Re: qemu-kvm and glusterfs support
Hello there, On Fri, Apr 24, 2015 at 07:50:59AM +0300, Risto Paavola wrote: In Debian Jessie, why qemu-kvm has been compiled without glusterfs network disk type support? Until glusterfs packaging is sane I think it's not a good idea to introduce glusterfs support. Also this is not really easy thing, because it brings us several non- trivial dependencies to the binary packages and users already complained about too much dependencies qemu has. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=775431 (actually, this bugreport came too late to affect Jessie; but this is where the maintainer states his position) However, if you need a Debian-based system with recent qemu-kvm and support for various network disk types, I'd recommend Proxmox. I haven't used it with glusterfs for quite some time (preferring ceph instead), but it appears well-integrated. HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20150424074558.gq25...@fernst.no-ip.org
Re: FW: hp server hardware monitoring
On Fri, Aug 15, 2014 at 09:30:24PM +, Bonno Bloksma wrote: [...] My apologies, I forgot that you need a newer version of cciss_vol_status than present in Squeeze/Wheezy. For that I had imported 1.10-1 into our local repository, which works just fine without any adjustments / backporting. Where did you get de the 1.10-1 version of cciss-vol-stat as a Debian package if not from the backports? I just looked and it is not in the Debian wheezy-backports. Did you compile and build it yourself? I took the version which was present at that time in Jessie/Sid. This version didn't need any adjustment / backporting to work in Wheezy. The version currently present in Jessie/Sid requires a newer libc6 than can be fulfilled in Wheezy, so I'd now recommend snapshot.debian.org to retreive a suitable version. HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140816152901.gl20...@fernst.no-ip.org
Re: hp server hardware monitoring
Hello, [ no need to CC me ] On Wed, Jul 30, 2014 at 07:50:28PM +, Bonno Bloksma wrote: [...] root@vz02:~# cciss_vol_status /dev/sda /dev/sda: (Smart Array P420i) RAID 5 Volume 0 status: OK. But I get... linein:~# cciss_vol_status /dev/sda cciss_vol_status: /dev/sda: Unknown SCSI device. My apologies, I forgot that you need a newer version of cciss_vol_status than present in Squeeze/Wheezy. For that I had imported 1.10-1 into our local repository, which works just fine without any adjustments / backporting. However, apart from that, Claudio provides good advice. If your iLO works reliably, that is. I had several systems where every once in a while it simply didn't react anymore / worked too slowly and had to be reset. And last I checked there wasn't even a viable patch to remedy that, but that might have changed in the meantime. Of course you need a new enough iLO to monitor disks, so at least in my environment I cannot fully rely on this (yet). Cheers, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140731073700.gh26...@fernst.no-ip.org
Re: hp server hardware monitoring
Hello, On Wed, Jul 30, 2014 at 05:28:10AM +, Bonno Bloksma wrote: [...] What may be relevant too is that on the g6 server Debian uses the CCISS drivers for the raid hardware, the volume shows up as /dev/cciss/c0d0 On the g7 and g8 hardware the raid volume simply shows up as /dev/sda cciss has been superseded by hpsa, The hpsa driver is intended to supplant the cciss driver for newer Smart Array controllers., cf. https://www.kernel.org/doc/Documentation/scsi/hpsa.txt. How can I get this to work under a g7 or g8 server. Do I need a newer version of the package of do I need a different package? I generally avoid installing third-party tools on the bulk of my servers, and for simple monitoring the packages cciss-vol-status and nagios-plugins-standard from Debian will suffice: root@vz02:~# lsmod | grep -e cciss -e hpsa hpsa 50787 2 root@vz02:~# cciss_vol_status /dev/sda /dev/sda: (Smart Array P420i) RAID 5 Volume 0 status: OK. root@vz02:~# /usr/lib/nagios/plugins/check_raid WARNING: cciss: No Smart Array Adapters were found on this machine root@vz02:~# modprobe sg root@vz02:~# /usr/lib/nagios/plugins/check_raid OK: cciss:[/dev/sda: (Smart Array P420i) RAID 5 Volume 0 status: OK, /dev/sda: (Smart Array P420i) RAID 5 Volume 0 status: OK]; smartctl:[/dev/sg0#0,/dev/sg0#1,/dev/sg0#2,/dev/sg0#3,/dev/sg0#4,/dev/sg0#5,/dev/sg0#6,/dev/sg1#0,/dev/sg1#1,/dev/sg1#2,/dev/sg1#3,/dev/sg1#4,/dev/sg1#5,/dev/sg1#6: OK] HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140730075116.ge26...@fernst.no-ip.org
Re: Chromium cannot access pages.
On Tue, Jun 17, 2014 at 05:30:44PM +0200, Sven Joachim wrote: On 2014-06-17 14:39 +0200, Florian Ernst wrote: On Tue, Jun 17, 2014 at 12:10:21PM +, Hendrik Boom wrote: [...] For every page I request, even the default Google-search page it normally displays in a new tab, chromium tells me : Aw, Snap! [...] Even the 'about chromium' page fails to display, so it won't tell me what version of chromium I'm using. Aptitude tells me it's version 35.0.1916.153-1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=751294 35.0.1916.153-2 contains a fix Yes. and will soon migrate to testing. No, it won't because it FTBFS on amd64. Based on previous experience¹, it's likely to take weeks before a fixed package reaches testing. :-( Ah, true, I failed to notice that. Yikes, ld terminated with signal 9, indicating external trouble rather that a problem with the source per se. However, in the meantime brahms managed to successfully build the package. The webpages haven't been fully updated yet, but chromium_35.0.1916.153-2_amd64.deb can be found in incoming. Cheers, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140618084541.gg27...@fernst.no-ip.org
Re: Chromium cannot access pages.
Hello there, On Tue, Jun 17, 2014 at 12:10:21PM +, Hendrik Boom wrote: [...] For every page I request, even the default Google-search page it normally displays in a new tab, chromium tells me : Aw, Snap! [...] Even the 'about chromium' page fails to display, so it won't tell me what version of chromium I'm using. Aptitude tells me it's version 35.0.1916.153-1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=751294 35.0.1916.153-2 contains a fix and will soon migrate to testing. HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140617123953.gd27...@fernst.no-ip.org
Re: Can't patch Heartbleed bug?
On Fri, Apr 11, 2014 at 08:33:38AM +1000, Charlie wrote: I'm already to g on Jessie. Is that good? openssl: Installed: 1.0.1g-2 Candidate: 1.0.1g-2 Version table: *** 1.0.1g-2 0 This is good as in if you have restarted all processes using vulnerable parts of OpenSSL after updating all OpenSSL packages then you are not vulnerable anymore by Heartbleed. Cheers, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140411074357.gh7...@fernst.no-ip.org
Re: Can't patch Heartbleed bug?
On Thu, Apr 10, 2014 at 10:14:59AM -0400, Dan Ritter wrote: On Thu, Apr 10, 2014 at 03:54:38PM +0200, Florian Ernst wrote: On Thu, Apr 10, 2014 at 09:18:00AM -0400, Brad Alexander wrote: I don't believe that Wheezy was vulnerable to Heartbleed. It was only the 1.0.1f (committed 31 Dec 2011) that incorporated the vulnerable heartbeat feature. My wheezy box has 1.0.1e: [...] So you shouldn't have anything to worry about. This is not accurate, OpenSSL 1.0.1 through 1.0.1f (inclusive) are vulnerable. Please see https://www.debian.org/security/2014/dsa-2896 Which says: For the stable distribution (wheezy), this problem has been fixed in version 1.0.1e-2+deb7u5. and then later this was upgraded to 1.0.1e-2+deb7u6. Looking at the 1.0.1e is not sufficient. True, this is https://www.debian.org/security/faq#version. I think the link to the DSA explains which Debian version is not vulnerable anymore, contrary to the general statement. Cheers, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140411074804.gi7...@fernst.no-ip.org
Re: Using wheezy-backports, advice needed
On Fri, Apr 11, 2014 at 04:58:56PM +0100, Martin Read wrote: As Lisi Reisz notes, findutils is not in wheezy-backports. A search on packages.debian.org reveals that a package of findutils 4.5.12 was created for experimental on 2013-09-28, but for whatever reason the maintainer has not acted to put any of the 4.5.x versions (the first 4.5.x version to land in experimental was 4.5.7, created on 2010-04-03) of findutils into unstable. findutils 4.5.x is considered alpha by its developers. As such, the Debian package maintainer understandably only uploads those releases to experimental. The latest upstream release still is 4.4.2, cf. http://ftp.gnu.org/gnu/findutils/ http://alpha.gnu.org/gnu/findutils/ @Paul: Has the bug you encountered already been reported to the BTS? If so, you might want to add some detail to it, e.g. the hint that it is supposedly fixed in 4.5.10. If not, you might want to file a bugreport. HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140411165302.ga...@fernst.no-ip.org
Re: Can't patch Heartbleed bug?
On Thu, Apr 10, 2014 at 03:00:52PM +0200, Erwan David wrote: Le 2014-04-10 14:56, Dr. Jennifer Nussbaum a écrit : I'm running Debian Wheezy 7.4 on a server in Amazon's EC2, that i installed, recently, from the official Debian AMI. I havent made any changes to the package infrastructure. I'm trying to fix the Heartbleed bug, but my system seems to think everything is up to date. [...] I've waited a day in case theres some issue with the m irrors not getting updated, but this still happens. Whats the right way to make my system safe? All the instructions Ive seen say apt-get update apt-get upgrade will do it, but not in my case! Add the security repositories to your mirror lists ? deb http://security.debian.org/ stable/updates main contrib non-free deb-src http://security.debian.org/ stable/updates main contrib non-free Ack, please see https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=743892 for background info. HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140410131323.gd7...@fernst.no-ip.org
Re: Can't patch Heartbleed bug?
On Thu, Apr 10, 2014 at 09:18:00AM -0400, Brad Alexander wrote: I don't believe that Wheezy was vulnerable to Heartbleed. It was only the 1.0.1f (committed 31 Dec 2011) that incorporated the vulnerable heartbeat feature. My wheezy box has 1.0.1e: [...] So you shouldn't have anything to worry about. This is not accurate, OpenSSL 1.0.1 through 1.0.1f (inclusive) are vulnerable. Please see https://www.debian.org/security/2014/dsa-2896 as well as http://heartbleed.com/ Cheers, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140410135438.ge7...@fernst.no-ip.org
Re: Can't patch Heartbleed bug?
On Thu, Apr 10, 2014 at 02:49:27PM +0100, Lisi Reisz wrote: lisi@Tux-II:~$ dpkg-query -l openssl Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad) ||/ Name Version Architecture Description +++--=-=-=== ii openssl 1.0.1e-2+deb7 amd64 Secure Socket Layer (SSL) binar lisi@Tux-II:~$ No u-anything. Try a wider window, or simply COLUMNS=200 dpkg-query -l openssl, or use apt-cache policy openssl. HTH, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140410135737.gf7...@fernst.no-ip.org
Re: gnutls security breach
Hello all, On Wed, Mar 05, 2014 at 03:10:59AM -0500, Ric Moore wrote: Anyone see this? http://arstechnica.com/security/2014/03/critical-crypto-bug-leaves-linux-hundreds-of-apps-open-to-eavesdropping/ Good thing Red Hat caught it: https://rhn.redhat.com/errata/RHSA-2014-0246.html For Debian, see https://security-tracker.debian.org/tracker/CVE-2014-0092. Yeow! I just did update / upgrade to Jessy, but didn't see the security fix come through yet. Ric Debian released a security advisory on the same day, cf. https://www.debian.org/security/2014/dsa-2869. The update for Jessie, which still is the current testing release and as such doesn't necessarily receive timely updates, is pending (currently blocked by not being built on all architectures). Cheers, Flo -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20140305082607.ga19...@fernst.no-ip.org
Persistent remote X applications with xpra (was: Remote xboard without Desktop Environment)
Hello there, On Wed, Nov 20, 2013 at 01:33:21PM -0200, Beco wrote: [...] (3) Now, how can I keep xboard running on lizard, after I turn off fox. TTBOMK (being former Debian package maintainer of xboard) xboard will need a DISPLAY to run interactively. ssh -X was already mentioned, but if the connection gets lost so does any X application running in ssh -X. screen won't help due to the X requirement. VNC and/or RDP (e.g. via xrdp) would allow persistent X applications, but both have some overhead you'd prefer to avoid (and I personally dislike both of them for the fixed display size of a running session). So far so bad. xpra might help, though. Consider the following sample session. fernst is my powerful (ahem...) workstation, c01n01v01 is a remote virtual server I just had running (still running Squeeze w/ backports, but still enough to illustrate my point). Starting xboard on c01n01v01: :) fe@fernst:~$ ssh fe@c01n01v01 Linux c01n01v01zfs 2.6.32-26-pve #1 SMP Mon Oct 14 08:22:20 CEST 2013 x86_64 The programs included with the Debian GNU/Linux system are free software; the exact distribution terms for each program are described in the individual files in /usr/share/doc/*/copyright. Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent permitted by applicable law. fe@c01n01v01zfs:~$ xpra start :8 Entering daemon mode; any further errors will be reported to: /usr/users/support/fe/.xpra/c01n01v01zfs-8.log fe@c01n01v01zfs:~$ DISPLAY=:8 xboard -ncp [1] 29158 fe@c01n01v01zfs:~$ ps auxwwkstart_time USER PID %CPU %MEMVSZ RSS TTY STAT START TIME COMMAND root 1 0.0 0.1 8360 804 ?Ss Nov20 0:00 init [2] root 2 0.0 0.0 0 0 ?SNov20 0:00 [kthreadd/150] root 3 0.0 0.0 0 0 ?SNov20 0:00 [khelper/150] root 340 0.0 0.2 60436 1492 ?Sl Nov20 0:00 /usr/sbin/rsyslogd -c4 root 412 0.0 0.1 21160 972 ?Ss Nov20 0:00 /usr/sbin/cron 102704 0.0 0.1 44144 1004 ?Ss Nov20 0:00 /usr/sbin/exim4 -bd -q30m root 728 0.0 1.4 53000 7672 ?Sl Nov20 0:14 /usr/bin/python /usr/bin/fail2ban-server -b -s /var/run/fail2ban/fail2ban.sock root 28523 0.0 0.2 49180 1136 ?Ss 09:23 0:00 /usr/sbin/sshd root 29136 0.0 0.6 70464 3288 ?Ss 09:27 0:00 sshd: fe [priv] fe 29138 0.0 0.3 70464 1596 ?S09:27 0:00 sshd: fe@pts/0 fe 29139 0.0 0.3 17712 1908 pts/0Ss 09:27 0:00 -bash fe 29148 0.8 4.1 185000 21664 ?S09:27 0:00 /usr/bin/python /usr/bin/xpra start :8 fe 29149 2.0 9.7 94236 51160 ?S09:27 0:00 Xvfb-for-Xpra-:8 +extension Composite -screen 0 3840x2560x24+32 -nolisten tcp -noreset -auth /usr/users/support/fe/.Xauthority :8 fe 29158 0.6 0.6 45828 3508 pts/0S09:27 0:00 xboard -ncp fe 29161 0.0 0.1 14820 1008 pts/0R+ 09:27 0:00 ps auxwwkstart_time fe@c01n01v01zfs:~$ logout Connection to xxx.x.xxx.xx closed. :) fe@fernst:~$ Now connecting to the still running xboard within my xpra session: :) fe@fernst:~$ xpra attach ssh:fe@c01n01v01:8 Attached (press Control-C to detach) The xboard window now appears in my regular DE on fernst. I can interact with it as usual (including resizing it to my liking), and finally detach (but keep running) as indicated: ^Creceived SIGINT, closing Connection lost :) fe@fernst:~$ I can reattach again at will. And if not required anymore the xpra session running on my server can be terminated via fe@c01n01v01zfs:~$ xpra stop :8 HTH, Flo -- 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/20131121083045.gb30...@fernst.no-ip.org
Re: Nullmailer and Linux Resource Containers
Hello all, On Thu, Oct 17, 2013 at 07:26:26AM -0500, Bryan K. Walton wrote: [...] So, my question is this: is this an issue with LXC? Or with Nullmailer? Is this a limitation of Linux Resource Containers or should I be able to do run daemons on both the host/guest at the same time? nullmailer: PID check fails on lxc host (and probably openvz and other VM) http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=687827 Yes, this is annoying, and should be fixed in nullmailer's initscript. HTH, Flo -- 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/20131017130145.ga14...@fernst.no-ip.org
Re: JSON::XS error
Hello all, On Wed, Oct 09, 2013 at 05:16:27AM -0400, Brad Alexander wrote: [...] This is on an i386 sid container (openvz) running on proxmox-ve (wheezy). When I try to install the thruk package (thruk_1.76-3_debian8_i386.deb), I get the following error message: Perl API version v5.14.0 of JSON::XS does not match v5.18.0 at [...] I have two JSON/XS.pm on the system, one from thruk and one from the thruk package, and the other from libjson-xs-perl. On my test system these both merely differ in perldoc. They don't need to be updated. Can anyone suggest how to get past this error? Consol needs to update their packages they provide for Jessie for Perl API v5.18.0. Most probably a rebuild will suffice, but that needs to be checked. I'm afraid there isn't much you can do at the moment but wait / prod Consol. Oh, and I'd refrain from needlessly installing CPAN packages. HTH, Flo -- 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/20131009102208.ga26...@fernst.no-ip.org
Re: JSON::XS error
Hello all, On Wed, Oct 09, 2013 at 08:58:20AM -0400, Brad Alexander wrote: Okay. So how do you rebuild the module? It looks like that is up to Consol. I just tried rebuilding the package myself but failed (first due to missing Build-Depends, then local libraries were not found during build, after which I stopped; cf. [0]). At the moment Consol provides packages that are uninstallable and don't build in a clean build environment. Anyone is free to debug their build setup, but I guess they could update the package best themselves. Maybe they just need a gentle prod. Cheers, Flo [0] relevant lines from my build log: 15:57:07 O: mv /tmp/buildd/thruk-1.76/debian/thruk/usr/share/thruk/local-lib /tmp/buildd/thruk-1.76/debian/thruk/usr/lib/thruk/perl5 15:57:07 O: mv: cannot stat '/tmp/buildd/thruk-1.76/debian/thruk/usr/share/thruk/local-lib': No such file or directory 15:57:07 O: make: *** [install] Error 1 15:57:07 O: dpkg-buildpackage: error: fakeroot debian/rules binary gave error exit status 2 15:57:07 O: E: Failed autobuilding of 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/20131009141137.gb1...@fernst.no-ip.org
Re: BIND RRL
Hello all, On Sat, Oct 05, 2013 at 03:05:02PM +0200, pch0317 wrote: Is DDoS Defense Module for BIND RRL available in Debian 7.1? If by that you mean the patches from http://www.redbarn.org/dns/ratelimits then yes, they are included in Wheezy and additionally available via squeeze-backports. HTH, Flo -- 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/20131009144406.ga10...@fernst.no-ip.org
Re: zfs not detecting drive failure
Hello there, On Mon, Sep 02, 2013 at 07:21:10PM +0500, Muhammad Yousuf Khan wrote: but there is a problem when i restart the machine folder on the root (name as pool name) of file system can b found however zfs partition is not mounted. i have to run zfs mount -a to remount the partition. what can be done to mount the zfs partition on boot. i know running this command /etc/rc.local would probably work but do you guys think is it a work around or proper way to mount the zfs on boot. Take a look at /etc/default/zfs and, for clarification, at /etc/init.d/zfs-*. However, this way zfs datasets will be mounted rather late during boot. That is fine for mere data storage, but won't suffice for storing e.g. /var. For that you could enter something along $local_fs +zvol +zfs to a new file /etc/insserv.conf.d/zfs.conf, followed by an insserv call. Hmm, you mentioned you were testing on Ubuntu. Methinks there was a zfs-mountall package available that might ease the task, but I'm not sure about its availability ... one more issue that i have found. i had smaller files 2 GB each. which i replace by 4GB files so that i could test if i need to increase the size of storage at some time later. eventually all the files has been replace with new one successfully however when i pass the command df -h it showed me the old partition size. how can i re claim the remaining empty partition. zpool online -e pool device..., and furthermore you can set the autoexpand property. See man zpool for details. BTW, I found the ZFS on Linux User Guide as linked from http://zfsonlinux.org/docs.html quite helpful, you might want to check it out. I'm running ZFS in production use on Squeeze and Wheezy, and it greatly helped me to get some things sorted. HTH, Flo -- 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/20130902173539.ga30...@fernst.no-ip.org
Re: Nicer way to manage installed build dependencies?
Hello all, On Sun, Mar 17, 2013 at 03:03:51PM +, Claudius Hubig wrote: [...] Are there any obvious other ways to solve this problem? I managed to put together a small shell script[1] that creates a dummy/meta package, but I’d really prefer something ‘official’. There is mk-build-deps in the package devscripts, and there once was sourcedeps.debian.net (now dysfunctional) which allowed to have a sources.list for Build-Depends-metapackages created using mk-build-deps. I don't know when and why the latter service was discontinued, so that doesn't help much for now, though ... Cheers, Flo -- 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/20130318083706.gc17...@fernst.no-ip.org
Re: Chromium update from 22 to 24 in testing
Hello all, On Tue, Feb 12, 2013 at 11:09:05AM +0100, Gernot Super wrote: Just upgraded wheezy today and noticed that chromium got updated from version 22 to 24. Can anybody give some hints why this happened (while in deep freeze)? Debian-devel gives no info about it. Updated as requested by jmm from the Debian Security Team, cf. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=700331 While it is true that in a deep freeze (as well as after release) such major updates usually won't take place, there are some exceptions. HTH, Flo -- 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/20130212103054.ga20...@fernst.no-ip.org
Re: multiarch - please do not force users to change a running system!
Hello all, just a rather minor nitpick amidst fine explanations: On Fri, Dec 14, 2012 at 11:21:56AM +0200, Andrei POPESCU wrote: [...] Compare 'apt-cache depends ia32-libs-i386' with the dependencies of the i386 package. The new ia32-libs (transition package) indirectly (via ia32-libs-i386) depends on *all* 32bit libraries previously contained in ia32-libs. Actually, some LDAP libraries the old-style ia32-libs used to depend on have been downgraded to a mere Recommends in order to avoid confusing debconf question on new installations. For details, see http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=684029. The corresponding changelog entry mentions: if you need them you should know about them anyway. :) It's possible you don't need them all, and even if you did, the direct dependency is usually better. True. True. Cheers, Flo signature.asc Description: Digital signature
Re: mariaDB in Debian
Hello there, On Sun, Nov 25, 2012 at 08:58:58PM +0100, Gour wrote: I've started using Tiki which supports both MySQL and MariaDB and seeing that the latter is not in the official Debian repos, I wonder what is the plan of including it and/or possibly replacing MySQL with it? The plans to include MariaDB are tracked in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565308 Given how much pain MySQL causes the security team (due to the inability to get the code changes corresponding with a single security issue) opinions have been voiced to fully replace MySQL with MariaDB, but I don't know of any definitive outcome of these discussions. Cheers, Flo -- 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/20121125213934.ga14...@fernst.no-ip.org
Re: Debian Package Version system
Hello all, On Thu, Nov 22, 2012 at 09:54:22AM +0100, Arnoud Tijssen wrote: After performing some vulnerability scans on some our systems one of the outcomes was that some software packages were out of date. We`re using the package management system of Debian and all packages were updated (apt-get update apt-get (dist-)upgrade) prior to the scan. Such scans often merely compare version numbers, which most often isn't quite appropriate to determine whether a certain vulnerability still exists. Please see The version number for a package indicates that I am still running a vulnerable version! in the Debian Security FAQ at http://www.debian.org/security/faq#version The remainder of that page provides further insight into some of the peculiarities involved. Cheers, Flo -- 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/20121122161341.gr14...@fernst.no-ip.org
Re: Using multiarch on wheezy
Hello there, On Sun, Nov 04, 2012 at 11:31:12AM +0100, Florian Weimer wrote: I'm trying to install the i386 version of openjdk-7-jre on an amd64 system. So I did: # dpkg --add-architecture i386 # apt-get update This appears to have been successful. But I still cannot install packages: [...] The following packages have unmet dependencies: emacs23 : Depends: libgif4 (= 4.1.4) but it is not going to be installed E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages. This does not make much sense. A proper version of libgif4 is available: [...] It is also installed for amd64: [...] Why does apt-get complain about dependency, which is already fullfilled? giflib hasn't been converted to multi-arch yet, cf. http://bugs.debian.org/647497. As such, you can now install either the amd64 version or the i386 version, but not both. The dependency apt-get complains about stems from the Conflicts of the package on itself, which would be broken if trying to have both versions installed at the same time. As such, once you install libgif4:i386, apt would try to remove libgif4:amd64 which in turn is something emacs23:amd64 doesn't quite like. Yes, it seems apt and friends need some more helpful output when dealing with multi-arch, and methinks some people work on that. Using aptitude myself, I most of the time didn't have much trouble seeing where it breaks, though ... HTH, Flo -- 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/20121104114651.gi29...@fernst.no-ip.org
Re: Catastrophe on Bootup
Hello all, On Thu, Nov 01, 2012 at 08:23:53AM +0200, David Baron wrote: [...] sudo /etc/init.d/udev [whatever...] Signal 11 (SEGV) caught by ps (procps-ng version 3.3.4). ps:display.c:59: please report this bug [warn] udev does not support containers, not started ... (warning). While I can get to my data, system is useless! What to do? Running Debian Sid, up-to-date. JFTR, this is dealt with in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=692063. Cheers, Flo -- 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/20121102090321.gc29...@fernst.no-ip.org
Re: How APT signs packages
Hello there, On Fri, Oct 19, 2012 at 01:14:44PM +0300, Lars Nooden wrote: On Fri, 19 Oct 2012, Darac Marjal wrote: [...] [1] http://wiki.debian.org/SecureApt Thanks. The weak point, relatively speaking, looks to be the MD5 checksums in Releases. The link above [1] says MD5 is now a broken hash function, and should be replaced for all security-minded usages. Out of curiosity, what are the plans then for moving up to SHA256 or better? There aren't any. That is, there aren't any such plans *anymore*, as SHA256 is already in use and that page is partially misleading, cf. - 8 - What does it mean for md5sum to be broken? Since it's a checksum, I thought the only way it can be broken is that it fail to compute the proper checksum. I have a feeling some other meaning is intended. --RossBoylan **it is broken as people were able to actually create a fake certificate that could sign anything and was trusted, they did this by finding a collision, they created a certificate that had the same md5 sum as the certificate they were issued, and where thereby able to give themselves right other than they were granted.--Scientes ***apt has supported sha256 checksums since version 0.7.7, so these will be used in lenny and future releases. --JoeyHess - 8 - in the comments of the very same page as well as check your /var/lib/apt/lists/*_{Release,Packages} for verification. Cheers, Flo -- 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/20121019102759.gd21...@fernst.no-ip.org
Re: Checking for kernel freshness
Hello all, On Thu, Apr 12, 2012 at 06:24:15PM +, Bernhard Schmidt wrote: [...] Fortunately there is /proc/version ... Thanks for everyone involved, I'm off writing checkscripts. FWIW, please check the attached script. HTH, Flo #!/usr/bin/perl -w # fe: ripped out from # # apt-dater - terminal-based remote package update manager # # $Id: apt-dater-host 352 2009-05-19 15:55:49Z liske $ # # Authors: # Andre Ellguth ellg...@ibh.de # Thomas Liske li...@ibh.de # # Copyright Holder: # 2008-2009 (C) IBH IT-Service GmbH [http://www.ibh.de/apt-dater/] # # License: # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this package; if not, write to the Free Software # Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA # use strict; my $CMD = shift; $ENV{'LC_ALL'} = 'C'; my $exitcode = do_kernel; exit $exitcode; sub do_kernel() { my $UNKNOWN = 3; my $infostr = 'KERNELINFO: Installed'; my $verfile = '/proc/version'; my $version = `uname -r`; chomp($version); unless(-r $verfile) { print $infostr $version, but could not read $verfile\n; return $UNKNOWN; } $verfile = `cat $verfile`; chomp($verfile); unless($verfile =~ /^\S+ \S+ \S+ \(Debian ([^\)]+)\)/) { print $infostr $version, but no comparison possible against apparent non-Debian kernel $verfile\n; return $UNKNOWN; } my $vers = $1; print $infostr $vers; my $reboot = 0; open(HDPKG, dpkg-query -W -f='\${Version} \${Status;20} \${Maintainer} \${Provides}\n' 'linux-image*'|grep 'install ok installed Debian Kernel Team'|grep linux-image|) or die Could not fork: $!\n; while(HDPKG) { next unless (/^(\S+)\s/); unless (system(dpkg, --compare-versions, $vers, lt, $1) 8) { $reboot = 1; print , available: $1; } } close(HDPKG); print \n; return $reboot; }
Re: Cron, 1 month but 2 months
Hello there, On Fri, Mar 09, 2012 at 05:09:23AM +, T o n g wrote: Is there any easy way to schedule a job for a duration that is greater than one month but shorter than 2 months? Any duration will do, as long as it is easy to schedule. Easy? Hmmm. 0 12 * * 1 [ $(expr $(/bin/date +\%s) / 60 / 60 / 24 / 7 \% 8) -eq 3 ] myscript works for me well enough: runs every Monday at noon and checks whether another eight weeks have passed, and if so executes myscript. By adjusting the -eq 3 the starting date for the eight week interval can be shifted around. HTH, Flo -- 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/20120309083201.ga3...@fernst.no-ip.org
Re: Cron, 1 month but 2 months
On Fri, Mar 09, 2012 at 12:08:38PM +0100, Jochen Spieker wrote: Florian Ernst: On Fri, Mar 09, 2012 at 05:09:23AM +, T o n g wrote: Is there any easy way to schedule a job for a duration that is greater than one month but shorter than 2 months? Any duration will do, as long as it is easy to schedule. Easy? Hmmm. 0 12 * * 1 [ $(expr $(/bin/date +\%s) / 60 / 60 / 24 / 7 \% 8) -eq 3 ] myscript Nice. Oops, forgot to paste the trailing ... || /bin/true which keeps newer cron from reporting the job as failed whenever [ ... ] doesn't exit successfully. Is there any reason why you don't use date '+%V' (or %W or %U)? That'd break with every new year: for instance, [ $(expr $(/bin/date +\%V) \% 8) -eq 1 ] would trigger on ISO week 1, 9, ..., 41, 49 and then again in the next year on ISO week 1, resulting in a shorter interval between the years than desired. Cheers, Flo signature.asc Description: Digital signature
Re: rabbitvcs missing from testing?
Hello all, On Sat, Feb 04, 2012 at 10:48:44AM +, Colin wrote: I would like to use rabbitvcs-nautilus but I can't find rabbitvcs on testing [1] although it's on unstable. What's the reason? All the information is there, you'd just have to dig a bit deeper ... :-) Following the links on the page you mentioned you can finally reach the developer information (PTS) page[0], whose most recent news entry[1] refers to a bugreport[2] mentioning that the current version is incompatible with GNOME 3 and was thus removed from testing to ease transitioning. HTH, Flo [0] http://packages.qa.debian.org/r/rabbitvcs.html [1] http://packages.qa.debian.org/r/rabbitvcs/news/2006T163911Z.html [2] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=644690 -- 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/20120204111324.ga20...@fernst.no-ip.org
Re: rabbitvcs missing from testing?
Hello all, On Sat, Feb 04, 2012 at 11:24:45AM +, Colin wrote: [...] Should I try the version from unstable and run it instead? I would prefer to use all packages from debian instead and can to do bug reports from there. The version from unstable won't work as it doesn't support the new APIs used in GNOME 3, an updated version is needed instead[0]. Apparently there was an attempt to update the package, but it looks like it didn't meet the criteria for entering Debian (there was once a package in NEW[1] that now cannot be found anymore, further data / background could possibly be provided by the package maintainer(s)). Cheers, Flo [0] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=635123 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=644690 [1] http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=7;bug=618934 -- 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/20120204185818.gc20...@fernst.no-ip.org
Re: Regression when I try to get connected to a network server
Hello all, On Sat, Feb 04, 2012 at 09:39:09PM +0100, th.lamotte wrote: About one week ago, I did an actualization with the actualization manager. Since then I got a regression with any connection to the network: any first attempts fail when using iceweasel, icedove, etc. I have to attempt them a few times to get finally connected to either a pop server, a web address, etc. This regression is now global on my system. I do not know where the weak package could be and how to track it. If you have any guess to help me, I would appreciate. This sounds like the regression in the eglibc resolver code as dealt with in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=658171 and http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=658424 Try downgrading as mentioned in the first bugreport and see whether that helps. Cheers, Flo -- 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/20120204215412.gf20...@fernst.no-ip.org
Re: what happened in sarge to lose 500 dd?
Hello, On Thu, Jan 12, 2012 at 03:21:05PM +0100, afuentes wrote: http://claudiocomputing.files.wordpress.com/2012/01/infographic_debian_history-en-v08.png what happened in sarge to lose 500 dd? For the most part of it this: Handling of (inactive) Debian Accounts http://lists.debian.org/debian-devel-announce/2007/02/msg8.html HTH, Flo -- 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/20120112152319.ga27...@fernst.no-ip.org
Re: Nonstandard .deb package management/apt problem: create archive with updateable packages for system not connected to internet
Hello, On Thu, Jan 12, 2012 at 04:12:40PM +0100, Fiedler Roman wrote: [...] A system isolated from internet should receive standard distribution package updates. Since the system cannot fetch the packages from public repositories or internal mirrors, some way to create an archive with all necessary .deb packages would be nice. The archive is then transferred to the machine via e.g. usb. [...] Are there tools or a howto available on how to extract the package information from source, use it for fetching ? $ apt-cache search apt fetch usb apt-zip - Update a non-networked computer using apt and removable media HTH, Flo -- 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/20120112152625.gb27...@fernst.no-ip.org
Re: How to get apt source code via VCS?
Hello, On Fri, May 13, 2011 at 03:29:45PM +0800, Edmond Halley wrote: I would like to get apt source code via VCS(git,bzr,svn...). I do not find the repository of apt. [...] Is there a repot of apt? Your first stop for checking something like this would be http://packages.qa.debian.org/packagename which in this case redirects to http://packages.qa.debian.org/a/apt.html Generally, if the package maintainers have included links to their VCS repo (via adding the appropriate headers to the package control file) then they are displayed here. If not, you could try checking the copyright file and/or the homepage of the software project in question, if available. Cheers, Flo -- 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/20110513075737.gb18...@fernst.no-ip.org
Re: rootfs on SSD
Hello Rainer, On Sun, May 08, 2011 at 04:14:53PM +0200, Rainer Dorsch wrote: [...] When I now run the mount command, I get all options listed for /home /dev/sdc2 on /home type ext4 (rw,noatime,discard,data=ordered) but not for root rootfs on / type rootfs (rw) Were the noatime, discard, and data options lost somewhere on the way? Can I verify that they work? Quoting Ted Ts'o from http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=9;bug=616317: | Debian simply doesn't support the mount options for the root file | system in /etc/fstab having any effect on how the root file system is | mounted. The root file system is mounted by the kernel, and the mount | options used by the kernel are specified by the rootflags= option on | the kernel's boot command line. See the remainder of the bugreport for a discussion on how to possibly remedy that in Debian. As for now, you need to go for kernel command line options passed by your bootloader. Cheers, Flo -- 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/20110508151754.gb6...@fernst.no-ip.org
Re: Cannot create raid5 with 4 disks corectly
Hello there, On Sat, Dec 04, 2010 at 01:56:36PM +0100, Reiner Buehl wrote: I would like to create another RAID5 array consisting of 4 2TB disks. When creating the array with /sbin/mdadm --create --verbose /dev/md5 --level=5 --raid-devices=4 /dev/sdm1 /dev/sdn1 /dev/sdo1 /dev/sdp1 the array is created but with one failed disk and a spare instead of 4 working disks: [...] Any ideas how I can fix this? Please see the manpage: | f...@host:~$ man mdadm | grep 'creating a RAID5' -A4 |When creating a RAID5 array, mdadm will automatically create a degraded |array with an extra spare drive. This is because building the spare |into a degraded array is in general faster than resyncing the parity on |a non-degraded, but not clean, array. This feature can be overridden |with the --force option. HTH, Flo signature.asc Description: Digital signature
Re: openvz
Hello all, quote restored See https://lists.debian.org/debian-release/2010/02/msg00115.html as well as the thread preceeding it. /quote On Sun, Feb 14, 2010 at 10:08:16AM -0500, Andrew Malcolmson wrote: While it's good to hear that OpenVZ will probably be in Squeeze after all, you could also check out a similar project Linux Containers (lxc) with support in kernels 2.6.29. http://delicious.com/andmalc/lxc Well, actually that's what the thread pointed to above was originally about, and a conclusion was that while it's really fancy to have lxc in kernel proper, it is no viable alternative for many business use cases as of now. Hopefully that will change soon. :-) Anyway, that being said, of the resources you have bookmarked I can heartily recommend Nigel McNie's notes as well as his lxc-debian script git repository; just in case anyone wants to experiment a bit ... Cheers, Flo signature.asc Description: Digital signature
Re: openvz
Hello all, On Fri, Feb 12, 2010 at 02:18:05PM +0100, Heiko Kokemoor wrote: does anybody knows, if openVZ will be available in sqeeze or what had happend to openVZ in debian? Lack of upstream support. Recently new plans have been announced: | I can now announce an half official statement from Kir | (who is the project manager of openvz) that they are now | dedicated to make a openvz [for Squeeze]. See https://lists.debian.org/debian-release/2010/02/msg00115.html as well as the thread preceeding it. HTH, Flo signature.asc Description: Digital signature
Re: vmlinuz-2.6.32-trunk-amd64?
Hello all, On Wed, Feb 03, 2010 at 07:20:29PM -0500, Tim Tebbit wrote: Thomas H. George wrote: I find I now have vmlinuz-2.6.32-trunk-amd64 in my boot directory. ---8--- I assume this will be corrected in due course but I am curious, what is the significance of the trunk designation? I wondered that as well. It's been sitting in Sid for weeks now like that. FWIW, trunk has no stable ABI yet. However, in the last uploads the kernel ABI had been set and thus the most recent image is 2.6.32-2, see http://packages.debian.org/changelogs/pool/main/l/linux-2.6/current/changelog. HTH, Flo signature.asc Description: Digital signature
Re: ssh-agent, keychain, xsession bash_profile scripting
Hello all, On Fri, Jan 15, 2010 at 04:07:44PM +, Adam Hardy wrote: I've been chasing my tail trying to work this one out following different examples off the web, but can't sort it out and keep getting the old Could not open a connection to your authentication agent. from ssh-add, and nothing but inaction from keychain. I know everything has to run as a child of ssh-agent to gain access to its envvars, but I don't how to achieve this. I used to have in $HOME/.bashrc: | AGENT_INFO_FILE=$HOME/.ssh/ssh-agent-info | if eval `cat $AGENT_INFO_FILE` 2 /dev/null \ | kill -0 $SSH_AGENT_PID 2 /dev/null | then | eval `cut -d'=' -f 1 $AGENT_INFO_FILE | xargs echo export` | else | eval `ssh-agent -t 86400` /dev/null | ( echo SSH_AUTH_SOCK=$SSH_AUTH_SOCK; | echo SSH_AGENT_PID=$SSH_AGENT_PID ) $AGENT_INFO_FILE | fi While SSH_AUTH_SOCK is set no agent will be started during X startup as per /etc/X11/Xsession.d/90x11-common_ssh-agent and the previously started one will be used. This way the agent will be started only once per boot and one only needs to ssh-add once for all gettys. HTH, Flo signature.asc Description: Digital signature
Re: Stale mirror for ftp.us.debian.org?
Hello all, debian-mirrors should be able to provide more insight into this matter, so CC'ing them. On Wed, Aug 19, 2009 at 09:08:33AM +1930, David Leon wrote: I'm currently having problems updating Lenny from ftp.us.debian.org. It seems one of the mirrors has stale information for its Release and Release.gpg file: Sometimes when I connect with a browser I get this: Index of /debian/dists/lenny [...] Release 27-Jun-2009 18:18 72K Release.gpg 27-Jun-2009 18:27 1.0K [...] Our current problem is that apt-mirror downloaded the Release file from one site and Release.gpg from another, so obviously the signature check fails and all packages are not authenticated. Checking ftp.us.debian.org with dig says ;; ANSWER SECTION: ftp.us.debian.org.300 IN A 128.30.2.36 ftp.us.debian.org.300 IN A 149.20.20.135 This one is it: mirrors1.kernel.org appears to not have fully sync'ed for a few days (at least). ftp.us.debian.org.300 IN A 35.9.37.225 ftp.us.debian.org.300 IN A 64.50.236.52 We've fixed this by manually using ctrl-F5 from a browser until our proxy cached the right Release file, but that is less than ideal. Are we getting old DNS information or are the mirrors out of sync? debian-mirrors, please comment (and keep debian-user in CC). TIA, Flo signature.asc Description: Digital signature
Re: may you can help me, xen-linux-system-2.6.26-1-xen-amd64
Hello all, On Sun, Mar 29, 2009 at 07:34:38PM +0200, Thomas Schneider wrote: I installed a xen system on my server, and it work fine (with xen-linux-system-2.6.26-1-xen-amd64). But i need to recompile the kernel to optimize some stuff. And the problem is I didn't found the source, [...] Then my question is: how did you make this pakage (xen-linux-system-2.6.26-1-xen-amd64), and how to get the source of it, to compile a kernel in the debian way to make .deb file for deploying on others server. The Debian kernel packages build is rather involved, but you can find some hints in http://wiki.debian.org/DebianKernelCustomCompilation and the pages linked therein. HTH, Flo signature.asc Description: Digital signature
Re: Question about proxy servers for debian packages
Hello all, On Mon, Mar 30, 2009 at 03:17:01PM -0600, Paul E Condon wrote: I see in the aptitude package list three packages that are proxy servers specifically designed to proxy the package files from a Debian repository/mirror. There may be more than three. I didn't work very [...] differ? Do the differences matter? And, are there any other proxy servers for packages that I haven't found, and are worth while? apt-cacher-ng and probably any non-specific/general proxy server like for example squid. Following up to an article in LWN this topic was recently raised on debian-mentors, please see the thread starting at http://lists.debian.org/debian-mentors/2009/03/msg00210.html and the links contained therein for further details. In my view nowadays approx and apt-cacher-ng seem to be the most recommendable solutions among the special-purpose servers (with the latter apparently having some problems in anything older than Squeeze). HTH, Flo signature.asc Description: Digital signature
Re: About creating debian custom ISO
Hello all, On Thu, Feb 26, 2009 at 03:52:21PM +0530, Kousik Maiti wrote: Can any-body help about the creation/customization of Debian install CD? A good starting point is the Debian Wiki page concerning the installer at http://wiki.debian.org/DebianInstaller. It seems you are most interested in the sections Creating CD images and CD/DVD related stuff and Using Debian Installer/Preseed. HTH, Flo signature.asc Description: Digital signature
Re: VPN connection using openvpn and openssl certs
Hello all, On Wed, Feb 25, 2009 at 01:00:56PM +1100, linuksos wrote: I have just simple question in regards to the OpenVPN connections. When using a Certificates and server directive in openvpn configuration file : server 192.168.0.0 255.255.0.0 the vpn tunnel is created is follows: SERVER: inet addr:192.168.0.1 P-t-P:192.168.0.2 Mask:255.255.255.255 CLIENT: inet addr:192.168.0.6 P-t-P:192.168.0.5 Mask:255.255.255.255 What is a logic behind a IP address given to the client from the server. Why first client connected to SERVER does not get 192.168.0.2 IP address but it gets 192.168.0.6 instead? I'm sure that there is a good explanation for that and I'm just missing something. See Why does OpenVPN's ifconfig-pool option use a /30 subnet (4 private IP addresses per client) when used in TUN mode? on http://openvpn.net/faq.html: | OpenVPN allocates one /30 subnet per client in order to provide | compatibility with Windows clients due to the limitation of the | TAP-Win32 driver's TUN emulation mode. | | If you know that only non-Windows clients will be connecting to your | OpenVPN server, you can avoid this behavior by using the | ifconfig-pool-linear directive. | | [...] Cheers, Flo signature.asc Description: Digital signature
Re: audacious-crossfade
Hello all, On Thu, Feb 05, 2009 at 01:58:38AM +, T o n g wrote: I am missing the audacious-crossfade plugin, after switching over from xmms. From the inet search I have an impression that audacious-crossfade has been a normal Debian package. True. It existed since uploading 0.3.11-3 of the source package xmms-crossfade and entered Lenny on 07 Sep 2007[0]. But it no longer exist from the official site. Does it ever exist? What happened to it? Why Debian doesn't have this package, since other distros like RedHat, Ubuntu, etc all have it. According to the last package maintainer audacious-crossfade is impossible to use reliably with audacious 1.5, so he requested it's removal[1] which actually happened on 12/13 May 2008. Thus, after all, audacious-crossfade was never part of a stable Debian release. I don't know what other distributions did to include this plugin. If they found a way and somebody transfers their solution to Debian the package might actually get resurrected. (NB: No, most probably I won't be available to sponsor this, even though I have once maintained that package myself.) Cheers, Flo [0] http://packages.qa.debian.org/xmms-crossfade [1] http://bugs.debian.org/479110 signature.asc Description: Digital signature
sa-exim still scans mails received via local smtp
Hello all, well, I'm far from being an exim expert, and thus I guess there is something (or maybe a lot) which I don't understand here, so please point me to any brainos I might have inserted. Given a Debian Etch installation of exim4 and sa-exim, all working fine, except for the tiny bit that mails received via local smtp are scanned even though they shouldn't. An excerpt from the exim4.conf.template on which the actual config is based: | # This access control list is used for every RCPT command in an incoming | # SMTP message. The tests are run in order until the address is either | # accepted or denied. | # | acl_check_rcpt: | | # Accept if the source is local SMTP (i.e. not over TCP/IP). We do this by | # testing for an empty sending host field. | accept | hosts = : | set acl_m0 = noscan | [...] | accept | hosts = +relay_from_hosts | set acl_m0 = noscan | control = submission/sender_retain | An excerpt from the sa-exim.conf: | SAEximRunCond: ${if !eq{$acl_m0}{noscan}} However, the following excerpts from a sample mail illustrate it doesn't quite work as intended: | Received: from $UID by $MAILHOST with local (Exim 4.63) | (envelope-from [EMAIL PROTECTED]) | id 1JUEb5-0001vz-GS | for $MAILADDRESS; Wed, 27 Feb 2008 06:12:53 +0100 | [...] | X-SA-Exim-Connect-IP: locally generated | [...] | X-SA-Exim-Scanned: Yes (on $MAILHOST) On the other hand, if a mail is sent from one of the relay_from_hosts it isn't scanned, just as intended. Any hints on what I might be missing here, please? Cheers, Flo signature.asc Description: Digital signature
Re: webmin in Etch?
On Sat, Oct 14, 2006 at 01:42:16AM -0400, Mark Grieveson wrote: Does anyone know if there are plans to include webmin as a part of Etch? Well, considering that webmin currently isn't even part of unstable the chances of including it in Etch are almost nonexistant. Please see http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=343897 for the former maintainer's reasoning for webmin removal from Debian. Cheers, Flo signature.asc Description: Digital signature
Re: help her
[... typical hoax, answering in German ...] On Tue, Aug 01, 2006 at 10:54:25PM +0200, yue zhang wrote: Sorry for none linux information, Eher »für überhaupt keine Informationen«. Dieser Hoax geht schon seit Jahren rum. urgent, Please help her, [...] Nein, hilf Dir selbst, auf dass Du keine irrelevanten Aufrufe ungehemmt weiterverbreitest. A germany women will be dead because of Leukamie searching for spender with AB-Rhesus negativ!!! Die »Blutgruppe« der weißen Blutkörperchen ist relevant, AB0 und Rhesusfaktor sind es hingegen nicht. Die Suche nach einem potentiellen Spender sollte man den Leuten überlassen, die davon auch Ahnung haben. Fehlinformationen und gefälschte Aufrufe hingegen können eigentlich interssierte Personen abschrecken. Weitere Info unter: http://www.tu-berlin.de/www/software/hoax/knochenmarkspende.shtml http://www.zkrd.de/spendersuche.html HTH, Flo signature.asc Description: Digital signature
Re: How to help translating
Hello Paolo, On Mon, Jul 31, 2006 at 11:05:07AM +0200, Paolo Pantaleo wrote: I would like to offer my help to transalate Debian packages in Italian. On How You Can Help page it says I can help translating, but it doesn't say how I do that. Can anyone tell me how can I do that? I bet the debian-l10n-italian list will gladly accept your help and explain in detail what you could do. :) Just write (in English or Italian) to debian-l10n-italian@lists.debian.org. HTH, Flo signature.asc Description: Digital signature
Re: OT: Port to minix3?
Hello *, On Thu, Jul 13, 2006 at 05:55:24PM -0600, Javier-Elias Vasquez-Vivas wrote: Just out of curiosity, is there any project trying to port debian to minix3? Yes. This was just recently covered in DWN, pointing to http://lists.debian.org/debian-devel/2006/05/msg02581.html. Additionally, this DWN issue is the first hit for http://www.google.com/search?q=debian+minix ... HTH, Flo signature.asc Description: Digital signature
Re: Wieso häng t netbase von inetd ab?
Moinmoin, On Fri, May 12, 2006 at 03:08:14PM +0200, Goran wrote: Ich verstehe nicht ganz warum bei einer Standardinstallation von Debian 3.1 auch der Internet Superserver inetd installiert wird und dieser auch noch Bedingung für netbase ist. historical reasons, und dies wird auch noch so bleiben, bis update-inetd allen Anforderungen entspricht, siehe auch bug#87240 (nebst diversen anderen Verweisen in netbase bugs). Obige Frage stellt sich mir gerade da ich ein frisches Debian 3.1 noch schlanker machen will. Bisher habe ich portmap, lpr, statd und atd entfernt da ich sie nicht auf dem System benötige. Im Bugreport findet sich eine Art Workaround. HTH, Flo signature.asc Description: Digital signature
Re: DNS-Server für alioth
Moinmoin, On Sat, Mar 25, 2006 at 04:29:19PM +0100, Wolfgang Jeltsch wrote: weiß jemand was da los ist: $ dig alioth.debian.org ;; reply from unexpected source: 217.237.149.225#53, expected 217.237.151.33#53 ;; Warning: ID mismatch: expected ID 19727, got 32959 ; DiG 9.2.4 alioth.debian.org ;; global options: printcmd ;; connection timed out; no servers could be reached Siehe bitte Downtime for some debian.org machines (people, db, buildd, etc.) http://lists.debian.org/debian-devel-announce/2006/03/msg00021.html und die folgenden Antworten unter http://lists.debian.org/debian-devel/2006/03/thrd3.html#01064. Gruß, Flo signature.asc Description: Digital signature
Re: GNOME 2.14 released :-)
Hello *, On Thu, Mar 16, 2006 at 04:16:52PM +, Magnus Therning wrote: I hope we'll see it in Debian Sid soon. Go GNOME team! Go! Anyway I can help? For starters, please see The GNOME team needs your help at http://lists.debian.org/debian-gtk-gnome/2006/01/msg00037.html, as recently mentioned in DWN. Cheers, Flo signature.asc Description: Digital signature
[EMAIL PROTECTED]: bug with the mailing list archive]
Hello *, this seems to be another instance of bug#208979, so forwarding this message there. Cheers, Flo - Forwarded message from Cedric BRINER [EMAIL PROTECTED] - From: Cedric BRINER [EMAIL PROTECTED] Subject: bug with the mailing list archive To: debian-user@lists.debian.org Date: Wed, 01 Mar 2006 16:52:17 +0100 Message-id: [EMAIL PROTECTED] User-Agent: Mutt/1.5.11 Mail-Followup-To: debian-user@lists.debian.org hello, I'm facing some uncomprehension with aptitude vs apt-get. And I remembered that there was a thread about this on the mailing list archive: http://lists.debian.org/debian-user/2005/02/msg00372.html but when I click on: Index(es): * Date * Thread --- this one I don't find this thread !!! I'd like to check the thread displayed like this because it is som much more convenient: *Set UID and GID for NFS purposes David Powell * Re: Set UID and GID for NFS purposes Alexander Schmehl * Re: Set UID and GID for NFS purposes Robert Brockway * Re: Set UID and GID for NFS purposes (Solved) David Powell o Re: Set UID and GID for NFS purposes (Solved) Miquel van Smoorenburg o Re: Setting up sound - alsa config Punit Ahluwalia + Re: Setting up sound - alsa config Sergio Basurto Juarez + Re: Setting up sound - alsa config Punit Ahluwalia but I repeat.. there is no such thread named 'aptitude vs. apt-get' on the page: http://lists.debian.org/debian-user/2005/02/threads.html#00372 and there is in the source of the previous link any reference to the anchor 372 I was suspecting aptitude to not work correctly.. but now my feelings are even stronger when I saw that there is some stuff hidden : ) any idea of what is wrong with the mailing list archive ?? Ced. -- Cedric BRINER Geneva - Switzerland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] - End forwarded message - signature.asc Description: Digital signature
Re: packages as per sources
Hello *, On Mon, Feb 27, 2006 at 07:10:02AM -0800, L.V.Gandhi wrote: May be I was not clear. I wanted to packages installed say from nerim source. How to get it? If there is an appropriate entry in /etc/apt/sources.list and the list have been updated (i.e. via aptitude update): | $ aptitude search '~i~OUnofficial Multimedia Packages' | wc -l | 18 So I have 18 packages from nerim.net installed. Unofficial Multimedia Packages as per the corresponding /var/lib/apt/lists/ftp.nerim.net_debian-marillat_dists_sid_Release. See /usr/share/aptitude/README for details about the search term. Further how to know what packages available from a source say nerim. How to go about? See above, just drop the ~i. HTH, Flo signature.asc Description: Digital signature
Re: packages as per sources
On Wed, Mar 01, 2006 at 03:59:56PM -0800, L.V.Gandhi wrote: But for Debian Unofficial, it didn't work. [...] All sun packages are installed from debian-unofficial. Works for me: | $ aptitude search '~ODebian Unofficial' | wc -l | 86 Cheers, Flo signature.asc Description: Digital signature
Re: ISDN-Trick: Ruf annehmen
Moinmoin, On Sun, Feb 19, 2006 at 05:19:52PM +0100, Sven Hartge wrote: Michael Hierweck [EMAIL PROTECTED] wrote: kann ich mit Kernel 2.6 und isdn4linux etwas tricksen, dass sich die ISDN-Karte für eine Runfummer zuständig fühlt, aber nicht entgegen nimmt? Du könntest vbox mit extrem hoher Klingelanzahl auf die Nummer ansetzen. ... oder gar RINGS=0 setzen, dann geht der AB gar nicht ran (wie sich mir kürzlich zu meinem Leidwesen eröffnete). Aber falls Du nicht einmal einen AB laufen hast, dann sollte es schlichtweg funktionieren, solange i4l korrekt eingerichtet ist, siehe | $ cat /var/log/kern.log | Feb 19 18:33:45 live kernel: isdn_net: Incoming call without OAD, assuming '0' | Feb 19 18:33:45 live kernel: isdn_net: call from 0 - 0 XX ignored | Feb 19 18:33:45 live kernel: isdn_tty: Incoming call without OAD, assuming '0' | Feb 19 18:33:45 live kernel: isdn_tty: call from 0, - RING on ttyI6 Gruß, Flo signature.asc Description: Digital signature
[EMAIL PROTECTED]: packages.debian.org]
JFYI, a blog entry from one of the ftp masters. Cheers, Flo - Forwarded message from Ganneffs little Blog [EMAIL PROTECTED] - From: Ganneffs little Blog [EMAIL PROTECTED] Subject: packages.debian.org To: [EMAIL PROTECTED] Date: Fri, 03 Feb 2006 22:42:50 - Message-Id: [EMAIL PROTECTED] User-Agent: rss2email Well, packages is still down, but there is a temporarily replacement until a new machine is running now on pdo.debian.net (anything except changelogs/copyright files). The new, final, place for it is scheduled to be up and running mid of next week, but until then you get the information, including searches (ie. pdo.debian.net/$PACKAGENAME) from that other host. debian-admin is informed that they can point the DNS for packages.debian.org there, to make it easier to access, lets hope one of them is doing this simple change. :) URL: http://ganneff.de/blog/2006/02/03#pdo - End forwarded message - signature.asc Description: Digital signature
Re: CC's fuer Bugs setzen
Moinmoin, On Tue, Jan 24, 2006 at 08:23:32PM +0100, Andreas Pakulat wrote: Kann ich bei einem Bug ein CC an meine Email-Addresse einrichten? Ich will ungern bugs.debian.org alle paar Tage auf Fortschritt pruefen muessen, habe aber fuer [EMAIL PROTECTED] kein entsprechendes Kommando gefunden... In der Art von »Subscribing to bugs« wie unter http://www.debian.org/Bugs/Developer#subscribe beschrieben? HTH, Flo signature.asc Description: Digital signature
Re: #foo mit bugs.debian .org verknüpfen
Hie Michelleh, On Mon, Jan 23, 2006 at 05:21:25PM +0100, Michelle Konzack wrote: Gibt es das auch in einer Version ohne 56 MByte KDE dran? Ich meine x/plain. Wurde von Wolf weiter unten in diesem Thread bereits vor Tagen erwähnt. Gruß, Flo signature.asc Description: Digital signature
Re: Nagios 2 als apt-Package?
Moinmoin, On Mon, Jan 23, 2006 at 12:31:00AM +0100, Philipp Flesch wrote: Gibt es irgendwo Nagios 2 schon als apt-Package?? Wurde eben zu Experimental hochgeladen, hängt aber noch in der NEW-Queue: http://ftp-master.debian.org/new.html Quellen außerhalb der Debian-Archive wären mir nicht bekannt. HTH, Flo signature.asc Description: Digital signature
Re: Pakete und Programmauffruf
Moinmoin, On Fri, Jan 13, 2006 at 01:45:07PM +0100, Gerhard Wolfstieg wrote: manchmal wird es mir zu bunt. Aus bestimmten Gründen wollte ich mir 3dchess anschauen. Habe das Paket installiert, 3dchess eingegeben = denkste, gibt's nicht, man 3dchess gibt's nicht, locate 3dchess kann [...] GRAUSIG! Hehe, die Strömung schlägt wieder zu. Oder heißt Du mitunter auch Michael? Gruß, Flo PS: Zur Klärung: http://lists.debian.org/debian-policy/2006/01/msg00053.html ff. signature.asc Description: Digital signature
Re: #foo mit bugs.debian .org verknüpfen
Moinmoin, On Fri, Jan 13, 2006 at 02:20:26PM +0100, Bastian Venthur wrote: Das finde ich ziemlich unhandlich. Bietet KDE dafür keine geschmeidigere Lösung an? Optimal wäre, wenn alles was nach Debian-Bug Aussieht direkt als Link zum jeweiligen Bug dargestellt würde. Das wird sicher nicht gehen, aber ein Systemweiter eintrag im Kontextmenü: #foo markieren - rechtsklick - bugs.debian.org öffnen Paket »klipper« installieren und starten, - »Configure« - »Actions« - Regular Expression »^#?\d{6}$« neu erstellen - Command »$yourbrowser http://bugs.debian.org/`echo %s | tr -d '#'`« hinzufügen - »OK« - Bug markieren (Beispiel bug#321123) - Enjoy. sollte doch machbar sein oder? Weis jemand Wie das geht oder hat einen besseren Vorschlag? *gespanntsei* HTH, Flo signature.asc Description: Digital signature
Re: #foo mit bugs.debian .org verknüpfen
Moinmoin, On Fri, Jan 13, 2006 at 03:55:47PM +0100, Gerhard Brauer wrote: * Florian Ernst [EMAIL PROTECTED] schrieb am [13.01.06 14:37]: »$yourbrowser http://bugs.debian.org/`echo %s | tr -d '#'`« Verbesserungsvorschlag: festverdrahteten $yourbrowser debian-konform durch x-www-browser ersetzen. Hmm, in der Tat, allgemein gesehen wäre das wohl angebrachter. Was ich eigentlich meinte war »$yourbrowserinvocation«, also z.B. »konqueror --profile profile«, »firefox -remote«, »opera -newpage«, ... Gruß, Flo signature.asc Description: Digital signature
Re: doxymacs configure error
Hello *, On Fri, Jan 13, 2006 at 08:44:39PM +1100, Yasir Assam wrote: I'm using Etch and I just tried installing doxymacs - I got the following error: [...] /usr/share/emacs21/site-lisp/doxymacs/doxymacs.el: !! File error ((Cannot open load file url)) [...] dpkg: error processing doxymacs (--configure): subprocess post-installation script returned error exit status 1 Errors were encountered while processing: doxymacs Does anyone have any ideas as to why I'm getting this error, and how to resolve it? Please always check the BTS first, see http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=344942 for details. HTH, Flo signature.asc Description: Digital signature
Re: Probleme mit Flash
Moinmoin, On Wed, Dec 14, 2005 at 11:42:47AM +0100, Rolf Leggewie wrote: Problem 2: Die Eingabe japanischer Zeichen z.B. in http://home.arcor.de/leggewie/editor.html scheitert daran, daß ich den notwendigen Front-End-Prozessor (z.B. anthy) nicht aktivieren kann. Dies geht auch auf dem Rechner meines Bekannten nicht. Keine Ahnung, ich verwende hier gemäß Dokumentation eingerichtete canna / kinput2, und da alles auf Anhieb funktionierte, reichen meine Kenntnisse nur zur Anwendung, aber nicht zum Troubleshooting... Du kannst im obigen Flash-Programm japanische Zeichen eingeben? Das ist ja geil. Geht das auch bei http://designer.spreadshirt.jp? Jetzt müssen wir nur noch rausbekommen, warum es bei Dir geht und bei mir nicht. »Eingeben« wäre etwas zuviel gesagt, da sich der Browser aufhängt, sobald ich anfange Kana einzutippen. Allerdings kann ich anderswo, z.B. in einem anderen Browser-Tab, meinen Text tippen und dann per Copy-and-Paste übertragen. Ich habe jedoch keine Ahnung, wo da der Fehler liegt, ob im IME, im Flashplayer, im Browser, im Karma... Ich verwende uim und anthy. Vielleicht liegt es schon daran. Mir Vielleicht, vielleicht auch nicht. Ich verwende canna / kinput2, weil es das erste war, das ich zu diesem Thema gefunden habe, und es hat meine eher kärglichen Ansprüche bislang erfüllt, so daß ich in meiner Bequemlichkeit bislang nicht mich nach Alternativen umgesehen habe. Das heißt aber noch lange nicht, daß diese Kombination besonders gut geeignet ist. scheint es aber sinnvoll, zunächst nach dem Grund der nicht dargestellten japanischen Zeichen zu forschen. Welches Release verwendest Du? Bei mir ist es testing mit einigen wenigen Paketen aus Umm, Release? oldstable/stable/testing/sid/experimental/unknown mit gewisser Betonung auf testing, wenn ich mir meine Paketliste so ansehe... ;) unstable. Das Schöne an http://home.arcor.de/leggewie/editor.html ist, daß es im Font-Pulldown-Menü alle Fonts auflistet, die dem Flash-Player zur Verfügung stehen. Kannst Du da mal schauen und die aufschreiben oder einen Screen-Shot machen? Kommt per PM, sowohl für den Editor als auch für die Shirts. Gruß, Flo signature.asc Description: Digital signature
Re: Probleme mit Flash
Moinmoin, On Tue, Dec 13, 2005 at 11:58:31PM +0100, Rolf Leggewie wrote: Problem 1: Die japanischen Zeichen werden nicht dargestellt. Auf dem SUSE-System eines Bekannten funktioniert zumindest dies problemlos. Gibt es jemanden in der Liste, der auf seinem Debian-System http://www.ola-ac.jp/ft_hitoe1.swf oder http://designer.spreadshirt.jp mit japanischen Zeichen im Flash-Teil angezeigt bekommt? *meld* Hast Du gsfonts-x11 installiert? Siehe auch http://bugs.debian.org/269478 und http://bugs.debian.org/340206. Problem 2: Die Eingabe japanischer Zeichen z.B. in http://home.arcor.de/leggewie/editor.html scheitert daran, daß ich den notwendigen Front-End-Prozessor (z.B. anthy) nicht aktivieren kann. Dies geht auch auf dem Rechner meines Bekannten nicht. Keine Ahnung, ich verwende hier gemäß Dokumentation eingerichtete canna / kinput2, und da alles auf Anhieb funktionierte, reichen meine Kenntnisse nur zur Anwendung, aber nicht zum Troubleshooting... Gruß, Flo signature.asc Description: Digital signature
Re: Ncurses
Hello *, On Wed, Dec 07, 2005 at 11:59:52AM -0800, Matt Jerdonek wrote: I'm trying to install Ncurses (for make menuconfig for Linux build). When I type apt-get install Ncurses, I'm told that Ncurses can't be found: Yep, there is no package that goes by this name. What you are searching for contains Ncurses, but is called slightly differently. [...] I looked at the Debian package list, and saw several packages when I searched for Ncurses. Which one should I install? When building a kernel you might want to check the kernel-package package which will help to do it the Debian way, i.e. by creating a deb package you can install. Incidentally, this kernel-package package suggests to also install libncurses-dev, which is a virtual package provided by libncurses5-dev and just what you are searching for. HTH, Flo signature.asc Description: Digital signature
Re: Text-Datei und Zeilenumbruch analysieren
Moinmoin, On Mon, Dec 05, 2005 at 01:31:09PM +0100, Al Bogner wrote: Dieses unix2dos finde ich aber nicht. Ist das ein eigenes Paket oder in einem Package? Noch befindet es sich im »sysutils«-Paket, es wird aber bald in ein eigenes Paket ausgelagert... Gruß, Flo signature.asc Description: Digital signature
Re: Videolan/vlc in Etch
Moinmoin, On Fri, Dec 02, 2005 at 01:10:56AM +0100, Agon S. Buchholz wrote: wie bekomme ich heraus, was mit einem Debian-Paket (in diesem Falle vlc bzw. Videolan) gerade passiert, also beispielsweise, ob das Paket umbenannt wurde oder ob in Testing gerade irgendwelche gravierende Änderungen in Zusammenhang mit dem betreffenden Paket durchgeführt werden? http://packages.qa.debian.org/vlc http://bjorn.haxx.se/debian/testing.pl?package=vlc http://ftp-master.debian.org/testing/hints/ Mir scheint, vlc wurde bereits Mitte Oktober entfernt, um einen Übergang der neuen wxwindows-Bibliothek zu erleichtern, und da vlc seit der in Sarge vorhandenen Version sich nicht mehr auf der arm-Architektur und zwischenzeitlich auch auf anderen Architekturen bauen liess und auch an diverse andere Übergänge seiner Abhängigkeiten gebunden ist, hat es noch nicht wieder geschafft, ins aktuelle Testing vorzudringen. HTH, Flo signature.asc Description: Digital signature
Re: Bezugsquelle fü r alte Pakete aus sarge?
Moinmoin, On Fri, Dec 02, 2005 at 04:27:58PM +0100, Sebastian Kayser wrote: [...] Jetzt stellen sich mir zwei Fragen: *) In erster Instanz wollte ich auf 1.3.5-9 downgraden. Wo bekomme ich denn das überhaupt noch her. Ich bin mir sicher, dass dergleichen schon öfter über die Liste gegeistert ist, ich finds aber aktuell einfach nicht ... http://snapshot.debian.net/ war früher die kanonische Antwort zu dieser Frage, aber leider hatte Ukai-san ein paar Hardware-Probleme... Derzeit sieht es in dieser Hinsicht ein wenig mau aus, da man in aller Regel nur die in den einzelnen Veröffentlichungen gerade aktuellen Versionen finden kann. Aber eventuell wird man auf einer alten Ausgabe einer debian-installer CD oder in einem alten Cache fündig. *) Warum fliesst prinzipiell 1.3.5-11 mit dem Fix für ein Problem wie in #314342 beschrieben nicht als Paket in sarge. zgrep ist de facto erstmal nicht nutzbar? zgrep ist nicht nutzbar, wenn man eine *andere* Shell als die bash verwendet. Nun ist aber die bash gerade die Standard-Shell in Debian, weswegen dieser Fehler zwar sehr bedauerlich ist, aber wohl kaum die strengen Anforderungen an eine Aktualisierung von Stable erfüllt, siehe http://people.debian.org/~joey/3.1r1/. Dort findet sich zwar ein Update für gzip, aber es berührt nicht Dein spezifisches Problem. Man müsste schon den Paketbetreuer und Joey als Stable Release Manager überzeugen, daß ein solches Update angemessen ist, und da sehe ich geringe Chancen... In Deinem Falle wäre es wohl am einfachsten, das Paket aus testing / unstable manuell zu installieren, da dessen Abhängigkeiten (debianutils (= 1.6), libc6 (= 2.3.2.ds1-21)) sich schon in Stable erfüllen lassen. Beachte dabei aber, daß Du damit dann keine weiteren automatischen Sicherheitsaktualisierungen für gzip mehr erhalten würdest, da Dein gzip dann eine höhere Versionsnummer als das Update hätte... Sollte dies untragbar sein, so könntest Du Dir ein Paket inklusive Fix selber bauen und installieren. So oder so wäre aber besondere Vorsicht vonnöten, wenn man sich von der Software entfernt, wie sie in den Debian-Veröffentlichungen enthalten ist, und besondere persönliche Aufmerksamkeit gegenüber Sicherheitsfragen angebracht. Viele Grüße, Flo signature.asc Description: Digital signature
Re: Bezugsquelle fü r alte Pakete aus sarge?
Moinmoin, On Fri, Dec 02, 2005 at 05:45:40PM +0100, Andreas Pakulat wrote: Wobei mir grad auffaelt, der Patch in 1.3.5-11 wird mit sed umgesetzt, aber gzip dependet nicht auf sed und auch nicht auf grep... | $ apt-cache show grep | head -n 2; apt-cache show sed | head -n 2 | Package: grep | Essential: yes | Package: sed | Essential: yes http://www.de.debian.org/doc/debian-policy/ch-binary.html#s3.5. HTH, Flo signature.asc Description: Digital signature
Re: update-grub on aptitude install of new linux image...
Hello *, On Wed, Nov 30, 2005 at 08:52:37PM -0600, Javier-Elias Vasquez-Vivas wrote: I'm wondering how to make update-grub automatically when ever a new linux image gets installed, whether from debian stock or generated through make-kpkg. Might be just editing a config file or something? /etc/kernel-img.conf is the one you want. If you have the kernel-package package installed you'll find further information in man kernel-img.conf and a sample configuration in /usr/share/doc/kernel-package/examples/sample.kernel-img.conf I personally use | do_symlinks = no | postinst_hook = /sbin/update-grub | postrm_hook = /sbin/update-grub | do_bootloader = no | do_bootfloppy = no HTH, Flo signature.asc Description: Digital signature
Re: apt preferences bzw. V erständnisfrage Paketsystem
Moinmoin, On Sat, Nov 26, 2005 at 01:27:10PM +0100, Andreas Pakulat wrote: On 26.11.05 10:39:57, Matthias Haegele wrote: Könnte ich das Paket fail2ban auch irgendwie gegen stable bauen?. Eventuell... Da es keine versionierte Abhaengigkeit auf irgendwas hat, sondern nur iptables und python braucht, sollte ein apt-get source fail2ban build-essential debhelper fakeroot cd fail2ban-... dpkg-buildpackage -rfakeroot -us -uc Dir ein .deb erzeugen. Hehe, Andreas, da weißt Du aber, daß das so definitiv nicht funktionieren wird... ;) Eher: 1. apt-get install build-essential fakeroot 2. apt-get build-dep fail2ban 3. apt-get source fail2ban 4. cd fail2ban-... etc NB: Das funktioniert allerdings nur, solange ein passender deb-src Eintrag in /etc/apt/sources.list gegeben ist. Eventuell sind womöglich noch zusätzliche Anpassungen vonnöten, wie es beim Rückportieren vorkommen kann. Gruß, Flo signature.asc Description: Digital signature
Re: Cron Job jeden Tag fr üher starten
Moinmoin, On Fri, Nov 25, 2005 at 05:29:32PM +0100, Matthias Houdek wrote: Was spricht gegen das Umschreiben der crontab per Script (wurde ja schon angesprochen). Sollte mit sed kein Problem sein. Oder einfach drauf achten, dass es immer die letzte Zeile ist. Die kann man einfach abschneiden und neu dranschreiben. Wer Debians Standard-cron verwenden und dabei seine crontab nicht automatisiert umschreiben möchte, der kann in einem Konstrukt à la 0 * * * * [ $(expr $(date +\%s) / 3600 \% 22) -eq 0 ] your_command eventuell eine Lösung finden... Gruß, Flo signature.asc Description: Digital signature
Re: Cron Job jeden Tag fr üher starten
Moinmoin, On Fri, Nov 25, 2005 at 06:14:08PM +0100, Matthias Houdek wrote: 0 * * * * [ $(expr $(date +\%s) / 3600 \% 22) -eq 0 ] your_command Interessante Lösung. Die Startzeit ist zwar undefiniert (gut, könnte man ausrechnen), aber ^^ das war ja in diesem Fall egal. Ansonsten müsste man noch ein Offset in das date-Kommando setzen (je nach gewünschter Startzeit). ...und dann Offset statt »-eq 0« eintragen. Gruß, Flo signature.asc Description: Digital signature
Re: Uptime ppp?
Moinmoin, On Fri, Oct 28, 2005 at 02:21:35AM +0200, Gerhard Brauer wrote: * Bertram Scharpf [EMAIL PROTECTED] schrieb am [27.10.05 20:16]: tut mir leid, ich find's nicht: Mein Gkrellm zeigt mir immer an, wie lang ich schon eingewählt bin. Wie ziehe ich mir diese Information am einfachsten auf der Befehlszeile? Die Tools wie gkrellm oder pppstatus starten wohl beim Online einen internen Timer, der diese Info dann zur Verfügung stellt. [...] Ich würde in /etc/pppd/ip-up.d ein Skript starten, welches mir die aktuelle Zeit (=Verbindungsaufbau) als Unix-Zeit (Sekunden seit 1970, siehe date +%s) in eine Datei schreibt. Dann ein Skript, welches mir die Onlinezeit berechnet. Und zwar: Wert von aktuellem `date +%s` minus Wert aus oben angelegter Datei = aktuelle Onlinezeit in Sekunden. Diese dann augengefällig in hh:mm:ss umrechnen und anzeigen. Im Grunde genommen arbeitet pppstatus auf diese Art und Weise: pppd schreibt beim Starten der Verbindung eine Datei /var/run/pppX.pid und pppstatus liest per stat deren mtime aus. Alles weitere läuft wie beschrieben... Gruß, Flo signature.asc Description: Digital signature
Re: DVD Sammlung verwalten
Moinmoin, On Mon, Oct 24, 2005 at 10:53:00PM +0200, Andreas Pakulat wrote: was nutzt man unter Debian so fuer die DVD Sammlung? Ich hab nur gcfilms und tellico gefunden, und beides unterstuetzt leider keinen automatischen Import der DVD's durch Auslesen derselben (die haben ja auch so ne Identifizierungsnummer). Genau das will ich aber, denn ich hab keine Lust alle DVD's per Hand zu erfassen. Tellico kann ausserdem keine dt. Jugendfreigabe-Level erfassen und bei gcfilms kann ich nicht verschiedene DVD-Versionen ein und desselben Films mit jeweils evtl. anderen Titeln verwalten (was schon cool waere). Gibts sowas? Oder muss ich selbst Hand anlegen dafuer? Da Du unten weiter von Python sprachst, fiel mir CeeMedia ein http://sponsors.debian.net/viewpkg.php?id=33. Es ist zwar (noch) nicht in Debian, und ich weiß auch nicht, ob es wirklich Deinen Anforderungen entspricht, aber Beidem ließe sich ja eventuell abhelfen... Gruß, Flo signature.asc Description: Digital signature
Re: KDE 3.4.3 als *.deb
Moinmoin, On Mon, Oct 17, 2005 at 07:20:41PM +0200, Thomas Weber wrote: Am Montag, den 17.10.2005, 15:12 +0200 schrieb Benny Stein: Ich suche die aktuelle KDE 3.4.3 als *.deb. Habe aber bislang nichts gefunden. Für sid stecken die ersten Pakete in incoming. Das wäre zum gegenwärtigen Zeitpunkt der entsprechenden transitions eher ungünstig. Folgerichtig sind die Pakete in incoming auch für experimental... Gruß, Flo signature.asc Description: Digital signature
Re: KDE 3.4.3 als *.deb
Moinmoin, On Mon, Oct 17, 2005 at 10:26:48PM +0200, Andreas Pakulat wrote: On 17.10.05 21:13:43, Florian Ernst wrote: On Mon, Oct 17, 2005 at 07:20:41PM +0200, Thomas Weber wrote: Für sid stecken die ersten Pakete in incoming. Das wäre zum gegenwärtigen Zeitpunkt der entsprechenden transitions eher ungünstig. Welche denn? Also gcc ist denke ich fast abgeschlossen und Xorg Transition sollte ebenso nahezu beendet sein. gcc ist zwar mittlerweile in einer entsprechenden 4er-Version in testing, aber die Nachwehen dieses Übergangs dauern noch an. Einen ungefähren Überblick darüber, was z.B. für den Übergang von KDE von unstable nach testing noch notwendig wäre, gibt Nathanael Nerode in http://lists.debian.org/debian-release/2005/10/msg00107.html (siehe auch den Kommentar von Steve Langasek in http://lists.debian.org/debian-release/2005/10/msg00170.html). Oder bezogst Du Dich eventuell mit »transition« nur auf Vorgänge, die in Sid ablaufen? Folgerichtig sind die Pakete in incoming auch für experimental... Toll, also demnaechst wieder KDE3.4 aus experimental installieren. Hauptsache sie laden diesmal auch alle KDE-Pakete nach experimental und nicht nur kdelibs+kdebase oder so... Derzeit finden sich in experimental bereits die Pakete arts (1.4.3), kdeaccessibility, kdeadmin, kdeartwork, kdebase, kdegraphics, kdelibs, kdemultimedia, kdenetwork, kdetoys und kdeutils (jeweils 3.4.3). Es fehlen derzeit noch kdegames, kdepim und kdewebdev, und das Quellpaket kdebindings muss sogar noch die g++ transition durchlaufen... Gruß, Flo signature.asc Description: Digital signature
Re: KDE 3.4.3 als *.deb
Moinmoin, On Mon, Oct 17, 2005 at 11:34:25PM +0200, Andreas Pakulat wrote: In dem Zusammenhang ja, mir war nicht klar, dass sie jetzt erst KDE 3.4.2 nach testing bringen wollen, anstatt gleich das neuere zu nehmen. IMHO kein sehr glueckliche Entscheidung, was ist wenn Etch eingefroren wird, bevor KDE 3.4.3 in testing ist? Nur der Vollständigkeit halber: das dürfte dann doch eher unwahrscheinlich sein, siehe die Pläne zu Etch à la http://lists.debian.org/debian-devel-announce/2005/10/msg4.html. Naja, soll nicht mein Problem sein, ich fahre Sid aufm Desktop und nach mir die Sinflut ;-) Hehe, warum musste ich dabei nur an ReleaseProposals und speziell an http://wiki.debian.org/DevelopersSatisfiedByUsingUnstable denken...? Gruß, Flo signature.asc Description: Digital signature
Re: Translation for package descriptions ?
Hello *, On Tue, Oct 11, 2005 at 09:10:58PM +0200, jjluza wrote: Do you think it could be possible to have translation for package descriptions in apt ? Yes, this was / is the aim of the Debian Description Translation Project (ddtp). However, their webpages have been defunct (IIRC after the incident in 2003) for a while, followed by some effort to revive the project in 2004. Unfortunately their pages are empty by now, so there seem to be no instructions on how to participate publically available, yet last I heard the mail address [EMAIL PROTECTED] acts as a gateway to the project. As this is just hearsay I cannot tell you how it is supposed to work. There is a start of the restart of the project at http://ddtp.debian.net, but I don't know any further about this either. AFAIK the organizer of the ddtp is still available for contact: Michael Bramer [EMAIL PROTECTED]. Sorry for being vague, but this all comes from way back in my mind... HTH, Flo signature.asc Description: Digital signature
Re: Translation for package descriptions ?
On Tue, Oct 11, 2005 at 09:57:15PM +0200, Florian Ernst wrote: On Tue, Oct 11, 2005 at 09:10:58PM +0200, jjluza wrote: Do you think it could be possible to have translation for package descriptions in apt ? Yes, this was / is the aim of the Debian Description Translation Project (ddtp). However, their webpages have been defunct (IIRC after the incident in 2003) for a while, [...] Just found an old report about how the project was supposed to work: http://lwn.net/Articles/14676/. Cheers, Flo signature.asc Description: Digital signature
Sid-Benutzer, bitte BTS benutzen (was: Wine / Xorg Fehler: BadAlloc)
Moinmoin, On Mon, Oct 10, 2005 at 11:23:51AM +0200, Tim Ruehsen wrote: ich bekomme nach dem apt-get dist-upgrade heute morgen beim Aufruf von Wine den Fehler: X Error of failed request: BadAlloc (insufficient resources for operation) [...] Ich spreche von einem SID System. Bitte bitte, oh Ihr Sid-Benutzer, benutzt das BTS: http://bugs.debian.org/wine erwähnt dieses Problem recht deutlich, und dort nachzusehen kostet nun wirklich nicht viel Zeit. Gruß, Flo signature.asc Description: Digital signature
[OT] Student? (was: Sid-Benutzer, bitte BTS benutzen)
Moinmoin, On Mon, Oct 10, 2005 at 03:20:14PM +0200, Tim Ruehsen wrote: Bitte bitte, oh Ihr Sid-Benutzer, benutzt das BTS: http://bugs.debian.org/wine erwähnt dieses Problem recht deutlich, und dort nachzusehen kostet nun wirklich nicht viel Zeit. Sorry, die Hektik läßt einen manchmal komische Dinge tun. Vor allem morgens ohne Kaffee. Morgens? | Date: Mon, 10 Oct 2005 11:23:51 +0200 Du musst Student sein... ;) Gruß, Flo signature.asc Description: Digital signature
Re: Lösche n von vielen Dateien in Verzeichnis
Moinmoin, On Thu, Sep 29, 2005 at 06:14:40PM +0200, Agon S. Buchholz wrote: gw:/var/lib/amavis/virusmails# rm *.gz bash: /bin/rm: Die Argumentliste ist zu lang Z.B. $ find /var/lib/amavis/virusmails -iname '*.gz' -print0 | xargs -0r rm -fv Weitere Details finden sich in den manpages zu find und xargs... HTH, Flo signature.asc Description: Digital signature
Re: Checking `bindshell'... INFECTED (PORTS: 3049)
Hello *, On Thu, Sep 29, 2005 at 11:12:27AM +0100, Jon Dowland wrote: On Thu, Sep 29, 2005 at 09:21:49AM +0400, [EMAIL PROTECTED] wrote: Bug in the chkrootkit or in the cfs ? Killing cfsd makes report clear,but i suppose it is not a solution . chkrootkit, but I'm not sure a false positive qualifies as a bug. /usr/share/doc/chkrootkit/README.FALSE-POSITIVES reads: | below is a (non-exhaustive) list of packages that are known to cause false | positives. before filing a bug report, please check this list. | | listens on well known ports | [...] | cfs: bindshell listens on port 3049. A similar warning is in README.Debian in Sarge's version... Cheers, Flo signature.asc Description: Digital signature
Re: DSL on boot bzw pppoeconf
Moinmoin, On Tue, Sep 27, 2005 at 09:27:40PM +0200, Aidan Pride wrote: Aha, jetzt kommen wir der Sache näher. Da gibt es keinen Eintrag für eth1, was sollte ich denn da eintragen damit das als ppp verbindung funktioniert?: Z.B. | iface eth1 inet manual | | auto myisp | iface myisp inet ppp | pre-up /sbin/ifconfig eth1 up | provider dsl-provider dsl-provider liegt in /etc/ppp/peers, gestalte es einfach nach den Beispielen aus /usr/share/doc/ppp/examples. iface dsl-provider inet static provider dsl-provider pre-up /sbin/ifconfig eth1 up # line maintained by pppoeconf name Unbekannter Schnittstellentyp hm, dann folgendes: :/etc/network# ifup dsl-provider Don't seem to be have all the variables for dsl-provider/inet. Failed to bring up dsl-provider Jup, für »inet« fehlen auch ein paar Infos. Bzw wenn wir gerade dabei sind, wo wäre denn die richtige stelle um ein script nach herstellen der Verbindung auszuführen, also z.b. den namensdienst bei no-ip.com zu aktualisieren? post-up in Analogie zu pre-up oben, näheres siehe man 5 interfaces. HTH, Flo signature.asc Description: Digital signature
Re: DSL on boot bzw pppoeconf
On Tue, Sep 27, 2005 at 09:48:29PM +0200, Florian Ernst wrote: On Tue, Sep 27, 2005 at 09:27:40PM +0200, Aidan Pride wrote: Bzw wenn wir gerade dabei sind, wo wäre denn die richtige stelle um ein script nach herstellen der Verbindung auszuführen, also z.b. den namensdienst bei no-ip.com zu aktualisieren? post-up in Analogie zu pre-up oben, näheres siehe man 5 interfaces. Uhh, oder latürnich /etc/ppp/ip-up.d/; zum Namespace bei der Skriptbenennung siehe auch man run-parts. HTH, Flo signature.asc Description: Digital signature
Re: hung upgrade
Hello *, On Tue, Sep 27, 2005 at 06:41:00AM -0600, Art Edwards wrote: I'm running a testing box. When I try to dist-upgrade, I get numerous errors shown below. [...] dpkg: error processing /var/cache/apt/archives/libdjvulibre15_3.5.15-1_i386.deb (--unpack): trying to overwrite `/usr/share/djvu/osi/de/libdjvu++.xml', which is also in package libdjvulibre1 dpkg-deb: subprocess paste killed by signal (Broken pipe) Errors were encountered while processing: /var/cache/apt/archives/libdjvulibre15_3.5.15-1_i386.deb E: Sub-process /usr/bin/dpkg returned an error code (1) When using testing/unstable please check the BTS, i.e. see http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=327190. You can force installation via some options to dpkg, but it's still a bug. HTH, Flo signature.asc Description: Digital signature
Re: Vollkommen OT aber wichtig *imho*
Moinmoin, On Fri, Sep 23, 2005 at 02:13:07PM +0200, Patrick Cornelissen wrote: Michael Bierenfeld wrote: http://oolite-linux.berlios.de/ mich tritt ein pferd. das ist elite Jau, ach die guten alten Zeiten :) Ach ja... Wann gibts das als Debian Paket? Hmmm, von http://developer.berlios.de/projects/oolite-linux/: »Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 2.0 license (http://creativecommons.org/licenses/by-nc-sa/2.0/).« Und von http://people.debian.org/~evan/ccsummary: »debian-legal contributors think that works licensed solely under licenses with the NonCommercial license element are not free according to the DFSG and should not be included in Debian. These licenses include: Attribution-NonCommercial-NoDerivs 2.0 license Attribution-NonCommercial 2.0 license Attribution-NonCommercial-ShareAlike 2.0 license« Gruß, Flo signature.asc Description: Digital signature
Re: Repackaging with different compile options.
Hello *, On Fri, Sep 23, 2005 at 10:15:57AM +0100, Jon Dowland wrote: On Fri, Sep 23, 2005 at 11:19:52AM +0300, Ben Sagal wrote: How would i go about creating a squid package with the --enable-ssl flag set. I would also want this package to override the default Debian one. $ apt-get source squid $ vi squid-*/debian/rules add desired configure flag to the ./configure call, probably under the configure: rule somewhere $ dpkg-buildpackage -rfakeroot -uc -us install the resulting .deb that appears in .. with dpkg: # dpkg -i ../squid*deb or similar mark it as 'hold' in aptitude This will only work if the build-essential packages are installed | apt-get install build-essential a deb-src line has been added to sources.list and the the database has been updated and the build-dependencies for squid are present | apt-get build-dep squid BTW, there is Roberto Sanchez's elaborate Debian Package Customization HOWTO at http://familiasanchez.net/~roberto/?page=debcustomize. HTH, Flo signature.asc Description: Digital signature
Re: apt-key
Hello *, On Thu, Sep 22, 2005 at 07:48:13PM +0200, Thomas Weinbrenner wrote: Rodney D. Myers wrote: What package has apt-key? apt-key is part of the apt package. Not in testing, which is what the OP is asking. apt-key has been part of apt since 0.6.0, and this branch currently can only be found in unstable/Sid. When using plain Debian, that is. Cheers, Flo signature.asc Description: Digital signature
Re: Opera-Browser jetzt ko stenlos erhältlich
Moinmoin, On Wed, Sep 21, 2005 at 11:08:39AM +0200, Ulrich Fürst wrote: Auf der Seite von opera.com gibt's auch ein deb-Verzeichnis deb http://deb.opera.com/opera/ stable non-free nur fehlen da Abhängigkeiten: Package: opera Version: 7.53-20040716.2 ^^^ Sicher, dass Du diese Version möchtest? Laut deren Homepage ist 8.5 aktuell, und die alten Versionen bis einschließlich 8.02 (oder so) sollen teils gravierende Lücken aufweisen... Kurzum, dieser Service scheint nicht mehr gepflegt zu werden. Priority: optional Section: non-free/web Depends: libc6 (= 2.1.3), xlib6g (= 3.3.6) | xlibs, libqt3-mt (= 2:3.0.3) apt-cache policy libqt3-mt libqt3-mt: Installed: (none) Candidate: (none) Version Table: Hm, wo krieg ich das jetzt wieder her? Oder lieber einen Bug-Report an opera.com senden? Das ist das Ergebnis unter Sarge und Etch, etwas vollständiger sieht es so aus: | libqt3-mt: | Installed: (none) | Candidate: 2:3.0.3-20020329-1woody2 | Version Table: | 3:3.3.5-1 0 | 50 http://ftp.de.debian.org unstable/main Packages | 2:3.0.3-20020329-1woody2 0 | 500 http://ftp.de.debian.org oldstable/main Packages | 990 http://security.debian.org oldstable/updates/main Packages HTH, Flo signature.asc Description: Digital signature
Re: Stable Updates zu DSA 815-1 (kdebase)
Moinmoin, On Tue, Sep 20, 2005 at 01:13:40AM +0200, Gerhard Brauer wrote: * Thomas Kosch [EMAIL PROTECTED] schrieb am [19.09.05 20:30]: On Day 43 of Bureaucracy 3171, Gerhard Brauer wrote: Ja, aber beim Bauen wird doch in einzelne Pakete (xterm,xfonts,...) aufgesplittet. Warum stellt man jetzt nicht nur _das_ Paket nach updates/stable was auch wirklich den Fix enthält? Weil das in der Paketverwaltung nicht vorgesehen ist. Hm, das ist mir irgendwie nicht nachvollziehbar. Wenn der build process meinetwegen unbedingt alle .debs mit einer höheren Minor-Nummer bauen muß, ist das ok. Jetzt könnte aber doch irgendjemand des SEC-Teams eingreifen und nur das/die Pakete aus dem build auch auf den SEC-Server laden, die wirklich verändert wurden. So ungern man es als Mensch eventuell auch hören mag, aber eine derartige manuelle Intervention würde sich als ausgesprochen fehlerträchtig erweisen. Es ist keineswegs so, daß auf gar keinen Fall ein anderes Prozedere _möglich_ wäre, es ist vielmehr das aktuelle Vorgehen sehr gut kalkulier- und überschaubar. Eine geringere Komplexität des Vorgangs wird in diesem Falle erkauft mit einem erhöhten Rechen- und Bandbreitenaufwand. Aber wenn ich mir meinen Absatz oben jetzt nochmal durchlese klingt das irgendwie reichlich naiv ;-) Die Wirklichkeit, die xfree und kde beim Bauen wahrscheinlich erfordern, auch aufgrund der anderen Plattformen, dürften anders sein als ich es halt gerne hätte. Trotzdem hat das Ganze, wahrscheinlich halt auch aufgrund der gewachsenen Strukturen für mich so etwas von: Ihr Rücklicht ist kaputt? Da müssen wir ihr Auto auswechseln. Ja, der Gedanke ist nachvollziehbar. Und die Parallele in gewissem Maße durchaus zutreffend: Viele Bauteile werden heutzutage nicht mehr separat oder manuell repariert / ausgebessert, sondern stattdessen lieber in toto ersetzt, da der Hersteller glaubt, nur so eine kontrollierte Produktqualität garantieren zu können. (Sicherlich spielt auch oft die Kundenschröpfung und damit ein kommerzielles Interesse eine Rolle, aber das lasse ich bei Debian mal außen vor...) Angenommen, der Fix würde später nach dem Bauen definitiv nur eine Änderung im xlibs-Paket bewirken. Dann müßte doch nur dieses Paket in der Packages von update/stable mit einer neuen Minor-Nummer erscheinen - So und jetzt denke das mal bis zu Ende durch. Für solche Monsterpakete wie Xfree/Xorg oder KDE. Und jetzt überlege mal was passiert wenn du das 2 Jahre lang machst. Dann hast du ein Xfree mit 15 unterschidlichen Versionen (von der Nummerierung her). Das willst du nicht wirklich. Also damit könnte ich leben. Für mich ist lediglich wichtig die Konsistenz der Pakete von sagen wir: xfree - (4.)3.0 kde - (4.)3.3.2 damit ich sehe, von welchem Entwickler-Zweig meine Version abstammt. Ob kdebase jetzt 4:3.3.2-1sarge1 oder 4:3.3.2-1sarge14 ist ist mir egal solange die Meta-Pakete und die Depends der Pakete unter sich (= 4:3.3.2-1) konsistent sind. Auch hier stellt sich die Frage, inwieweit dies nach einer solchen Abfolge von Aktualisierungen garantiert werden kann. Es könnten sich auf irgendwelchen verworrenen Pfaden Fehler eingeschlichen haben, die niemand hat vorhersehen können oder wollen. Bereits jetzt gibt es bedauerlicherweise schon unangenehm viele Fehler im Aktualisierungsprozess für Stable, die zumeist auf eine Nachlässigkeit eines Entwicklers zurückzuführen sind... :( Und auch die changelogs würden für mich nachvollziehbarer sein, da, sagen wir der X bug, wirklich in xlibs-x.y.z-12 gefixt wurde. Und eben nicht in den xfonts-Paketen da er dort ja garnicht auftrat. Deswegen könnten die ruhig bei xfonts-x.y.z-11 bleiben. Im gegebenen Falle sind Deine Bedenken ausgesprochen zutreffend. Es sollte jedoch nach Möglichkeit eine Einzelfallunterscheidung unterbleiben und lieber ein festes, kalkulierbares und überschaubares Regelwerk verfolgt werden, auf daß eine gewisse innere Konsistenz so gut als irgend möglich garantiert werden kann. Der in meinen Augen einzig sinnvolle Weg das zu lösen, ist der, den z.B SuSE geht. Die erstellen keine komletten Binär-Pakete sondern patch-rpms. Die Enhalten nur die Teile, die sich geändert haben (und zwar auf Dateiebene und nicht auf Paketebene) und setzen dann die Versionierung beim Einspielen für das Gesamtpaket entsprechend. Aber das läßt weder dpkg noch die Infrastruktur zu. Hm, die Infrastruktur mag sein. Aber warum eigentlich nicht dpkg/apt? Gerade dieses Paketsystem sollte das doch prima können. Nehmen wir mal Alexanders Beispiel der fehlerhaften man page in xterm. Wenn ich jetzt ein .deb nur mit der geänderten man page habe dann müßte ich dieses Paket doch mit dpkg/apt einspielen können und es würde aus dem Content nur diese man page auf meinem System aktualisiert werden. Voraussetzung eines solches Patch-debs ist natürlich, das die Liste der bereitgestellten Files etc. auch die des Ursprungs-Paketes
Re: [SOLVED] Re: AltGr-Taste
Moinmoin, On Tue, Sep 20, 2005 at 01:15:10PM +0200, Thomas Schönhoff wrote: Es scheint, dass sich da eine nie von mir zuvor benutzte XKB-Option alt:meta_win eingeschlichen hatte. Nach Löschen des Eintrags funktioniert alles normal (grummel..). Jetzt muß ich wohl dem Debian-Maintainer schonend beibringen, dass dies ein Fehlalarm gewesen ist. Kann man seinen Fehlerbericht eigentlich auch wieder zurückziehen, wenn ja wie? Schick einfach eine E-Mail an [EMAIL PROTECTED], also z.B. an [EMAIL PROTECTED] (-close existiert als alias für -done). Füge allerdings nach Möglchkeit eine Erklärung ein, warum Dein Fehler zurückgezogen werden kann. Wenn Du Dir hingegen mal nicht sicher sein solltest, ob ein Fehler wirklich in allen Versionen behoben ist, dann schicke nur eine Erklärung an [EMAIL PROTECTED], also z.B. an [EMAIL PROTECTED], und überlasse die weitere Handhabung dem entsprechenden Paketbetreuer. Gruß, Flo signature.asc Description: Digital signature
Re: Mirrors for security.debian.org?
Hello *, On Tue, Sep 20, 2005 at 08:49:27AM +0200, Kjetil Kjernsmo wrote: [...] This policy is quite understandable, but as an aside: Anybody know if the developers have started preparing a point release? Yes, please see the massages labeled Preparation of the next stable Debian GNU/Linux update on debian-release, the latest is at http://lists.debian.org/debian-release/2005/09/msg00132.html. HTH, Flo signature.asc Description: Digital signature