz.priv.at writes:
r...@vs152058:~# (
env -i
date -R testfile1
ls -ldog testfile1
echo
rm -f testfile2
echo
cp testfile1 testfile2
ls -ldog testfile2
)
- -r-Sr-x--- 1 32 30. Dez 20:22 testfile1
This is really wierd. Your testfile1 should not have been created with
these permissions. I dont know what would cause this apart from bugs or
some sort of corruption.
stat64(testfile1, {st_dev=makedev(144, 109), st_ino=37590572,
st_mode=S_IFREG|S_ISUID|0450, st_nlink=1, st_uid=0, st_gid=0,
st_blksize=4096, st_blocks=8, st_size=32, st_atime=2010/12/30-20:22:29,
st_mtime=2010/12/30-20:22:29, st_ctime=2010/12/30-20:22:29}) = 0
stat64(testfile2, 0xbfffd624) = -1 ENOENT (No such file or
directory)
open(testfile1, O_RDONLY|O_LARGEFILE) = 3
open(testfile2, O_WRONLY|O_CREAT|O_EXCL|O_LARGEFILE, 0) = 4
does this tell you something?
It tells me why your testfile2 has zero permissions - because cp(1)
created it that way. The third argument to open(2) when used with
O_CREAT is the permissions to use on the new file (the kernel will apply
the umask).
cp(1) should have used a value of 0450 for that (since it uses the
source file permissions). I don't know why it used 0.
I'm sorry that I can't provide any answers. Maybe by pointing out these
anomalies, someone else may see what is wrong.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/51f.4d1e8489.ba...@getafix.xdna.net
Re: how to kill a process that is defunct?
Robert Brockway rob...@timetraveller.org writes: On Sun, 21 Nov 2010, François TOURDE wrote: The zombie process don't use any resources in general. No need to reboot at this point, because nothing is wrong. Right. I can't see how the OP's process is a zombie as a zombie won't consume CPU (or any other resource). It exists solely to hand back the exit code to the parent when it can. It is likely that the OP did not understand that the CPU usage stats reported by ps(1) is the percentage of time spent running during the entire lifetime of the process [1][2]. The zombie was not using CPU while it was a zombie - as you state. [1] ps(1) man page [2] Recent ubuntu-user post by Carl Smoot-Mitchell (in response to this same question) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/e24.4ceaf64b.5f...@getafix.xdna.net
Re: map '-' to '_' and '_' to '-'
Kamaraju S Kusumanchi raju.mailingli...@gmail.com writes: Yes. This works and it is just great. Is there a solution when the machine does not run X? I don't know the specific incantations, but it should be possible with loadkeys(1) from the console-tools package. The man pages for dumpkeys(1) and keymaps(5) will probably be necessary reading. From: Cameron Hutchison li...@xdna.net Newsgroups: gmane.linux.debian.user Subject: Re: map '-' to '_' and '_' to '-' References: ibis8b$4d...@dough.gmane.org 20101112133310.ga26...@aurora.owens.net 5b5d.4cddb78d.c6...@getafix.xdna.net ibmvp2$v9...@dough.gmane.org Kamaraju S Kusumanchi raju.mailingli...@gmail.com writes: Yes. This works and it is just great. Is there a solution when the machine does not run X? I don't know the specific incantations, but it should be possible with loadkeys(1) from the console-tools package. The man pages for dumpkeys(1) and keymaps(5) will probably be necessary reading. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/400d.4cdfb349.52...@getafix.xdna.net
Re: map '-' to '_' and '_' to '-'
Rob Owens row...@ptd.net writes:
On Fri, Nov 12, 2010 at 03:02:00AM -0500, Kamaraju S Kusumanchi wrote:
While working on my Debian box (which btw is a mix of Lenny and Squeeze), I
find that I type the underscore character ('_') more often than the dash
character ('-'). It could be file names, variable names while writing code
etc.,
Typing '_' involves holding the shift key while typing '-' does not involve
any. I am wondering if there is a way at the OS (or shell) level solution to
remap '-', '_' one to another so that typing '_' does not involve holding
shift key but typing '-' does.
Use xmodmap and/or a .Xmodmap file in your home directory. You man need
to use xev to find the keycodes.
The X11 keysyms for these characters are minus and underscore. If
you run xmodmap -pke | grep underscore you should see which keycode has
these keysyms mapped.
For me, this is:
$ xmodmap -pke | grep underscore
keycode 20 = minus underscore minus underscore
I suspect that's a standard keyboard keycode so you'll get the same
results, but run it anyway to be sure.
You can remap this like so:
$ xmodmap -e keycode 20 = underscore minus underscore minus
That swaps minus and underscore.
You can put the part between quotes in your ~/.Xmodmap file and GNOME
should auto-load it (although I think I remember reading something
lately that this may change).
Finally, I suggest you not do this. I've make changes to my standard
keyboard in the past and when you switch to someone elses keyboard you
just get annoyed :-) I used to map CapsLock to Ctrl, but if you get too
used to that, YOU JUST END UP SHOUTING when you only meant to press
Ctrl.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/5b5d.4cddb78d.c6...@getafix.xdna.net
Re: SNMP and MIBs
Mark Kamichoff p...@prolixium.com writes:
Ever since the snmp package stopped shipping MIBs due to licensing
problems, I've been unable to get snmpwalk and friends to correctly
resolve OIDs to names (and vice versa), at all.
[...]
I've got the following in my snmp.conf, which I believe adds the
downloaded and included MIBs to the search path:
% grep -v ^# /etc/snmp/snmp.conf
mibs ALL
mibdirs /var/lib/mibs /usr/share/mibs
I think this may be your problem (two problems actually).
The mibs that are downloaded are installed in subdirectories of
/usr/share/mibs (or /var/lib/mibs). Mine are in
/usr/share/mibs/{iana,ietf}.
Secondly I think that mibdirs needs to be a colon separated list of
directories.
mibdirs /usr/share/mibs/ietf:/usr/share/mibs/iana
That's what I have in my /etc/snmp/snmp.conf and I can resolve the
standard OIDs.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/6a3e.4cce1703.16...@getafix.xdna.net
Re: Fascinating problem with bash
Oliver Schneider borba...@gmxpro.net writes: Both cases can occur in several places (outer while loop). Since the paths can contain blanks, I resorted to a while read loop because for simply would tokenize the file names more than desirable. As soon as I read this paragraph I saw the problem. I confirmed it looking at the code. It's a common problem. This construct: some_cmd | while read var ; do OTHER_VAR=... done will result in OTHER_VAR being unset at the completion of the loop. That is because the while command is on the right-hand side of the pipe meaning it runs in a subshell. At the end of the while loop, the subshell exits and any vars it sets will go away with it. Since your some_cmd is just an echo, a solution that will work for you is: while read var ; do OTHER_VAR=... done EOF $SCRIPTCONF /etc/$SCRIPTCONF EOF This uses what is called a here document. You can read more about it in the bash man page if you need to. If a here document is not sufficient (some_cmd is more complex), the redirection (some_cmd) can be used. It appears in the bash 3.2.39 man page. It may not be in bash version 2. while read var ; do OTHER_VAR=... done (some_cmd) You may find that clearer than a here document. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/6da1.4c7305f1.3c...@getafix.xdna.net
Re: mdadm without initramfs
Boyd Stephen Smith Jr. b...@iguanasuicide.net writes: In 20100815190053.ga4...@gandalf.home.lxtec.de, Elimar Riesebieter wrote: How do I set up mdadm to create the root array witout an initramfs? You can't. When did this change? I have a box (my NAS) that is running Debian stable (lenny) with a custom 2.6.30 kernel. I never build initramfs images. Currently: $ df / Filesystem 1K-blocks Used Available Use% Mounted on /dev/md0 3842296 1578104 2225156 42% / $ cat /proc/cmdline root=/dev/md0 ro My /dev/sd[bc]1 partitions are of type 0xFD. /dev/md0 has a version 00.90 superblock. If I upgrade this kernel, or upgrade to the new stable when it's released, is this box going to stop booting? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2fc2.4c69d2df.bf...@getafix.xdna.net
Re: Shell Expansion in Bourne Shell Script Question
vogelke+deb...@pobox.com (Karl Vogel) writes:
On Wed, 28 Jul 2010 23:58:11 +0200,
Mart Frauenlob mart.frauen...@chello.at said:
M One might be better of with some like this:
M find /DIR -regextype posix-egrep -regex '.*\.(zip|ZIP)' -exec \
M some_command {} +
If the filelist is potentially too big for the max argument list on the
system, I would do something like this:
find $MAGDIR -print | grep -i '\.zip$' | tr0 | xargs -0 some command
find $MAGDIR -iname '*.zip' -print0 | xargs -0 some-command
-iname matches names case insensitively. Since you then dont need grep,
you also dont need tr0.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/788e.4c50d39b.7c...@getafix.xdna.net
Re: Restarting network
Johann Spies jsp...@sun.ac.za writes: Apparently '/etc/init.d/networking restart' is depricated. It is not doing the job any more on squeeze. '/etc/init.d/ifplugd restart' ignores virtual interfaces defined in /etc/network/interfaces. So how do I get my virtual interfaces active after a reboot or restart of the network without having to do 'ifup eth0:0' by hand? One way is to stop using the long-deprecated interface aliases and instead add secondary addresses to the single interface: iface eth0 inet dhcp up ip addr add W.X.Y.Z/N dev $IFACE down ip addr del W.X.Y.Z/N dev $IFACE That way, you can let ifplugd start eth0, and the above commands will add/remove the additional IP address you had set up on eth0:0 The ip(8) command is in the iproute package. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/573e.4c33c014.de...@getafix.xdna.net
Re: problem running kernel compiled on different machine
H.S. hs.sa...@gmail.com writes: Here are the grub.cfg stanzas for the current running kernel and for my compiled kernel respectively: #the default debian kernel initrd /initrd.img-2.6.32-trunk-686 #kernel compiled by me the Debian way initrd /initrd.img-2.6.32-100528-firewire Two things I can think to do next: 1) Change the initrd of your kernel to point to the original initrd. I dont know much about initrds (because I build all I need to boot into my kernels), but their purpose it to set up the kernel to get the right modules loaded to bootstrap the root filesystem. Since it looks like your problems _may_ be there, I'd try the original initrd (I say _may_ because it looks to me like the rootfs is mounted, but it is failing to mount your other filesystems listed in /etc/fstab) 2) Boot your new kernel and get to the emergency shell. diff the two kernel configs (the working one and your new one) and see if they differ (run diff -u). If they do differ, you can start to look into why. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/113d.4c0353c8.2c...@getafix.xdna.net
Automatic partial archive mirroring
I have a few debian unstable boxes that I like to keep up-to-date. Currently, I run apt-cacher-ng (a proxy for apt-get which stores packages so they don't need to be downloaded again) on my gateway box (lenny 32-bit), so I don't double download from my couple of other boxes (sid 64-bit). This works well enough, but I want to make it better. My internet connection has an off-peak period (2am-8am) where downloads are not counted in the monthly quota. I want my downloads to occur during this window, automatically. The problem is that the sid boxes are not powered-up during this window so I cannot just simply schedule a cronjob to do an apt-get -d dist-upgrade. I would like my lenny box to do this on behalf of my sid boxes. I dont really want to mirror the whole archive. Is there some way to do an intelligent partial mirror of the archive of what is installed on other boxes, without too much overhead of managing a package list? (i.e. I dont want to have to manually update some package list on my lenny box when I install a new package on a sid box). I've looked at apt-proxy, apt-cacher, apt-cacher-ng and approx, but they all appear to download on demand, not according to a schedule. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/85e.4c0313ab.76...@getafix.xdna.net
Re: How do I fsck an XFS file system in Squeeze
lrhorer lrho...@satx.rr.com writes: lrhorer put forth on 5/20/2010 6:09 PM: How can I obtain the XFS file utilities - particularly xfs-repair - under Squeeze? The simple answer to my original question was, xfsprogs. Doing a synaptic search for xfs returns far, far too many results through which to easily sift, and searching for xfs_repair returns none at all. The meta-answer to your original question is http://packages.debian.org from which you can search the contents of packages if you know what file you're looking for. At the bottom of that page is a section titled Search the contents of packages. It's so useful that I have a iceweasel/firefox smart bookmark set up for it, so I can just type dpcs xfs_repair into my address bar and I get the answer you were looking for (dpcs == debian package content search) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/b15.4c032b08.16...@getafix.xdna.net
Re: Automatic partial archive mirroring
Celejar cele...@gmail.com writes: Cameron Hutchison li...@xdna.net wrote: Is there some way to do an intelligent partial mirror of the archive of what is installed on other boxes, without too much overhead of managing a package list? (i.e. I dont want to have to manually update some package list on my lenny box when I install a new package on a sid box). You could use use the output of 'dpkg --get-selections' on the Sid boxes, pruning it using grep or similar to keep only the lines ending with 'install'. ok. that's easy. Combine the results from the various Sid boxes, so's that. and then have the gateway box get those packages. hmmm. How? Is there an existing tool that will do this? Doing it manually (in a script) would require too much work (essentially implementing apt-get -d dist-upgrade against a specified package index and package list). -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/c07.4c03366f.8a...@getafix.xdna.net
Re: Automatic partial archive mirroring
Celejar cele...@gmail.com writes: Cameron Hutchison li...@xdna.net wrote: and then have the gateway box get those packages. hmmm. How? Is there an existing tool that will do this? Doing it manually (in a script) would require too much work (essentially implementing apt-get -d dist-upgrade against a specified package index and package list). How about using 'aptitude download', possibly feeding the package list to aptitude with xargs? The 'download' action downloads to the current directory, so I suppose that you'd use a temp directory, and that would have the side effect of getting a copy stored by the cacher. The problem is the the dpkg architecture (and distribution) of the host is different to the targets. 'aptitude download' is going to use the native dpkg architecture, and is going to look in /etc/apt/sources.list for the sources to use. It looks like it may be possible to use something like 'aptitude extract-cache-subset' on the targets and set Apt::Architecture to the target on the command line. apt.conf(5) has a number of details of how I can override most of the host environment. Hmmm, I can see how this might work now. * periodically run aptitude extract-cache-subset on the targets and send that to the host * every night on the host, run 'aptitude -d dist-upgrade', but overriding all the directories for configs, indexes, pkgstates, etc and architecture to point to what has been captured by extract-cache-subset. Also point aptitude to the local apt-cacher-ng proxy so all downloads go through it. * delete the downloaded files, since apt-cacher-ng has already cached them. Thanks for the ideas. Perhaps I can knock this up fairly easily. I'll certainly learn more about apt than I probably want to :-) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/eb9.4c0344bd.22...@getafix.xdna.net
Re: problem running kernel compiled on different machine
H.S. hs.sa...@gmail.com writes: I am testing a patch for the Debian kernel (regarding the firewire bug I reported earlier in this list). I have been sent the patch and I followed the Debian way of compiling the kernel after patching the source. ( Here's is the installation error -- ) Hmm. There is a symbolic link /lib/modules/2.6.32.100528-firewire/build However, I can not read it: No such file or directory Therefore, I am deleting /lib/modules/2.6.32.100528-firewire/build ( Here's is the installation error -- ) Hmm. The package shipped with a symbolic link /lib/modules/2.6.32.100528-firewire/source However, I can not read the target: No such file or directory Therefore, I am deleting /lib/modules/2.6.32.100528-firewire/source You can ignore these two errors. You are installing the kernel on a different machine to the one you built on, so those links do not work. That is why it deletes the links. Anybody know why I am getting this error? Also AFAIK, a 32 bit kernel can be compiled on a 64 bit machine as long as it is running a 32 bit OS. Correct? Correct. Use menuconfig to load the current config from /boot and save and exit $ cp /boot/config-`uname -r` ./.config This looks like you are copying the config from the build machine, not your target. Can you confirm you actually copied the config from the target machine? Then apply the patch $ patch --verbose -p1 patches-hs/firewire_trv25_bug.patch I assume there's a typo there? Should be ../patches-hs/... and you should be in the linux directory if you are using patch -p1 Otherwise it looks you've done it right. In your later post, you show errors trying to mount /dev/hda4. The /dev/hd* driver is the legacy PATA driver (CONFIG_IDE). The newer driver based on libata (CONFIG_ATA - both SATA and PATA) uses /dev/sd* . It looks like your fstab is set up to mount partitions based on the legacy driver, but you are probably using the new driver. I'd say you are not using the same config from the target when building the new kernel. Is your old target kernel also 2.6.32? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1031.4c034c89.19...@getafix.xdna.net
Re: How to run a script when pppoe connection comes up?
Hendrik Boom hend...@topoi.pooq.com writes: I have a script that runs at boot time. It should really be rerun every time my pppoe connection has come up and has created device ppp0 for me. Evidently I have to invoke this script from someplace different from what I'm doing now. Where? After a ppp connection is established, pppd(8) runs the scripts in /etc/ppp/ip-up.d . Check the man page for pppd(8) under the section SCRIPTS. This describes which scripts are run when and what environment variables are available. The standard ppp runs the script /etc/ppp/ip-up when the link comes up. Debian implements this script to run all the scripts in /etc/ppp/ip-up.d -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1b84.4bf0cb0f.d3...@getafix.xdna.net
Re: cannot type power of 1 or 2, 4-9 are typeable.
jeremy jozwik jerjoz.for...@gmail.com writes: im trying to type [copy from character map] power of 2. i can read power of 2 on webpages but if i were to cope paste from that page, the power displays as a normal character 2. is this a dpkg-reconfigure locales issue? how can i gain the ability to type a power of 2? I can tell you how I do it - there may be other ways. I set up a compose key through the GNOME settings: System - Preferences - Keyboard - Layout Tab - Layout Options button From there you can set up a key to be the compose key in the Compose Key position section. I use Caps Lock since I do not use that key otherwise, but you can choose any of the keys available. Once you have a compose key defined you can enter the power of two character by typing: compose ^ 2 This is three separate key presses. You do not hold down the compose key. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/15ed.4bc2be5a.3b...@getafix.xdna.net
Re: Making make-kpkg quieter
Mart Frauenlob mart.frauen...@chello.at writes: On 08.04.2010 01:59, Cameron Hutchison wrote: Is there any way to make make-kpkg (kernel-package 12.033) quieter? When I run a make-kpkg clean it spits out lots of lines about unlinking files in debian/... On a slow link, this is very annoying (if I forget to run screen) I have RTFM but I cannot see anything about making make-kpkg less verbose (as opposed to the kernel makefiles). ok, there's no parameter available, but how about: make-kpkg clean 1/dev/null Does one really need an option for that? The problem is when I build a new kernel, make-kpkg starts with a clean, followed by the build. I definately want to see the output of the build. It just seemed like needless verbosity (debug output really) that I figured there might be a way to turn it off. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/6098.4bbdac15.23...@getafix.xdna.net
Re: Making make-kpkg quieter
Manoj Srivastava sriva...@ieee.org writes: On Wed, Apr 07 2010, Cameron Hutchison wrote: Is there any way to make make-kpkg (kernel-package 12.033) quieter? When I run a make-kpkg clean it spits out lots of lines about unlinking files in debian/... Please file a wishlist bug. I have usually looked for more insight into what is happening, and might not have considered the desire to only see error with the level of effort it needs (though I think the unlink verbiage comes from the underlying utility, not make-kpkg itself) Thanks Manoj. I've filed the bug. If I get time this weekend I'll have a look into it and see if I can track it down. I figured there was a simple echo/print somewhere and it would be an easy change. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4d9.4bbe6608.8e...@getafix.xdna.net
Re: Need help installing an alternative
Ron Johnson ron.l.john...@cox.net writes: # update-alternatives --install x-www-browser firefox \ /usr/local/firefox/firefox 3 update-alternatives: error: alternative link is not absolute as it should be: x-www-browser What am I doing wrong? The easiest way to see how this stuff works is to look at a postinst script that already does it. Pick an existing alternative and look at /var/lib/dpkg/info/$PACKAGE.postinst. I think you want this: # update-alternatives --install x-www-browser /usr/bin/x-www-browser \ /usr/local/firefox/firefox 3 That sets up /usr/local/firefox/firefox as an alternative for /usr/sbin/x-www-browser . -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/9e9.4bbe8530.5c...@getafix.xdna.net
Re: Making make-kpkg quieter
Mart Frauenlob mart.frauen...@chello.at writes: You want to see the output of the build, but the clean process is too much? The build output will be multiple times greater anyways. This is not a problem over the slow link? Ok. It looks like you haven't run the latest make-kpkg. I just did a rebuild of an existing built source tree. The make-kpkg output contained 8839 lines of output just to describe unlinking things in the debian/... directory. That 8839 lines is about 980kB. There was later objcopy output from the make-kpkg process amounting to 180kB. The rest of the output is 73kB. It's a significant difference. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/acb.4bbe8b7c.83...@getafix.xdna.net
Re: Need help installing an alternative
Tom Furie t...@furie.org.uk writes: On Fri, Apr 09, 2010 at 01:38:56AM -, Cameron Hutchison wrote: Ron Johnson ron.l.john...@cox.net writes: # update-alternatives --install x-www-browser firefox \ /usr/local/firefox/firefox 3 update-alternatives: error: alternative link is not absolute as it should be: x-www-browser I think you want this: # update-alternatives --install x-www-browser /usr/bin/x-www-browser \ /usr/local/firefox/firefox 3 That sets up /usr/local/firefox/firefox as an alternative for /usr/sbin/x-www-browser . According to the manpage (on lenny) the arguments to --install are genname symlink altern priority, so that would be # update-alternatives --install /usr/bin/x-www-browser x-www-browser \ /usr/local/firefox/firefox 3 Absolutely right. Apologies for the incorrect information. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/b05.4bbe8be4.c4...@getafix.xdna.net
Making make-kpkg quieter
Is there any way to make make-kpkg (kernel-package 12.033) quieter? When I run a make-kpkg clean it spits out lots of lines about unlinking files in debian/... On a slow link, this is very annoying (if I forget to run screen) I have RTFM but I cannot see anything about making make-kpkg less verbose (as opposed to the kernel makefiles). -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/48c8.4bbd1c72.3e...@getafix.xdna.net
Re: /boot partition changes when it should not
Clive McBarton clivemcbar...@web.de writes: Stephen Powell wrote: For example, the boot loader may be updating the mount count or updating the last referenced date/time, if there is such a field in the filesystem, for the kernel image or the initial RAM disk image. I assume you mean atime, which exists in ext3. And no, it was not updated. I checked with ls, it has the same value it had since the last kernel update. An ext3 filesystem has a last mount time in the superblock. This is different to any of the POSIX times in the root directory of the filesystem. The Linux kernel does not update this field on a read-only mount. If it were me trying to diagnose this, I would be diffing the images that should be the same and seeing where they are different. There's plenty of info out there on the data structures of the ext3 filesystem, so it shouldn't be too difficult to determine which attributes or other parts of the filesystem are being modified. At that point, then you can start speculating on what application is making the modification. Trying to guess what is making the modifications without knowing what was modified is going about it backwards. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/6585.4b96fc84.21...@getafix.xdna.net
Re: Bluetooth A2DP
Mark Kamichoff p...@prolixium.com writes:
It's got me wondering, do folks out there who have the appropriate
headsets actually use Bluetooth hi-fi audio on Linux? Perhaps there is
a workaround for this problem that everybody's using, that doesn't
appear on any Google searches?
I have been using my Philips SHB6100 headset with Debian sid (i686 and
x86_64) for the last 6 months or so, with some issues along the way, but
otherwise working pretty well.
I am using pulseaudio and bluetoothd to drive the headset, not alsa. I
have alsa configured to use pulseaudio for output, so all alsa is doing
is routing the default output to pulseaudio. I also have USB speakers,
which are driven by alsa when I chose that pulseaudio output.
I'm not completely sure what is needed to get this running as it is
something that I've played with quite a lot in the early days, but
looking at the packages I have manually installed, these seem to be the
related ones:
gnome-bluetooth
bluez-utils
gstreamer0.10-pulseaudio
pulseaudio
pulseaudio-esound-compat
pulseaudio-module-bluetooth
pulseaudio-module-gconf
pulseaudio-module-hal
pulseaudio-module-x11
pulseaudio-module-zeroconf
bluetooth, bluez and bluez-alsa were automatically installed.
My /etc/asound.conf looks like this:
pcm.!default {
type pulse
}
ctl.!default {
type pulse
}
I'm quite happy with this setup. I can use pavucontrol to change
application audio streams between the USB speakers and the BT headset
without needing to restart any applications. It works with mplayer,
gnome applications (gstreamer) and iceweasel with nonfree flash. That
is, it works with all the apps I use.
Hope this helps.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Trouble with flash player plugin
T o n g mlist4sunt...@yahoo.com writes: I am having trouble installing flash player plugin package from debian-multimedia. First, there is no version suitable for my Debian Testing. $ apt-cache policy flashplayer-mozilla Try flashplugin-nonfree instead. That's what I have installed and have working. This downloads and installs the nonfree flash plugin from Adobe. It is in unstable/contrib - it may also be in testing but I dont have that installed to check. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Flash plugin problem - cant click - amd64 unstable
Cameron Hutchison li...@xdna.net writes: Tyler Smith tyler.sm...@eku.edu writes: Cameron Hutchison li...@xdna.net writes: Is anyone else having problems with the flash plugin (flashplugin-nonfree) on unstable on a 64-bit platform? Yes, I'm having the same problem you described (below) on testing on a 32 bit machine. We're not alone, when I tried to trouble shoot this I found a bunch of forum posts ( http://bbs.archlinux.org/viewtopic.php?pid=650257 ) detailing the issue. I forgot to read that before my last reply. Seems other's have traced it back to the WM too. A workaround that seems to fix the problem for me is to add the following to ~/.mozilla/firefox/rc, or /etc/iceweasel/iceweaselrc for a global setting: --- # This is to fix flash where left-clicks dont work: # See https://bugs.edge.launchpad.net/ubuntu/+source/flashplugin-nonfree/+bug/410407 GDK_NATIVE_WINDOWS=1 export GDK_NATIVE_WINDOWS --- Thanks for you help, Tyler. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Flash plugin problem - cant click - amd64 unstable
Cameron Hutchison li...@xdna.net writes: Is anyone else having problems with the flash plugin (flashplugin-nonfree) on unstable on a 64-bit platform? On some sites, the flash videos load but the plugin seems to ignore my mouse clicks so I can't start the video playing. This does not happen on all sites - snotr.com works ok, but the video on this page: http://www.heatbeads.com.au/bbq-tips/44-bbq-tips/149-how-to-light-weber-bbq-heat-beads.html does not work. I cannot interact with the player at all. All my clicks are ignored. I can right click to get the context menu, but if I select Settings... I get a dialog that I cannot interact with. I forgot to mention that I am using Iceweasel, although that was probably assumed. Thanks to the info reported by Kumar, Camaleón and Tyler, I've isolated the problem to my window manager - fvwm2. If I start a gnome session with metacity, the flash plugin works properly. When running with fvwm2, I get the problems described above. Camaleón's workaround works for me too - that is, right click for the context menu, then double left-click on the button you want to activate. Why a left-click seems to go through the window, but a right-click does not is beyond me. I'll fiddle with my fvwm config and see if I can make a difference. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Flash plugin problem - cant click - amd64 unstable
Tyler Smith tyler.sm...@eku.edu writes: Cameron Hutchison li...@xdna.net writes: Is anyone else having problems with the flash plugin (flashplugin-nonfree) on unstable on a 64-bit platform? Yes, I'm having the same problem you described (below) on testing on a 32 bit machine. We're not alone, when I tried to trouble shoot this I found a bunch of forum posts ( http://bbs.archlinux.org/viewtopic.php?pid=650257 ) detailing the issue. I forgot to read that before my last reply. Seems other's have traced it back to the WM too. A 'fix' that works for me is to right-click on the flash screen, then double-left click on the flash button (play, pause, full-screen etc). This works on Hulu, the only way I can view videos full screen. I incorrectly attributed this fix to Camaleón in my last reply. Sorry. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Flash plugin problem - cant click - amd64 unstable
Is anyone else having problems with the flash plugin (flashplugin-nonfree) on unstable on a 64-bit platform? On some sites, the flash videos load but the plugin seems to ignore my mouse clicks so I can't start the video playing. This does not happen on all sites - snotr.com works ok, but the video on this page: http://www.heatbeads.com.au/bbq-tips/44-bbq-tips/149-how-to-light-weber-bbq-heat-beads.html does not work. I cannot interact with the player at all. All my clicks are ignored. I can right click to get the context menu, but if I select Settings... I get a dialog that I cannot interact with. This started happening not so long ago (sometime in the last month I guess, but I dont use flash sites much so I dont know exactly). I've tried uninstalling and reinstalling flashplugin-nonfree but that made no difference. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Can't tweak anything about synaptics touchpad
Matteo Riva mura...@gmail.com writes: After the full-upgrade of my testing system (with kernel 2.6.30) I can't do anything about my touchpad anymore, with the main problem being that I can't disable it. Neither synclient nor xinput have any effect on it, [...] From /usr/share/doc/xserver-xorg-input-synaptics/NEWS.Debian.gz: * All the shared memory code has been removed from synclient and syndaemon. The SHMConfig option is still available but the only use is for monitoring the hardware state data. If you have any script that changes the synaptics driver configuration make sure that you don't use the -s option to synclient. Or even better, configure the driver via hal. Option SHMConfig true That looks like the problem. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: How to replicate the behaviour of /usr/local
Klistvud quotati...@aliceadsl.fr writes: I have a shared directory on my system; what I'd like to achieve is making every newly created (or copied from elsewhere) file belong to the group owner users. # chgrp users /path/to/shared/directory # chmod g+s /path/to/shared/directory The set-group-id bit on a directory causes all files created in that directory to take the group id of the directory. Any directories created in that directory also take the group id of the parent directory and automatically have the set-group-id bit set on that directory. This causes the group id to propogate down the hierarchy as new directories are created. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: How to replicate the behaviour of /usr/local
Klistvud quotati...@aliceadsl.fr writes: Dne, 16. 10. 2009 12:19:03 je Cameron Hutchison napisal(a): Klistvud quotati...@aliceadsl.fr writes: I have a shared directory on my system; what I'd like to achieve is making every newly created (or copied from elsewhere) file belong to the group owner users. # chgrp users /path/to/shared/directory # chmod g+s /path/to/shared/directory Thanx. Your solution, though, only works for newly created files. Files *copied to* my shared dir from elsewhere still retain their original group ownership(s)... That never happens in usr/local. I think you need to look again and perhaps do some tests. What I described is exactly how /usr/local is set up. I don't know what you are doing wrong, but without you explaining exactly what you are doing, no-one can really help you. Let me demonstrate: $ touch /tmp/testfile $ ls -l /tmp/testfile -rw-rw-r-- 1 camh camh 0 2009-10-17 09:24 /tmp/testfile $ mkdir /tmp/shared $ chgrp staff /tmp/shared $ chmod g+s /tmp/shared $ ls -ld /tmp/shared drwxrwsr-x 2 camh staff 4096 2009-10-17 09:25 /tmp/shared $ cp /tmp/testfile /tmp/shared $ ls -l /tmp/shared/testfile -rw-rw-r-- 1 camh staff 0 2009-10-17 09:26 /tmp/shared/testfile Notice how the file /tmp/shared/testfile now has group staff where it originally had camh? Are you copying files into a directory nested underneath the top-level shared directory? You need all your directories under your top-level directory to be set up the same way. That is, the actual directory you are putting files into needs to be set-group-id - higher levels in the hierarchy do not matter. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: bi-directional file-synchronization tool
T o n g mlist4sunt...@yahoo.com writes: Anyone knows a good bi-directional file-synchronization tool that can synchronize changes to files and directories in both directions on different hosts, propagating the changes between them? syrep is too limited, unison seems to be the exact tool that I'm looking for, just I want to avoid its dependency (OCaml) if possible. Unison does not have a dependency on OCaml - just libc6. It may be written in OCaml, but that's only a build dependency. Do you plan on modifying a synchronisation tool? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Manoj Srivastava sriva...@ieee.org writes: On Thu, Sep 10 2009, Cameron Hutchison wrote: Version 3 (below) is properly written, in a functional style. It's much longer, but much easier to read. The main() function is very simple, as is each individual function. It's written in such a way that you can add extra filters if you want to extend it to get extra information (like the -v bit you asked about). What kind of license are you distributing this under? I would like to put this into my toolkit (nice work, BTW), but only if you choose to license it out. Public Domain. (same as sqlite) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Javier Barroso javibarr...@gmail.com writes:
On Thu, Sep 10, 2009 at 11:33 PM, Cameron Hutchison li...@xdna.net wrote:
/proc/pid/cmdline usually has ASCII NUL separated fields, which awk does
not split, so usually you have to use xargs -0. I noticed some cases
where the args were space separated (perl script), so I needed awk for
that case. I'll look more into awk and see if it can handle NULs in
some way. It doesn't by default.
Ok, I didn't know that. Thank you for the explication
awk -F '\000' '{print $1;exit}' /proc/$pid/cmdline
do the trick then
Well, no that does not handle the case where there are spaces separating
the fields. This was the case with one process in particular on my
system.
If gawk is installed I could do
gawk -F '[ \000]' '{print $1; exit}'
But that doesn't work with mawk, which is the default awk on Debian:
$ awk -F '[ \000]' '{print $1}' /proc/1663/cmdline
awk: line 0: regular expression compile failed (bad class -- [], [^] or [)
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Israel Garcia igalva...@gmail.com writes:
On 9/9/09, Cameron Hutchison li...@xdna.net wrote:
Israel Garcia igalva...@gmail.com writes:
I have more than 10 debian (etch and lenny) servers and I want to find
a way to know remotely on every server:
1. Name of running daemons and ports (tcp/udp) they're using.
2. Version of the package (installed by APT) used by these daemons.
3. Version of the latest package (from deb mirros) used by these daemons.
I tried to make a script but didn't resolve my problem.
That's really nice. It gives what I'm looking for..BUT, I have other
daemons installed from source, so dpkg -S returns an error. In my
case ruby. See below:
Ok. Here's version 2. Fixes are:
* Sorted the output by port number and removed duplicates. Duplicates
happen when a daemon listens on multiple IP addresses (samba is one).
* Skip non-existent processes
* remove (delete) from the end of readlink paths. This may happen if a
package has been upgraded and the old exe deleted.
* Use argv[0] if its an executable instead of /proc/pid/exe. This
makes daemons that are running under interpretters (perl, ruby, etc)
identified properly. In my case, postgrey failed, as a perl process.
* Ignore dpkg -S errors, and write a shorter line if there is no
package for the process.
Ruby was compile from source, How can I modify this script to remove
this error or better run -v option on daemons not installed by APT.
What do you mean by -v option? If you mean run the exe with -v to get
the version, that could easily fail and do unpredictable things, as -v
is not standardised as a way to get the version of a program.
netstat -lntup \
| awk '/^tcp/ { print $4/$1, $7 } /^udp/ { print $4/$1, $6 }' \
| sed -n 's|^[^ ]*:\([^ ]*\) \([0-9]*\)/.*|\1 \2|p' \
| sort -nu \
| while read port pid ; do
[ -d /proc/$pid ] || continue
bin=$(xargs -n 1 -0 echo /proc/$pid/cmdline | awk '{print $1
; exit}')
[ -x $bin ] || bin=$(readlink /proc/$pid/exe | sed 's/
(deleted)//')
pkg=$(dpkg -S $bin 2/dev/null | cut -d: -f1)
[ -n $pkg ] || { echo $bin on port $port; continue; }
version=$(dpkg-query -W --showformat='${Version}' $pkg)
latest=$version
latest=$(apt-cache show -a $pkg | grep ^Version: | { while
read x ver ; do
if dpkg --compare-versions $latest lt $ver ; then
latest=$ver
fi
done ; echo $latest; } )
echo -n $bin on port $port from package $pkg (version $version
if [ $latest != $version ] ; then
echo -n , $latest available
fi
echo )
done
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Cameron Hutchison li...@xdna.net writes:
Ok. Here's version 2. Fixes are:
One more iteration before I go to bed.
Version 2 was the quickly knocked together script that looks ugly and
hard to read, but is nice and compact. Maybe nice isn't the right
word.
Version 3 (below) is properly written, in a functional style. It's much
longer, but much easier to read. The main() function is very simple,
as is each individual function. It's written in such a way that you
can add extra filters if you want to extend it to get extra information
(like the -v bit you asked about).
If you dont want the result pretty-printed, just remove the
| map_lines pretty_print in main.
The only fix is to ignore errors from xargs (in get_pid_from_exe) which
I noticed coming out on a faster machine.
It runs under bash for the local keyword, but I think that should also
work under /bin/sh on Debian. It probably needs gnu xargs for -0 (gnu
xargs is standard on Debian, but busybox xargs may be different).
#!/bin/bash
main()
{
ports_and_pids \
| map_lines add_pkg_info \
| map_lines pretty_print
}
ports_and_pids()
{
netstat -lntup \
| awk '/^tcp/ { print $4/$1, $7 } /^udp/ { print $4/$1, $6 }' \
| sed -n 's|^[^ ]*:\([^ ]*\) \([0-9]*\)/.*|\1 \2|p' \
| sort -nu
}
add_pkg_info()
{
local port=$1 pid=$2 bin pkg version newer
bin=$(get_exe_from_pid $pid) || return
pkg=$(get_pkg $bin) {
version=$(get_installed_version $pkg)
newer=$(get_latest_version $pkg)
[ $newer != $version ] || newer=
}
echo $port $pid $bin $pkg $version $newer
}
pretty_print()
{
[ -n $1 ] [ -n $3 ] || return
echo $3 on port $1 ${4+from package $4} \
${5:+(version $5${6:+, $6 available})}
}
get_exe_from_pid()
{
[ -d /proc/$1 ] || return
local bin=$(
xargs -n 1 -0 echo /proc/$1/cmdline 2/dev/null \
| awk '{print $1 ; exit}'
)
[ -x $bin ] || bin=$(readlink /proc/$1/exe | sed 's/ (deleted)//')
echo $bin
}
get_pkg()
{
local pkg=$(dpkg -S $1 2/dev/null | cut -d: -f1)
[ -n $pkg ] || return
echo $pkg
}
get_installed_version()
{
dpkg-query -W --showformat='${Version}' $1
}
get_latest_version()
{
apt-cache show -a $pkg \
| awk '/^Version:/ {print $2}' \
| foldl_lines latest_version
}
latest_version()
{
dpkg --compare-versions $1 gt $2 echo $1 || echo $2
}
# map_lines func
# evaluate func for each line of input
map_lines()
{
while read line ; do
eval $1 $line
done
}
# foldl_lines func lhs
# evaluate (func (func (func lhs line1) line2) line3) ... for lines of input
foldl_lines()
{
func=$1
lhs=$2
while read line ; do
lhs=$(eval $func $lhs $line)
done
echo $lhs
}
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Javier Barroso javibarr...@gmail.com writes:
is this xargs: echo: terminated by signal 13 the output it should be?
Probably, substituting:
bin=$(xargs -n 1 -0 echo /proc/$pid/cmdline | awk '{print $1 ; exit}')
with
bin=$(awk '{print $1; exit}' /proc/$pid/cmdline)
will solved the issue
But I'm not sure why Cameron used xargs in this case.
/proc/pid/cmdline usually has ASCII NUL separated fields, which awk does
not split, so usually you have to use xargs -0. I noticed some cases
where the args were space separated (perl script), so I needed awk for
that case. I'll look more into awk and see if it can handle NULs in
some way. It doesn't by default.
The simple fix is to dump erors to /dev/null:
bin=$(xargs -n 1 -0 echo /proc/$pid/cmdline 2/dev/null | awk '{print $1 ;
exit}')
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Israel Garcia igalva...@gmail.com writes: On 9/10/09, Cameron Hutchison li...@xdna.net wrote: Version 3 (below) is properly written, in a functional style. [...] Well, in version 3 I see no output when I run the script...I double check but I dont know where the problem is. Hmmm, work for me (tm). Try isolating the failure by stripping down the main function. Just run ports_and_pids and see if you get the expected output. Then try the first filter (map_lines add_pkg_info), then the second (map_lines pretty_print). I did some last minute renaming of functions, so maybe I broke something, but I thought I tested it before posting. I'll have another look in a little while. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Cameron Hutchison li...@xdna.net writes: Israel Garcia igalva...@gmail.com writes: On 9/10/09, Cameron Hutchison li...@xdna.net wrote: Version 3 (below) is properly written, in a functional style. [...] Well, in version 3 I see no output when I run the script...I double check but I dont know where the problem is. Hmmm, work for me (tm). Doh!. Cut'n'paste error. I left one line off the end. Add this line: main It's no good having functions if there's nothing to call them. :-) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Israel Garcia igalva...@gmail.com writes:
[...] it seems when the script found
duplicate lines, like named/tcp and named/udp it only show one, se
below:
vps204:/usr/local/bin# netstat -lntup
tcp0 0 67.212.94.125:530.0.0.0:* LISTEN 23874/named
tcp0 0 127.0.0.1:530.0.0.0:* LISTEN 23874/named
tcp6 0 0 :::53 :::* LISTEN 23874/named
udp0 0 67.212.94.125:530.0.0.0:* 23874/named
udp0 0 127.0.0.1:530.0.0.0:* 23874/named
udp6 0 0 :::53 :::* 23874/named
And the script output:
vps204:/usr/local/bin# check2.sh
/usr/sbin/named on port 53/tcp from package bind9 (version 1:9.5.1.dfsg.P3-1)
As you can see the output only show named 53/tcp.
Good catch. Change the sort in the ports_and_pids function to
sort -u -t/ -k1n -k2
It should look like this now:
ports_and_pids()
{
netstat -lntup \
| awk '/^tcp/ { print $4/$1, $7 } /^udp/ { print $4/$1, $6 }' \
| sed -n 's|^[^ ]*:\([^ ]*\) \([0-9]*\)/.*|\1 \2|p' \
| sort -u -t/ -k1n -k2
}
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: looking for packages versions of running daemons
Israel Garcia igalva...@gmail.com writes:
I have more than 10 debian (etch and lenny) servers and I want to find
a way to know remotely on every server:
1. Name of running daemons and ports (tcp/udp) they're using.
2. Version of the package (installed by APT) used by these daemons.
3. Version of the latest package (from deb mirros) used by these daemons.
I tried to make a script but didn't resolve my problem.
Here's a script I just wrote to do what you want (it was an interesting
diversion).
For requirement #3, I'm not sure exactly what you wanted, so I took the
easy way out. I assumed you wanted the latest version for the
distribution you have in your /etc/apt/sources.list. To make the script
work, run apt-get update first so that your apt-cache has the latest
versions from your mirror.
netstat -lntup \
| awk '/^tcp/ { print $4/$1, $7 } /^udp/ { print $4/$1, $6 }' \
| sed -n 's|^[^ ]*:\([^ ]*\) \([0-9]*\)/.*|\1 \2|p' \
| while read port pid ; do
bin=$(readlink /proc/$pid/exe)
pkg=$(dpkg -S $bin | cut -d: -f1)
version=$(dpkg-query -W --showformat='${Version}' $pkg)
latest=$version
latest=$(apt-cache show -a $pkg | grep ^Version: | { while
read x ver ; do
if dpkg --compare-versions $latest lt $ver ; then
latest=$ver
fi
done ; echo $latest; } )
echo -n $bin on port $port from package $pkg (version $version
if [ $latest != $version ] ; then
echo -n , $latest available
fi
echo )
done
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: List Ettiquette
Chris Jones cjns1...@gmail.com writes: On Sat, Aug 29, 2009 at 06:53:41PM EDT, ghe wrote: :0Hfhw * ^Return-Path: bounce-debian-user=ghe=slsware@lists.debian.org | $FORMAIL -i Reply-To: debian-user@lists.debian.org Thanks for the idea. I will take a look at my .procmailrc and see how I can conversely eliminate these noxious Reply-to headers before they cause further damage. I feed all my mailing list subscriptions to local news groups, and part of that is to feed all incoming list emails through a python script. I just recently added the feature you're looking for, so maybe this is useful. Here's a snippet that should work as a self-contained script on stdin and stdout: #!/usr/bin/env python import email, sys m = email.message_from_file(sys.stdin) # Remove Reply-To if it is to the list if m.has_key(Reply-To) and m.has_key(List-Post): listpost = m['List-Post'].replace(mailto:;, ) if m['Reply-To'].find(listpost) != -1: del m['Reply-To'] print m -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Post to this list delayed over 15 hours
John Magolske listm...@b79.net writes: * John Magolske listm...@b79.net [090830 11:25]: My last post to this list took over 15 hours to get through, looks like the issue has to do with greylisting. I seem to recall having this problem a while back where after a few posts the response time improved...will see how long it takes for this message to get through. hmmm...this one came through right away. That's how greylisting works. The first message you send will be delayed, later messages will go straight through. After some period of time it will forget you (defaults to about 30 days I think), so you have to go through the delay again if you dont post for a while. The initial timeout is based on two things: 1. The timeout set in postgrey, which defaults to 5 minutes. Postgrey will not accept another email from you before that timeout expires. 2. The retry time of the sending mail server. When the sending mail server gets told of a temporary delivery failure (greylisting), the mail server will have to retry later. How much later is up to it. Sometimes it will retry within the postgrey 5 minute timeout and be rejected again which causes it to back off for much longer. That may be what happened to you. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
GNOME: How to automount pluggable drives?
How to I get drives to be auto-mounted in GNOME, without having a nautilus window open? I am using sid. It used to be that gnome-volume-manager would auto-mount media devices and auto-run specified programs for specific media types (ipod, CDs, DVDs, etc). This functionality of g-v-m has been removed in Debian since nautilus now does this. I don't use nautilus to manage my desktop, and I cannot see any way to run nautilus as a daemon so it can perform the auto-mount functionality as required. It seems I have to have a nautilus window open for auto-mount to work. In particular, I want rhythmbox to start when I plug in my ipod. Can this be done without having a nautilus window open? This is on my little eeePC 701, so memory, CPU and screen real estate are constrained. I dont want it chewed up by an unwanted nautilus window. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: usb hsdpa modem only detected after using it on windows
Umarzuki Mochlis umarz...@gmail.com writes: the modem is Huawei E1762 which is provided by local ISP when registering for their wireless broadband service. this modem, weirdly, can only be detected when i used it recently from windows xp, while the device still connected into the usb port, reboot into debian, then it will be detected and can be used with Vodafone Mobile Connect Card driver for Linux Look into a utility called usb-modeswitch. Some Huawei HSPA modems (maybe all) initially show up as a USB storage device, containing drivers for Windows and maybe Mac. Once the drivers are installed, they switch the mode of the modem to serial mode so the modem can be used. It sounds like this is what is happening with your modem. usb-modeswitch is a userland tool that switches the mode from storage to serial. On later kernels where the device is known, this utility is not needed as the kernel driver will switch the mode automatically. For instance, I have a E169 which uses the option driver (CONFIG_USB_SERIAL_OPTION). This driver knows about the E169 mode and switches it automatically for me. Debian unstable has the package usb-modeswitch, but this is not in lenny (stable). The home page for usb-modeswitch is: http://www.draisberghof.de/usb_modeswitch/ -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: APT configuration file
Kc9EYE kc9...@gmail.com writes: Following the thread on apt defaulting to install recommends, I would like to turn this option off. A previous poster stated to add this line to the /etc/apt/apt.conf file: APT::Install-Recommends 0; . I would love to do that but I am unable as yet to find a file named /etc/apt/apt.conf. Just create it with the line you want. No other contents are needed. The man page for apt.conf(5) describes the syntax of the file and the configuration parameters used by apt. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Reliable remote logging with rsyslog on lenny
I'm trying to set up a couple of boxes that both run lenny (stable) for remote logging. I have installed rsyslog on both boxes. One is to be the log server, the other I want to log remotely to the log server. According to the package description of rsyslog, it does reliable syslog over TCP. According to the rsyslog homepage, this reliable logging is RELP and can be enabled on the server like this: $ModLoad imrelp $InputRELPServerRun 2514 and on the client like this: *.* :omrelp:192.168.0.1:2514 (where 192.168.0.1 is the address of the log server) I cannot get this to work. From what I can tell with netstat, rsyslogd is not listening on port 2514 on the server. The package also looks like it does not contain the modules imrelp and omrelp - but I'm not sure about this. There aren't any .so files for them in the module directory, but they may be built in instead. How do I enable the reliable logging feature of rsyslog? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Program for quoting text like in email?
Thomas Anderson andersontho...@gmail.com writes: OK, so I guess it should be sed 's/^/ /'. However I noticed another problem. When I quote a text that is already quoted, the result gets the characters moved around. I would like to get orfor text that is quoted twice. What do you mean by moved around? Can you post an example? The sed script quoted above is very simple. It just puts at the start of each line and should not move anything around. If you are piping the result to fmt(1), then I guess you'd see the moving around you might be talking about. If this is the problem, you may want to use par(1) instead (from the par package). par knows about quoting and preserves it properly when reformatting. It's got quite a few options, but I typically use it as par 72q - that uses 72 chars as the max line width and handles quoting. $ sed 's/^/ /' | par 72q A slightly more complicated sed script to use not put an extra space after the quote on already quoted text would be: $ sed -e 's/^//' -e '/^\([^]\|$\)/s/^/ /' | par 72q -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: sed/awk/fmt question hijacked from Re: Program for quoting text like in email?
Tony Baldwin photodha...@gmail.com writes: Why is it that with sed, stuff like sed -e /searchterm/d I have to do sed -e /searchterm/d infile outfile, and can't just do sed -e /searchterm/d file, without having to generate another file? GNU sed (which is what you are most likely running) has the -i option to make the changes in the source file. sed -i -e /searchterm/d inoutfile Likewise, I have to the same with grep like grep -v ^$ filename newfilename I can't just to grep -v ^$ filename and have grep do its magic on filename, without generating a second output file. You cannot do the same with grep, since grep is usually used to print out pattern matches. You would not normally want to kill the file you are grepping. In your specific case, you can use sed: sed -i -e /^$/d filename As such, while manipulating my script today, I kept having to do stuff like sed -e /.gconf/d script script2 mv script2 script rm script2 sed -e /.java/d script script2 mv script2 script rm script2 You can combine these into a single sed invocation: set -i -e /.gconf/d -e /.java/d script -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
gnome-panel: Major breakage in unstable
A heads-up to those running unstable/sid. The latest gnome-panel that has just entered unstable is severely broken. It is the first revision of GNOME 2.26. When logging in, you get an infinite number of Starting File Manager entries on the panel, and CPU usage runs very high, continuously. It renders the desktop effectively unusable. I had to downgrade to the squeeze versions of: gnome-panel gnome-panel-data gnome-session gnome-settings-daemon It looks like only the amd64 architecture is affected at the moment, as the other architectures have 2.24 or older. This is reported in: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=531100 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: executing udev rules on _un_plug
cr...@got.net (i'll teach you to turn away.) writes: Cameron Hutchison li...@xdna.net wrote: CH How can I write a udev rule to be run/matched when a device is CH unplugged? CH CH I have a 3G modem that I wrote a rule for to run ifup ppp0 when the CH modem is plugged in. I would like to have ifdown ppp0 be automatically CH run when the modem is unplugged. when i was using a modem, i used a cronjob that checked for /var/run/ppp0.pid. when it's not detected, you can have it execute ifdown ppp0. Thanks, but unfortunately, a cron job is not responsive enough for me, as I want to be able to unplug the modem and plug it back it and have that work. I found the issue(s) in the end. Udev does support the action remove, but you cannot use all the match rules you might use on an add event. I guess that this is because the device is no longer there so the sysfs attributes are gone. Once I tweaked my rules I got half way there. The other half way there was realisin that there was no remove event generated for ttyUSB0 - I assume that it is because the device is still in use just at that time. Thankfully the modem also generates events for ttyUSB1 and ttyUSB2, so I could trigger the removal off one of those. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
executing udev rules on _un_plug
How can I write a udev rule to be run/matched when a device is unplugged? I have a 3G modem that I wrote a rule for to run ifup ppp0 when the modem is plugged in. I would like to have ifdown ppp0 be automatically run when the modem is unplugged. Without a rule to run ifdown ppp0, the network state of ppp0 is remembered as up, so if I plug in the modem again, ifup ppp0 then fails because it says the interface is already configured. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: tar up a symbolic linked directory
Douglas A. Tutty dtu...@vianet.ca writes:
On Sat, May 02, 2009 at 09:04:38PM +, T o n g wrote:
Hi,
I want to tar up a symbolic linked directory as if it is a real
directory. Is there any easy way to do it?
Let me explain with an example (that you can try):
mkdir d1
touch d1/{a,b,c}
ln -s c d1/d
ln -s d1 d2
I want that the result tar file looks like this:
tar -tvzf d2.tgz
drwxrwx--x tong/tong 0 2009-05-01 09:38 d2/
-rw-rw tong/tong 0 2009-05-01 09:37 d2/a
-rw-rw tong/tong 0 2009-05-01 09:37 d2/b
-rw-rw tong/tong 0 2009-05-01 09:37 d2/c
lrwxrwxrwx tong/tong 0 2009-05-01 09:38 d2/d - c
Any easy way to do it?
add -h to the tar parameters. It dereferences the symbolic lyinks.
However, then you won't get the d2/d - c reference.
I can't see that it would be possible to dereference the top-level
symlink but no others. No commands that I know of support selective
symlink dereferencing, except find(1) with -H.
That lead me to try:
$ find -H d2 | cpio -o -L -H ustar d2.tar
This comes close, storing d2/d as a link, but as a hardlink, not a
symlink.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Suggestions for multilevel backup of single machine?
James Youngman j...@gnu.org writes: (2) It would be useful to have a historic backup capability too (e.g. the way the filesystem looked yesterday, last week, last month and a year ago), at least for filesystems like /home. What are good solutions for doing (2)? (Please only recommend software you're using yourself :) I do this using rsync with the --link-dest option. The idea is that you sync A to B with reference to C. If a file in A and C are the same, a hard link is created in B instead of a copy of the file. This means the only new files in B are files that have changed in A relative to C. When I start the backup I create a directory for the new backup, named with the date of the backup. If a symlink latest exists, I run rsync with the argument --link-dest=.../latest. The new backup will be performed relative to the latest backup. When rsync is complete, I update the latest link to point to the just-performed backup. I periodically purge old backups, leaving behind backups ending in 01. This leaves one backup per month (e.g. 20090301, 20090201, etc) for a longer term historical backup. I have a script that does this for me which you are welcome to. Just email me. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: xterm font sizes choices?
zhang zhengquan zhang.zhengq...@gmail.com writes: Thanks, then maybe 10x20 is just small for me... To verify that the correct resources are being used, run xterm -fn 10x20. This will start an xterm with that font, or display an error that it cannot find the font. If this gives you a different font to what you normally get then you will need to see what is wrong with your loading of resources. One thing that comes to mind is that you may be using uxterm, with a resource class of UXTerm. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Unison not syncing files under ~/.kde
A. F. Cano a...@shibaya.lonestar.org writes: Hi, I've been using unison to keep home directories syncronized for quite a while now, but this little bit is starting to aggravate me. It appears that, unless I'm missing something in the configuration below, files under a .directory are ignored. Note that at the bottom there is a blanket ignore = Name .* to leave alone most .files which are system specific, but certain ones I want to propagate, thus the ignorenot lines. The relevant bits on the ~/.unison/default.prf file are: ignorenot = Path afc/.procmailrc ignorenot = Path afc/.fetchmailrc ignorenot = Path afc/.signature ignorenot = Path afc/.mutt ignorenot = Path afc/.profile ignorenot = Path afc/.bashrc ignorenot = Path afc/.vimrc All these get propagated just fine ignorenot = Path afc/.kde/share/apps/kpilot #this is a directory ignorenot = Path afc/.kde/share/apps/karm/karm.ics #individual file ignorenot = Path afc/.wine/user_files But these do not propagate. ignore = Path afc/Desktop ignore = Name .* ignore = Name *.o ignore= Name .* ignorenot = Path afc/.kde ignore= Path afc/.kde/* ignorenot = Path afc/.kde/share ignore= Path afc/.kde/share/* ignorenot = Path afc/.kde/share/apps ignore= Path afc/.kde/share/apps/* ignorenot = Path afc/.kde/share/apps/kpilot ignorenot = Path afc/.kde/share/apps/karm ignore= Path afc/.kde/share/apps/karm/* ignorenot = Path afc/.kde/share/apps/karm/karm.isc ignorenot = Path afc/.wine ignore= Path afc/.wine/* ignorenot = Path afc/.wine/user_files Its annoying, but I think its the only way to work around this. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: ifconfig data
Tzafrir Cohen tzaf...@cohens.org.il writes:
On Thu, Feb 19, 2009 at 02:02:36PM +0100, Tanco . wrote:
Hi Hugo,
this will give you the IP :)
ifconfig ppp0 | grep inet addr: | awk '{ print $2}' | tail -c14
ifconfig ppp0 | awk '/inet addr:/{ print $2}' | cut -d: -f2
ifconfig ppp0 | awk '/inet addr:/{ split($2, a, :); print a[2] }'
Once you invoke awk, you may as well use it for all your string
manipulation.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: stuff in ~/bin won't run
Tzafrir Cohen tzaf...@cohens.org.il writes: On Wed, Jan 14, 2009 at 04:16:57PM -0800, Mike Castle wrote: I don't think .bash_profile gets sourced when you log in via an XDM/GDM type session. (After all, when would it, since you don't really have a login shell.) Possible fix: echo 'if [ $SHELL = /bin/bash ]; then . $HOME/.bash_profile; fi' /etc/X11/Xsession.d/91bash_profile This looks all too simple so someone must have thought of it previously and decided not to use it. Why? Probably because of this line: #!/bin/sh at the start of /etc/X11/Xsession, which is the master script that will source your 91bash_profile script. So, $SHELL will not be /bin/bash and $HOME/.bash_profile will not be sourced. If you try to source .bash_profile unconditionally from there, any bashisms in it will cause Xsession to fail. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: hard crash on leap second
Travis Crump pretz...@techhouse.org writes: I had a hard crash of my lenny system precisely when the leap second was added. While X has flaked in the past, I've never had a hard crash before. I have no other evidence they were related, but I wasn't doing anything unusual at the time. Any ideas? I have a lenny and sid system, oth of which are still running OK. I am running openntpd on both, synced to my border router (openwrt), also still running and running openntpd. My logs show openntpd drifting the time back into sync from the leap second. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: vi -- Turn off Line Number Colors and Screen Buffering
Sven Joachim [EMAIL PROTECTED] writes: On 2008-11-04 06:28 +0100, Matt Miller wrote: Also, I don't want the editor to do whatever special screen buffer swapping, or whatever it is, that prevents me from scrolling back in my terminal history when the editor is open, and then clears away the screen and redisplays what was there before after the editor is closed. That's still an annoyance. This whole take-over-the-screen-because-that- must-be-what-you-really-want thing has got be turn-offable. This is a feature of your terminal (and ncurses). Please consult your terminal's manual how to turn the alternate screen off. In xterm, you need to set the `titeInhibit' resource to true. Alternatively, set TERM=vt100 as that terminal does not have the escape codes defined for this feature. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: lynx upgrade issue?
[EMAIL PROTECTED] (i'll teach you to turn away.) writes: right, i tried putting exactly that in ~/.lynx/colors, but it ignored me. only editing /etc/lynx-cur/lynx.lss changes the colors - they're laid out ENTIRELY differently from my old lynx color entries, which match what you've dumped here (except for the exact colors themselves). what's going on with my system?! heh. Aha, found it. I also have: COLOR_STYLE: in /etc/lynx-cur/local.cfg It looks like it defaults to COLOR_STYLE: lynx.lss so by turning that off, it looks like thats how I got it to pay attention to my ~/.lynx/colors file. Hopefully there was nothing else I changed - it was a while ago so I've forgotten all about it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: lynx upgrade issue?
[EMAIL PROTECTED] (i'll teach you to turn away.) writes: ok, so i added the directory ~/.lynx i dumped the color settings into ~/.lynx/colors but it's still green. however, i did find that /etc/lynx-cur has a .lss file in it which apparently controls color in a totally different layout than old lynx. wow, that's irritating. time to edit manually for a half hour i guess. The contents of my ~/.lynx/colors are: COLOR:0:lightgray:black COLOR:1:cyan:black COLOR:2:yellow:blue COLOR:3:green:black COLOR:4:green:black COLOR:5:brightcyan:black COLOR:6:red:black COLOR:7:black:white I think I just copied and pasted that from my old config, but that was a while ago, so I might be wrong. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: lynx upgrade issue?
[EMAIL PROTECTED] (i'll teach you to turn away.) writes: so in fixing my pine issue (see pine thread), i reinstalled/upgraded lynx. the new lynx removed lynx-ssl installed lynx-cur. the problem there is that it completely ignores my color preferences, which are fairly important considering i generally hate graphics color spent some time coming up with a nice lynx layout that doesn't make me insane. i checked both local system .cfg, while it retained my color settings, it just ignores them. new lynx is primarily green horrible. The lynx-cur config is the /etc/lynx-cur directory for system wide stuff and ~/.lynx for per-user stuff. One of the per-user files is ~/.lynx/colors where you can put your color preferences. I have added this file myself and have been able to change the colors. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
apt-get/aptitude bustage with gnucash
I have just attempted a dist-upgrade with bost apt-get and aptitude, and both of them want to remove gnucash from my system (sid). This seems to be busted. At the moment there is a new gnucash-common (2.2.4-2) but no new gnucash package to match. Normally that's ok and apt will not upgrade gnucash-common, but for some reason, this time it wants to upgrade gnucash-common and remove gnucash to resolve the conflict. gnucash-common 2.4.4-2 has the following deps: Replaces: gnucash ( 1.8.8-5) Recommends: gnucash (= 2.2.4-2) Conflicts: gnucash ( 2.2.4-2) So, this conflicts with gnucash 2.2.4-1 (currently installed) but does not replace it. So why would apt remove gnucash? I thought apt was only meant to remove a package itself when another replaces it. Is this a bug in apt? BTW. I'm not looking for any solutions here - i'll just wait until tomorrow or the next day and the problem will go away. I've just never seem apt try to remove something I want installed without there being a replacement. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: apt-get/aptitude bustage with gnucash
Eugene V. Lyubimkin [EMAIL PROTECTED] writes: Cameron Hutchison wrote: At the moment there is a new gnucash-common (2.2.4-2) but no new gnucash package to match. Normally that's ok and apt will not upgrade gnucash-common, but for some reason, this time it wants to upgrade gnucash-common and remove gnucash to resolve the conflict. I can say wrong, but dist-upgrade (or full-upgrade) can remove the packages. Try 'aptitude safe-upgrade'. I am aware that dist-upgrade can remove packages, but I thought this is only when other packages replace them. In this case, there is no replacement, only a conflict. Usually I see this state as broken with aptitude giving options to work around the breakage - often holding back or removing packages. There was no conflict given for gnucash[-common] today and aptitude/apt-get was just going to remove gnucash without prompting me for a resolution. This seems wrong, and was the point of my original posting. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Why isn't Gnome 2.22 in sid yet?
Why isn't all of Gnome 2.22 in sid yet? I'm most interested in the panel and the applets which are still at 2.20.3. 2.22 is in experimental but has not yet gone into sid? What's holding it up? What needs to be done to get it into sid and is there anything I can do to help it along? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Gnome international clock in sid
In Gnome 2.22 released a few months ago, a new clock applet was introduced where you can set it up with multiple timezones. The About dialog box for the clock in sid shows the version is 2.20.3. I assume that I need version 2.22 of this applet. Where can I get this for sid? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Gnome international clock in sid
[EMAIL PROTECTED] writes: Am 15.07.2008 um Uhr haben Sie geschrieben: Where can I get this for sid? you can get the gnomepanel 2.22 from experimental the nautilus 2.22 is there, too Last time I tried something from experimental I borked my system. Any idea when it will make it to sid? Or what's holding it up? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bash: pipe once more
Stefan Schmidt [EMAIL PROTECTED] writes: Hello, if I pipe the output of a cd command the working directory doesn't change. That's because all elements of a pipeline except the last are run in different processes to the main shell that starts the pipeline. As such, the cd command is running in a subshell which exits when cd exits. This has no effect on your main shell, which will keep the original working directory. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bash: pipe once more
Stefan Schmidt [EMAIL PROTECTED] writes: if I pipe the output of a cd command the working directory doesn't change. That's because all elements of a pipeline except the last are run in different processes to the main shell that starts the pipeline. As such, the cd command is running in a subshell which exits when cd exits. This has no effect on your main shell, which will keep the original working directory. Is there any way to change this behaviour? Actually, my explanation is wrong. I was confusing it with the behaviour of zsh. According to the bash man page: Each command in a pipeline is executed as a separate process (i.e., in a subshell). This means you cannot run cd in a pipeline. I don't think you can change this behaviour. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Building an i386 kernel on amd64 host with make-kpkg
Steve Kemp [EMAIL PROTECTED] writes: On Sat Jul 12, 2008 at 04:12:56 -, Cameron Hutchison wrote: I just needed to add that --cross-compile - argument and it worked. I spoke too soon. It does not quite work. It builds an amd64 arch package, so I cannot install it on an i386 arch. I use this: setarch i386 make-kpkg --initrd --arch=i386 --revision $revision binary Thanks for that info, Steve. However I still cannot get it to work. When I run make-kpkg with the --arch parameter, it seems to always try to use a cross-compiler - i486-linux-gnu-gcc (or something like that - I've forgotten exactly and its scrolled off the screen now). This happens whether I prefix make-kpkg with setarch i386 or linux32 or not. If I supply the argument --cross-compiler - it uses the correct compiler but always builds a package for the host arch not the target arch. I got around it by hacking make-kpkg itself. It seems if you specify --arch, it will build a deb for the target arch, but only if you dont specify --cross-compiler -. I remove the check for that, and now it works fine. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Building an i386 kernel on amd64 host with make-kpkg
I am trying to build a kernel for my i386 box on my amd64 host. I want to do this because the amd64 box build at least 10 times faster than the target box. I am using make-kpkg to build the kernel package. I am having the following problem: $ make-kpkg --rootcmd fakeroot --arch i386 kernel_image /bin/sh: i486-linux-gnu-gcc: command not found make: *** [debian/stamp-kernel-conf] Error 2 $ I thought that gcc on an amd64 box can build 32-bit binaries without needing a separate cross compiler. Can someone give me some pointers on how to do this? Thanks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Building an i386 kernel on amd64 host with make-kpkg
Cameron Hutchison [EMAIL PROTECTED] writes: I am trying to build a kernel for my i386 box on my amd64 host. I want to do this because the amd64 box build at least 10 times faster than the target box. I am using make-kpkg to build the kernel package. I am having the following problem: $ make-kpkg --rootcmd fakeroot --arch i386 kernel_image Following up with a solution to my own problem: $ make-kpkg --rootcmd fakeroot --cross-compile - --arch i386 kernel_image I just needed to add that --cross-compile - argument and it worked. Information came from: http://wiki.debian.org/DebianAMD64Faq (last question) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Building an i386 kernel on amd64 host with make-kpkg
Cameron Hutchison [EMAIL PROTECTED] writes: Cameron Hutchison [EMAIL PROTECTED] writes: I am trying to build a kernel for my i386 box on my amd64 host. I want to do this because the amd64 box build at least 10 times faster than the target box. I am using make-kpkg to build the kernel package. I am having the following problem: $ make-kpkg --rootcmd fakeroot --arch i386 kernel_image Following up with a solution to my own problem: $ make-kpkg --rootcmd fakeroot --cross-compile - --arch i386 kernel_image I just needed to add that --cross-compile - argument and it worked. I spoke too soon. It does not quite work. It builds an amd64 arch package, so I cannot install it on an i386 arch. The build itself seems to be done correctly for the target arch, but the package is created for the host arch. I would have thought that -arch i386 would produce an i386 arch package, but apparently not. Any ideas? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: X apps don't launch from Xterm
Eric d'Alibut [EMAIL PROTECTED] writes: If I then start an xterm and attempt to launch an X app, such as gqview, or evince, from the command prompt in that xterm, no app launches; this is what I get: --snip-- [EMAIL PROTECTED]:~$ gqview (gqview:26337): Gtk-WARNING **: cannot open display: [EMAIL PROTECTED]:~$ evince (evince:28047): Gtk-WARNING **: cannot open display: [EMAIL PROTECTED]:~$ echo $DISPLAY 0:0. That DISPLAY variable is wrong. It should be :0.0 If you can launch an xterm, then at that point in time $DISPLAY must be correct. Once at the command line, it is wrong though. I would be looking in your .bashrc or equivalent shell rc file (if you use a different shell) and see if you have any manipulation of $DISPLAY in there. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: OpenOffice.org Quick Start
David Staer [EMAIL PROTECTED] writes: Cameron Hutchison wrote: How do you use the OpenOffice.org quick start functionality? I have installed the package openoffice.org-gtk [...] You need to enable it in OpenOffice. Tools Options OpenOffice.org Memory - then check Enable systray Quickstarter. The next time you log in, or possibly sooner, it should be in the notification area. Thanks David. That worked a treat. As soon as I enabled the checkbox and OKed the dialog, the quickstarter icon appeared in the notification area. No need to log out/in. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
OpenOffice.org Quick Start
How do you use the OpenOffice.org quick start functionality? I have installed the package openoffice.org-gtk which says in its description: It also contains a QuickStarter for the notification area. How do I make this QuickStarter work? I could not see any reference to it in the README files that came with the package. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: script makes kernel panic
Dexter Filmore [EMAIL PROTECTED] writes: Got a laptop here (Samsung X22), WinXP Pro and data partition in /dev/sda[23]. Wrote this script to backup both partitions 1:1 to an external USB disk. Teh script itself works absolutely as intended. BUT: I added an entry to GRUB's menu.lst like that: title Windows XP Backup root(hd0,4) kernel /boot/vmlinuz-2.6.18-6-686 root=/dev/sda5 rw init=/sbin/windows_backup initrd /boot/initrd.img-2.6.18-6-686 savedefault When I select that entry, kernel comes up, script is executed alright, but after the script called halt it just sits there and eventually throws a kernel panic - and I have no clue why. By default, halt calls shutdown(8), which signals init(8) to shutdown the system. Since you are not running a standard init, that signal is never received by anything, so halt effectively does nothing. When your script exits, since it is running as process 1 (init) you get a kernel panic when it exits. Init should never exit. Try using halt -f. Check the man page for halt to see if there are any other options you want to use. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: changing xterm colors
Jan Willem Stumpel [EMAIL PROTECTED] writes: Jamie Griffin wrote: Jan Willem Stumpel wrote: You can edit a text file called .Xresources in your home directory (or create it if it does not exist). Put the following lines in the file: xterm*VT100*foreground: green xterm*VT100*background: black xterm*VT100*cursorColor: red I've tried that and it hasn't worked. Not sure what to try next - does anyone have any other ideas? This definitely *should* work. If it does not work something weird is the matter. Or it could be that the line: allow-user-resources is not no /etc/X11/Xsession.options. The .Xresources file will only be loaded if that option is present. To test if your .Xresources are working without logging out and back in, run: $ xrdb -load .Xresources and then start up an xterm. . -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Cannot remove Google maps sidebar
Raj Kiran Grandhi wrote: Cameron Hutchison wrote: When I visit Google maps (http://maps.google.com/) using either Epiphany or Iceweasel on my Debian sid system, I do not get the little triangle button in the map sidebar that collapses it. This only happens for me with Debian. On my systems with Ubuntu, google maps works as expected. Does anyone know why this may be happening? It is a problem with the user agent setting. Google maps does not seem to realize that iceweasel and firefox are practically the same. You can work around this by visiting about:config in iceweasel and replacing iceweasel with firefox for the general.useragent.extra.firefox string. Thank you. I just added Firefox/2.0.0.14 to the end of the string in Epiphany and it is working now. I guess Ubuntu must have changed Epiphany to add the Firefox useragent. Thanks again. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Cannot remove Google maps sidebar
When I visit Google maps (http://maps.google.com/) using either Epiphany or Iceweasel on my Debian sid system, I do not get the little triangle button in the map sidebar that collapses it. This only happens for me with Debian. On my systems with Ubuntu, google maps works as expected. Does anyone know why this may be happening? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Cannot remove Google maps sidebar
Rich Healey wrote: Cameron Hutchison wrote: When I visit Google maps (http://maps.google.com/) using either Epiphany or Iceweasel on my Debian sid system, I do not get the little triangle button in the map sidebar that collapses it. This only happens for me with Debian. On my systems with Ubuntu, google maps works as expected. Does anyone know why this may be happening? At a guess, one of the debian patches is breaking it. Possibly. It is a little strange that the Ubuntu version works though, being based on the Debian version. Try a binary release, or build you own and compare. I'd rather not do that. I prefer to keep to Debian packaged components. As a last ditch effort, I may try that to see if I can isolate the issue, but for now I'll see if anyone on the mailing list knows anything about this. Are you using FF2 or 3b? I am using Epiphany 2.22.1.1-1, and Iceweasel 2.0.0.14-2. These are the versions currently in Debian sid. I've had the problem since December last year (2007) when I re-installed Debian and dropped Ubuntu, so it is not something just introduced. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: ip forwarding woes
David Zelinsky wrote: With this setup, I expect to be able to ping 10.0.0.2 from 192.168.0.2 (and vice versa), with packets routed through the firewall, but it doesn't work. What am I overlooking? It looks like that 10.0.0.2 does not have a route to 192.168.0.0/24 or that 192.168.0.2 does not have a route to 10.0.0.0/24 or both: 192.168.0.2# ip route add 10.0.0.0/24 via 192.168.0.1 10.0.0.2# ip route add 192.168.0.0/24 via 10.0.0.1 If the firewall is the main router for the network then you can set up a default route to it on each of the other hosts: 192.168.0.2# ip route add default via 192.168.0.1 10.0.0.2# ip route add default via 10.0.0.1 You will need to have the iproute package installed to get the ip(8) command. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: chown all files on a data drive
dave N [EMAIL PROTECTED] wrote: I used to run Fedora and now all the files on my data drives are uid 500 and gid 500. Now under Debian the same user name and password I'd previously had are uid 1000 and gid 1000. Though I can access the files on the drive I can't do anything with them except as root. How can I rectify this? chown -R 1000:1000? This'll cause problems with the lost+found as well as any .Trash folders, should I then change the uids and gids back? chown -R will work, but may be a little too indiscriminate. You can be more discriminating by find(1) and only changing the UID of files that are 500, and a GID of 500. $ find /path -uid 500 -print0 | xargs -0 chown 1000 $ find /path -gid 500 -print0 | xargs -0 chgrp 1000 You can combine all this into one command if all files with UID 500 also have a GID of 500, but if not, the above is safer leaving you to remap other IDs as you need to. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Correct way to add multiple secondary IPsto one dev/interface in /etc/network/interfaces?
Adrian Levi [EMAIL PROTECTED] wrote: AFAIK ip, route et al are called my ifconfig to do the work. This is not right. ifconfig uses the old ioctl interface to control the network interfaces. ip uses the new netlink protocol. Creating an alias in the interfaces file is the correct way to do it. Why do you not want to use an alias? According to various ip(8) documents, aliases are deprecated in favor of secondary addresses. This may just be wishful thinking on the part of the authors of the iproute2 utility suite and kernel implementation, since most distros still use ifconfig as the primary utility to configure network interfaces. However, if we are to have any chance of moving forward, we need to drop the old ways of doing things for the new. I assume that this is what the original poster is trying to do. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Correct way to add multiple secondary IPsto one dev/interface in /etc/network/interfaces?
Orig-To: Jonathan Wilson [EMAIL PROTECTED]
Here's the most important example from the original post. It works as-is.
If I uncomment the commented up and down lines, everything breaks.
Can you describe in more detail what everything breaks means?
The internet still works for me, so it cannot be everything :-)
When you run ifup eth0 on the command line, do you get any output?
Can you run ifup -v eth0 to get verbose output?
What scripts do you have in /etc/network/if{,-pre}-up.d
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Correct way to add multiple secondary IPsto one dev/interface in /etc/network/interfaces?
Jonathan Wilson [EMAIL PROTECTED] wrote: Can you describe in more detail what everything breaks means? The internet still works for me, so it cannot be everything :-) But it is everything, ... I was trying to make my point in a funny way. I'm assuming your fridge still works and your cat didn't keel over and die, so not everything broke. So, just what is the scope of everything breaks? ... even then /etc/init.d/network start won't work. In all my years of using debian, I've never manually run /etc/init.d/network. I have always just run ifup/ifdown on the interfaces I am playing with. I dont know what side-effects the init script will have. Of course, when you're happy that it seems to work, its worth rebooting to ensure that everything comes up properly from the init script, but that's usually a last step for me (if I do it at all. If ifup works, its always worked on reboot for me). I guess what I'm trying to say is to test just using ifup/ifdown so you can isolate any problems to the smallest action needed to cause them. I asked on freenode IRC and simonrvn told me not to use up and down but to use post-up and pre-down instead. These are just aliases for up and down, I guess to make it clearer just when they run. That's according to the sid source code anyway. I can't see how it can make any difference using one or the other. allow-hotplug eth0 iface eth0 inet static address nn..179.107 netmask 255.255.255.224 broadcast nn..179.127 gateway nn..179.97 post-up ip addr add nn..179.108/27 dev eth0:0 post-up ip addr add nn..179.109/27 dev eth0:1 pre-down ip addr del nn..179.108/27 dev eth0:0 pre-down ip addr del nn..179.109/27 dev eth0:1 # dns-* options are implemented by the resolvconf package, if installed dns-nameservers 192.168.0.8 192.168.0.16 dns-search graphics claborn.net It seems to be working, at long last. The only difference between that and your previous config is that you have removed the 'label' argument and the broadcast argument (and the post-up/pre-down bit). Have you tried running the commands manually that ifup would run? (running ifup -v should show you what it runs). This would show you where it fails. Cameron: From your statement The internet still works for me I assume you use something like this. May I see an example of your conf please? Where did you find documentation for it? The internet still works for me was just a joke based on your everything breaks statement. However, I did try a similar setup to yours and I had no problems. I am running sid though. When you next get a chance, its worth running that ifup -v command to see just where it fails. Right now, I can't see what might have been wrong. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: ntpd restart on IP address change the Debian way
Bob wrote:
Bob wrote:
Is there such a thing?
When my firewall / dhcp server / ntp server gets a fresh IP address
from my ISP the ntp daemon stops responding to requests.
Is the silence because it's a stupid question or because there isn't a
preferred work around for this?
How does your firewall get its IP address?
If it is by PPP then you can scripts in /etc/ppp/ip-{up,down}.d to start
and stop the ntp daemon. This should get triggered when a new IP address
is negotiated.
Check the man page for pppd(8) and search for ip-up to see the details
of how it's used, but just dropping scripts in those directories should
be sufficient.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: how to build from _modified_ source package
Daniel B. wrote: Are there any instructions for proceeding from having downloaded the source package files and _not_ having unpacked things? (I think my current state is as if I had done apt-get source --download-only xfree86 (I didn't actually do --download-only, but from cleaning up after trying a couple things I think that's the state I'm in.)) Just run apt-get source xfree86 in the directory where you've downloaded the existing files. apt-get will not download them again if it does not need to, and will unpack the source files. You can then edit the source and build it as described in other messages in this thread. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Recommended MPLAYER config for old computers
Amit Uttamchandani [EMAIL PROTECTED] wrote: What happens if you run cdck to check the quality of the DVD? I installed cdck from the deb repo and ran it. A lot of errors. There are a lot: ! unable to read sector 195007, reason: Input/output error ! unable to read sector 195008, reason: Input/output error ! unable to read sector 195009, reason: Input/output error ! unable to read sector 195010, reason: Input/output error ! unable to read sector 195203, reason: Input/output error ! unable to read sector 195204, reason: Input/output error ! unable to read sector 195205, reason: Input/output error This may be because of new forms of copy protection they are adding to DVDs. A DVD player navigates through the disc based on the structures in the VOB files, so it can be told to skip blocks on the disc. If you try to read the disc as a UDF filesystem, you'll hit the blocks that the VOB navigation skips. The copy protection works by making those blocks bad so you cannot do a block by block copy of the disc. Can you turn the DVD into an iso or something file and play from hard disk? How do I do this? mkisofs? Or can I use mencode or dvdbackup? What do you recommend? If you get a recent version of vobcopy (http://vobcopy.org/), it should be able to make a copy of the disc which can then be played off the HDD. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: LVM partition full (was:what is /command directory?)
Richard Lyons [EMAIL PROTECTED] wrote: I also notice a number of directories I've never heard of before under root: command, package, service Still, it was late and I was panicking rather unnecessarily. After sleeping on it, I am tending to the view they must be from that tinydns episode. I'm going to remove them on that assumption and see what happens. I never persuaded tinydns to run, so it is no loss. They will be from tinydns. The /command, /package and /service directories are what Dan Bernstein uses in his software (qmail, djbdns, daemontools, etc). If you dont have any of his software installed, you can safely remove these directories. DJB software typically has service installation software included and that creates those directories. The debian packages also include a -fhs variant that puts the files in FHS compliant places. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Good fdisk Practices
David Brodbeck [EMAIL PROTECTED] wrote: I'd always heard that swap files are slower than swap partitions. Is that a myth? Not a myth, just old information. It used to be the case that swap files were slower than swap partitions, but this stopped being true sometime around kernel 2.4 Also, is there any good reason to have a separate /boot on a modern system? I always thought /boot was just a kludge to get around old BIOSes that couldn't load anything that wasn't on the first part of the disk. I tend to just combine /boot and / on my newer systems -- am I taking some kind of risk by doing so? I do the same and have had no problems. This may restrict the type of filesystem you can use on your root partition though. I dont think GRUB can load a kernel from an XFS filesystem, so by separating root and /boot, you can put ext2/3 on /boot and something else on root. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: undeleteable file
Frank McCormick [EMAIL PROTECTED] wrote: This is the output of ls -l [EMAIL PROTECTED]:/lost+found/#573699/drivers/i2c/busses$ ls -l * ---s---r-t 1 993200132 3086322235 0 1931-09-13 15:22 i2c-nforce2.ko What is the output of lsattr ? It could be that it has the immutable bit set. To fix this, use chattr(1): # chattr -i i2c-nforce2.ko # rm i2c-nforce2.ko -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Comparing files in two directories
L.V.Gandhi wrote: On 6/13/07, Keith Christian [EMAIL PROTECTED] wrote: L.V.Gandhi wrote: I have two directories A and B. In each directory, I have nearly 1000 files with same names. I would like to compare both directories and find out which files differ more than say 5 lines. I use kompare and see manually. How to do it in command line easily? Try these diff commands. The verbosity is greater in the ones with the u option. Thanks for your effort and time. All these leads to get list of lines differing in each file. Once again I have to go through all lines to find out total number of lines differing in file X ib both folders A and B. You can then pipe the output of diff into the diffstat program (in the package of the same name) which will tell you how many lines have been added and removed (modified too, but that always seems to come up zero). That will give you a summary of differences, one line per file changed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: delete lines that contain duplicated column items
Jeff Zhang wrote: On 4/3/07, Ron Johnson [EMAIL PROTECTED] wrote: On 04/03/07 07:39, Jeff Zhang wrote: [snip] I just care about duplicated ones in column 2, if so, to delete the line. Does it matter which line is deleted? no, have a unique column 2 and keep undeleted lines to its original format will be ok. $ sort -u -k2 filename -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
