Re: hard disk access on every keystroke in console mode!
On Sat, 20 Dec 2003, michelle wrote: ... The pauses in KDE3 must be due to something else. Parts of KDE3 will freeze when kbuildsycoca runs. e.g., upgrading changes the menues, which gets noticed by FAM, which starts kbuildsycoca, which results in konqueror freezing until KDE sorts out what is on the filesystems it is monitoring. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: XFree86 4.3 in stable sarge - how likely
Daniel Stone's packages have migrated from penguinppc.org to Sid, so you no longer need those sources.list entries you mentioned (but you did Are you sure? I just did an apt-get update, but it still shows 4.2.1 : They are in experimental. Apt will not fetch them unless you explicitly choose experimental as the default release... either APT::Default-Release experimental;, or -t experimental, or package/experimental. If you use dselect you need to use APT::Default-Release, then (un)comment the experimental line in sources.list to control whether experimental packages are seen or not. deb http://http.us.debian.org/debian ../project/experimental main contrib non-free - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Microsoft good press over Longhorn
...we have plenty of free software, and as we all know, Linux will get you through times of no money better than money will get you through times of no Linux... :-) ditto -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: fsck seg. fault Please Help!
On Wed, 29 Oct 2003, matt wrote: hi, the other day my woody machine just kinda stopped working...upon a reboot i learned via beep-codes that the ram had failed. i broke out the shop-vac and cleaned it out and it started to boot...here's where the troubble began... ... Unable to andle kernel paging request at virtual address blabla hex I've seen this message (or something very similar) when I had a bad stick of RAM... did you replace your RAM or just vacuum the dust off? ... but seriously, what other utils can i use instead of fsck to repair the drive? There is a low level tool (can't remember the name and a quick search through packages.debian.org didn't find it, may also only be in unstable or testing) but you'd need to know the structure of the fs at the bits level to use it. If fsck is able to recover a block (or sequence of blocks) but doesn't know where they belong it will place them in lost+found (each partition has one) -- you then need to manually look at the data, figure out what it is, and cp it to where it should be... not fun, and really only worth it for stuff you can't just download again or configure up. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Xfree86 4.3
oopps On Tue, 28 Oct 2003, Bruce Sass wrote: 2) explicitly tell apt you want to fetch from experimental, either: # apt-get -t experimental install ... or --- /etc/apt/sources.list --- s/b /etc/apt/apt.conf APT::Default-Release experimental; --- The sources.list method is a must if you use dselect, and you probably s/b apt.conf method want to do an update/upgrade sans experimental immediately before updating with experimental so that you can put everything else in experimental on hold. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Xfree86 4.3
On Tue, 28 Oct 2003, John Holland wrote: Can anyone say how stable the experimental XFree86 4.3 packages are? Is there any reliable way to install this into debian? I'm running a mix of stable and unstable. Thanks, I've been using Xfree-4.3 for awhile (first Daniel Stone's packages and now 4.3.0-0pre1v4 from experimental) without any serious problems... on-the-fly resolution switching (CTRL-ALT-+|-) messes up the display and X needs to be restarted to fix it, SDL apps cause the monitor to go into standby mode (switch to text then back to X to fix it). To install you will need to: 1) add experimental to your sources.list deb http://http.us.debian.org/debian ../project/experimental main contrib non-free 2) explicitly tell apt you want to fetch from experimental, either: # apt-get -t experimental install ... or --- /etc/apt/sources.list --- APT::Default-Release experimental; --- The sources.list method is a must if you use dselect, and you probably want to do an update/upgrade sans experimental immediately before updating with experimental so that you can put everything else in experimental on hold. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: zmodem file transfer w/ minicom over serial connection
... HDD IRQs are unmasked: hdparm -u1 dev # READ the hdparm manpage first! At which end? What does this do? both, if possible (may break things with some chipsets), it ensures serial IRQ events are handled in a timely manner -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Wierd apt-get error?
remove the stable entries from your sources.list... it is very unlikely you have any stable packages installed because unstable contains newer versions get more RAM create another swap partition create a swap file don't post HTML... nothing to do with the update problem, but making us downloading a 17K message when a couple of K would do is just not very nice On Mon, 11 Aug 2003, j2 wrote: Running a unstable/stable system and i got the below and would apreciate any hints? cookiemonster:~# apt-get update Hit http://ftp.du.se unstable/main Packages Hit http://ftp.du.se unstable/main Release Hit http://ftp.du.se unstable/non-free Packages Hit http://ftp.du.se unstable/non-free Release Hit http://ftp.du.se unstable/contrib Packages Hit http://ftp.du.se unstable/contrib Release Hit http://ftp.du.se unstable/main Sources Hit http://ftp.du.se unstable/main Release Hit http://ftp.du.se unstable/non-free Sources Hit http://ftp.du.se unstable/non-free Release Hit http://ftp.se.debian.org stable/main Packages Hit http://security.debian.org stable/updates/main Packages Hit http://security.debian.org stable/updates/main Release Hit http://security.debian.org stable/updates/contrib Packages Hit http://security.debian.org stable/updates/contrib Release Hit http://security.debian.org stable/updates/non-free Packages Hit http://security.debian.org stable/updates/non-free Release Hit http://ftp.du.se unstable/contrib Sources Hit http://ftp.du.se unstable/contrib Release Hit http://non-us.debian.org stable/non-US/main Packages Hit http://non-us.debian.org stable/non-US/main Release Hit http://non-us.debian.org stable/non-US/contrib Packages Hit http://ftp.se.debian.org stable/main Release Hit http://non-us.debian.org stable/non-US/contrib Release Hit http://non-us.debian.org stable/non-US/non-free Packages Hit http://non-us.debian.org stable/non-US/non-free Release Hit http://non-us.debian.org stable/non-US/main Sources Hit http://non-us.debian.org stable/non-US/main Release Hit http://non-us.debian.org stable/non-US/contrib Sources Hit http://non-us.debian.org stable/non-US/contrib Release Hit http://ftp.se.debian.org stable/non-free Packages Hit http://ftp.se.debian.org stable/non-free Release Hit http://ftp.se.debian.org stable/contrib Packages Hit http://ftp.se.debian.org stable/contrib Release Hit http://ftp.se.debian.org stable/main Sources Hit http://ftp.se.debian.org stable/main Release Hit http://ftp.se.debian.org stable/non-free Sources Hit http://ftp.se.debian.org stable/non-free Release Hit http://non-us.debian.org stable/non-US/non-free Sources Hit http://ftp.se.debian.org stable/contrib Sources Hit http://non-us.debian.org stable/non-US/non-free Release Hit http://ftp.se.debian.org stable/contrib Release Hit http://non-us.debian.org unstable/non-US/main Packages Hit http://non-us.debian.org unstable/non-US/main Release Hit http://non-us.debian.org unstable/non-US/contrib Packages Hit http://non-us.debian.org unstable/non-US/contrib Release Hit http://non-us.debian.org unstable/non-US/non-free Packages Hit http://non-us.debian.org unstable/non-US/non-free Release Hit http://non-us.debian.org unstable/non-US/main Sources Hit http://non-us.debian.org unstable/non-US/main Release Hit http://non-us.debian.org unstable/non-US/contrib Sources Hit http://non-us.debian.org unstable/non-US/contrib Release Hit http://non-us.debian.org unstable/non-US/non-free Sources Hit http://non-us.debian.org unstable/non-US/non-free Release Reading Package Lists... Error! E: Dynamic MMap ran out of room E: Dynamic MMap ran out of room E: Error occured while processing x3270 (NewVersion1) E: Problem with MergeList /var/lib/apt/lists/ftp.du.se_debian_dists_unstable_non-free_binary-i386_Packages E: The package lists or status file could not be parsed or opened. cookiemonster:~# apt-get -u upgrade Reading Package Lists... Error! E: Dynamic MMap ran out of room E: Dynamic MMap ran out of room E: Error occured while processing x3270 (NewVersion1) E: Problem with MergeList /var/lib/apt/lists/ftp.du.se_debian_dists_unstable_non-free_binary-i386_Packages E: The package lists or status file could not be parsed or opened. cookiemonster:~# -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: looking at deborphan's output first
On Sat, 9 Aug 2003, Dan Jacobson wrote: Is part of deborphan's job to clean up for sloppy package maintainers' bad Depends: etc. fields? no. A depends on B, A gets removed, B is still there -- can't remove B because you may have compiled something against it. Or during normal usage does a debian system accrue useless packages? Useless, no, not needed anymore... yes. ,... Any sure tests? Would I have linked to one of them if I never compile anything on my own? no. no. Installing from a tarball is another way to get a dependency which dpkg doesn't know about. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: problems with kde3
On Wed, 13 Aug 2003, [EMAIL PROTECTED] wrote: Are there kde3 or gnome2 log files anywhere? at the bash prompt... $ startx /usr/bin/kde3 /tmp/kde-log -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: Obnoxious autoresponders was:Re: Out of Office AutoReply: how NOT to work with debian
On Mon, 11 Aug 2003, Joyce, Matthew wrote: * iain d broadfoot ([EMAIL PROTECTED]) [030811 04:12]: * Petrisor Marian ([EMAIL PROTECTED]) wrote: [an entirely blank message with a semi-informative subject line] This pisses me off majorly There's more important things in life... My d key deals with messages like his auto-response just fine -- and it doesn't affect me emotionally one bit. I agree with Hall's post. Hooray for Hall! It takes so much less energy to delete emails than to complain about them. I'll 1/2 agree. Complaining to the list is as much a waste as the auto-response; telling [EMAIL PROTECTED] is more satisfying, may get the problem fixed, and perhaps even a thanks. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Wierd apt-get error?
On Mon, 11 Aug 2003, j2 wrote: get more RAM 2GB is enough for most people. create another swap partition 1GB should suffice.. Especially since i am only using about 600MB of RAM. create a swap file See above. Ya, I didn't know about the apt config thing... still learning something new everyday. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: d-u / Usenet gateway (was Re: Challenge-response mail filters
On Thu, 7 Aug 2003, Alan Connor wrote: ... What other choice do they have? There arguments have been shown to be utter nonsense or outright disinformation. As a reader who has been following the threads simply for the entertainment value (I don't do spam filtering or CR, or have an opinion one way or the other on either issue)... Alan, it is your posts which have been shown to contain utter nonsense or outright disinformation. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Using XFree86 4.3 server in /usr/local
On Sat, 2 Aug 2003, Russell Shaw wrote: Hi, I'm using mostly testing and have XFree86 4.2.1 installed via apt-get. If i compile XFree86 4.3 from source, what's the easiest way to find all the compile options suitable for it to work well in a debian system, yet have it installed in /usr/local? Probably, swipe them from Daniel Stone's packaging of XF86-4.3 for Sid: http://www.penguinppc.org/~daniels/sid/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: upgrade turned off lpd?
On Fri, 1 Aug 2003, Bob Hilliard wrote: Qian Gong [EMAIL PROTECTED] writes: My lpd stopped several times with unknown reason. Command /etc/init.d/lpd restart will let it work again. On a debian system, the canonical command is (as root, of course): invoke-rc.d lpd restart caveat: will only work for services that have been configured to start in the current runlevel. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: dselect apt-get coordination
On Thu, 31 Jul 2003, R Ransbottom wrote: How do you coordinate apt-get and dselect so that they want the same packages? If I understand the apt-get man page running apt-get dselect-upgrade will set up the system per the setting last made using dselect. Correct? yes How do you the inverse? That is how do you alter the dselect database to reflect packages installed with apt-get or, put another way, to reflect the current state of the system? (So that a casual use of dselect does not materially alter the system.) Not necessary, APT and dselect use the same DB... /var/lib/dpkg/status To figure out what is available dselect uses /var/lib/dpkg/available, APT uses (iirc) /var/cache/apt/pkgcache.bin which is constructed from the Packages files it downloads to /var/lib/apt/lists. So, while APT and dselect can have different ideas about what is available, they will always think alike with respect to what is installed. Doing a dselect-upgrade will cause APT to do whatever dselect has flagged in /var/lib/dpkg/status instead of figuring out what to download. Loading the status file into a text editor for more info. HTH - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: konqueror
On Tue, 29 Jul 2003, gerard wrote: I am using openbox and I like to use konqueror for my file browser. Generally it works very well but sometimes it will come up with an error. The error is Could not start process. Can't talk to klauncher Any ideas why this happens or how to fix it? Some of the KDE infrastructure is not getting started. Whether it is a bug or something you have to live with when running Konqueror sans KDE is best asked on a KDE specific mailing list ([EMAIL PROTECTED] perhaps). Try starting klauncher manually then doing whatever causes the error. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: local user halt
On Mon, 28 Jul 2003, matt zagrabelny wrote: is there a method for allowing (besides root) local users (ordinary users sitting at the keyboard of the computer) the ability to use the shutdown command? i dont want those logged in via ssh or other remote method having this capability. CTRL-ALT-DEL see /etc/inittab -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: local user halt
On Mon, 28 Jul 2003, Stephen Touset wrote: Bruce Sass wrote: On Mon, 28 Jul 2003, matt zagrabelny wrote: is there a method for allowing (besides root) local users (ordinary users sitting at the keyboard of the computer) the ability to use the shutdown command? i dont want those logged in via ssh or other remote method having this capability. CTRL-ALT-DEL see /etc/inittab And if you have CTRL+ALT+DEL pointing to /bin/false like a sane sysadmin, using sudo is a good alternative ;) I was going to mention sudo, but don't know the answer to this... Can a ssh-ed in user do sudo shutdown...? - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Moving a harddrive to newer machine
On Wed, 16 Jul 2003, Kjetil Kjernsmo wrote: Hi all! I was just given a box with a PIII processor, and I intend to use it to replace my main server, which currently has a Pentium PRO 180 MHz. ... The hardware is obviously quite different on this new box, but I'm running a unmodified Debian kernel-image-2.4.18-686 2.4.18-5. When the disk will suddenly find itself in a quite different machine, is it something I should remember, or should the kernel boot effortlessly on the new hardware? have an alternate means of booting the machine (rescue floppy or CD, the local LUG swears by Knoppix for this sorta thing) download (and install if possible) packages you will need on the new box before moving the HDD. e.g., xserver, kernel source, anything that supports hardware which is new to the system on the HDD reconfigure what you can beforehand e.g., if you use SVGATextMode, make sure you tell it to use generic VGA settings instead of the video card in the old box. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: The nature of testing and where can others help (Was Re: HowTofor Gnome2??)
Hi, I accidentally deleted all the messages in my debian-user folder sigh. However, I do remember enough of your original post to (hopefully) enlighten you. I have also done the nasty cross-post thing to -devel because I conclude with a thought on how to get the package pool living up to its potential. There is no way to show anyone what the next Debian will look like until it is released . ^^^ You can see which packages are being worked on for inclusion in the next release (those in unstable), but until the release is made there is no way to tell if a package will get enough of its known bugs fixed to be included. Even a package being in testing when a freeze comes along does not provide enough information with which to make that determination. Debian unstable, testing, and stable archives are not development, pre-release, and released archives... until a freeze comes along. Maybe this will help... Flow of new software into Debian's archives: 1 23 --- --- --- N) upstream+ --- unstable --- testing stable F) upstream+ --- unstable testing stable R) upstream+ --- unstable testing --- stable 1, 2, and 3 represent the movement of packages; N, F, and R indicate Debian's normal, in a freeze, and release modes of operation; upstream+ represents the original source plus modifications made by a Debian developer; unstable, testing and stable are Debian archives. 1 can happen at any time and is controlled by the autobuilder; 2 can only happen between freezes and is controlled by the bug tracking system (BTS); 3 is a manually triggered operation whose timing is determined with input from the BTS and developers. To put it into perspective... once (only one or two releases ago) there was just unstable and stable, testing only existed during a freeze. Release cycles were too long and developers didn't like having to put development of new software on hold when a freeze came along, so they decided to do away with separate archives and move to a package pool system -- all packages actually exist in a common pool and specific versions are flagged as being included in one or more virtual archives. The idea being that developers could continue to develop at their own pace and relatively stable packages would automatically accumulate in testing, at some point testing would look good enough to freeze and then release as the current stable. Speculation/opinion: Why is testing in such bad shape... It seems to be the case that developers are moving packages through unstable too fast for testing to reach a point where it looks good enough to freeze and polish into a release. What could be done to fix the situation... Create a permanent frozen archive, fed from a consistent set of packages in testing which have been flagged as release candidates. By set of packages I mean all the packages which make up, for example, Gnome or Gnome2 or KDE2 or KDE3, etc. If only one version of a set is allowed to be a release candidate at any point during a release cycle, and only packages which have achieved stability are allowed to move into the frozen archive... testing+frozen will quickly become what users want testing to be (a peek at the next Debian), and the release manager only needs to look at the quantity of packages in frozen to determine if it is time (are all or enough of the release candidates here?). Useful Debian systems with the existence of a frozen archive... (where your sources.list lines point to) unstableDevelopers developing and users who don't mind bleeding from time to time. testing/unstableUsers who don't mind getting hurt as long as first aid is likely to be available. testing Crazy people, this could contain (for example) Gnome, Gnome2, KDE2 and KDE3, all at the same time! Basically a holding area for packages transitioning from development to pre-release. frozen/testing Developers polishing and users wanting a peek at the next release. frozen Anyone interested in how far along the next release has progressed. Stable quality, but not a complete distribution. stable most users Flow of new software within a Debian with a frozen archive: | --- development ---|--- pre-release ---|- released -| | phase| phase|| unstable -- testing - frozen --- stable, archived 1 23 0 - (not shown) flow into unstable, controlled by the autobuilder 1 - controlled by the BTS 2 - BTS (tighter restrictions than 1) and developer input via a release candidate flag 3 - BTS and release manager (who could be a program which
Re: Tab-Completion in gnuplot
On Sat, 21 Jun 2003, Colin Watson wrote: On Fri, Jun 20, 2003 at 04:53:56PM -0600, Gary Hennigan wrote: Joey Hess [EMAIL PROTECTED] writes: Quoting /usr/share/doc/gnuplot/README.Debian: libreadline --- [...] Of course I don't fully understand all the GPL implications. I believe what the Debian gnuplot maintainer means is that it's ok to use GNU readline library with gnuplot, gnuplot just can't be distributed that way and so (s)he doesn't distribute it that way with Debian. Correct. You can't take GPL code, link it against non-GPL code (more specifically code that cannot be distributed under the terms of the GPL; this means that BSD-minus-advertising-clause and LGPL are not a problem, for example, but gnuplot's licence is since it imposes additional restrictions over and above those in the GPL), and distribute the resulting binary. What other .debs are in the same situation and would therefore benefit from a local tweak and rebuild? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: amaya was Re: Good Open Source Web Development software
I don't know; a home built gtk-gl version also bombs, but a build against lesstif works fine. You could do: cd /usr/src ; apt-get source amaya modify the debian/rules file to say --without-g... then build a package (debian/rules -b) sans gtk-gl. I've used ../configure --with-x --without-gtk --without-gl to build a lesstif version which lives in /usr/local. HTH - Bruce -- On Tue, 10 Jun 2003, Emma Jane Hogbin wrote: On Tue, Jun 10, 2003 at 01:42:07AM -0700, Paul Johnson wrote: On Tue, Jun 10, 2003 at 02:10:45AM -0400, lists1 wrote: I read a while back that Amaya development/updating wasn't very rapid. It supports the current standard. People who make the complaint that it doesn't get rapid development don't realize that people don't mess with a UI that just works. It kind of reminds me of the old Windows program HoTMetaL. Ok, I tried to give amaya a try but got the following error when I tried to start the program: [EMAIL PROTECTED]:/web$ amaya Error creating GtkGLArea! Any one have any ideas on what I need to do to get it to work? thanks! emma -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: The myth of aptitude simplicity
On Sun, 16 Feb 2003, Nathan E Norman wrote: There's one problem: dselect is retarded WRT Recommends. That is, ... you. Annoying. Fortunately, you can tell dselect No really, I want you to not install package foo by typing 'Q' to exit the resolution screen, but it's still a PITA. I would say persistent rather than retarded. Dselect works best if you have a good grasp of the packaging system, and consequently know when to use Q or X and when to fall back to dpkg -i... - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: KDE 3? or 4?
On Fri, 14 Feb 2003, Bill Webster wrote: My laptop was running KDE3 from unstable. About 3 days ago I did an upgrade that upgraded it to what appears to be a partial upgrade to version 4. But in the process it removed something that is needed in ... Does anyone have any idea what might be going on? No. There is no KDE4 yet, KDE3 uses kdelibs4 though. You will need to be more specific about what you are seeing. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
passing mem= with a bootdisk, mkrescue
Hi, I have a box with hda=30G and hdb=854M. The 30G is not recognized so I boot from a floppy and mount root on hdb (history and changing plans, soon to be fixed). I would like to move either of these drives to a different box, which one gets moved depends on if the box they are in can see more than 64M of RAM. Of course I would like to find out before I start ripping out drives and start messing with jumpers. How do I pass mem=72M when using a bootdisk created during the installation of a kernel built with make-kpkg? Is there a way to tell the kernel it has more RAM after it is running? While I am fiddling with kernels and boot disks I may as well make a rescue disk, eh... but mkrescue doesn't work because /proc/cmdline is empty (result of the BIOS booting the kernel directly from a floppy?). I'm not even sure it would work to create a self contained rescue disk. Is mkrescue the right tool for this task? Any advice on making a rescue disk will be appreciated. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Some myths regarding apt pinning
Missed one... ++ | | | Recommended | | | User type | Criteria | Debian | Note | | | | Flavor | | |---+---+-+--| | | | | Don't use APT pinning, | | | Uses Debian, | | prefer the lowest level | | Debian| has some | testing + | interfaces, and only | | Hacker| computing and | unstable| reinstall broken stuff | | | GNU/Linux | | using testing if you | | | experience. | | can't fix or live with | | | | | the breakage.| ++ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: debian rookie trying to get his bearings...
On Sat, 25 Jan 2003, Shyamal Prasad wrote:
David == David Z Maze David writes:
David Jeff Hahn [EMAIL PROTECTED] writes:
I'm setting up a test debian server (contemplating a move of
several redhat boxes)
One quick question to get me going a little better... How do
you install services (apache, samba, whatever) and NOT have
them start on system startup?
David Probably the easiest way is to 'rm /etc/rc2.d/S20apache',
David etc. as root.
IMHO this is almost the Debian Way to do this. I would actually
suggest 'mv /etc/rc2.d/S20apache /etc/rc2.d/K20apache' and so on for
each service you want to shutdown in runlevel 2.
Have a look in rc2.d and rc6.d, you will notice that some services
have Ssmallnum and Kbignum or the reverse situation. It may not
be a good idea to just change the S to a K to disable those services.
The most Debian way to do it is via update-rc.d,
the easiest way is to just rm the entry.
The best way (imo) is to use rm, but leave one of the low numbered
runlevels (2-5) and rc6.d untouched... so you can always see at a
glance what the default sequencing is for both bringing up and taking
down installed services. Remember, you can also setup rc{7,8,9}.d,
which will not be touched by automated runs of Debian tools.
- Bruce
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Desktop Performance Issue
Hello Alvin,
On Sat, 28 Dec 2002, Alvin Oga wrote:
/dev/hda:
setting using_dma to 1 (on)
HDIO_SET_DMA failed: Operation not permitted
using_dma= 0 (off)
Do it as root
of course
sometimes .. you have to make sure that the chips and
the drive supports DMA ...
- ( check the kernel IDE/dma options )
I think it is the chips which don't do DMA.
circa 1990 hardware
ah ... first problem. those circa drives supports
multiword dma ... not ultra-dma
which exactly is supported...
hdparm -iv /dev/hda will tell you
bms:~# hdparm -iv /dev/hda
/dev/hda:
multcount= 0 (off)
IO_support = 0 (default 16-bit)
unmaskirq= 0 (off)
using_dma= 0 (off)
keepsettings = 0 (off)
readonly = 0 (off)
readahead= 8 (on)
geometry = 59570/16/63, sectors = 60046560, start = 0
Model=FUJITSU MPG3307AT, FwRev=02B9, SerialNo=VG13P1201HK8
Config={ HardSect NotMFM HdSw15uSec Fixed DTR10Mbs }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=off
CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=60046560
IORDY=yes, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 *mdma2
UDMA modes: udma0 udma1 udma2 udma3 udma4 udma5
AdvancedPM=yes: disabled (255) WriteCache=enabled
Drive conforms to: ATA/ATAPI-5 T13 1321D revision 1: 1 2 3 4 5
...
hmmm...
# CONFIG_IDEDMA_AUTO is not set
# CONFIG_DMA_NONPCI is not set
and if that is the kernel you're booting...
hdparm options in the bootups will not work
# echo ... /proc/... could fix that?
I'm curious about the PIO modes, and if one of those is what is being
used. Since the drive is defaulting to a DMA mode but the OS isn't
doing DMA, is the system falling back to PIO and should explicitly
selecting the best PIO mode be expected to improve performance.
the list of dma options from hdparm -iv wil tell you which
one ( marked w/ * ) your machine is currently using
The drive is defaulting to mdma, which is reasonable for the
hardware... so it probably doesn't matter that hdparm fails to set
a DMA mode, or the kernel thinks DMA isn't happening, the hardware has
it all sorted out?
gotta lot of reading and fiddling to do
pio vs udma vs dma vs ...
http://www.Linux-1U.net/Disks/
and for system fiddling and tuning
http://www.Linux-1U.net/Tuning
Thanks
- Bruce
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Desktop Performance Issue
On Sat, 28 Dec 2002, Ron Johnson wrote: On Sat, 2002-12-28 at 15:25, Bruce Sass wrote: [snip] I think it is the chips which don't do DMA. circa 1990 hardware Wow, that's what? A 486/20 w/ all ISA slots and a 100MB HDD??? 486DX2-25, 64M RAM, all ISA(PNP), 850M + 30G HDD running KDE3.1 (animations turned off) Not bad for hardware that didn't expect to see more than 16M of RAM and was probably designed for win31, eh. lol - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Desktop Performance Issue
On Sat, 28 Dec 2002, Alvin Oga wrote: On Sat, 28 Dec 2002, Antoine Jacoutot wrote: From: Bruce Sass[EMAIL PROTECTED] ... /sbin/hdparm -d 1 /dev/hd[abc] Works like a treat! # hdparm -d 1 /dev/hda /dev/hda: setting using_dma to 1 (on) HDIO_SET_DMA failed: Operation not permitted using_dma= 0 (off) Do it as root of course sometimes .. you have to make sure that the chips and the drive supports DMA ... - ( check the kernel IDE/dma options ) I think it is the chips which don't do DMA. circa 1990 hardware and if hdparm was in your bootup files, it should have worked... ...it should not matter if done in /etc/init.d/hwtools, or from a command line after boot, right? hmmm... # CONFIG_IDEDMA_AUTO is not set # CONFIG_DMA_NONPCI is not set ...and I don't have the kernel source installed to check why I didn't select these (I had no probs using DMA for sound, the only other DMA a grep through the .config turned up). but, aside from maybe installing a different kernel and trying DMA again... (got a BIOS recognizing hda prob to fix first ;) I'm curious about the PIO modes, and if one of those is what is being used. Since the drive is defaulting to a DMA mode but the OS isn't doing DMA, is the system falling back to PIO and should explicitly selecting the best PIO mode be expected to improve performance. gotta lot of reading and fiddling to do - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: lyx-1.2
On Thu, 12 Dec 2002 [EMAIL PROTECTED] wrote: At Wed, 11 Dec 2002 08:33:41 -0700 (MST), Bruce Sass wrote: On Tue, 10 Dec 2002, Craig Dickson wrote: That's nice, but I'm really waiting for lyx 1.3 with the Qt front-end. XForms is vile. Qt lyx-1.3.0cvs is worth trying out, feature frozen even. What Qt version is it supposed to link to? I prefer the leanness of pure gtk (v.1 without the gnome fluff). I want to create a well-structured document not an eye-candy filled desktop environment. Qt2 or Qt3, needs automake1.5 or better. I've built against the Qt in Woody, and been told it works better with Qt3. Turning off any KDE feature with animation or hover in its description helps a lot. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: lyx-1.2
On Tue, 10 Dec 2002, Craig Dickson wrote: That's nice, but I'm really waiting for lyx 1.3 with the Qt front-end. XForms is vile. Qt lyx-1.3.0cvs is worth trying out, feature frozen even. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: apt-get install -f fail, please help
did you do
# apt-get update
before the upgrade, and what is the output of:
# cat /etc/{debian_version,apt/sources.list}
--
On Thu, 5 Dec 2002, eric wrote:
Dear Bruce or any linux er:
Hi Bruce, thanks your reply and hint, it work most of it ,
except
progeny:/var/cache/apt/archives# apt-get upgrade
Reading Package Lists... Done
Building Dependency Tree... Done
The following packages have been kept back
... lots ...
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: rebuilding available
On Wed, 4 Dec 2002, Michael P. Soulier wrote: Is it possible to rebuild the dpkg available file? I have on one a P75 that's so big it's impossible to efficiently parse anymore. A simple dpkg -l takes minutes. I don't think so, dlocate (from the pkg with the same name) is faster. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: KDE sound
On Wed, 4 Dec 2002, Bruce Park wrote: I'm using KDE3 and I had a few questions regarding the sound server aRts. I need to edit the system volume but when I go into sounds in the control system, I don't see anything that pertains to volume. It states something about using a sound server at start up and I'm thinking that this should be arts. I do NOT have the arts plugin for XMMS nor do I really want to run a useless process and hog memory. Can someone recommend a way to manipulate the system volume? As always, any help or suggestion is greatly appreciated. kmix, or any other mixer I'd imagine. artsd is mainly concerned with getting the need to make sound from an app to the soundcard, not with controlling how the sound is generated. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Proposal - non-free software removal
On 16 Nov 2002, John Hasler wrote: Bruce writes: If I was a company I would certainly be hesitant to do anything with Debian because it seems to have a problem with people making money off software. Baffle. One of the most common reasons for packages to be in non-free is that their licenses forbid the making of money from them. You may be able to convince them by pointing to non-free, and the most common reason stuff is in non-free, or by showing them the relevent sections of the DFSG--at least until non-free disappears and the DFSG gets changed (if that is what actually happens). I'm not referring to any specific technical reading of the DFSG, just the impression a company trying to sell software is likely to get when looking at supporting a distribution with an odd-ball packaging scheme, GPL-like guidelines, and no support infrastructure for commercial software. Note: they [Debian's developers] seem to have a problem with making money off Linux is a comment from the owner of a local Linux VAR. I don't necessarily agree with it, but couldn't really argue against it because I've never tried to deal with Debian on a company - organisation level. Unfortunately, the most vocal Debian proponent on the list, who has his own consulting business, was unable or unwilling to counter the claim. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Proposal - non-free software removal
On Sat, 16 Nov 2002, Nicolaus Kedegren wrote: On Fri, Nov 15, 2002 at 11:36:48AM -0700, Bruce Sass wrote: I think a much better solution would be for Debian to find a multi-national commercial partner to take over non-free before it gets dumped... maybe HP. This is a typical _drone_ argument. There is no such thing as a multi-national corporation that works for the public good. Trust me, I work for one of 'em. The best solution for Debian is to completely drop non-free and let the control freaks fight it out amongst themselves. And just to satisfy my curiosity, why is HP so especially suited to take on the task of maintaining non-free? Ya, ok, partner was a bad choice, ally would have been better; I would expect them to act in their own self interest (which should coincide with that of Debian's users). HP just kinda sprung to mind as a Debian friendly entity (DWN piece about Debian being added to their TestDrive program, probably, iirc) with a broad enough operation to support what would probably be a loss leader. joke Maybe MS will run non-free to show they really do support competition in the market. /joke - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Proposal - non-free software removal
On Fri, 15 Nov 2002, Travis Crump wrote: ... There is a long long ongoing debate in debian-devel[1] on this. Please don't start another debate here. This seems like the proper place for a discussion, they are wanting to change the Social Contract, and we are the society they have the contract with. I'm worried that if Debian totally ignores the non-free software world, it will ignore Debian. I think a much better solution would be for Debian to find a multi-national commercial partner to take over non-free before it gets dumped... maybe HP. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Use of telinit by regular users
On Wed, 13 Nov 2002, Cameron Dale wrote: I am not a member of this mailing list, so please copy me on all replies. I have a 4 year old laptop running potato (2.2r6). I use the various runlevels to control the services that are running. I want a regular user to be able to switch runlevels, but telinit seems to be able to only be run by root. Is there a way to get around this? sudo, or maybe calife or super, perhaps with a gtk or gnome frontend... presumably you know what you are doing and have a rescue floppy handy, so I won't go into the hazards (probably couldn't do 'em justice anyways :) - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: KDE 3.0.4 update various errors struck....
On Wed, 13 Nov 2002, Haralambos Geortgilakis wrote: Hi Yall, from the Help dang the update did not work dept What does this mean Unpacking libkcal2 (from .../libkcal2_4%3a3.0.4-1_i386.deb) ... dpkg: error processing /var/cache/apt/archives/libkcal2_4%3a3.0.4-1_i386.deb (--unpack): trying to overwrite `/usr/lib/libkcal.la', which is also in package kdepim-libs The libkcal.la file exists in (at least) two packages, and dpkg don't like that so it skips over the problem. You can force the issue with: dpkg -i --force-overwrite /var/cache...libkcal2...deb IF you are confident that doing so will not mess anything up. Selecting previously deselected package libkgantt0. Unpacking libkgantt0 (from .../libkgantt0_4%3a3.0.4-1_i386.deb) ... dpkg: error processing /var/cache/apt/archives/libkgantt0_4%3a3.0.4-1_i386.deb (--unpack): trying to overwrite `/usr/lib/libkgantt.so.0.0.2', which is also in package kdepim-libs ditto for this also Seeing how they are both KDE packages from the same release, I would go ahead and --force-overwrite, then file a bug report. - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [OT] literate programming
On Sun, 20 Oct 2002, Sean 'Shaleh' Perry wrote:
On Sunday 20 October 2002 13:36, Bruce Sass wrote:
Whether LyX:File-Build-Program generates a file set that adheres to
GNU standards is something you would have to check for as the last
part of the build procedure.
rant please oh please just don't format your code according to GNU
standards. Brance at beginning of line, hanging brace on line above fine.
BUt the brace does not belong half indented. Ick
/rant
Ya, it should be fully indented {
like
this
}
;)
- Bruce
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kde system wide file associations
On Wed, 12 Sep 2001, joe golden wrote: I have a small network of seven machines. Most students log on to the kde desktop. Is there any way as root that I can alter file associations for all users on all machines so that all *.doc files and *.rtf files are automatically opened with abiword? I know how to do this one by one, but I have 20 users. see: /usr/share/mimelnk either edit manually, or use the per-user GUI tool and cp - Bruce
Re: Latest woody upgrade to causes fonts in X to disappear
On Mon, 10 Sep 2001, Bob Nielsen wrote: I did an 'apt-get upgrade' to my testing system yesterday and got 50 MB of upgraded packages(!) This upgrade included version 4.10 of X. I notice now that on a large number of applications, such as the Gnome desktop and applications which have pop-up or drop-down menus, that the text has been replaced by squares (in fact, this makes Gnome essentially unusable; switching to kde restored much of the functionality). I assume this means that a particular font got messed up by the upgrade, but I really don't know where to look. Has anyone else seen this (and found a solution)? I tried Gnome for the first time in a long time last night and had the same problem, it was using blackbox - which looked OK, just the Gnome bits were un-usable. I fired up XFCE, started the gnome-panel (which came up ok), used the gnome control center to change the gtk theme and font, logged out, restarted gnome-session... and all is well[1]. I'm using the svga server from X-3.3.6. - Bruce [1] ...except that Gnome pauses for 2-3 second before it catches up with the pointer when using the menus, and another 2-3 second pause before it opens up a submenu... I think I'll stick with KDE, it may be a little slower starting up apps, but at least it responds immediately when the pointer is moved.
Re: How to handle unofficial package upgrade
On Thu, 6 Sep 2001, Greg Wiley wrote: Good day all- On a Debian Potato, I am using KDE 2.1 packages that are, obviously, not part of Potato but are from a fairly common source. If you are using stuff Ivan Moore has done... Since KDE 2.2 is slated for inclusion in the upcoming Debian release, what is the best way to prepare for the upgrade? I cannot assume that the new packages will be aware of the old and will upgrade them automatically ( will they?). Sure you can. So, am I best off finding every trace of the non-Debian KDE and eradicating or will things just sort of work out if I leave it all alone? Any suggestions? If things don't `just work' Ivan will be getting some bug reports. If you are running a KDE not built by Ivan, best be safe than sorry and go do a search and destroy on KDE before upgrading. - Bruce
Re: ALERT: XFree86 4.1.0-3 maintainer scripts hosed; please wait for 4.1.0-4
On Thu, 30 Aug 2001, Branden Robinson wrote: Folks might want to wait for 4.1.0-4. I'm preparing it now. Several bugs have already been filed; no one needs to add to them. The problem is understood, and the fix has been written and tested. If you already have 4.1.0-3 installed successfully, there is nothing to worry about. Data points: * the problem is in the preinst and postinst scripts of several of the packages, including ones that almost everyone has installed * if your /bin/sh is ash, you will likely have this problem and just in case it is not obvious (or too scary)... If you have sh linked to ash: $ rm /bin/sh $ ln -s /bin/bash /bin/sh dselect - install, or whatever ...is now working away to install the 58 broken packages I accumulated over the last couple of upgrades. Branden, is there anyone who should not do the above? - Bruce
Corel Linux (was: [ELUG] Interesting story in the news) (fwd)
Someone was wondering what's up with Corel Linux... http://cbc.ca/cgi-bin/templates/view.cgi?/news/2001/08/29/corel_010829
Re: ALERT: XFree86 4.1.0-3 maintainer scripts hosed; please wait for 4.1.0-4
On Fri, 31 Aug 2001, Zephaniah E. Hull wrote: On Fri, Aug 31, 2001 at 12:33:08AM -0600, Bruce Sass wrote: snip and just in case it is not obvious (or too scary)... If you have sh linked to ash: $ rm /bin/sh $ ln -s /bin/bash /bin/sh dselect - install, or whatever ...is now working away to install the 58 broken packages I accumulated over the last couple of upgrades. Branden, is there anyone who should not do the above? Yes, almost everyone. The proper command is 'dpkg --purge ash'. Seems a little heavy handed to me. Ash is not broken, it just needs to be moved out of the way. Dselect is done... ln -s /bin/ash /bin/sh and all is well again. - Bruce
Re: PrettyGoodPrivacy
On Fri, 31 Aug 2001, greg wrote: I've tried to install and use PGP but I'm having problems finding a suitable package for my Debian/GNUlinux 2.2 system. Would anyone point me in the right direction ? You've heard about gpg already... The international version of PGP 5 is available in a package named: pgp5i, in the non-US archive. - Bruce
Re: no mouse with X
On Thu, 30 Aug 2001, Cliff Sarginson wrote: On Thu, Aug 30, 2001 at 01:51:15PM -0400, Eric Cheney wrote: Hello. I'm trying to introduce a debian box to my officeI am not a sysadmin, so I stumble a little bit with these things. Anyway, I want to make a good impression of deb. I've installed woody. I'm getting weird things with the mouse. Can somebody help? Here's what's up. I loaded up X server and it connects and all that. If I start X with gpm, I get a mouse, but it is very erratic and unacceptable. So, if I remove gpm, and then restart X, the mouse works fine under X. Ok, if I then reboot, there's no connection to the mouse without gpm; so after reboot (after removing gpm), the mouse is dead under X. In the XF86Config-4 setup file I have the mouse on /dev/psaux. I checked, and there is a link from /dev/psaux - gpmdata. I'm using a ps/2 mouse. That is correct. Hmmm... try X looking at /dev/gpmdata, gpm looking at /dev/psaux and repeating `raw' --- /etc/gpm.conf --- device=/dev/psaux responsiveness= repeat_type=raw type=ps2 append=-l \a-zA-Z0-9_.:~/\300-\326\330-\366\370-\377\ - Console mouse (gpm) and window managers don't always get on well, I recall this is often a particular problem with gnome. Works for me with twm, mwm, blackbox, icewm, xfce, and KDE... I don't use Gnome though. - Bruce
can't open /var/lib/dhelp/titles
Hi, Would someone please post the output of ls -l /var/lib/dhelp. TIA - Bruce
Re: can't open /var/lib/dhelp/titles
On Wed, 22 Aug 2001, allen wayne best just ramblin in his amx wrote: On Wednesday 22 August 2001 17:08, Bruce Sass wrote: Would someone please post the output of ls -l /var/lib/dhelp. [EMAIL PROTECTED] Build IA32 [/root]! #ls -l /var/lib/dhelp total 842 -rw-r--r--1 root root 815104 Aug 15 10:01 dbase -rw-r--r--1 root root28672 Jul 18 17:05 dbase_fsstnd -rw-r--r--1 root root10240 Aug 15 10:01 titles -rw-r--r--1 root root 1280 Jul 18 17:05 titles_fsstnd :-/ Darn, and I'd have put money on the problem (dhelp_parse failing when installing stuff) being a result of inadvertently doing a cp -R on /var instead of cp -a. sigh Oh well, I guess I'll have to actually use a brain cell or two to solve this one. ;-) Sorry 'bout bugging the list. - Bruce
Re: can't open /var/lib/dhelp/titles
On Wed, 22 Aug 2001, Eric G. Miller wrote: On Wed, Aug 22, 2001 at 06:08:37PM -0600, Bruce Sass wrote: Would someone please post the output of ls -l /var/lib/dhelp. dhelp is broken in unstable if that's what you're referring to. There's already a bug report on it. That's it... darn coincidences. [I.O.Eric one brain cell] :) - Bruce
Re: How do I make python2 work with readline?
On Thu, 16 Aug 2001, Oleksandr Moskalenko wrote: Hello, I'm puzzled as to why python2 doesn't use readline. I'm learning Python and really liked command line capabilities that 1.5.2 + readline provided. However, python2 doesn't seem to utilize readline leading to some horrible experience for someone like me who makes a lot of mistypings. Does anybody know how to wrestle python2 into using readline? I think it is a bug; python2 is Python-2.0.1, so readline should be uncommented in Setup because 2.0.1 is supposed to be GPL compatible. shrug The Python-2.1.1 packages in http://people.debian.org/~flight/ have been working ok for me, or you can build 2.0 yourself under /usr/local. - Bruce
Re: FW: Careful. This is for information only.
On 7 Aug 2001, John Hasler wrote: Ian writes: Seems like a pretty grey area though. He is ultimately responsible for it. There has been enough publicity about CR to ascertain that the system admin was negligent in his duty. Someone on Advogato just pointed out another risk. What if you only popped up a message but someone else installed a nasty trojan? How do you prove that you didn't do it? The burden of proof should be on the accuser, especially when dozens (hundreds, thousands?) of computer owner's log files show the box was engaged in a potentially damaging activity before the warning dialog popped up. Perhaps checking with a registry of infected/infested machines first would help. If someone put a vermicide package in Debian, I would probably install and use it... for sure if it had a sequence of actions that were triggered by sucessive probes from the infested machine (ranging from least to most intrusive, and taking steps to minimise the equivalent of a DDOS attack on the infected machine). I think the time is right for software antibiotic and vermicide packages, and I think anything that helps stop the spread of infectious agents would be welcome by everyone using the 'net for legitimate purposes... as long as the steps taken are not proactive, there should be little room for the owners of systems that get medicated to accuse the users of such packages of engaging in nefarious activities. - Bruce
Re: IP Masquerading
On Tue, 7 Aug 2001, Vineet Kumar wrote: * [EMAIL PROTECTED] ([EMAIL PROTECTED]) [010807 10:35]: What is a good program for Windows 98 that will allow me to set up IP Masquerading to share my internet connection with some Linux boxes? The Right Way to do this is to make one of the Linux machines do the masquerade. A windows 98 machine should never used as any type of server / gateway / anything except a desktop system. Additionally, this list is the wrong place to ask questions about windows programs. If you are implying he should have asked on a Windows list... would he have got the same Right answer? Maybe he did the right thing. ;-) - Bruce
Re: FW: Careful. This is for information only.
On 8 Aug 2001, John Hasler wrote: Bruce writes: If someone put a vermicide package in Debian,... Such a package cannot go into Debian. Besides, such a thing should not be used by anyone not prepared to build it from source. I think the time is right for software antibiotic and vermicide packages, and I think anything that helps stop the spread of infectious agents would be welcome by everyone using the 'net for legitimate purposes... as long as the steps taken are not proactive, there should be little room for the owners of systems that get medicated to accuse the users of such packages of engaging in nefarious activities. I agree with you completely, but the NatCops are unlikely to agree with you at all. There is safety in numbers... hence the call for packages (binary, source, or installer, using Debian's terminology). The NetCops (Nat was a typo?) obviously don't know what to do except threaten legal action, and I believe they would be seen as the bad guys for harassing people trying to minimise or fix something they themselves recognize as a serious problem. It would be a good test of `Good Samaritan' laws, if nothing else. - Bruce
Re: FW: Careful. This is for information only.
On 8 Aug 2001, John Hasler wrote: Bruce Sass writes: The NetCops (Nat was a typo?) No. 'NatCops' == 'National Police': FBI, BATF, DEA, Treasury Agents, etc. It comes from an old alternate-worlds sf story I disremember the name of. In that world one did _not_ call a NatCop 'NatCop' to his face. Ok, thanks. Here in Canada it would be the RCMP (and probably CSIS). Has anyone asked the NatCops what they think of vermicidal and antibiotic software? (i.e., has it come up before and elicited an official response from any organisation) It would be kinda like doctors doing house calls, everyone old enough probably misses that, eh. Like everything else in life, good salesmanship would go a long way. - Bruce [off to find an email address for the RCMP]
jurisdictions (was: FW: Careful. This is for information only.)
On 8 Aug 2001, John Hasler wrote: Bruce Sass writes: Here in Canada it would be the RCMP Except that the US purports to have no national police, police power being one of those powers supposedly reserved to the States by the Constitution. Searching the RCMP site reveals they consider computer crimes to be a matter for the local police. The only mention of viruses, etc., I have found (so far) is unconnected to any program or service they offer. Not encouraging. - Bruce
Re: FW: Careful. This is for information only.
On Wed, 8 Aug 2001, Dave Sherohman wrote: On Wed, Aug 08, 2001 at 02:08:57PM -0600, Bruce Sass wrote: Has anyone asked the NatCops what they think of vermicidal and antibiotic software? (i.e., has it come up before and elicited an official response from any organisation) Late last month, news reports say that there was much discussion within FBI, etc. regarding what to do about CR and antiworms were dismissed on the basis that it would be sinking to the original author's level. Ya, releasing more self-replicating code into the environment is probably a bad idea, no better than biological warfare. I'm thinking of an automated and distributed response system. The idea is to identify infested machines, take increasingly intrusive measures to notify the sysadmin of the problem, finally either cleaning the box or disconnecting it from the 'net. The course of action taken in the last step would be determined on the basis of whether or not the owner of the machine had opted-in to the auto-cleaning service. Betcha it would `fly' in the corporate world if you charged money for the cleaning service, and got rid of the notifications... - Bruce
Re: jurisdictions (was: FW: Careful. This is for information only.)
On 8 Aug 2001, John Hasler wrote: Bruce writes: Searching the RCMP site reveals they consider computer crimes to be a matter for the local police. Sounds good to me. Sure, once an offender has been identified. Not encouraging. Why? I've seen no evidence that the bungling boobs at the FBI have been any help. I have confidence that my local police force can collaborate effectively with a jurisdiction on the other side of the country, but once you start crossing international borders... also, taking Canada and the US as examples, any cooperation between the two nations that involves `problem areas', rather than specific incidents, seems to be handled by the RCMP and one of the organizations you mentioned (Drug Enforcement Administration; Bureau of Alcohol, Tobacco and Firearms, etc.) Essentially, this international issue is not on the RCMP's radar; viruses, trojans, etc., are seen as a matter of keeping ones virus scanning software up to date (too bad that paradigm doesn't seem to be working). So far, it looks like the federal government will need to pass explicit laws making viruses, etc., illegal before the RCMP can/will even consider the issue... that is not encouraging because governments tend to be slow and too cut'n'dried/black'n'white with their remedies. - Bruce
Re: Python 2.1 .debs (for sid)
On 7 Aug 2001, João Alfredo wrote: Does anyone knows where I can find it?!? http://people.debian.org/~flight/
Re: PINE
On Wed, 18 Jul 2001, User zos wrote:
Compiling Pine on a debian system doesn't work out of the box in my
experiance. There are .debs of Pine 4.22 last time I checked, but its been
a while. I have compiled Pine on a few debian boxen and if you are
interested in making a compile from source work (the better option vs. a
debianized source package IMHO) you should look at the makefile structure
of pine and start looking at why it won't compile. I think I remember it
complaining about not seeing libraries and going and changing those
references manually. Hope that helps.
This may help someone compiling Pine...
The Pine people confirmed it as a known problem and said it would be
fixed next release.
---
From [EMAIL PROTECTED] Wed Jul 18 02:17:41 2001
Date: Tue, 26 Jun 2001 23:51:42 -0600 (MDT)
From: Bruce Sass [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: build lnp sys/time.h problem
Hi,
Here is what I have...
Pine 4.33
libc6: glibc-2.2.3
gcc: gcc-2.95.4
on a Debian testing/unstable system (a.k.a. Woody/Sid)
Here is what I saw (close enough)...
in tz_sv4.c, function rfc822_timezone, line 30:
tzname undeclared, daylight undeclared,
dereferencing pointer to incomplete type
The last line was common to all the time.h related errors that came up
when these where compiling:
imap/src/osdep/unix/{os_lnx.c,news.c,phile.c}
I also fixed mh.c and mx.c, just 'cause they were probably next.
The fix was s!sys/time.h!time.h!
pointed out by a note in the sys/time.h file provided by libc6.
---
- Bruce
Re: python2 and qt
Yes. I've built sip and PyQt (2.5pre stuff) in /usr/local against Debian's Python-1.5.2 and 2.0.1, and 2.1 in /usr/local... in all cases, I used --with-python=/path/to/py/executable in the ./configure ... command and everything just worked. - Bruce -- On Sun, 15 Jul 2001, Sergio E. Schvezov wrote: sorry for not saying i have Sid, I've installed that package and can import it with no problems with python (from python-base), but not with python2 (python2-base) So do i have to build pyqt against python-2.x? TIA * Bruce Sass ([EMAIL PROTECTED]) wrote: On Sun, 15 Jul 2001, Sergio E. Schvezov wrote: hi, simple q, maybe hard answer, here goes: how can i work with python2 and qt in debian? Right now, for stable... build PyQt and Python-2.x yourself. Sid has python-pyqt, built against Python-1.5.2.
Re: python2 and qt
On Sun, 15 Jul 2001, Sergio E. Schvezov wrote: hi, simple q, maybe hard answer, here goes: how can i work with python2 and qt in debian? Right now, for stable... build PyQt and Python-2.x yourself. Sid has python-pyqt, built against Python-1.5.2. - Bruce
Re: startx -- :1
On Tue, 10 Jul 2001, Kent West wrote: harsha wrote: gave startx -- :1 but this is what happens ... I noticed this a few months ago on my machines also (running Sid). I don't know if the startx script/sequence is broken, or if the procedure has been changed, but some one a few weeks back mentioned that you can get around the problem by adding the vt number you want. For example, you'd type: startx -- :1 vt8 I missed that, thanks. I hope it's just a bug and not a new method; I know it's been filed as a bug, but I haven't checked on the status lately. ditto. The command line has gone from, e.g., startx blackbox -- :1 to startx /usr/bin/blackbox -- :1 vt6 which means, to get the same functionality you need to do... startx `which blackbox` -- :1 `nextfreevt` ...but I don't think there is a nextfreevt program available. - Bruce
Re: migrating debian unstable to debian testing ?
On Tue, 3 Jul 2001, Bostjan Muller wrote: I'd like to know if there is any way of downgrading all packages from Debian unstable (sid) version to Debian testing (woody) version? Hmmm, I seem to recall a message (a month or so ago) stating that you could force a downgrade by pinning all the packages to testing, then doing a dist-upgrade. I'd backup everything first and have a working rescue disk on hand before trying it... well, if I didn't like living dangerously. ;) - Bruce
Re: Re. Total Confusion
On Mon, 2 Jul 2001, D-Man wrote: On Sun, Jul 01, 2001 at 02:49:43AM -0700, Sidney Brooks wrote: | For those of you who tried to help with my problem, several weeks | ago, here is a statement of the problem and solution. | After following all suggestions offered here and consulting with a | computer technician, the conclusion was that it had to be the modem. | I bought a new Diamond modem and both versions of linux now get me | online. Our guess is that I had a line surge that knocked out a part | of the modem that linux requires, but that Windows can do without. It is always good to find a solution :-). | I still have two minor problems that I may be able to work out myself. In | order to get on line with Debian, I must use ppp. Minicom and wvdial | connect but fail to authenticate. As Wayne mentioned, minicom and wvdial aren't supposed to authenticate or maintain a ppp connection, that is pppd's job :-). minicom is an _interactive_ dialer. Minicom is a terminal program or comm program... as in dial up over a serial line, login, use your shell account. It is only inteneded to dial the modem, no more, no less. No, it is intended as a comm program. Also, because it is interactive, it is only really useful when determining what the dialog with the ISP should be, and then it is essential. It is interactive because a comm program would be useless if it was not. ... I used minicom to see what my ISP sends and what it expects. With the knowledge of these expect-send pairs I set up a chat script (chat controls the modem and is driven by a set of expect-send strings in a config file) and use 'pon' to dial. Sounds like a good use of minicom if you don't have serial access to a box. ... Minicom is a great tool for determining how your ISP handles an incoming call, then after that it isn't really useful because (AFAIK) it isn't scriptable. Yes it is, but if you don't have a dialup shell account the feature is kinda useless (it simulates keypresses), eh. Youngsters! What is this world coming to, never heard of a comm prg, probably don't know what x/y/zmodem and kermit are either. ;) just for the fun of it... I can dial in and read my mail/surf-the-web using a C64 and a comm program, and if my ZX81 was still working I could hook up a home-built low-speed modem I built many years ago and use it. Maybe I'll go have a nap now, I'm feeling old all of a sudden. :) - Bruce
Re: editing /etc/init.d
On Mon, 2 Jul 2001, john gennard wrote: I need to edit a script in /etc/init.d to allow a program, which I haven't been able to get to run, to log what is happening. So far I've not touched this area and find the manpage for 'update-rc.d' a little confusing. The script to be edited is linked to /etc/rc0/ /etc/rc6/ (as defaults with runlevels and sequence code). Do I need to remove all of these links with -f, so running 'update rc.d -f xx defaults' ? I presume I could run the same command with -n instead of -f just to have a 'look-see' -but I'm wary of touching something I don't yet understand. In any event this seems too simple. Would someone be good enough to advise me. Don't use update-rc.d It is great for what it was intended, package maintenance scripts... but is just an extra layer of potential confusion for the sysadmin who just wants to stop xdm from starting in rl3 (whatever). I customize rl's 2-4 and 7-9 with rm and ln commands, and leave rl5 alone - don't touch _one_ of rl's 2-5 and Debian will honour your configs through upgrades. - Bruce
Re: Which web server for multiple domains?
On Thu, 28 Jun 2001, Craig Dickson wrote:
I've been using the web server Boa, and generally it has done a good job
without using up too much memory or CPU time. The one thing it doesn't
do that I now need is to host multiple domains as logically-distinct
sites. E.g. if someone requests www.domain1.com (not my real domain
name, btw), I want a different page to come up than if they request
www.domain2.com, even though the DNS addresses for those sites are
identical. From the docs, I think Boa could do it if the two sites were
served through separate network interfaces, but that's not the case
here. There's one network interface, and the only difference between
www.domain1.com and www.domain2.com is the domain in the URL (which I
believe HTTP does provide to the server, so the server should be able
to differentiate based on it).
Does anyone know if Boa can do this, or if any of the other servers
could? (I daresay AOLServer or Apache could, but I'd rather go with a
smaller, simpler server if it will do what I need).
Hmmm, it sounds like you want the virtual host thing...
- add VirtualHost to /etc/boa/boa.conf
- edit /etc/hosts to look like, say...
127.0.0.1 you localhost
127.0.0.2 other1 localhost
127.0.0.3 other2 localhost
- setup /var/www to have dirs named, 127.0.0.{1,2,3}
- restart boa
requests for:
http://you/, http://other1/, and http://other2/ will be served out of
localhost/var/www/127.0.{1,2,3}, respectively.
- Bruce
Re: Unix administrator
On Fri, 29 Jun 2001, Matthew Garman wrote: - set up X - configure and use network devices: ethernet, serial analog modem, etc. - install and use a printer - download a program's source code, compile it, and install it on your system (i.e., do *not* use the system's package management tool to install software) - do figure out how the package management system(s) work (because they are starting to turn up everywhere) - configure system services, especially common ones such as: - FTP (e.g. proftpd) - a web server (e.g. Apache) - a mailer (e.g. postfix, or sendmail if you're brave) - secure shell (e.g. OpenSSH) - familiarize yourself with shell scripts (and maybe perl scripts)---at *least* enough that you can understand what's going on in them (and that's a bare minimum) - configure and recompile your kernel All worth repeating. :) - Bruce
Re: big IBM harddisk (fwd)
On Wed, 27 Jun 2001, Frank Zimmermann wrote: virtanen wrote: I tried, but if I connect my IBM, the machine doesn't boot at all... This sounds pretty much like a hardware failure to me. I had a problem when one of my harddrives was broken. I always got a drive error and the PC wouln't boot at all. No matter where I put the HD. If you have the possibility to check the HD on another PC do it ASAP. Not necessarily. I picked up a 30G drive a couple of weeks ago, jumped through a hoop to install it as hda with the MBR on hda, everything was fine... until I thought it would be a good idea to put a small DOS partition at the start of the disk. DOS couldn't handle the big HD and left the BIOS in a state where it doesn't recognize the master HD anymore and refuses to boot (but I can still boot from a floppy); I may need to swap hda and hdb if I can't re-convince the BIOS that there is actually a drive hd0. - Bruce
Re: OT: C++ Newbie and KDE/QT
On Tue, 26 Jun 2001, Brendon wrote: This summer holiday I took on the task of learning C++ with (shamefully :) the help of C++ For Dummies. Having tried to learn C++ in the past I'm now reasonably familar with it's synax so i thought i'd also try learning QT/KDE programming at the same time. But the tutorials I've been through on the doc.trolltech.com site have left me a little disappointed. Does anyone know of a good site where QT/KDE programming is explained? And what did you start with when learning C++? Have you installed the kdevelop package. iirc, it comes with a tutorial and some other KDE programming material, their site should also have some good stuff. - Bruce
Re: help? dpkg db screwed up
On Mon, 18 Jun 2001, Iain wrote: I seem to have put the dpkg database into a bad state. I was attempting to install phpgroupware from unstable onto a potato system. Anyway somewhere along the installation process it screwed up so I tried to remove it. dpkg -r phpgroupware gives the following: dpkg: error processing phpgroupware (--remove): Package is in a very bad inconsistent state - you should reinstall it before attempting a removal. Errors were encountered while processing: phpgroupware Reinstalling it gives the following: Selecting previously deselected package phpgroupware. (Reading database ... 16354 files and directories currently installed.) Preparing to replace phpgroupware 0.9.10v-6 (using phpgroupware_0.9.10v-6_all.deb) ... Unpacking replacement phpgroupware ... dpkg: warning - old post-removal script returned error exit status 10 dpkg - trying script from the new package instead ... dpkg: error processing phpgroupware_0.9.10v-6_all.deb (--install): subprocess new post-removal script returned error exit status 10 dpkg: error while cleaning up: subprocess post-removal script returned error exit status 10 Errors were encountered while processing: phpgroupware_0.9.10v-6_all.deb Hmmm, there should be something in the ... bit that tells you what the actual problem is, and (about?) which line of the post-removal script is causing the problem -- comment out that line (in /var/lib/dpkg/info/phpgroupware.postrm) and try using the pkg mgt tools again (make note of what the line is supposed to do, just in case you need to do it manually). As a last resort, or if it is a small (few files) package, you can manually remove it: - run any necessary commands from the pre and post removal scripts - rm all the files listed in /var/lib/pkg/info/phpgroupware.list that do not also belong to another software package - edit out the entry for phpgroupware in /var/lib/dpkg/status I have tried dpkg -r --force-remove-reinstreq phpgroupware but it doesn't work either. Is there some way I can get things back to normal? Please help. There are usually a few options, but without seeing the full text of the messages you are getting it is tough to say which is the best fix. HTH - Bruce
Re: KDE on a slow machine (was: Best WM for slow machine?)
On Sun, 10 Jun 2001, Margarete Hans wrote: Mmh. My laptop has 20 MB of RAM. I don't think that I'll be using it very extensively - it is rather a test to decide if I'm going to install debian on my main computer, which by now is also starting to get old (166 with 32 MB of RAM and 3 GB HD - still running windows With 20M (or even 24) you would want to keep an eye on how much and what is being swapped, just so you can tell the difference between poor system performance and an overtaxed system. I usually run top on a 132x60 text console, or do something like alias snap=top -n3 -d1 ~/top.txt to keep an eye on this stuff. ( ). Does gnome use as much memory as KDE? shrug... Besides, what does KDE give you more than these smaller WMs? Most noticeable would be the Control Center style configuration handling of the KDE and kapps, the apparent embedding of one app in another (e.g., previewers in konqueror), and a framework for doing the desktop shortcuts and mimetype magic things you want. Aside from the embedding stuff, you can probably do everything KDE does via xsession, etc., with a non-DE window manager. ...I ran a little experiment. Procedure: for wm in none,twm,blackbox,kde2 reboot text login top -n3 -d1 ... start wm wait top -n3 -d1 ... then for wm in none,twm,blackbox,kde2 reboot start wm wait text login top -n3 -d1 ... Results (representative): --first top after a reboot 23:15:20 up 2 min, 1 user, load average: 0.35, 0.36, 0.15 30 processes: 29 sleeping, 1 running, 0 zombie, 0 stopped CPU states: 35.9% user, 12.9% system, 0.0% nice, 51.2% idle Mem: 62864K total,35172K used,27692K free, 1344K buffers Swap:55664K total,0K used,55664K free,22112K cached --twm 23:19:26 up 3 min, 2 users, load average: 0.01, 0.18, 0.12 32 processes: 31 sleeping, 1 running, 0 zombie, 0 stopped CPU states: 19.3% user, 7.5% system, 0.0% nice, 73.2% idle Mem: 62864K total,38984K used,23880K free, 1440K buffers Swap:55664K total,0K used,55664K free,23804K cached --blackbox 23:30:40 up 3 min, 2 users, load average: 0.13, 0.21, 0.13 32 processes: 31 sleeping, 1 running, 0 zombie, 0 stopped CPU states: 20.1% user, 8.0% system, 0.0% nice, 71.9% idle Mem: 62864K total,37464K used,25400K free, 1396K buffers Swap:55664K total,0K used,55664K free,23232K cached --kde2 00:04:50 up 8 min, 2 users, load average: 0.53, 1.19, 0.67 50 processes: 49 sleeping, 1 running, 0 zombie, 0 stopped CPU states: 58.1% user, 9.8% system, 0.0% nice, 32.1% idle Mem: 62864K total,61432K used, 1432K free, 2048K buffers Swap:55664K total,8K used,55656K free,32416K cached Notes: The uptimes show the time taken to startup the X environment via kdm, switch to a tty, login, then do a top command (with a 486DX2-25); I ended up doing kde2 4 times 'cause I got hit with `maximal mount count reached...' twice, both times no swap was used. Conclusions: Compared to twm or blackbox, KDE2 takes about 6 times longer to startup and needs about 64M of RAM to avoid swapping, plus whatever your apps need if you want to actual use KDE without swapping (although, from experience, switching between apps usually only takes a few seconds and switching desktops or giving focus to large apps that got swapped out, takes quite a few seconds, even at 25MHz). Not for the impatient, but what do you expect from something that was probably rescued from the scrap heap or landfill. HTH, Bruce
Re: Xemacs Debian packaging systems
On 10 Jun 2001, Glyn Millington wrote: ... Well I ran Xemacs as root and did some upgrading with it's packaging system and it doesn't appear to have done any harm ;-) Maybe we should just have assumed that apt-get is just too damned good to be thrown by Xemacs' antics! The only time you should run into a problem doing this sorta thing is if: the app managed package contains the same files as a Debian package, but they do something different; or you purge the Debian package and it takes out something the app installed package needs. - Bruce
Re: gnome/KDE
On Sun, 10 Jun 2001, Phillip Deackes wrote: On Sun, 10 Jun 2001 01:16:40 -0400 Margarete Hans [EMAIL PROTECTED] wrote: What are the advantages/disadvantages of gnome and KDE? Basically, which one should I install? Depends. We all have our favourite window manager (although KDE and Gnome are more than window managers, they are more 'window environments'). desktop environments, actually. Personally, I don't like KDE very much - the 'KDE' angle is very prominent and when using it, 'KDE' is in your face all the time - KDE this, KDE that , Kwhatever and so on. Running non-KDE apps in the KDE environment almost ...and with gnome it seems to be Gwhatever shrug. makes you feel you are doing something wrong! I also don't like the look that qt (the library on which KDE is based) gives to windows. I don't like the default KDE setup either, but a couple of clicks gets you a nice BeOS(?) window style, or motif, or windows, or... and if that is not enough there are themes and icon themes to play with... but it is just eye-candy kinda stuff. Gnome seems to me to be less pervasive - OK, there are plenty of Gnome apps around but they look good - less plastic and more functional, more like other non-gnome apps. A Gnome environment looks good with Gnome and non-Gnome apps. Gnome is based on GTK libraries and there are plenty of other GTK Linux apps around which have nothing to do with Gnome. There is nothing stopping someone running Gnome or GTK based apps in KDE, and they should look the same in either environment - if that is what you want. Both are attempts to give Linux that conformity of look feel that Windows has. Both are fairly easily configurable as regards the look. Look'n'feel is a small part of it; interoperability between apps and the technical infrastructure you need to do the ActiveX and COM kinda stuff you see with Windows is more important -- you can't get that with the traditional X based window managers, and it is still not quite there for either Gnome or KDE. Obviously, these are personal observations and you will almost certainly get the complete opposite opinion from someone else. The answer really is to try both and see which you prefer. Definately, but you will need to try both for _awhile_ to get a true picture... there is a lot more to the DEs than just looks'n'feels. I suggest installing and using both (if you have the room), then reporting back to the list in a few weeks with the results of the test. :) - Bruce
KDE on a slow machine (was: Best WM for slow machine?)
On Sun, 10 Jun 2001, Margarete Hans wrote: Woops... I was just planning on installing KDE on my old computer (486, 40kHz CPU). I guess I'll try something else if there are already problems running it on a 166... The problem is not the 166, it is the 16 (Meg of RAM). I'm running KDE on a 486DX2-25 and it works fine... once it gets running. KDE startup is, and some apps (e.g., konquorer in a largish dir) can be, very slow. I wouldn't try it with less than 32M of RAM, and if I had less than the 64M I have I would probably use something else (KDE would run, but it would be too painful to run the 4 or 5 apps I usually have going at the same time). Someone just interested in checking out KDE to see what is there could get by with 24M of RAM. - Bruce
Re: debian on a 486
On Mon, 11 Jun 2001, Robert Ruzbacky wrote: Hi. I want to put debian on a 486 that I recently got from my school. . for the boot disk, or find an older debian boot disk and hope it works with the machine and still works with the debian installation process. You could try using an old version of debian like debian 1.3 (Bo). I think you can get it in the archivesit should work fine for a 486. Mine came I'd suggest ver2.0 (get it from archive.debian.org) for any floppy based intallation on a resource limited box. There is a low mem install (4Meg or better), the boot floppy uses the slow-safe-and-stupid driver (it was the only one that would work on the 486 I installed a couple of weeks ago), there are static versions of apt-get and dpkg available for it (easy upgrade path), and it doesn't need much HD space for the base system. - Bruce
Re: (dup) serial port: LSR safety check engaged (fwd)
On Mon, 4 Jun 2001, J.A.Serralheiro wrote: well, cant tell exactly whats the problem if you have a pnp card, then running pnpdump will tell you what configurations are supported by that card. Then, all you have to do is run isapnp with the particular configuration you want. I looked at that... pnpdump spits out the same as what I used to config the card, well over a year ago. Be sure of one thing, if your kenel has isapnp built in, then you will have many problems. serial.o is builtin, isa-pnp.o is a module. before configure isapnp.conf with the start up configuration of you pnp modem (I think) ytou must recompile the kernel without isapnp built in ( but with plug and play supports still built in). Probably you kernel identifies the board and then isapnp tries to reconfigure it. THe thing is that you get LSR blah blah ISAPNP is allways run at start up. usually isapnp.conf is empty so there ae no problems. If you put something there you must check what I told you. I've moved isapnp.conf out of the way and rebooted, it matters not. But why would it stop working all of a sudden??? It must be something in testing/unstable... everything worked well, until setserial started playing around with that AUTOSAVE crap. I had to do this because of another matter ( similiar). I couldnt configure my modem with the exact configuration I wanted AND was supported by the card. This is not a new card, kernel, or config setup... later, Bruce
pinning down... serial port: LSR safety check engaged
It has worked with an explicit 'skip_test' and the implicit test, with both the 2.2.15 and 2.4.3 kernels. I wouldn't be surprised if it sometimes works without skip_test. Has it ever failed with it? Is skip_test in /etc/serial.conf? Yes, during this and the previous round. Not right now. This is the best I can do to pin it down... Looking at the timestamps of older serial.conf-s I can tell that I ran with skip_test until problems cropped up with irq7 on the same port (turned out to be the sound card), skip_test went out while I redid isapnp to use irq3. At some point I put skip_test back in until the first episode of LSR..., by then I had moved from Potato/2.2.15|17 to Woody/2.4.3, then to Sid/Woody (although I don't think there are many woodies on the system anymore). The problem first hit a few weeks after adding unstable to sources.list, closer to when I started to fiddle with setserial via the debconf interface, not clearly associated with an upgrade (i.e., I have toggled the connection a few times without problems in both instances), and only after a reboot. I may have triggered an unlikely bug with some combination of manual and automated serial.conf fiddling at just the wrong time, coupled with setserial changing the infrastructure for handling serial.conf and isapnptool upgrades. Looking at the changelogs shows the most recent upgrade which would directly affect ttyS3 to be that of isapnp (May 18), but if that borked things I should have noticed earlier 'cause libc6 has been through a few quick upgrades since then (I reboot whenever libc gets an upgrade). To make things interesting... isapnptools was also upgraded (Apr 17) a little before I recall having LSR problems the first time (before the end of April, but I destroyed the timestamp evidence, no backup, oops), and I would have rebooted when xlibs got an upgrade (Apr 29). I don't have a clue as to exactly when or how I fiddled with serial.conf during this timeframe, but I did fiddle with it. So... the best correlation I can come up with is a new isapnp followed by a reboot, but muddied up with a changing serial.conf. I hope anyone who made it this far got something out of it... like maybe keep more backups, logs, and scribbled notes. :) Thanks for all your help and suggestions. - Bruce
Re: serial port: LSR safety check engaged (fwd)
-- Forwarded message -- Date: Mon, 4 Jun 2001 17:14:15 -0600 (MDT) From: Bruce Sass [EMAIL PROTECTED] To: John Hasler [EMAIL PROTECTED] Subject: Re: serial port: LSR safety check engaged On 4 Jun 2001, John Hasler wrote: I just rooted around a bit in the kernel sources. Although I find no mention of the 'LSR safety check' in the logs, it seems to be missing from 2.2.14. Perhaps you have uncovered a kernel bug. Do you have an old kernel to try? Ya, but I can get it to fail on that port also. I dug up an old 2.2.15 boot floppy, it didn't like that there was no support for itself on the fs (also sans isapnp and ppp modules) but serial.o was built in - and it worked. I removed the link to isapnp in rcS.d and rebooted the 2.2.15 kernel (since there is no isapnp.o, it shouldn't matter(?))... it didn't work. Anyways... I ended up fiddling with setserial (eventually going back to manual), rebooting a few times (flip-flopping between 2.4.3 and 2.2.15)... and all of a sudden it works again. :/ I'm suspecting setserial's autosave code, maybe it gets confused easily if all the bits (stuff in /etc and /var) are not consistent; although I wouldn't rule out some side-effect like interaction between isapnp and setserial. - Bruce
Re: (dup) serial port: LSR safety check engaged (fwd)
On 4 Jun 2001, John Hasler wrote: The LSR test will not be performed if ASYNC_BUGGY_UART is set, and include/linux/serial.h says: #define ASYNC_BUGGY_UART 0x4000 /* This is a buggy UART, skip some safety * checks. Note: can be dangerous! */ Try configuring the port with 'skip_test'. It has worked with an explicit 'skip_test' and the implicit test, with both the 2.2.15 and 2.4.3 kernels. I think it is going to take something more rigorous than the simple 'change one variable at a time' approach I've (mostly) been using. The last time I noticed really strange behaviour like this[1] was with KDE (the restart KDE or rm ~/.kde days), and it appeared that inconsistent configs was contributing the problem (hard-coded, system-wide, per-user, and DB could all be different, depending on what you had been upto), simply because there was always wider agreement between the pieces as to the way things should be when they worked than when they didn't work. All this user friendly config handling is nice... but it appears to be pretty fragile at times. - Bruce [1] I reboot whenever unstable gives me a new libc or xlibs (to ensure upgrade related problems bite me sooner rather than later), and tend not to fiddle with it if it is not broken. The only thing I can think of that may have had a delayed effect was if something like setserial came along and fiddled (or let me fiddle) with things I would only see the actual effect of when I rebooted (like maybe changing from manual to autosave-once when that nice S-lang debconf screen comes up 'cause setserial got upgraded).
Re: dselect, can't do update
On 6 Jun 2001, Daniel Wagner wrote: since yesterday i've a strange problem with dselect. i tried to look for new packages at the mirror ( i use sid ), and did a update in dselect, everything seemed to work, but at the end of operation the following errormessage came. |Reading Package Lists... Done |Building Dependency Tree... Done |Merging Available information |Replacing available packages info, using /var/cache/apt/available. |dpkg: parse error, in file `/var/cache/apt/available' near line 1020 package `blt': | `Replaces' field, reference to `blt-dev': error in version: version string has embedded spaces There are two or three instances of this in the blt and blt-demo packages; it looks like: ( 2.4i-1 )s/b ( 2.4i-1) or with j instead of i. | |update available list script returned error exit status 2. |Press enter to continue. any suggestions what to do. i tried already to remove the available list to enforce the rebuild of it, but this didn't work. There is probably a better way to do this... /var/cache/apt/available is built from the stuff in /var/lib/apt/lists, which is where the problem lies (in the file(s) for the main distribution). If you try to edit /var/cache/apt/available and rerun `update' your fixes get over written when the available file is built. If you edit the files in /var/lib/apt/lists to fix the problem (in the entries for the blt and blt-demo packages), apt notices that you don't have the same as what is in the remote archive and fetches the files again -- first downloading them into /var/lib/apt/lists/partial. If you interrupt the connection, apt will abandon the stuff in the partial directory and rebuilt the available file from what it has in lists (the `using old packages lists' message). So... do the edits in /var/lib/apt/lists, restart the update then interrupt it, wait for apt/dselect to timeout (ignore the error message), choose select ...all should be well (until next time, maybe). HTH - Bruce
Re: dselect, can't do update
On 6 Jun 2001, Dave Carrigan wrote: Bruce Sass [EMAIL PROTECTED] writes: If you edit the files in /var/lib/apt/lists to fix the problem (in the entries for the blt and blt-demo packages), apt notices that you don't have the same as what is in the remote archive and fetches the files again -- first downloading them into /var/lib/apt/lists/partial. Actually, I fixed the problem yesterday for myself by editing /var/lib/apt/lists/http.us.debian.org_debian_dists_unstable_main_binary-i386_Packages and apt did not download a new one on the next dselect update. I think maybe it compares timestamps and does not download a new packages file if the local one is newer. That's odd... not that it didn't re-fetch the packages files, but that it didn't for you and did for me. - Bruce
Unidentified subject!
Hi, Anybody know what an LSR safety check is, and how to disengage it? I've been getting this message... ttyS3: LSR safety check engaged! ...when I do pon or minicom (which also says /dev/ttyS3 doesn't exist), the modem doesn't dial. ttyS3 does exist (an isapnp card), the card's config is good (worked in the past), the permissions are correct (root dialout crw-rw), and the user is a member of dialout. It doesn't seem to matter if setserial knows about the card or not; restarting setserial gets me: /dev/ttyS0 at 0x03f8 (irq = 4) is a 16450 /dev/ttyS1 at 0x02f8 (irq = 3) is a 16550A ttyS3: LSR safety check engaged! ttyS3: LSR safety check engaged! ttyS3: LSR safety check engaged! /dev/ttyS3 at 0x02e8 (irq = 3) is a 16550A (ttyS0 and ttyS1 are not used) I'm running testing/unstable (current as of last thursday), and a 2.4.3 kernel. I can't find anything in the docs for setserial, pppd, or the kernel and aside from being able to check my mail (from an 8M 486-25, running little more than a Debian 2.0 base system), I'm pretty much offline :( - Bruce
(dup) serial port: LSR safety check engaged
Hi, Anybody know what an LSR safety check is, and how to disengage it? I've been getting this message... ttyS3: LSR safety check engaged! ...when I do pon or minicom (which also says /dev/ttyS3 doesn't exist), the modem doesn't dial. ttyS3 does exist (an isapnp card), the card's config is good (worked in the past), the permissions are correct (root dialout crw-rw), and the user is a member of dialout. It doesn't seem to matter if setserial knows about the card or not; restarting setserial gets me: /dev/ttyS0 at 0x03f8 (irq = 4) is a 16450 /dev/ttyS1 at 0x02f8 (irq = 3) is a 16550A ttyS3: LSR safety check engaged! ttyS3: LSR safety check engaged! ttyS3: LSR safety check engaged! /dev/ttyS3 at 0x02e8 (irq = 3) is a 16550A (ttyS0 and ttyS1 are not used) I'm running testing/unstable (current as of last thursday), and a 2.4.3 kernel. I can't find anything in the docs for setserial, pppd, or the kernel and aside from being able to check my mail (from an 8M 486-25, running little more than a Debian 2.0 base system), I'm pretty much offline :( - Bruce
Re: Prog. Languages (was: question?)
On Fri, 1 Jun 2001, Romain Lerallut wrote: Usually, you choose a language depending on what you want done. YMMV. If you want to learn a bit about computer languages *in general* you may want to: 1) start with interpreted languages, such as Perl. ( not Python which is strongly object-oriented). It's easy to create a proglet that is useful , and very satisfying. This would be a mistake. Perl code can be hard to read and the language itself is biased towards text processing (Practical Extracting and Reporting Language; Python is a general purpose language with clear syntax and semantics, just what a beginner needs. 2a) then learn about object-oriented languages (Java, Python). I would advise *against* starting with an OO language, since it *might* be harder to come back to non-OO languages. Python does not require you to use obviously OO techniques (classes, etc.) for everything, you can write in pretty much any style you like. *OR* 2b) then learn about compiled languages such as C. C is harder to program than interpreted languages, mostly because of memory management issues, but it is also IMHO very elegantly written. It has a lot of balance in its conception. You can understand more of the internals with C than with Java. *After* being decently fluent in C, you may want to learn C++ or others. Though I like C better for its simplicity and elegance. Sure, if you want to learn about computer internals also (hard to get away from that with a low level language like C)... I think it is the best way to start. Steep learning curve, though. - Bruce
Re: Qt2-gl - help!
Somethings not right here... On Tue, 29 May 2001, Michael O'Brien wrote: -- On Mon, May 28, 2001 at 02:42:07PM +0200, % wrote: The latest QT2 is compiled with GL support by default, thus obsoleting qt2-gl. I have the following installed: ii libqt22.3.0-final-5 Qt GUI Library (runtime version). I haven't upgraded yet today, but I have GL in the std libqt, and no _nogl version. ~$ dpkg -s libqt2 Package: libqt2 Status: install ok installed Priority: optional Section: libs Installed-Size: 4912 Maintainer: Ivan E. Moore II [EMAIL PROTECTED] Source: qt-x11 Version: 2:2.3.0-final-5 Replaces: libqt2.1, libqt2.2-gl, libqt2.2, libqt2-gl Provides: libqt2-gl Depends: libc6 (= 2.2.3-1), libgl1, libjpeg62, libmng1 (= 1.0.1-0), libpng2 (= 1.0.10), libstdc++2.10-glibc2.2, xlibs ( 4.0.3), zlib1g (= 1:1.1.3) Suggests: anti-aliasing-howto Conflicts: libqt2.2-gl, libqt2.2, libqt2-gl, libqt2.1, kde-designer ( 2.2.3-4)Description: Qt GUI Library (runtime version). This package contains the files necessary for running applications that use Qt. ~$ dpkg -L libqt2 /. /usr /usr/lib /usr/lib/libqt.so.2.3.0 /usr/share /usr/share/doc /usr/share/doc/libqt2 /usr/share/doc/libqt2/README.Debian /usr/share/doc/libqt2/copyright /usr/share/doc/libqt2/changelog.gz /usr/share/doc/libqt2/changelog.Debian.gz /usr/lib/libqt.so.2 /usr/lib/libqt.so.2.3 ~$ locate lib/libqt /usr/lib/libqt.a /usr/lib/libqt.so /usr/lib/libqt.so.2 /usr/lib/libqt.so.2.3 /usr/lib/libqt.so.2.3.0 /usr/lib/libqtmcop.la /usr/lib/libqtmcop.so /usr/lib/libqtmcop.so.0 /usr/lib/libqtmcop.so.0.0.0 ~$ ls -lA /usr/lib/libqt.* -rw-r--r--1 root root 7821450 May 25 01:28 /usr/lib/libqt.a lrwxrwxrwx1 root root 14 May 27 12:13 /usr/lib/libqt.so - libqt.so.2.3.0 lrwxrwxrwx1 root root 14 May 27 12:14 /usr/lib/libqt.so.2 - libqt.so.2.3.0 lrwxrwxrwx1 root root 14 May 27 12:14 /usr/lib/libqt.so.2.3 - libqt.so.2.3.0 -rw-r--r--1 root root 4943852 May 25 01:28 /usr/lib/libqt.so.2.3.0 The package contains libqt.so.2.3.0 and libqt.so.2.3.0_nogl (which are both ~5meg). libqt.so was pointing to libqt.so.2.3.0_nogl, so you will need to change the link to point to libqt.so.2.3.0 to get gl support. It would great if the installation allowed you to configure this package so that only one or both versions of libqt are installed. Also, it would be wonderful if the configure script let you choose which version of libqt libqt.so is pointing to. It looks like you have some cruft from an old libqt installation(s). Ivan may be interested in your libqt installation history, just in case it is a packaging bug that is likely to hit others. - Bruce
Re: Packaging WM themes - question
On Sun, 27 May 2001, Viktor Rosenfeld wrote: ... It might not be fast, but this is a 386 we're talking about. It simply isn't fast by todays standards [1]. But for some purposes it's good enough. today's could be, your's, the people you know, the region you live in, the economy you are part of... today's standard really only applies to the possible-world you live in. It's a modal logic thing. Now, how is that lessened by the fact, that Debian takes ages to install on such a machine? Not at all. It's obviously a load-intensive job, so you get a bigger machine. A bigger machine is not always an option, sometimes even for those living next door to a computer store. And making that process less dependent on CPU power is not an option when this means that core functionality of apt is sacrificed (ie the ability to figure out dependencies). Simply because I don't think that is true, or at least not something that is unsolvable with some new software. the vast majority of Debian users has no problem using it (at least speed-wise). No one forces you to use apt. If it's to slow for you, than don't use it, there are alternatives (eg Slackware, installing from sratch). Sure. The thing that gets me is that it is possible for software to accommodate anything (after all, it is software), yet there is steady pressure to drop support for older machines because it is bloat, etc. -- even though the systems it is bloat for would probably not even notice the extra resource usage, and compile time options could be used to tailor the build. ... [1] IIRC, the 386 I installed recently had roughly more than 1 BogoMIPS. The 486 I tried a few days later had 7.88 BogoMIPS. This was the stock A 1Mhz 386 and an 8MHz 486? potato kernel 2.2.17pre-something I believe. The machine I'm sitting in front of right now, is a Celeron 333 -- not exactly the fastest machine in the world, but the fastest I have in my home. It has 680 BogoMIPS. I know that BogoMIPS are ... well, bogus, but I think it proves my point. What do you expect from that kind of performance? How often do you hear people with oldslow machines gripping because a menu takes 0.5s to come up... the issue is not speed, it is one of not shutting people out just because they don't live up to today's standards in some possible world. - Bruce
Re: Kde Sid directory problem
On Sun, 27 May 2001, Matthew Gibbins wrote:
And yo was Bruce Sass heard to yodel:
On Sat, 26 May 2001, Matthew Gibbins wrote:
I'm running konqueror in Sid and am encountering problems loading1 some
modules
for konqueror configuration.
Particularly those under the directories:
/usr/share/applnk/Settings/WebBrowsing
/usr/share/applnk/Settings/FileBrowsing
No problems here. I can get at them via:
K - Control Center
K - Preferences - Web Browsing | File Browsing
and by pointing Konqueror at
/usr/share/applnk/Settings/{File,Web}Browsing
Just to clarify only ebrowsing.desktop, crypto.desktop and nsplugin.desktop
are loaded.
The rest fail with 'The diagnostics is:' where there is unfortunately no
diagnostic information.
Hmmm, I'd start looking closely at the system: version of all KDE
related packages, dpkg -C, cruft, deborphan, etc. It appears there
is something strange about _your_ setup (mismatched versions, old libs
shadowing newer versions, ...), and the list may not be of much help.
Good Luck.
- Bruce
Re: Kde Sid directory problem
On Sat, 26 May 2001, Matthew Gibbins wrote:
Hi,
I'm running konqueror in Sid and am encountering problems loading1 some
modules
for konqueror configuration.
Particularly those under the directories:
/usr/share/applnk/Settings/WebBrowsing
/usr/share/applnk/Settings/FileBrowsing
No problems here. I can get at them via:
K - Control Center
K - Preferences - Web Browsing | File Browsing
and by pointing Konqueror at
/usr/share/applnk/Settings/{File,Web}Browsing
- Bruce
Re: Multi-platform software development
On Thu, 24 May 2001, James Leigh wrote: from what I have seems is is easy to port from linux - win32 and very hard to port from win32 - linux. So the strategy for world domination should be: port apps to win32, when they get hooked... tell'em the apps run better on the OS they originally were designed on. ;) - Bruce
Re: for i in *
On Sun, 20 May 2001, Viktor Rosenfeld wrote: Martin Fluch wrote: Of course, some people argue, that spaces in filenames is a Bad Thing(tm), but I fail to see why. Could exactly this be the reason, why spaces in filenames are considered as a bad thing, since they easily lead into trouble? But, this trouble is easily avoided with double quotes and on the flip side, spaces make things much more readable. IMeanIt'SNotLikeWeDon'tUseSpacesInNormalWriting. And-I-have-yet-to-see-somebody-who-replaces-all-spaces-with-dashes-or-dots. See.what.I.mean? I/don't/think/you/would write/text like/this Is it one path or three? It may just be many years of not using spaces in filenames that has me seeing three paths, even though I've known for the same amount of time that filenames can contain spaces and the quotes would seem to indicate that it is one path... but I think it has more to do with our wetware naturally wanting to break things up into groups, and a blank space being a natural candidate to base a division on. It could be an efficiency issue. Also, along with bash's tab completion or GUI file managers, spaces don't cause me any trouble. Ya, computers don't care about efficiency or readability. ;) - Bruce
who gets this bug, pppconfig or ash?
Whenever I do pon, /etc/resolv.conf get rewitten with, -e \nnameserver 198.73.176.2 (i.e., I can only get to hosts listed in /etc/hosts) which I've traced back to, echo -e \nnameserver $DNS1 $TEMPRESOLV in /etc/ppp/ip-up.d/0dns-up. The problem goes away if I tell 0dns-up to use bash instead of sh, sh is a symlink to ash on this box. I originally thought it was related to a critical bug against ash (to do with the IFS), but that has been closed and the problem persists. :( - Bruce
Re: Multi-platform software development
On Wed, 30 May 2001 [EMAIL PROTECTED] wrote: Well, I think qt is the better choice. I didn't realize there was an windows version. I'm intending to develop mostly base and client-server software. So, Java won't fit. It didn't get mentioned, but there are Python bindings for Qt -- PyQt. - Bruce
Re: User-Created Menu Entries
On Mon, 7 May 2001, Mark Wagnon wrote: ... but I'm still unsure about how to add a menu entry. Has anyone created any that they wouldn't mind sharing to be used as a template for creating other entries? /usr/lib/menu has lots of files you can use as templates cp one to /etc/menu, tweak it, run (as root) update-menus With KDE you may need to force the issue with... rm -rf ~/.kde/share/applnk to force the DB to be rebuilt, or maybe just restart KDE. - Bruce
Re: dpkg - restoring working packages
On Tue, 8 May 2001, Tomasz Wzietek wrote: suppose i have a working package. i decide to upgrade it, so i download the deb and use 'dpkg -i' to install it. let's say, there are some unsatisfied dependencies that i don't want to resolve (let's assume that it would require some further drastic upgrading). how do i go back to the previously working older version of the package without reinstalling it? ... is there a switch or another deb-* command that would do the trick? or do i need to download the old version and reinstall it? If you have already installed the .deb, the old version is gone. dpkg-repack will create a .deb from an installed package. - Bruce
Re: Error message
On Fri, 4 May 2001, Eileen Orbell wrote: I keep getting this error message about 20 times a day emailed to me via my server from Cron Daemon: /bin/sh: rnews: command not found I am not even sure why? Or how to get rid of it? I do not want the news server anyway.. Snoop around /etc (probably in cron.daily) for the script cron is trying to run, then use dpkg -S script-path to find out which package it belongs to, followed by dpkg --purge package to get rid of it. Maybe run deborphan (if you are using testing/unstable) to see what got left behind. - Bruce
Re: I have instaledl too many packages
Hello,
hi,all
I am a debian newbie.I have installed too may packages.
I want to use 'dselect' to delete some,but show many errors
I want to use 'apt-get remove filename' but not know exactly the
filename,and there are so many package to remove.
How can I reinstall but keep the smallest base system so that I
not need to reboot and can use apt-get,deselect,netconfig etc.
dselect, apt-get, and the other stuff is great, but a little confusing
if you don't really know how the package management system works. A
new user needs to learn two things at the same time, the basic system
and the user friendly wrapper.
Instead of reinstalling, you may want to look in /usr/share/doc (every
package creates a directory with the same name as the package) and
start removing _applications_ you don't want.
Use:
dpkg --purge package-names
instead of dselect or apt, they call dpkg to do the removal anyways,
and you will get to see what kind of output dpkg produces - this is
important because it allows you to discriminate between actual errors
with packages, and cases where the user friendly tool is trying to
impose it's idea of the right thing to do (which I suspect is the
case with the many errors you mention in relation to dselect).
When you run into something you can't solve... use the script
command to make a transcript of what you are doing (start script, do
the dpkg command, do CTRL-D, mail the output to the list).
[warning: may require installing yet another package :)]
If you don't mind doing a little editing... you can have access to
dpkg and a reasonably user-friendly interface. Is git (a text based
filesystem browser) installed.
When it is, do the following (as root)...
# backup the config file, it maybe in /usr/share/git
cp /usr/lib/git/{.gitrc.common,.gitrc.common.orig}
edit .gitrc.common so it includes this line in the
[GIT-Keys] section:
^C^Dp = DPKG-purge; dpkg --purge %s{Purge package: ,%d}
What this does is define the key press sequence
CTRL-C CTRL-D p
to dpkg --purge package, where package is whatever directory the
cursor is at, and you get a chance to edit the package name (most
likely to change it into a list of package names).
Personally, I prefer:
^C^Dp = DPKG-purge; dpkg --purge %i
because it allows you to select multiple directories and purge the
associated packages with a single dpkg command, no option to edit what
is being purged though.
Here are a few others (to get you started ;)...
^C^Ds = DPKG-status; dpkg -s %d
^C^DS = DPKG-search; dpkg -S %f
^C^Dl = DPKG-list; dpkg -l | $GIT_PAGER
^C^DL = DPKG-listfiles; dpkg -L %s{List files in package: ,%d} | $GIT_PAGER
^C^Dc = DPKG-configure; dpkg --configure %s{Configure package(s): ,--pending}
^C^Dr = DPKG-remove; dpkg -r %s{Remove package: ,%d}
Putting:
export GIT_PAGER=less
eval `lesspipe`
in ~/.bash_profile makes the ^Xv GIT key sequence more useful than the
default settings.
I've found the combination of GIT + dpkg specific key sequence
definitions (essentially using /usr/share/doc as a menu) as the most
effective way to snoop around the system with an eye to getting rid of
installed packages.
Have fun.
- Bruce
