Re: kdm, desktop-base and theme across two heads
On 10 November 2007 at 11:58, Andrew Sackville-West wrote: | On Sat, Nov 10, 2007 at 11:50:35AM -0600, Dirk Eddelbuettel wrote: | > | > Hi all, | > | > Using the nice desktop-base package, I prettified my two-head desktop to | > greet me under kdm with the same theme as my laptop had. | > | > Unfortunately, on a dual-head machine (using nvidia and twinview, if that | > matters) the theme sticks only to one screen. I can't seem to find a hint to | > change that. Any ideas? | | please provide relevant data like: what hardware, contents of | xorg.conf. I think you misunderstand. Xorg.conf works as advertised -- I've had the dual heads for many months. But the 'theme for kdm' activated via desktop-base is only only screen 1 of 2 whereas screen 2 still has the default kde background. I would like the theme to extend to screen 1 and 2. Thanks, Dirk | | Also, what do you mean by: sticks to only one screen. Are you able to | use both screens? is it just the desktop wallpaper that won't go to | both screens? just the panels? we need more specific descriptions as | we can't see what is happening. | | A -- Three out of two people have difficulties with fractions. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
kdm, desktop-base and theme across two heads
Hi all, Using the nice desktop-base package, I prettified my two-head desktop to greet me under kdm with the same theme as my laptop had. Unfortunately, on a dual-head machine (using nvidia and twinview, if that matters) the theme sticks only to one screen. I can't seem to find a hint to change that. Any ideas? Please CC me on replied, and I look forward to suggestions, incl a maybe more pertinent place to post this question. Cheers, Dirk -- Three out of two people have difficulties with fractions. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: checkinstall lacking on stable
On 23 June 2007 at 09:31, Ross Boylan wrote: | On Fri, 2007-06-22 at 00:32 -0700, Francesco Pietra wrote: | > I am now at such cases for amd64 (dual core opterons) | > with OpenMPI (a parallelization support) and Amber (a | > molecular dynamics package), which I wish to compile | > with my installed intel fortran and c. | I can't help with your problem, but your mail raises a couple of | questions. I heard a couple years ago that Intel had made some changes | to their compilers that made them not work (or maybe not work well) with | chips from other vendors. Is that information inaccurate? | | Also, some of us are working on getting R and OpenMPI working together. | If anyone has done that, or would like to help, just speak up. Yes, I made a first stab and have something really rudimentary. But it has been moved 'back down the queue' as I (as part of the pkg-openmpi team on Alioth) am trying to get an all new, all shiny OpenMPI into Debian. Maybe this weekend. If and when it happens, I *will* post in debian-science (and possibly debian-devel) to reach other MPI maintainers and users. And yes, once we have a properly maintained and non-buggy OpenMPI package, it will be worthwhile coordinating with other the MPI implementation packages, and updating the packages using MPI (and particularly lam) such as Rmpi. Help would be welcome on any and all of these projects. For the R/MPI intersection, I am currently the only worker bee so this is bound to be slow. Dirk -- Hell, there are no rules here - we're trying to accomplish something. -- Thomas A. Edison -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [R] How to install R 2.2.0 Debian 'unstable' package in otherwise 'sarge' system
Dan, First off, there is even a r-sig-debian list in the universe of R mailing lists (c.f. the R FAQ). That probably provides a more focused readership than the union of r-help (where many won't know Debian) and debian-user (where many won't know R). On Wed, Oct 12, 2005 at 04:04:50PM -0500, Dan Davison wrote: > I would like to install the latest version of R (the statistical computing > software). This is package r-base version 2.2.0 and is in the debian > 'unstable' repository. Otherwise my system has 'sarge' packages, including > r-base 2.1.0. What is the best way to do this? Short answers: i) In general, and especially between 'testing' and 'unstable', use apt-pinning, explained in the apt-howto packages, esp apt-howto-en for English; and on various places across the Net; try Google'ing for apt-pinning. That way you get the option of installing selected parts of unstable without being forced to upgrade wholesale. Myself and countless other have used this for years between testing and unstable. However, I am not so sure how well it works between stable and unstable. It may work now as stable is fairly recent, but may fail further down the road. ii) In this particular case, and as explained in the R FAQ, the CRAN archives have an apt-get'able section for Debian stable. However, Christian, the (CC'ed) maintainer of this backport was traveling between conferences in the US and has not yet provided R 2.2.0 (which was released less than a week ago, after all). It should appear shortly. > to unstable, must I compile R myself, or is there some way to install this > as a debian package? You can also recompile locally using one of two ways: iii) as a local Debian package, and apt-get source makes that almost automatic (provided you have source URIs in /etc/apt/sources.list, and that the Build-Depends are actually satisfiable under Debian stable), or iv) as non-Debian compile into /usr/local If you're in a hurry, iv) is your way. If you're intrigued by iii), try it. If you can wait a few days, ii) is probably your best bet. > I have tried: > > (i) pointing /etc/apt/sources.list at unstable, apt-get updating and then > apt-get install r-base. This results in > > [... full output at bottom of email ...] > > E: This installation run will require temporarily removing the essential > package e2fsprogs due to a Conflicts/Pre-Depends loop. This is often bad, > but if you really want to do it, activate the APT::Force-LoopBreak option. > E: Internal Error, Could not early remove e2fsprogs > > which scares me into desisting with this course of action. I'd agree. Don't force things against warnings like this. > (ii) apt-get install -t unstable r-base, but it replies that > r-base is already the newest version. Have I not invoked this command > correctly, or does the -t switch not do what I was thinking it did? Did you run 'apt-get update' after altering /etc/apt/sources.list ? Try 'apt-cache policy r-base-core' which will tell you about the versions it knows, where they are from, and how they are prioritized (aka "pinned"). > Thanks very much for any help, Pleasure. Let me know in private mail if this is clear enough, and we could even follow up on a local phone call. Greetings from across town to Hyde Park, Dirk > > Dan > > > dd:/home/dan# apt-get install r-base > Reading Package Lists... Done > Building Dependency Tree... Done > The following extra packages will be installed: >e2fslibs e2fsprogs gcc-3.4-base libblkid1 libc6 libc6-dev libg2c0 > libgcc1 libss2 >libuuid1 locales r-base-core r-base-dev r-recommended > Suggested packages: >gpart parted e2fsck-static glibc-doc ess libpaper-utils > The following NEW packages will be installed: >e2fslibs gcc-3.4-base libblkid1 libss2 libuuid1 > The following packages will be upgraded: >e2fsprogs libc6 libc6-dev libg2c0 libgcc1 locales r-base r-base-core > r-base-dev >r-recommended > 10 upgraded, 5 newly installed, 0 to remove and 536 not upgraded. > Need to get 0B/21.4MB of archives. > After unpacking 5886kB of additional disk space will be used. > Do you want to continue? [Y/n] Y > E: This installation run will require temporarily removing the essential > package e2fsprogs due to a Conflicts/Pre-Depends loop. This is often bad, > but if you really want to do it, activate the APT::Force-LoopBreak option. > E: Internal Error, Could not early remove e2fsprogs > > __ > R-help@stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Statistics: The (futile) attempt to offer certainty about uncertainty. -- Roger Koenker, 'Dictionary of Received Ideas of Statistics' -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: XEmacs octave mode (font lock problem)
On Thu, Feb 05, 2004 at 12:49:15PM +0900, Victor Munoz wrote: > > > Hello. Lately I've had a problem with editing octave files with > XEmacs. I'm using Debian woody, and this may be an issue with the > distribution or with XEmacs. So I'm posting this to the octave help > and debian user lists, in case someone has a clue. > > When I open a file, XEmacs loads octave mode. However, when I try > to edit it (for instance, I add a space in a line such as 'if fig==8,', > XEmacs complains: > > (warning/warning) Error caught in font-lock-pre-idle-hook': (error No such > face font-lock-builtin-face) > > Now, searching in the Internet I found discussions about this issue, > but they're very old. In fact, I got this same warning when I used > Debian potato. Last year I switched to woody, and I was glad to see > it disappear, so I understood the patch was available in the woody > distribution. However, now I installed a new machine, installed the same > XEmacs/octave packages I had before, and the problem is here again! > > In case it helps, I use xemacs 21 (mule version, not-gnome version), and > the octave 2.0.16.92-7 package. Octave mode works better with emacs 21, > but I still prefer what I used to have a few months ago with xemacs. > > Any ideas? Thanks in advance, Well, 2.0.16.92-7 is from Debian stable and therefore almost two years old. Another bug report regarding XEmacs and the very latest octave2.1-emacsen package was filed recently [1], and addressed with a small patch by JWE. You could try the newer .el files from a current octave2.1-emacsen package -- as this isn't binary code, it can't interfere with your stable system. But you'd need to manually tweak the installation or simply overwrite the .el files and then rebuild the .elc (see what the postinst does). In the long run, as the 'code gap' between stable and the newer upstream widens, it would be nice if someone started to provides backports of current testing/unstable packages to stable -- www.backports.org would be a place to start. Hth, Dirk [1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=229336 -- The relationship between the computed price and reality is as yet unknown. -- From the pac(8) manual page -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Quantian 0.3 released
Version 0.3 of Quantian is now available. o Now based on clusterKnoppix: this adds support for OpenMosix, including terminalserver support for booting cluster nodes off the machine running quantian o Bowing to popular demand, we added the comedi control and measurement device interface, the grass geographical information system, the gri language for scientific graphics programming and the ipython python shell. We also re-introduced the lyx LaTeX document processor, added XEmacs which clusterKnoppix removes and fixes a few small bugs. The Quantian pages are at http://dirk.eddelbuettel.com/quantian.html and Quantian iso images are available at http://software.biostat.washington.edu/edd/quantian/ http://franz.stat.wisc.edu/~edd/quantian/ Regards, Dirk -- Don't drink and derive. Alcohol and analysis don't mix. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[ANN] Quantian: A Knoppix remastering for Scientific Computing
[ If there is sufficient interest, this may morph into a Debian subproject. Please contact me if you would want to be part of it. --edd ] Announcing the "Quantian Scientific Computing Environment" A Knoppix / Debian variant tailored to numerical and quantitative analysis Quantian is a remastering of Knoppix, the self-configuring bootable cdrom that turns any pc or laptop into a full-featured Linux workstation. This version is based on Knoppix 3.2 (2003-05-20), an earlier release based on Knoppix 3.1 (2003-01-20) is still available; see below for URLs. Quantian differs from Knoppix by adding a set of programs of interest to applied or theoretical workers in quantitative or data-driven fields. The added programs include o R, including several add-on packages (such as tseries, RODBC, coda, mcmcpack, gtkdevice, rgtk, rquantlib), out-of-the box support for the powerful ESS modes for Emacs as well as the Ggobi visualisation program o Octave with add-on packages octave-forge, octave-sp, octave-epstk, matwrap and Inline::Octave o Maxima, including the x11 front-end and emacs support o the Gnu Scientific Library (GSL) incl example binaries o the Pari/GP, Gap, Ginac and Yacas computer algebra systems o the Quantlib quantitative finance library incl. the Python interface o the OpenDX and Mayavi data visualization systems o TeXmacs for wysiwyg (La)TeX editing o and various other programs such apcalc, aribas, autoclass, euler, evolver, freefem, gambit, geg, geomview, glpk, gnuplot, gperiodic, gmt, gretl, lp-solve, mcl, multimix, rasmol, plotutils, pgapack, pspp, pdl, rcalc, yorick and xlispstat while at the same time retaining programs and features already in Knoppix o Auto-configuration of graphics, sound, disks, networking, auxiliary devices which is second to none among computer installations o The current version 3.1 of the KDE desktop environment o The GNU compiler suite comprising gcc, g77, g++ compilers in releases 2.95, 3.2 and 3.3 as well as gcj in version 3.2 o Perl and Python with loads of add-ons, plus ruby, tcl, ... o The Emacs and Vim editors, as well as kate, joe, kate, nedit and zile o A complete teTeX TeX and LaTeX setup for scientific publishing o Gnumeric, Abiword, Koffice, ... office tools o a Swiss-army knife collection of networking tools allowing access to wired and wireless lans, covering ethernet, isdn or dial-up modems In total, over 1200 Debian packages are included. Quantian iso images are available at http://software.biostat.washington.edu/edd/quantian/ http://franz.stat.wisc.edu/~edd/quantian/ and a Quantian overview page, including a recent paper, is at http://dirk.eddelbuettel.com/quantian.html Plans for future versions are not set in stone but may entail support for OpenMosix clustering. Comments, questions, feedback are more than welcome! -- Dirk Eddelbuettel <[EMAIL PROTECTED]> Tue, 27 May 2003 21:27:40 -0500 -- Don't drink and derive. Alcohol and analysis don't mix. -- Don't drink and derive. Alcohol and analysis don't mix. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: nfs-root booting works for 2.2.* client, but not 2.4.19
On Sat, Sep 21, 2002 at 10:42:53PM -0300, Jorge L. deLyra wrote: > > I now crash right here. The last good lines are > > VFS: Mounted root (nfs filesystem) > > Freeing unused kernel memory: 76k freed > > Undable to handle kernel paging request at vritual address 14d58d54 > > printing eip: > > [ gooblehoo ... ] > > > > Ans you're right -- that is exactly where 2.2.2* pass control to init. The > > next line is 'Activating swap'. It starts to smell like a memory issue. > > Yes, this is where diskless nodes with bad memory crash more often. Since > 2.2 and 2.4 probably use memory in different ways, it might be that you > are just lucky to get through with 2.2. I guess a standalone memtest run > is in order. You can run it from floppy and it is a sure-fire thing... No, the machine should be fine in and by itself. Two computers ago, it was my main workstation. I just changed it again to boot in 2.2.21 and copied /usr/sbin/memtest (from the sysutils package) over; memtest is currently running merrily and I'd be surprised if it showed something. I suspect that I am falling victim to either a resource starvation (that would be rather bad as 2.4.* ought to boot in 32mb) or, more likely, a misconfiguration. The odd thing is that by now, the larger of the two working 2.2.2* bzimages is 711k, whereas the non-functional 2.4.19 is 737k, or only 26k more. That can't be it. I must be missing one dreaded configuration. Darn. Thanks again, Dirk -- Good judgement comes from experience; experience comes from bad judgement. -- Fred Brooks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: nfs-root booting works for 2.2.* client, but not 2.4.19
On Sat, Sep 21, 2002 at 08:54:15PM -0300, Jorge L. deLyra wrote: > This is a bit confusing. This is neither a fail-to-mount-root panic nor a > cant-find-init panic, it's an Ooops, a processing error within the kernel. Right. And fiddling some more, I actually get a little further: > The partition check is before mounting the root and before init comes in. > Here is the sequence from my box, with 2.4.19: > > - > ... > Partition check: > hda: hda1 hda2 hda3 hda4 < hda5 hda6 > > hdb: hdb1 > hdc: hdc1 > hdd: hdd1 hdd2 > > NET4: Linux TCP/IP 1.0 for NET4.0 > IP Protocols: ICMP, UDP, TCP, IGMP > IP: routing cache hash table of 2048 buckets, 16Kbytes > TCP: Hash tables configured (established 16384 bind 16384) > Linux IP multicast router 0.06 plus PIM-SM > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > VFS: Mounted root (ext2 filesystem) readonly. > I now crash right here. The last good lines are VFS: Mounted root (nfs filesystem) Freeing unused kernel memory: 76k freed Undable to handle kernel paging request at vritual address 14d58d54 printing eip: [ gooblehoo ... ] Ans you're right -- that is exactly where 2.2.2* pass control to init. The next line is 'Activating swap'. It starts to smell like a memory issue. The 2.4.* kernel is only barely larger than the 2.2.* ones, 770k to 710k. > - > > Looks like the problem is with setting up the network. The message looks I think I got over that. > like a memory problem. Are you sure the kernel is guessing the correct > amount of RAM? In any case, here is a method we use here which will allow It wasn't, but specifying the (correct, and only) 32mb doesn't help. It could still be a memory issues as I don't have a swap partition or file on the thin client. But then 2.2.2(0|1) work... I also see the authenticated nfs request in the server's log. > you to cut-and-paste the kernel boot messages into an email message. It is > a bit complicated but usefull for a lot of things: use a serial console. > > 1) Enable the serial console on the kernel, boot with the parameter >console=ttyS0,9600 (or some other speed that works for you). > > 2) Have available some other machine with X11 running. Run seyon in an X11 >session, attached to some serial port. Configure seyon for the correct >speed, 1 bit, no parity, CR translations, etc. You can test this using >it as a terminal in some working machine where you put a getty in one >of the serial ports. You have to change /etc/inittab for this: > > # Example of how to put a getty on a serial line (for a terminal) > T0:23:respawn:/sbin/getty -L ttyS0 9600 vt100 > > 3) Build a simple 3-wire crossed serial cable and interconnect the two >ports. Boot the node, you will be able to see the whole kernel boot >procedure within you X11 seyon window. Just cut and paste... I used to have one of those, but I don't think I still do... Thanks for your help. Dirk > Note that your X11 session could be anywhere, not necessarily in the > console of the machine with the serial port connected to the node. You log > into the machine via the network and open the seyon window anywhere. Lilo > and Etherboot also can be configured to use the serial port. You can use > seyon as a terminal, reboot a node in the server root and look at the > whole boot process (except the node's hardware cycling, of course) from > the confort of your office. > Cheers, > > > Jorge L. deLyra, Associate Professor of Physics > The University of Sao Paulo, IFUSP-DFMA >For more information: finger [EMAIL PROTECTED] > > -- Good judgement comes from experience; experience comes from bad judgement. -- Fred Brooks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: nfs-root booting works for 2.2.* client, but not 2.4.19
On Sat, Sep 21, 2002 at 02:02:46PM -0400, Jonathan D. Proulx wrote: > > To point out the obvious... > > are you sure you flipped the kernel-autoconfig bit in networking and > > the allow nfsroot bit (somewhere else I forget, probably network > file systems) Yes, I checked that: CONFIG_IP_PNP=y this should be the kernel-autoconfig CONFIG_IP_PNP_ENABLE=y CONFIG_IP_PNP_DHCP=y CONFIG_IP_PNP_BOOTP=y CONFIG_IP_PNP_RARP=y [...] CONFIG_NFS_FS=y CONFIG_NFS_V3=y CONFIG_ROOT_NFS=y this is the root-NFS part CONFIG_NFSD=y CONFIG_NFSD_V3=y > have the right NIC driver builtin to the boot kernel Yep. It is, IIRC, an old ISA card. I musr assume that 2.4.* does support them as the options are still there. CONFIG_NET_ETHERNET=y CONFIG_RTL8139TOO=y this is I still need to install (100mbps) CONFIG_8139TOO_8129=y CONFIG_NET_ISA=y CONFIG_NE2000=y this one is installed, and works for 2.2.* CONFIG_NET_EISA=y CONFIG_EEXPRESS_PRO100=ythese two are used in other boxen here CONFIG_NE2K_PCI=y Those 8129/8139 variants shouldn't trample on each other, should they? > I presume you're getting a can't find init message or a panic about > the root device. Something like that. Retyping from the screen next to me: All well until including to hda: ... hdb: ... Uniform CD-ROM driver Rev Partition check: hda: and the trouble starts on that last line with <1>Unable to handle kernel paging request at virtual address 22c476e8 printing eip [ screenful of gooblygoo relating to the registers, stack and trace ] <0>Kernel panic: Attempted to kill init! > One issue I had with 2.4 kernels is that I sometimes made them too > big. I'm not sure too big for what as the one that works is bigger > than 640k and the ones that failed fit on the floppy with syslinux and > config stuff... Yes. The 2.4.* ones that fail are around 820 to 840kb, the 2.2.* ones that work are around 620 to 695kb, depending on what other (unused) stuff I left enabled. Dirk -- Good judgement comes from experience; experience comes from bad judgement. -- Fred Brooks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: nfs-root booting works for 2.2.* client, but not 2.4.19
On Sat, Sep 21, 2002 at 02:32:10PM -0300, Jorge L. deLyra wrote: > > I going nuts here. In June, I set up a scheme to have an old PC boot from a > > local image (floppy and/or hd) and run of a nfs-root partition on my > > server. Works very well with 2.2.20, and I replicated it with 2.2.21. > > > > But for the life of me, I cannot get it to work with 2.4.19. Neither the > > stock kernel (from the sources .dep) nor with the open-mosix patch. I tried > > numerous config option, and all I get, invariably, is a kernel panic relating > > to init right after the local hd* devices are found. At this point 2.2.* > > seems to find and use the eth0 device and the remote partitions. > > > > Any suggestions welcome. Please CC me as I'm not subscribed on -user or > > -beowulf. > > How about posting in detail the kernel messages just before the crash? Sure, I was just being lazy as I would have to type those manually, and it is a screen full of them. Also, the server end does work (as 2.2.20 and 2.2.21 run just fine off it) and I am fairly certain that my problem is with the 2.4.19 configuration. > What are the contents of that nfs-root (which distribuition)? Since you Debian testing, installing using debootstrap, and recently updated using a simple chroot call to start a session on the server, then the usual apt-get gymnastics. The nfs-root partition is down to 70-ish packages which is nice. All is does is to the query a remote xdm (well, kdm) session on the server. So the nfs-root partition acts as the hd for the 'thin client' pc which uses it to bootstrap itself to be a remote x11 terminal. That all works, and I would like it to be an openmosix client too. Hence the need for 2.4.19. > get to the init message, I presume it is mounted OK by the kernel? What is > the set of boot line parameters you use for that floopy kernel? Same for both 2.2.20/2.2.21 (which works) and 2.4.19 (which fails). I currently use loadlin, and it passes the set of network boot parameters. > Are you using DHCP to configure the network? Can you diff the kernel No, static assignment as I have only one thin client so far... > configuration files for 2.2.20 and 2.4.19 to check the differences? One I avoided that so far as many other things have changed between the kernels. But I guess I need to go there, given that I have the working setup... > funny thing I remember is that in the 2.4 kernels you must turn on the > dhcp option in the kernel-level autoconfiguration section even you are not > actually going to use it... Yes, I activate the lot: CONFIG_IP_PNP=y CONFIG_IP_PNP_ENABLE=y CONFIG_IP_PNP_DHCP=y CONFIG_IP_PNP_BOOTP=y CONFIG_IP_PNP_RARP=y Thanks, Dirk -- Good judgement comes from experience; experience comes from bad judgement. -- Fred Brooks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
nfs-root booting works for 2.2.* client, but not 2.4.19
I going nuts here. In June, I set up a scheme to have an old PC boot from a local image (floppy and/or hd) and run of a nfs-root partition on my server. Works very well with 2.2.20, and I replicated it with 2.2.21. But for the life of me, I cannot get it to work with 2.4.19. Neither the stock kernel (from the sources .dep) nor with the open-mosix patch. I tried numerous config option, and all I get, invariably, is a kernel panic relating to init right after the local hd* devices are found. At this point 2.2.* seems to find and use the eth0 device and the remote partitions. Any suggestions welcome. Please CC me as I'm not subscribed on -user or -beowulf. Thanks, Dirk -- Good judgement comes from experience; experience comes from bad judgement. -- Fred Brooks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Announcement: Automatic ATLAS support under Debian GNU/Linux
ecompilation of the Atlas libraries. This way the behaviour will be tuned exactly to the specific CPU rather than the broader class of CPUs. It has been reported that this can increase performance by a further 12% on the examples above. Detailed instructions are in /usr/share/doc/atlas2-base/README.debian.gz but the process is essentially the following [ courtesy of Doug Bates ] apt-get source atlas2-base cd atlas2-$VERSION fakeroot debian/rules/custom # wait for a *very* long time dpkg -i ../atlas2-base*.deb IV. See also The Atlas packages have a very detailed README.Debian file which should be consulted; it also details local recompilation. Sources and documentation for Atlas are at http://www.netlib.org/atlas. V. Acknowledgements Camm Maguire developed the scheme of overloading Atlas over the default blas libraries and deserves all the credit. Many thanks to John Eaton for helping debug some errors in the initial setup, and to Doug Bates for work on the R package. Special thanks to Ben Collins for providing a patched ldconfig as part of the libc6 package. Initial version -- Dirk Eddelbuettel <[EMAIL PROTECTED]> Tue, 21 Aug 2001 21:37:15 -0500 First updated -- Dirk Eddelbuettel <[EMAIL PROTECTED]> Sun, 11 Nov 2001 11:03:19 -0600 -- Better to have an approximate answer to the right question than a precise answer to the wrong question. -- John Tukey
BBDB problem with VM and Xemacs 21
(Please CC me on replies as I am currently unsubscribed.)
The more recent bbdb versions in testing seemed to have broken my setup. Did
anybody else experience a similar problem, and better still, fix it? I am
using bbdb_2.2-1 along with xemacs21-{bin,nomule,support}_21.1.10-5 as
currently in 'testing'.
Thanks, Dirk
--
According to the latest figures, 43% of all statistics are totally worthless.
Problems with Matlab 6.0 (R12) on potato
[ Please CC me on replies.] At work, and on a computer which runs Debian 2.2, I upgraded Matlab today. The new Matlan version just some Java issues, which are probably libc6 interaction issues -- in fact, Matlab dies with a seg fault and leaves a java.log.PID logfile. The Java runtime environment comes shipped with Matlab and is (IIRC) 1.1.8_v3 of the blackdown.org port. The only help which the Mathworks could suggest was to start matlab with the '-nojvm' switch, thus forgoing the tools provided in their Java envinronment. Would anybody here have experienced, and possibly solved :-), the same issue? Best regards, Dirk -- According to the latest figures, 43% of all statistics are totally worthless.
Pinging Andreas Franzen && ITA task-science
It appears that Andreas Franzen is unreachable via his Maintainer: address. I would like to get in contact with him, or failing that, will adopt his task-science package which needs reworking. It also has a bug report with > 250 days on it to which he never replied. db.debian.org has no trace of him past May 2000. If anybody knows where he is, please drop me a line. Dirk -- Three out of two people have difficulties with fractions. --- Begin Message --- Hi, The BTS uses the same address -- you'll have to ask on -devel or somewhere else for Andreas Franzen's real e-mail address... - Forwarded message from Mail Delivery System <[EMAIL PROTECTED]> - Delivery-date: Wed, 25 Oct 2000 06:22:46 +0200 X-Failed-Recipients: [EMAIL PROTECTED] From: Mail Delivery System <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Subject: Mail delivery failed: returning message to sender Date: Tue, 24 Oct 2000 23:18:04 -0500 This message was created automatically by mail delivery software. A message that you sent could not be delivered to one or more of its recipients. The following address(es) failed: [EMAIL PROTECTED]: unknown local-part "andreas.franzen" in domain "debian.org" -- This is a copy of the message, including all the headers. -- Return-path: <[EMAIL PROTECTED]> Received: from gecko by master.debian.org with local (Exim 3.12 1 (Debian)) id 13oI0q-aS-00; Tue, 24 Oct 2000 23:18:04 -0500 Subject: Bug#75519: task-science: Wrong maintainer field Reply-To: <[EMAIL PROTECTED]>, [EMAIL PROTECTED] Resent-From: <[EMAIL PROTECTED]> Orignal-Sender: Dirk Eddelbuettel <[EMAIL PROTECTED]> Resent-To: debian-bugs-dist@lists.debian.org Resent-CC: Andreas Franzen <[EMAIL PROTECTED]> Resent-Date: Wed, 25 Oct 2000 04:18:02 GMT Resent-Message-ID: <[EMAIL PROTECTED]> Resent-Sender: [EMAIL PROTECTED] X-Debian-PR-Message: report 75519 X-Debian-PR-Package: task-science X-Debian-PR-Keywords: X-Loop: [EMAIL PROTECTED] Received: via spool by [EMAIL PROTECTED] id=B.9724470851144 (code B ref -1); Wed, 25 Oct 2000 04:18:02 GMT From: <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] X-Mailer: bug 3.2.10 Message-Id: <[EMAIL PROTECTED]> Sender: Dirk Eddelbuettel <[EMAIL PROTECTED]> Date: Tue, 24 Oct 2000 23:11:25 -0500 Delivered-To: [EMAIL PROTECTED] Package: task-science Version: N/A Severity: normal Mail to '[EMAIL PROTECTED]' bounces. A new maintainer: field in debian/control might help. -- System Information Debian Release: 2.2 Kernel Version: Linux master 2.2.17pre15 #1 Wed Aug 2 23:29:13 CDT 2000 i686 unknown - End forwarded message - -- Digital Electronic Being Intended for Assassination and Nullification --- End Message ---
Re: [Q] Can Samba mount 'shared' (not 'served') Win drives ?
On Fri, Aug 04, 2000 at 09:58:43AM -0500, Kent West wrote: > Dirk Eddelbuettel wrote: > > Yes, as I wrote, that works fine for NT servers providing a share. I use > > that with the corresponding entry in /etc/fstab so that I can simply say > > mount /mountpoint > > and the rest happens automatically. > > > > I now would like to access the C:\ I declared as 'shared' on my desktop. I > > can't figure out what that would be. Whatever I try yields 'session request > > to DESKTOP failed'. > > > > Any idea? > > You said you "would like to access the C:\ I declared as 'shared' > on my desktop." By "desktop", do you mean your workstation Yes. > computer, or do you mean your Windows Desktop (shell program)? If No, I mean C:\ as the main partition on the 'desktop' computer. > the latter, you can't mean "C:\", yet the error message that you > mention indicates the latter. > > So, what directory have you shared, "C:\" or > "C:\WINNT\PROFILES\YOURUSERNAME\DESKTOP", and with what > permissions? Permissions are read-access for everyone. Can I read those from Linux via Samba? Thanks, Dirk -- According to the latest figures, 43% of all statistics are totally worthless.
Re: [Q] Can Samba mount 'shared' (not 'served') Win drives ?
On Fri, Aug 04, 2000 at 08:44:01AM -0500, Phil Brutsche wrote: > A long time ago, in a galaxy far, far way, someone said... > > At work, in a predominantly NT environment, I use Samba to mount drives of > > the NT servers on the Lan. However, I'd also love to access files on my > > (vanilla NT 4.0) desktop at work which is set to let other 'share' its > > files. I tried mounting these from a Linux box but failed. Is there a way > > to get to these files so that I could access the files from Linux? > > mount -t smbfs //server/share /mountpoint -o > username=,password=,uid= > > That should all be one line, of course. Yes, as I wrote, that works fine for NT servers providing a share. I use that with the corresponding entry in /etc/fstab so that I can simply say mount /mountpoint and the rest happens automatically. I now would like to access the C:\ I declared as 'shared' on my desktop. I can't figure out what that would be. Whatever I try yields 'session request to DESKTOP failed'. Any idea? -- According to the latest figures, 43% of all statistics are totally worthless.
[Q] Can Samba mount 'shared' (not 'served') Win drives ?
This might be a trivial questions with a quick No! as the answer ... At work, in a predominantly NT environment, I use Samba to mount drives of the NT servers on the Lan. However, I'd also love to access files on my (vanilla NT 4.0) desktop at work which is set to let other 'share' its files. I tried mounting these from a Linux box but failed. Is there a way to get to these files so that I could access the files from Linux? CC's welcome as I am not currently subscribed here... Thanks, Dirk -- According to the latest figures, 43% of all statistics are totally worthless.
Re: [q] koffice-cvs doesn't run -- any idea?
"Ivan" == Ivan E Moore, <[EMAIL PROTECTED]> writes: Ivan> it's because of some changes in kdelibs/kdebase and I hadn't gotten Ivan> around to rebuilding koffice so that it would work again. :) Ivan> Ivan> This evening there will be a clean set of .deb's for all packages Ivan> based on current (as of a few hours ago) upstream source. Thanks, that sounds great! Guess I can stop scratching my head now, and send some virtual beers over to you. Dirk Ivan> Ivan> Ivan Ivan> Ivan> On Sat, May 20, 2000 at 04:24:04PM -0400, Dirk Eddelbuettel wrote: Ivan> Content-Description: forwarded message >> Date: Sat, 20 May 2000 11:10:21 -0400 (EDT) From: Dirk Eddelbuettel >> <[EMAIL PROTECTED]> Subject: koffice-cvs unresolved libs To: Debian Users >> >> >> >> (Please CC me on replies as I am not debian-user.) >> >> Just had a first look at the koffice work-in-progress using Ivan's fine >> kde archive, but there is something wrong with the dependencies: >> >> [EMAIL PROTECTED]:~> kword kword: error in loading shared libraries: >> /usr/lib/libkofficecore.so.1: undefined symbol: >> setDocument__13KXMLGUIClientRC12QDomDocumentb >> >> even though 'ldd $(which kword)' shows all libraries. This is on an >> up-to-date frozen machine. What am I missing? >> >> Thanks for any insight, Dirk >> >> -- According to the latest figures, 43% of all statistics are totally >> worthless. Ivan> Content-Description: .signature >> -- According to the latest figures, 43% of all statistics are totally >> worthless. Ivan> ---end quoted text--- Ivan> Ivan> -- Ivan E. Moore II [EMAIL PROTECTED] Ivan> http://snowcrash.tdyc.com GPG KeyID=90BCE0DD GPG Fingerprint=F2FC Ivan> 69FD 0DA0 4FB8 225E 27B6 7645 8141 90BC E0DD Ivan> -- According to the latest figures, 43% of all statistics are totally worthless.
koffice-cvs unresolved libs
(Please CC me on replies as I am not debian-user.) Just had a first look at the koffice work-in-progress using Ivan's fine kde archive, but there is something wrong with the dependencies: [EMAIL PROTECTED]:~> kword kword: error in loading shared libraries: /usr/lib/libkofficecore.so.1: undefined symbol: setDocument__13KXMLGUIClientRC12QDomDocumentb even though 'ldd $(which kword)' shows all libraries. This is on an up-to-date frozen machine. What am I missing? Thanks for any insight, Dirk -- According to the latest figures, 43% of all statistics are totally worthless.
New Octave .deb packages for installing both 2.0.* and 2.1.*
I prepared the text below to announce this to the Debian dupload program, but I think it was taken. CC'ed to the Octave list as well. Feedback by Debian and Octave users on the new packages would be appreciated. Announcing Octave 2.0 and 2.1 packages for Debian This is to announce the upload of new Octave packages for Debian, taken from both the stable octave-2.0.* as well as the development octave-2.1.* branch, to Debian's master archive. Using Debian's update-alternatives mechanism, as well as minor reorganisation of the Octave file tree which John Eaton kindly provided in the upstream sources, permits to install *both* at the same time providing two main binaries octave2.0 and octave2.1 as well as the alternative octave which defaults to the stable 2.0 branch. The adjunct packages for Emacs support, info docs, html docs and ps docs are also versioned. With this mechanism, Debian allows for the stability of the 2.0.* release (and thereby provides continued support for "legacy" applications in Octave which could break by subtle, but important, changes in Octave 2.1). At same time, it finally introduces the newer Octave 2.1.* which is the centre of current development into Debian. The transition might introduce a few bugs, but the package layout and build process has undergone some testing over the last few weeks. As these are "new" packages as far as the Debian archive is concerned, it might be a few days until they show up in "unstable" aka "woody", Debian's current development distribution. -- According to the latest figures, 43% of all statistics are totally worthless.
[Off-topic] Scanner advice sought
I'd appreciate advice on selecting a simple Linux-compatible scanner. Nothing fancy, just to scan the odd picture of our daughter etc. Lowish resolution is fine. I wouldn't want to spend more than 100 USD. I have no SCSI adapter (yet ?). I've studied the SANE webpage as well as the Hardware HOWTO, but have difficulties locating any of the supported models in stores. Also, real-life feedback might be quite illuminating. So any helpful information would be greatly appreciated. Please email me privately as I am not currently subscribed to debian-user. I will summarise back to the list. Thanks, Dirk -- According to the latest figures, 43% of all statistics are totally worthless.
Re: Cannot print from StarOffice
Julio> You're lucky - my current potato install doesn't even install Julio> StarOffice.. Feel free to send me your `dpkg -l' so that we can diff ours. It was really no problem here (potato apt-get'ed up tp speed two days ago, SO51 re-downloaded yesterday). -- According to the latest figures, 43% of all statistics are totally worthless.
Re: Cannot print from StarOffice
Dirk> I am at wit's end here -- I have a `network' installation of Dirk> StarOffice 5.1 here at home for use by myself and my wife. It works Dirk> fine, but just won't print. I even re-downloaded 5.1 again, and Dirk> reinstalled, but no luck. The machine is current potato with lpd and Dirk> magicfilter installed and working. Julio> What distribution are you using, slink or potato? "Current" Potato (see end of penultimate line above). Julio> It seems that some StarOffice functionalities hang when running over Julio> potato with libc6 2.1. It does work fine with slink (libc6 2.0.7) Julio> indeed. Yes, that's like it. Whenever it attempts to print, it hangs and starts several soffice.bin processes which I have to kill hard (kill -9, sometime even xkill). So then it's now me but rather glibc and staroffice? -- According to the latest figures, 43% of all statistics are totally worthless.
Cannot print from StarOffice
(Please CC: me on replied. Thanks.) I am at wit's end here -- I have a `network' installation of StarOffice 5.1 here at home for use by myself and my wife. It works fine, but just won't print. I even re-downloaded 5.1 again, and reinstalled, but no luck. The machine is current potato with lpd and magicfilter installed and working. Does anybody out here have an idea? Thanks, Dirk -- According to the latest figures, 43% of all statistics are totally worthless.
Re: ftp1.us.debian.org and FTP mirrors
Lukas> Once someone posted this script to find the fastest debian mirror,
Lukas> which could be used in some way in the postinst of apt:
IIRC Chris Lameter wrote this as part of one of his misc. utility packages.
I rewrote it in perl (and added a thing or two) as part of the mirror package
which installs it under the name 'debian-mirrors'. As it's short and sweet, I
include it here:
#!/usr/bin/perl
#
# debian-mirrorsmeasure ping time to all mirrors in README.mirrors
#
# downloads README.mirrors from ftp.debian.org (but not if a local file is
# pointed to), runs fping on all Debian mirrors, sorts the result by ping
# time and output the 'n' (default is 20) fastest
#
# Written by Dirk Eddelbuettel <[EMAIL PROTECTED]> and released under the GPL
# $Id: debian-mirrors.pl,v 1.3 1998/02/19 02:12:02 edd Exp $
## Yes, Net::FTP is much cooler, but we don't want to depend on libnet-perl
## 'ftp.pl' is provided by the mirror package
unshift( @INC, "/usr/lib/mirror" );
require 'ftp.pl';
#use strict;# doesn't work with ftp.pl
use English;
use File::Basename;
use vars qw($opt_h $opt_f $opt_v);
use Getopt::Std;
use IPC::Open2;
my $tempfile = "/tmp/".$file."-".$$;
my $readfrom = $tempfile;
my $max = 20;
$PROGRAM_NAME =~ s|.*/||; # strip everything before last slash
getopts('hf:n:v') or die("Try `$PROGRAM_NAME -h` for help screen.\n");
if ($opt_h or $#ARGV != -1) {
print "Usage:\n $PROGRAM_NAME [options]\n";
print "Options:\n";
print " -f file\tpoint to local version README.mirrors\n";
print " -n max\tmaximum number of mirrors to show\n";
print " -v\t\tverbose operation\n";
print " -h\t\tshow this help\n";
exit 0;
}
$max = $opt_n if ($opt_n);
print "Max is set to $max\n" if $opt_v;
if ( ! $opt_f ) {
my $hostname = "ftp.debian.org";
my $account = "ftp";
my $password = "[EMAIL PROTECTED]";
my $directory = "/pub/debian";
my $filename = "README.mirrors";
print "Retrieving ", $filename, " for analysis.";
ftp::debug(1) if $opt_v;# for debugging output
ftp::open($hostname,21,0,1) or die "No ftp connection\n";
ftp::login($account, $password) or die "Couldn't login\n";
ftp::cwd($directory) or die "Cannot cd to $directory\n";
ftp::get($filename, $tempfile, 0) or die "Could not get $file\n";
ftp::quit();
} else {
die "File $opt_f does not exist." unless -f $opt_f;
$readfrom = $opt_f;
}
print "Measuring ping times to all Debian mirror sites. " if $opt_v;
print "Please be patient.\n" if $opt_v;
my ($mirror,$path,$time);
open(DATA, $readfrom) or die "Cannot open $filename\n";
open2("PINGOUT", "PINGIN", "fping -ae") or die "Cannot start fping(1)\n";;
while () {
if (($mirror,$path) = ($ARG =~ m|^\s*(\S*\.\S*\.\S*.*):(/\S*)\s.*|)) {
print "Pinging $mirror \n" if $opt_v;
print PINGIN "$mirror\n"; # if ($mirror =~ m/.*\.edu$/);
}
}
close(DATA);
close(PINGIN);
open2("SORTOUT", "SORTIN", "sort -n") or die "Cannot start sort(1)\n";
while () {
($mirror,$time) = ($ARG =~ m|^(\S*)\s*\((\d*) msec\)|);
print "fping reports $time for $mirror\n" if $opt_v;
print SORTIN "$time $mirror\n";
}
close(SORTIN);
close(PINGOUT);
my $i=1;
while () {
print "$_";
last if ($i++ == $max);
}
close(SORTOUT);
--
Linux is not only free; it is, arguably, a better operating system, offering
a degree of stability and an ability to scale up that NT cannot match.
-- The Economist, Oct 3, 1998
Re: Using VM with Emacs19 and XEmacs20
Dirk> (Please CC: me as I'm no longer on this list.) Dirk> Dirk> I'm scratching my head: Have been using VM with Emacs19 for years, Dirk> and as XEmacs20 supports VM as well (with its own copy), I thought I Dirk> could migrate. Doesn't work (yet). Matthew> It might well be that having emacs19 and xemacs20 is causing a Matthew> problem. Try upgrading emacs to 20 Matthew> That would imply uninstalling vm, as vm does not work with FSF Emacs 20. So I'd loose my working setup. Hmpf. -- Linux is not only free; it is, arguably, a better operating system, offering a degree of stability and an ability to scale up that NT cannot match. -- The Economist, Oct 3, 1998
Using VM with Emacs19 and XEmacs20
(Please CC: me as I'm no longer on this list.) I'm scratching my head: Have been using VM with Emacs19 for years, and as XEmacs20 supports VM as well (with its own copy), I thought I could migrate. Doesn't work (yet). Anybody have any idea? I use ii emacs19 19.34-21 The GNU Emacs editor. ii xemacs20-bin20.4-7 Editor and kitchen sink -- support binaries ii xemacs20-nomule 20.4-7 Editor and kitchen sink -- Non-mule binary ii vm 6.62-2 A mail user agent for Emacs Thanks, Dirk -- Linux is not only free; it is, arguably, a better operating system, offering a degree of stability and an ability to scale up that NT cannot match. -- The Economist, Oct 3, 1998
[?] 2.0.35 doesn't boot
2.0.35 hangs on my plain vanilla Pentium (approx 3 yrs old, Asus board, AMI bios). Right after Partition check: hda: hda1 hda2 hda3 hda4 hdc: hdc1 hdc2 hdc3 hdc4 < hdc5 hdc6 > hdd: [PTBL] [621/64/63] hdd1 hdd2 hdd3 < hdd5 hdd6 > VFS: Mounted root (ext2 filesystem) readonly. nothing happens. 2.0.34 starts fine, as did all other kernels I tried. Do I now have to turn a new chipset switch on ? Please CC: me on replies as I'm not on this list. Thanks, Dirk -- mailto:[EMAIL PROTECTED] According to the latest official figures, http://rosebud.ml.org/~edd 43% of all statistics are totally worthless. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
Re: Bug#21412: tob deletes system files
[ You edited citations and appended/deleted material without showing that you did so. That is considered to be very Bad Taste (TM) as it implies that the person who you cited (ie, me) said things in a different context or order. That is bad style. And I am not amused. --edd ] Tim> With tob_0.14-4 no longer available for download, I am happy. If I Tim> spend my time tracking down and reporting a bug that has the potential Tim> to wipe out someones system, ... then you should be happy as you just participated in what lead to Infoworld giving the award for 'Best Support in 1997' to all of us -- the Linux communitty -- by providing the feedback and support needed to improve Linux (Debian in this case). Tim> and the response I get for doing so if there is a next time is to Tim> "chill out", I will not be happy. Slowly, please. I was not trying to paternalise you, I was merely trying to point out that your classification of 'severe, release critical' was a little out of place given that - the bug was already fixed - the bug only occurred because you (unknowingly) pushed the enveloppe by putting an 'unstable' package on 'stable'. Further, and as you had said [ same email, few lines above ] Tim> Not being a Debian developer, I have no way of knowing that uploading Tim> tob_0.14-5 implies tob_0.14-4 is "gone, vanished, nothing left". Again, I recommend that you simply take things a little easier and try to learn how Debian works. It is obvious to everybody but you that uploading a version 1.2-4 of a package replaces version 1.2-3 of the same package. Regards, Dirk -- mailto:[EMAIL PROTECTED] According to the latest official figures, http://rosebud.ml.org/~edd 43% of all statistics are totally worthless. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#21412: tob deletes system files
Tim,
Tim> Thanks for looking into this.
Well, that's what we're here for ;-)
Tim> You are correct that my system is Debian 1.3. I went to
Tim> "http://www.debian.org/packages.html";, did a search for tob, found
Tim> tob_0.14-4 at
Tim>
"http://cgi.debian.org/www-master/debian.org/Packages/unstable/utils/tob.html";,
Tim> downloaded and installed it.
You searched under 'unstable'. You should have searched under 'stable' (ie
selected the radio button for it) as 'stable' corresponds to Debian 1.3.
That would have lead you to tob_0.14-2 which is part of Debian 1.3 CD.
Tim> No mention is made on the web page or from "dpkg --info
Tim> tob_0.14-4.deb" of any package or Debian version (hamm or bo)
Tim> dependencies, and it installs without error or warning messages.
Well, for binary packages, the dependency on libc6 ensures that you don't mix
libc5 (Debian 1.3) and libc6 (Debian 2.0).
As Joey Hess just pointed out, binary-all packages as tob need an explicit
dependency. So I just made tob_0.14-6 which I am uploading as we speak.
Tim> So my question(s) would be:
Tim>
Tim> 1. How is one to know that a hamm system is required for this package?
It's implicit as you took it from the 'unstable' tree.
Tim> 2. I notice there is currently no mention of bug#21412 in the
Tim> bug-tracking system. Will this bug be considered closed?
Yes, I closed it via a mail to [EMAIL PROTECTED] You can reopen
it [ who shouldn't, see below ], see the documentation in the debian/doc
directory on the mirror sites, or in /usr/doc/debian/ if you have the
'doc-debian' package installed.
Tim> My view would be that it should be considered Severity:Critical, the
Tim> bug report should stay open indefinetly and/or tob_0.14-4 should be
Tim> completely removed from all Debian sites. After all, the next "bo"
Tim> user that installs this risks taking out their entire system.
Please chill out a little and read the emails you got yesterday. I released
tob_0.14-5 yesterday which added the following test
cleanup ()
{
message 'Cleaning up.'
# add a safety check here --edd 20 Apr 98, regarding #21412
-> if [ "$TMPLIST" != "" -a "$FILELIST" != "" ] ; then
$RM -f $TMPLIST* $FILELIST*
fi
postcommand
}
which already prevents the behaviour you experienced. That version was
installed last night into the archive. This means that 0.14-4 is gone.
Vanished. Nothing left. [ It's still in the mirrors, but no longer on
master.debian.org and will be replaced in the mirrors. ] Are you happy ?
Further, and as I just said, I adopted a better fix upon a suggestion from a
fellow developer and made tob depend explicitly on debianutils. Ie you cannot
run it without having debianutils installed.
I am sorry for the grieve that the package caused you. In retrospect, I
should have added the dependency on debianutils in tob_0.14-4 when tempfile
was added for safer creation of temporary files. However, we use the
'unstable' release to iron such bugs out.
You helped us in finding the bug, and hence helped other users from being
bitten by it. Thanks !
Regards, Dirk
--
mailto:[EMAIL PROTECTED] According to the latest official figures,
http://rosebud.ml.org/~edd 43% of all statistics are totally worthless.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Problem with a new Mach64 3D Xpression 2MB
I am trying to set up a new machine for someone else. The machine came with an ATI 3D Xpression 2MB card. Using Debian 1.3.1, I can't get any X11 server other than vga16 to work. The Mach64 server seems to splits the screen vertically a couple of times. X -probeonly says unknown ATI (82304x) rev 154, Aperture @ 0xe00, Block I/0 @ 0x6100 Memory type: 3 Clock type: Internal Ramdac is Internal I would be grateful if anyone could help me out. Please email me directly as I am not following this list very regularly. Thanks, Dirk -- [EMAIL PROTECTED] According to the latest official figures, http://rosebud.ml.org/~edd43% of all statistics are totally worthless. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Resent-Sender header missing from mailing list
Jim> Could you add the Resent-Sender header back onto the mailing lists Jim> again? It was much easier to filter e-mail with them rather than Jim> filtering by the To: and Cc: fields. If you use procmail, consider the ^TO meta-expression which is *much* more general. Works for me all debian- sublists; see this concrete example for debian-devel: * [EMAIL PROTECTED](debian.org|pixar.com) FYI here's a piece from procmailrc(5): MISCELLANEOUS If the regular expression contains `^TO' it will be sub stituted by `(^((Original-)?(Resent-)?(To|Cc|Bcc)|(X- Envelope|Apparently(-Resent)?)-To):(.*[^a-zA-Z])?)', which should catch all destination specifications. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: /dev/ttyS1 root only.....
Ed> I'm trying to configure my modem and, as recommended in the Ed> Serial-HOWTO, I've been using kermit to test it out. Problem is that Ed> /dev/ttyS1, which is the serial port my modem is on, seems to only Ed> allow root access. Add yourself to the "dialout" group in /etc/group. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Debian 1.1 (Kernel v2.0) Questions
James> I was curious if it's possible to yank the Linux Kernel v2.0.20 off James> sunsite.unc.edu and install it on a Debian v1.1 (Kernel v2.0) James> machine. Yes, that should work. Just untar it somewhere, have /usr/src/linux point to that directory, install the Debian's kernel-package, read it's doc and say cd /usr/src/linux make xconfig make-kpkg --revision local.1 kernel_image and you should see shiny new kernel-image_2.0.20-2.0.20-local.1.deb in .. Or you can patch up from kernel-source-2.0.6 (I'm currently at patch-2.0.17). -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: dvips top margin
Billy> (other than the inflexible configuration, some features like Billy> previewing documents with ps fonts with xdvi are broken in Debian). Not so. I am both LaTeX'ing and xdvi'ing with psfonts on Debian boxes with no problems. You just have to add gsftopk, and modify MakeTeXPK (I send a patch in a mail archived with bug#3414) and psfonts.map slightly. I have been planing to package this, but find no time to do it. If anybody wants to do this, I'd be more than happy to help. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Re^2: dvips top margin
Susan> I had another problem with the latest version of TeX which I could Susan> not seem to fix by adjusting dvips/config.ps. It was that an extra Susan> 1" seemed to have been added to the margins on all sides. I fixed Susan> this using vmargin, but that wasn't necessary before. Weird. Do you get those margins even when you say $ latex testpage to run this LaTeX2e-updated utility that originates from Lamport? On my box the rulers go right to the edges. Susan> What additional config file should I be looking at? I don't know. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: as86?
Marcus Hightower writes: Marcus> Does anybody know where I can find this as86 thing. See the Debian FAQ (from the doc-debian package, or the ftp site) 7.4. How can I find out what package produced a particular file? To identify the package that produced the file named foo execute either: o dpkg -S filename. This searches through the lists of installed files. This is (currently) equivalent to searching all of the files having the file extension of .list in the directory /var/lib/dpkg/info/. o grep foo Contents, or zgrep foo Contents.gz. This searches for files which contain the substring foo in their full path names. The files Contents and Contents.gz reside in the major package directories (Debian-1.1, non-free, contrib, etc.) at a Debian FTP site. A Contents file refers only to the packages in the subdirectory tree where it resides. Therefore, a user might have to search more than one Contents files to find the package containing the file foo. Hence: miles:~ [root] # zgrep as86 /mirror/debian/rex/Contents.gz usr/bin/as86 bin86 so you have to install the bin86 package. Marcus> Also does the zimage automaticly gets installed when you do a Marcus> make zimage or does it just makes it without installation. Use the kernel-package from the rex/binary-all/misc directory. It will build a .deb file with your kernel and install it all for you (you need to read some doc, though). -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: dvips top margin
Nick> I installed Debian 1.1 and found that my tex files were no longer Nick> leaving a margin at the top of the page. I traced it to Nick> /usr/lib/texmf/dvips/config.ps where a4size was set instead of Nick> letter. Correct. Also the app-defaults file for xdvi sets a4 paper and deskjet mode. We really need a script that sets these. Any takers? -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: curses.h?
Marcus Hightower writes: Marcus> I need a header file named curses.h Markus, it's documented in the FAQ as I wrote you some hours ago. So please do use the Contents or Contents.gz file. miles:~ [root] # zgrep "curses.h" /mirror/debian/rex/Contents.gz usr/i486-linuxaout/include/bsd/curses.h libc4-dev usr/i486-linuxaout/include/curses.h libc4-dev usr/include/curses.h ncurses3.0-dev usr/include/ncurses.hncurses3.0-dev usr/include/slang/slcurses.h slang-devel In this case you want the ncurses3.0-dev package. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: kernel-package
Chuma> Anyway after make config , make-kpkg refuses to run telling me there Chuma> is nothing to make. What do I need to do to have make-kpkg rebuild Chuma> with the new config ? Run make-kpkg clean *before* you do a make xconfig -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: recompiling kernel and ppp
Kevin> what am I missing?!?!?!? You didn't tell us what you were selecting for PPP in "make xconfig". If I am not mistaken, you need modules. I always use them where I can, and they work great for me. I select: # # Network device support # CONFIG_NETDEVICES=y CONFIG_DUMMY=m # CONFIG_EQUALIZER is not set # CONFIG_DLCI is not set CONFIG_PLIP=m CONFIG_PPP=m # # CCP compressors for PPP are only built as modules. # CONFIG_SLIP=m CONFIG_SLIP_COMPRESSED=y CONFIG_SLIP_SMART=y and # # Character devices # CONFIG_SERIAL=m which, using kerneld with the "auto" option in /etc/modules, and "alias dummy0 dummy" in /etc/conf.modules gives: miles:~ [root] # lsmod Module:#pages: Used by: ppp51 (autoclean) dummy0 11 (autoclean) slip 22 (autoclean) slhc 2[ppp slip] 3 (autoclean) serial 74 (autoclean) (Ppp is used to the ISP, and slip is loaded as I also use cslip over a null-modem cable as an el-cheapo network to a laptop running Debian). -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: llseek error & large partitions
Johnny> I do not know why this happens, only that the same thing happend Johnny> to me when I tried to format my Quantum Sirrocco 2.5 Gig hard Johnny> drive. Around about the 2.1 gig it starts reporting errors. I can Johnny> only assume that the system cannot handle partitions over 2 gig. You need a version of fdisk that as recent as 2.0 or better to support disks bigger than 2 GB, for example the fdisk from util-linux-2.5 which says miles:~ [root] # fdisk -v fdisk v2.1 (>4GB) As for maximal partition size, read (using altavista or dejanews) the newsgroup article with referenced below, it shows how partitions can be as large as 959 GB (yes, just under one TB, a million MB). You have to increase the block size which becomes a little inefficient at that size ... I cannot recommend few big partitions anyway. Disks do fail, and you're better off when it happens on a limited partition, /usr/local say, rather than on /. Subject: Re: Really 2 GB file size limit? From: [EMAIL PROTECTED] (Kristian =?ISO-8859-1?Q?K=F6hntopp?=) Date: 1996/07/28 Message-Id: <[EMAIL PROTECTED]> References: <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> Organization: =?ISO-8859-1?Q?entf=E4llt?= Newsgroups: comp.os.linux.development.system -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: newaliases script?
Kevin> I've got an application I'm trying to install on my new Debian Kevin> system and it is giving me a warning about a missing script (called Kevin> newaliases). [...] Kevin> Anyone have any ideas about where I can find this script? miles:~ [root] # zgrep newaliases /mirror/debian/rex/Contents.gz usr/bin/newaliases sendmail usr/man/man1/newaliases.1sendmail -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
RE: texbin postinst in unstable fails
Casper> I wonder if you might possibly translate that for the rest of us. My excuses! That mail was not meant to also go to the list. "texbin" fails to configure because version mfbasfnt-1.0-5 is missing the file manfnt.mf (a special font for something in Knuth's book). Downgrading to mfbasfnt-1.0-3 from "buzz-fixed" overcomes the problem. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: texbin postinst in unstable fails
Miro Torrielli writes: Miro> I installed debian 1.1.8 on another pc, using dpgk-ftp to retrieve Miro> all necessary packages, from stable & unstable. Firstly, I noticed Miro> that when installing a large number of packages on the system, the Miro> unpacking of some of them fail, complaining that some lib or another Miro> is missing. This has already happened twice before, on two other Miro> machines. Il m'est impossible de voir ce que a echoue. Plus de detail? Miro> Second, and for the first time, texbin postinst fails, complaining of Miro> missing man.fmt. I had to manually comment out the postinst script. C'est manfnt.mf. Install mfbasfnt-1.0-3 de "buzz-fixed" et ca ira. Miro> Finally, elm, in this new installation, has no ispell option in Miro> mailing menu, whereas on my other machines it does. Aucune idee, je n'utilise pas elm (mais emacs avec vm). Amities, Dirk -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: unsubscribing these lists
Michael> How does one unsubscribe these lists? $ echo unsubscribe | mail [EMAIL PROTECTED] -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Creating deb packages
Algirdas> Could someone please mail me a complete step-by-step instruction
Algirdas> on creating deb packages?
Vadik> Please post it to the list,
New version of dpkg (<< 1.4.0) or dpkg-dev (>= 1.4.0) come with two html'ed
manuals in /usr/doc/dpkg/{programmer,policy}.html/ that pretty much cover it
all.
--
Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: re:D-Link DE-220P driver needed
Todd> Some friends and I all have a bunch of D-Link DE-250CT's, which are Todd> NE2000 compatible. They all work fine under Linux. I had to use Todd> their setup program (under M$-DOS) to set the IRQ, I/O address, BNC, Todd> RJ-45, and a few other things. After that, yer set! Looking around at Don Becker's site, I once found two really handy C source files for us NE2000 users. The first, atlantic.c, allows me to set these irq, i/o, bnc or rj-45, ... under Linux rather than the unspeakable os. The other, ne2k.c, display the ne2000 cards status. I seem to have gotten them from ftp://cesdis.gsfc.nasa.gov/pub/linux -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: D-Link DE-220P driver needed
Steve> I like Bruce Perens's idea of splitting up the mailing Steve> list because, even if I did have a mail filter (which I assume Steve> parses the subject line), Wrong assumption. Install the procmail package and see that it can parse _anything_ from mail-header fields to actual mail contest. Mail-readers that can "thread" as the emacs VM or GNUS modes are also useful. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Babel installation fix?
David> Do I have to completely remove TeX and then reinstall it? Just downgrade to mfbasfnt-1.0-3 from the "stable" aka buzz-fixed release. It contains a manfnt.mf so that TeX builds. "dpkg --configure babel" should then set babel straight. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
corrupt filesystem superblock
My laptop doesn't boot Linux on hda3 anymore. I just installed a new system by overwriting the swap partition on hda2. My first attempt of running e2fsck gives: The filesystem superblock is corrupt. Try running e2fsck with an alternate superblock using the -b option. (8193 is commonly as alternate superblock; Henc 'e2fsck -b 8193 ' may recover the filesystem.) I tried this, but it didn't work either. Any help as to what I might do now would be welcomed. The partition table is as follows (in sectors) # TypeFirstLast Offset Length Filesystem Type Flags 1 Primary 0 131039 63 131040 Dos 16bit06 None 2 Primary 131040 180431 0 49392 Linux (83) Boot 3 Primary 180432 1032191 0 851760 Linux (83) None and was created by the Debian install disks around February or March. Thanks in advance, Dirk -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: char-major-10
I assume that Bruce meant $ ls -l /dev | grep "10," which yields: [EMAIL PROTECTED]:~> ls -l /dev|grep "10," crw-rw 1 root sys 10, 134 Dec 31 1969 apm_bios crw-rw-rw- 1 root sys 10, 3 Dec 31 1969 atibm crw-rw 1 root audio 10, 128 Dec 31 1969 beep crw-rw 1 root sys 10, 133 Dec 31 1969 exttrap crw-rw 1 root sys 10, 132 Dec 31 1969 hwtrap crw-rw-rw- 1 root sys 10, 2 Dec 31 1969 inportbm crw-rw-rw- 1 root sys 10, 4 Dec 31 1969 jbm crw-rw-rw- 1 root sys 10, 0 Dec 31 1969 logibm crw-rw 1 root sys 10, 129 Dec 31 1969 modreq crw-rw-rw- 1 root sys 10, 1 Dec 31 1969 psaux crw-rw 1 root lp10, 138 Dec 31 1969 qcam0 crw-rw 1 root sys 10, 135 Dec 31 1969 rtc crw-rw 1 root sys 10, 131 Dec 31 1969 temperature crw-rw 1 root sys 10, 130 Dec 31 1969 watchdog -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Debian (not Linux) newbie
I can't help you there. While I use (and maintain) tob and afio, I never use muti-volume backups as I do them off-line anyway. I just define my backup "volumes" such that they will fit on one 170 MB tape. That holds /etc, /home and /usr/local, which is all I need. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: mirror of a site
Miguel> Cna anynone please tell me how do I mirror a site? Is there any Miguel> howto? No, but I've put a README with some documentation, as well as an example file for ftp.debian.org, into the Debian mirror package. The manual page for mirror is quite good. Miguel> what are the procedures? Which files shall I configure? Install ftp://ftp.debian.org/pub/debian/rex/binary-all/net/mirror_2.8-7.deb and look at the example in /etc/mirror/packages/ftp.debian.org. I recommend creating a package file per site in /etc/mirror/packages and then to start the mirror as mirror /etc/mirror/packages/ -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: iBCS-problem
Dolgopolsky Igor writes: Igor> Hello. I have a problem with iBCS from 1.2.13 don't match 2.0.16. Igor> I downloaded iBCS 2.0 but cannot compile this. You can just ignore the errors. I usually do "make" after un-tar-ing the source, with no confih whatsoever and it works. The make breaks in the 2nd or 3rd subdirectory it visits --- but it does create the file "iBCS" in one of the subdirs, and this is the module you are looking for. I usually move it to /lib/modules/`uname -r`/ibcs.o, and have a script insmod it when needed. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Latex: where to install new styles? Mf: how to muzzle?
Jean> 1) Where/how am I supposed to install latex styles for all users? Below /usr/local/lib/texmf/tex/latex/ As kpathsea searches recursively (see /usr/lib/texmf/web2c/texmf.cnf), I usually keep some subdirs (that match the CTAN structure). Jean> 2) Kpathsea is a nice and powerful thing. However, when viewing a dvi Jean> from netscape, chances are it will contain unsual fonts and thus Jean> produce a lot of output like [...] Hm, I presume netscape just calls xdvi as you would from the command line. So you have to configure it anyway. One file that needs some configuration is /usr/lib/texmf/dvips/config.ps for things as resolution, papersize, ... This should probably be in /etc and softlinked back to /usr/lib/texmf/dvips/config.ps -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: kernel with APM
Giuseppe Vacanti writes: Giuseppe> I would like to try out the Advanced Power Management features Giuseppe> in my PC. However I haven't been able to enable APM in the Giuseppe> kernel. When I do make xconfig I do nt see any obvious switch Giuseppe> referring to APM. APM is in the "Character Devices" menu, after mice and tape and before watchdog. Giuseppe> What am I missing? Could anybody who's ben using APM tell me if Giuseppe> they're happy with it? Quite, especially with the apmd package. The suspend and sleep features are very nice. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: latex
Giuseppe Vacanti writes: Giuseppe> The LaTeX in Debian seems to be broken. Revision 2e-4 didn't Giuseppe> seem to find the standard classes and packages. Revision 2e-7 Giuseppe> dies on installation. Giuseppe> Giuseppe> Has anybody been successful with LaTeX? I never had any problem whatsoever with LaTeX under Debian. And I update to the most recent "rex" releases every two or three days (dselect rules !!). I'll send you my "dpkg -l" output in private mail. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Upgrading to un-debianized kernel
Miro> I recently upgraded to the 2.0.17 kernel, although this problem Miro> occured with other non-debian kernels. I now get, on boot, and in Miro> my daemon log when use diald and pppd, Miro> Can't locate module net-pf-4 Miro> """""" -5 Boris> it's aliases Boris> ignore it Or set up your /etc/conf.modules: #alias eth0 #alias off alias net-pf-4 off alias net-pf-5 off -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Two Packages Missing
Richard> I taper no longer supported or is there an updated package that Richard> includes it? I kind of liked it. It is orphaned. If you really like taper, you could maintain it. There have been new upstream releases. :-) As for tape backups, we have "tob" which most people who tried it seem to be rather happy with, your's truly included. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Debian install: Thinkpad560 Dual Scan
Derek> Hi I am contemplating installing debian on an IBM thinkpad 560 Dual Derek> scan. Are there any special things to do regarding this computer? No idea, check the Linux Laptop home page (http://www.redhat.com/llhp/) Derek> Are there Linux utilities to test the battery level etc? Yes, apmd. Also available as a Debian package. Requires that the laptop respects APM Bios 2.0 or 2.1, I believe (all new ones do). -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: popclient
Miro> Does anyone know how to use procmail with popclient? When Miro> configured to use procmail, popclient exits saying that it can't exec Miro> a few things. Any hints? :-) The two are unrelated. I use popclient as in popclient -3 -P to grab the mail. Procmail gets into the picture via the ~/.forward file once popclient has the mail delivered locally (which uses smail on my box). I explained the .forward and .procmail setup in a couple of mails here recently, email me if you can't find a copy. Otherwise, the procmail doc included in the package is pretty good. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: inn installation question
Erik> Can't really seem to figure out what is wrong. Indeed I do not have Erik> a /usr/bin/mail, could it be my use of sendmail instead of smail? If Erik> anyone has any hints on how to solve this one, I'll be very grateful! Install the "mailx" package, it provides the simple /usr/bin/mail that is so handy for scripts. If the inn package you use does not depend on mailx, you should report this as a bug. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: a4.sty and a4wide.sty
>> Hakan Ardo writes: > Sure it does, the only trouble with that one is
>> that it only uses half of > the paper :) The rest is margins...
>>
>> That's true. And I don't know how to solve it correctly. I just entered
>> a latex preamble:
I always use "vmargin.sty" which allows to define the margins sizes
explicitly, and also allows me to switch from letter to a4 paper with one
command, eg
\usepackage{vmargin}
\setpapersize{A4}
\setmargrb{3cm}{3cm}{3cm}{3cm}
selects a4 paper (default is letter), and then sets left, top, right and
bottom margins to 3cm. Vmargin is on CTAN hosts as
tex-archive/macros/latex/contrib/other/misc/vmargin.sty
--
Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: module cdu31a fails on insert
Todd> Is there any chance you have checked the bootparm-howto? Todd> Todd> http://sunsite.unc.edu/mdw ?? If you install Debian's doc-linux package, they are also at /usr/doc/HOWTO/* As the doc-linux package is updated monthly with the upstream HOWTOs, you should always look for the newest one in unstable/binary-all/doc. Safe to install on any system as it's only text. Hard to put a bug into that :-) -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Not keeping correct time
Shaya Potter writes: Shaya> I just noticed that my linux box dosen't seem to be keeping correct Shaya> time. File that under "advantages of being IBM PC compatible" :-) Shaya> Does anyone know of an efficient way to make my linux box keep Shaya> correct time? I have these two lines in a script that cron runs nightly for root netdate tcp clock -u -w Netdate is in netstd, you already have it. I am sure that you can find a timeserver in your domain, if not, pointers to do them are in the xntp doc. Afterwards, clock -u -w writes (-w) the correctly set time into the bios using GMT (-u) time. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: ftape format warning!!!
Ken Gaugler writes: Ken> I had a nasty surprise using ftape to dump my filesystems for Ken> archival right before repartitioning my hard drive. I wanted to let Ken> people know about this in case they were not already aware. Ken> Ken> If you plan to dump several filesystems to one tape (that is, more Ken> than one session on a tape) you need to erase the tape using mt Ken> FIRST!!! Otherwise you will not be able to read any sessions beyond Ken> the first. You could have read the Ftape-HOWTO before attempting to do this. It says: 6.4. Putting more than one tar file on a tape To put more than one tar file on a tape you must have the mt utility. You will probably have it already, if you got one of the mainline distributions, e.g. Slackware or Debian. tar generates a single Tape ARchive (that's why it is called `tar') and knows nothing about multiple files or positioning of a tape, it just reads or writes from/to a device. mt knows everyting about moving the tape back and forth, but nothing about reading the data off the tape. As you might have guessed, tar and mt in conjunction, does the trick. By using the nrft[0-3] (nftape) device, you can use `mt' to position the tape the correct place (`mt -f /dev/nftape fsf 2' means step over two ``file marks'', i.e. tar files) and then use tar to read or write the relevant data. Ken> There were only two sessions on my tape. Sadly, the second session Ken> contained the /usr filesystem, so I lost everything :( Moreover, you can simplify your life by using a program as tob which reads from several filesystems. Debian tob's package is in the admin section. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: can't ifconfig dummy
Bill Roman writes: Bill> Why doesn't `ifconfig dummy songdog' work? It tells me: Bill> SIOCSIFADDR: No such device Try ifconfig dummy0 route add and if you use kerneld to load modules, add this in /etc/conf.modules: alias dummy0 dummy -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: All my memory disappears, rather too quickly
Jay> In addition to previous mail, would like to add that I've started Jay> receiving messages like this too: [...] Jay> Aug 16 16:11:31 named[145]: NSTATS 840208291 840075082 A=102 PTR=2 Jay> MX=3 ANY=3 Aug 16 16:11:31 named[145]: XSTATS 840208291 840075082 Jay> RQ=110 RR=44 RIQ=0 RNXD=1 RFwdQ=37 RFwdR=30 RDupQ=0 RDupR=0 RFail=0 Jay> RFErr=0 RErr=0 RTCP=0 RAXFR=0 RLame=0 ROpts=0 SSysQ=38 SAns=73 Jay> SFwdQ=37 SFwdR=30 SDupQ=0 SFail=43 SFErr=0 SErr=0 RNotNsQ=110 Jay> SNaAns=72 SNXD=1 It means you are are running the named daemon from the bind package. Sure you need an internet name server on your machine? -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
HOWTOs (Was: Problems with aha1542.o ...)
Christopher> It took me a while to search this down, but the documentation Christopher> (thanks to LDP) is available at: Christopher> Christopher> http://sunsite.unc.edu/mdw/HOWTO/BootPrompt-HOWTO-4.html#ss4.2 If you install the doc-linux package, it is also at /usr/doc/HOWTO/* As the doc-linux package is updated monthly with the upstream HOWTOs, you should always look for the newest one in unstable/binary-all/doc. Safe to install on any system as it's only text. Hard to put a bug into that :-) -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Netscape and XKeysymDB
Ken Gaugler writes: Ken> I am getting lots of messages about the key symbols not being right. Ken> I remember from BSDI that you had to have the nls directory set up in Ken> a special place. Tried to hack this but still not getting the desired Ken> result (i.e. no more messages about missing keysyms). There are several ways out: - read the README that comes with netscape :-) - install Debian's motifnls package, it creates the required nls directory - use Debian's "netscape" installer package and there are still more possibilities. I simply put netscape's binary into /usr/local/lib/netscape, along with the moz3_0.zip and the XKeysymDB file, and then use this script which is saved as /usr/local/bin/netscape #!/bin/sh XNLSPATH=/usr/lib/X11/nls/ \ XKEYSYMDB=/usr/local/lib/netscape/XKeysymDB \ exec /usr/local/lib/netscape/netscape "$@" -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: bash PATH, dselect and ftape
Chris> 1. This question is more of a general bash question and if I Chris> shouldn't be posting it here someone let me know and I won't pester Chris> this group about it any more. All the following applies to the root Chris> account, I haven't checked it for an ordinary user yet. [lots of stuff deleted] Read the man pages for bash and xterm (or rxvt) and try to understand what happens. You can have shells that are not login, they use ~/.bashrc. Every "normal" (ie non-login) xterm reads this file _unless_ you tell xterm to be start a login shell via the -ls option. Login shells read ~/.bash_profile. Chris> 2. I'm trying to add packages to my installation, among them lyx and Chris> mailcrypt and taper. Dselect complains that lyx needs the package Chris> xforms and mailcrypt needs pgp. However, I have downloaded both of Chris> these packages and told dselect about them. When it asks to When you tell dselect about the directories, it wants to find the Packages[.gz] file. You might have noticed the error message when that failed. If you only want to install three packages, why don't you just use dpkg -i ? Chris> 3. I can't find ftape anywhere. Does anyone know where to find this? Since around 1.3.72, it is in the kernel sources in the directory /usr/src/linux/drivers/char/ftape. Just recompile your kernel (use "make menuconfig" or "make xconfig" instead of make config). Hope this helps, Dirk -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: fileutils 3.13-3 seems to be messed up
(Follow-up only to debian-user. Please don't write to debian-devel and debian-user at the same time.) Shaya Potter writes: Shaya> I just installed fileutils 3.13-3, and now my 'ls' is screwed up. Shaya> I removed color-ls as it said I was supposed to. I remember reading Shaya> about how color-ls is now included in fileutils. However there is Shaya> no color-ls file in my /usr/bin directory anymore, there is a Shaya> dircolors file though, doing a dpkg -S dircolors tells me its from Shaya> the fileutils package. Why did fileutils install dircolors and not Shaya> color-ls? Am I missing something here? You are missing a discussion we had about a ago on this. I append my mail below. Erick added documentation to fileutils-3.13-3, you can find it in /usr/doc/fileutils/color-ls.gz >From edd Wed Jul 31 11:19:20 -0400 1996 From: [EMAIL PROTECTED] To: Mark Phillips <[EMAIL PROTECTED]> Cc: debian-user@lists.debian.org, Erick Branderhorst <[EMAIL PROTECTED]> Subject: Re: color ls Mark Phillips writes: Mark> Hi, Something has happened to stop ls giving color output. I used Mark> to be able to just run: Mark> Mark> eval `dircolors` Mark> Mark> and ls would work in color - even without specifying the "--color" Mark> option. (And no, ls was not aliased) Oh yes, it was aliased to do that. Just start the dircolors (the old one from the color-ls package, that is) from the shell and you'll see. Mark> Now it seems I need to type "ls --color" to get color? The old dircolors created the LS_COLORS env variable with your colour selection the default or specified rc file _and_ created the aliases. Now it only builds LS_COLORS so that I changed the code in /usr/local/etc/profile (which I source from /etc/profile) to eval `dircolors -b /usr/local/etc/colour-ls.rc` alias ls='ls --color=auto '; alias dir='ls --color=auto --format=vertical'; alias vdir='ls --color=auto --format=long'; alias d=dir; alias v=vdir; alias ols='/bin/ls ' Before, I only needed the "eval 'dircolors -b `" line and the aliases were built automagically (the format was slightly different, though, there was also --8-bit or some such). Mark> What is the problem? I've changed a number of things of late - moved Mark> from tcsh to bash (but same thing happens in both shells), and Mark> upgraded a number of packages. So I don't know what has caused the Mark> change. Any ideas anyone? Your upgrade to the newest fileutils package which replaced the now redundant color-ls package. Erick, is there a way that you can persuade/hack dircolors to do what the old one did? Or put a note in the package to ease transition? -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: diald
Tim> Does anyone know how to get diald to work? I'm almost there; I can Tim> get it to do it's thing, dial, connect to my ISP, start sending data, Tim> but after about 30 seconds, diald kills pppd, and reports that it has Tim> timed out waiting for ppp to connect. :/ Since I was in the middle of Tim> receiving an html document from the web, I'd say it's safe to assume Tim> that pppd was already connected. Anyone got any ideas how I can cure Tim> this VERY annoying problem? Diald can be configured to do almost everything. Using it "out of the box" with Eric Schenk's original /etc/diald.conf as packaged for Debian, I also noticed that I could die after some seconds, claiming inactivity. You can either modify to diald.conf so that it always stays up longer. What I did was to add a this to /etc/ppp/ip-up # send a couple of pings ping -c 20 -s 2048 -i 15 my.isp.provider >/dev/null Recently, I created a fifo /etc/diald.fifo, set the option fifo /etc/diald.fifo in /etc/diald.options and wrote the following script to issue commands as $ dial force etc. This is quite handy as a one-command interface for other scripts. Hope this helps, Dirk - /usr/local/bin/dial --- #!/bin/sh # # dial --- communicate via fifo `/etc/diald.fifo' with the diald daemon # # edd 12 Jul 96 if [ $# -ne 1 ] then echo "Usage: dial " exit 1 fi case "$1" in block|unblock|force|unforce|down|up|delay-quit|quit|reset|queue) echo $1 >> /etc/diald.fifo ;; *) echo "Unknown (or unsupported) command: $1" exit 1 ;; esac exit 0 - -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: mail-delivery-agents?
Hi Lamar I hope you don't mind that I put this back onto the list as other people might wonder about this too. Lamar> Hi, Dirk. I seem to be having some trouble getting posts to the Lamar> debian-user list, so I'm mailing you directly. Lamar> Lamar> You posted something about using procmail _and_ pop to sort your Lamar> mail. I was wondering how you invoke procmail? I, too, use Lamar> popclient to get my mail, and I'd _love_ to be able to have procmail Lamar> (or slocal or something) filter it for me. It is all in the procmail manpages --- but as those are quite detailed it hard to find as first sight. Note also that Debian procmail package has, as many other packages, a lot of documentation in /usr/doc/ and /usr/doc/example/. 1. I start popclient as (indented by a TAB for readability) popclient -s -3 -P ~/.file-with-password host.that.has.mail to get my mail to my local machine. 2. One needs a file ~/.forward of the following form "|IFS=' '&&exec /usr/bin/procmail -f-||exit 75 #edd" where the end must #" as a fallback strategy. This passes the mail to the procmail program. 3. A file ~/.procmailrc describes the sorting rules. There are lots of examples in the manpage, and the /usr/doc/examples/procmail directory. But as a concrete example, here are some pieces. If it looks all to strange, than it's probably time to review a Unix book with something on regular expressions. I just show some entries as it is mostly repetitive ~/.procmailrc -- # edd 26.10.95 installed from adapted version from /usr/doc/examples/procmail PATH=$HOME/bin:/usr/bin:/bin:/usr/local/bin:. MAILDIR=$HOME/mail # use `usual' default and not this one: DEFAULT=$MAILDIR/mail.in LOGABSTRACT=all LOGFILE=$MAILDIR/log/procmail :0: # mail To or CC ctan-ann * (^To|^CC|^Resent-).*ctan-ann@(shsu.edu|RELAY.URZ.UNI-HEIDELBERG.DE) in.ctan-announce :0: # mail To or CC debian-announce * (^To|^CC|^Resent-)[EMAIL PROTECTED](debian.org|pixar.com) in.debian-announce :0: # mail To or CC debian-bugs, debian-bugs-done * (^To|^CC|^Resent-).*debian-bugs(-done)[EMAIL PROTECTED](debian.org|pixar.com) in.debian-bugs :0: # Another one for new bugs system * (^To|^CC|^Resent-)[EMAIL PROTECTED] in.debian-bugs :0: # mail To or CC debian-changes * (^To|^CC|^Resent-)[EMAIL PROTECTED](debian.org|pixar.com) in.debian-changes :0: # mail To or CC debian-devel * (^To|^CC|^Resent-)[EMAIL PROTECTED](debian.org|pixar.com) in.debian-devel :0: # mirror logs * ^To: edd * ^Subject: mirror update$ in.mirror-update - :0: always starts a new rule section. Line with * describes rules, if there are several they are ANDed together. The last line shows the folder into which a message is put. Hope this helps, Dirk -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: color ls
Robert> hmm ok well i have checked 5 sites now and fileutils-3.12-4.deb IS Robert> the newer version. if i am wrong about this please tell where there Robert> is a newer version. You have to look in the "rex" aka "development" aka "unstable" tree, not in "Debian-1.1-fixed" aka "Debian-1.1.3" aka "buzz-fixed" aka "stable". You'll find fileutils-3.13-2 in the base section. Maybe wait a few days until fileutils-3.13-3 gets moved in which Erick uploaded this morning to the developers site. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: mail-delivery-agents?
Dale Scheetz writes: Dale> The current way I manage my e-mail involves using popclient to get Dale> incoming e-mail from my ISP (mail.polaris.net) to dwarf's incoming Dale> mail folder on my machine (dwarf.polaris.net), Same here. I advocate [EMAIL PROTECTED], which forwards to [EMAIL PROTECTED], my wife's Debian box on campus, from where I popclient everything to the house ([EMAIL PROTECTED], dynamic IP). Dale> outgoing mail is handled by Pine using SMTP. I use smail, locally and for outgoing mail. _Very_ easy to setup and change all the time because Ian Jackson wrote a fantastic /usr/sbin/smailconfig. The only problem is that you have to "fake" another email address for outgoing mail. There are several ways to do it, directly with emacs, pine (as you compiled the switch in, on my hint :-) and elm, or via smail indepently of the MUA as was recently discussed on debian-user. I also use procmail to sort my mail into 21 (!!!) different mail folders for different mailing lists. Dale> I would like to try some other mail readers like elm and emacs but Dale> none of these seem to be able to do SMTP and so require the presence Dale> of a mail-delivery-agent. I can really recommend emacs with the vm mode. Really. Dale> I know I have heard pro and con discussions about smail vs sendmail Dale> but am still confused. Also I hear hints that qmail (?) will soon be Dale> a Debian package. I am a complete idiot when it comes to mail Dale> systems, so please be simple and clear in your answer ;-) Goes with smail. Use the "smarthost" feature of the smailconfig for outgoing mail. Dale> What do I need to understand to install and use any of the available Dale> packages? Nothing, basically :-) Dale> If I am going to go to the trouble o1f learning to install a Dale> mail-delivery-agent I will want to be able to filter mail to various Dale> mail-folders based on the source of that mail. You can already do that with your existing setup. Just install procmail, and email me for examples. Dale> If that is not available in the mail-delivery-agent, then point me Dale> to a package that will work with the target mail-delivery-agent. Dale> Dale> Thanks in advance for any help, Pleasure. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: mirror
Ricardo> Hi! First, thanks for the pointers... But I'm still having Ricardo> problems. I've done an unpack and install on mirror-2.8-0.deb Ricardo> (that's the version that I have here for 0.93), but there's Ricardo> nothing under /etc/mirror (as a matter of fact, that directory Ricardo> didn't even exist). There's also no README.debian under Ricardo> /usr/doc/examples/mirror. I was referring to the newest mirror package, mirror-2.8-6. As it's written entirely in perl, this is an "Architecture: all" package so that you can fetch it from Debian-1.1.3/binary-all/net/mirror-2.8-6.deb That should work with 0.93R6, but you should update to Debian-1.1 as many, many things have improved. See the "upgrades" directory for hints and notes. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: dpkg-ftp or installing via FTP
Ricardo> Hi guys, I grabbed and installed dpkg-ftp. But I see no mention Ricardo> of any ftp usage when I run dpkg --help. Is there supposed to be Ricardo> a separate binary called dpkg-ftp? Because I can't find any... It is meant to be used from within _dselect_. It gives you a new line in the * 0. [A]ccess Choose the access method to use. menu. Ricardo> I'm trying to download the latest Debian distribution via FTP, but Ricardo> I've had problems both with ncftp and with dftp. I was hoping Ricardo> dpkg-ftp would be a better solution... What's wrong with ncftp? Maybe your networking is not setup correctly. Ricardo> Another suggestion was to mirror debian. How much space is Ricardo> required to mirror? Depends on how many files and directories you select in the mirror parameter file /etc/mirror/packages/ Ricardo> Can a mirror be partial? Sure, see the example /etc/mirror/packages/ftp.debian.org, but please not that it was written with the old directory structure in mind. Ricardo> Lastly, once I've gotten all that's required from FTP, is dselect Ricardo> a good method to use for performing the install? Yes. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: color ls
Robert> same here. i installed fileutils-3.12-4. the color-ls package still
^^
Robert> remains and wasn't removed
Dirk> Sure, 3.12-4 is not 3.13-{1,2}. Try a newer version of fileutils.
Michael> Sorry Dirk but
I never said that _you_ were using 3.12. Robert said he did, and I was only
pointing out that wasn't the right one to begin with.
--
Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: color ls
Dirk> Oh yes, it was aliased to do that. Just start the dircolors (the old Dirk> one from the color-ls package, that is) from the shell and you'll Dirk> see. Michael> I like to see that the standard ls is now colorized. But I wonder Michael> why the old one isn't removed automatically? Is there a reason for Michael> this? color-ls was removed here when I installed fileutils-3.13-1. [EMAIL PROTECTED]:~> dpkg -l color-ls fileutils Desired=Unknown/Install/Remove/Purge | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad) ||/ NameVersionDescription +++-===-==- pn color-ls (no description available) ii fileutils 3.13-2 GNU file management utilities. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: color ls
Michael> I like to see that the standard ls is now colorized. But I wonder Michael> why the old one isn't removed automatically? Is there a reason for Michael> this? Wichert> Some monitors make text unreadable (cheap laptops for example) Wichert> when you use colors. Maybe a postinst-script could ask if color-ls Wichert> should be the new default, move ls to ls-normal and make ls a link Wichert> to color-ls. Don't worry. It is one and the same binary, bin/ls, which has no color as the default. You have use the --color option to enable color. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: color ls
Guy> I will add that `--color' is equivalent to `--color=yes', which is
Guy> probably not what you want. If you make any aliases, use
Guy> `--color=tty'. You don't want nasty escape codes if it's not a tty.
Those are the old style options implemented by color-ls (which is a patched
version based on fileutils-3.12). The new form is, according to ls --help
[...]
--color[=WHEN] control whether color is used to distinguish file
types. WHEN may be `never', `always', or `auto'
[...]
By default, color is not used to distinguish types of files. That is
equivalent to using --color=none. Using the --color option without the
optional WHEN argument is equivalent to using --color=always. With
--color=auto, color codes are output only if standard output is connected
to a terminal (tty).
By the way, the man page distributed with fileutils-3.13-2 still documents
the old form (ie --color={yes,tty})
Guy> Try `ls --color | less' to see what I mean.
That would now be 'ls --color=always | less' and is why --color=auto is a
better choice.
Guy> And fileutils does have a bug with regards to color-ls. It should
Guy> conflict and replace with it so that it will be removed automatically.
Guy> I'll file a report.
Yes, please add that it has Pre-Depends: and not Depends:, and that the
manpage needs an update.
--
Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: color ls
Dirk> > color-ls was removed here when I installed fileutils-3.13-1.
^^
Michael> Strange. On my site it wasn't.
Robert> same here. i installed fileutils-3.12-4. the color-ls package still
^^
Robert> remains and wasn't removed
Sure, 3.12-4 is not 3.13-{1,2}. Try a newer version of fileutils.
--
Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: digested mailing list available?
John Cook writes: John> Does this mailing list offer the option of receiving periodic John> message digests as e-mail instead of every inidividual message? No, but you can use mailagent or procmail (both available as Debian packages) to filter you mail into different folders which IMHO is better than digests. I use procmail the following regexp in ~/.procmailrc to filter debian-user: :0: # mail To or CC debian-user * (^To|^CC|^Resent-).*debian-user@(.*pixar.com|.*yggdrasil.com|.*debian.org) in.debian-user It is not as complicated as it looks. The ":0:" line starts a new rule, and "in.debian-user" ends it by saying that debian-user is my incoming mail folder. Any mail that has a To:, CC: or Resent: which contains debian-user@ followed by any of the three possible lists (our old one, our new one, and yggdrasil's gateway) get sorted into that folder. I use similar ones for other mailing lists. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Adding a Lilo boot
Noam Rettig writes: Noam> I have msdos on /dev/hda1 and linux on /dev/hda2. I know that lilo Noam> can be configured to boot with either one at startup by pressing the Noam> TAB key. How can I add the DOS image to my startup. You have to read the LILO documentation in /usr/doc/lilo, or in a decent book as "Running Linux". If you are _really_ lost, email me in private. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Default Pager
Noam Rettig writes: Noam> Currently the default pager on my system is 'more'. I want to Noam> change it to 'most'. How should I do that? Add this line to /etc/profile (or ~/.bash_profile) export PAGER=most -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Will Debian 1.1 have emacs-19.31?
David> Where is the .deb file? Oh, sorry, I must have taken emacs-19.31 from the developers system. The public archive lags a little, mostly a day, sometimes a little longer. Expect to see emacs-19.31 in unstable/binary/editors any time now. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Will Debian 1.1 have emacs-19.31?
David M Cooke writes: David> Will the officially release Debian 1.1 include emacs-19.31? [EMAIL PROTECTED]:~> emacs --version 19.31.1 [EMAIL PROTECTED]:~> cat /etc/debian_version 1.1 Debian 1.1 is in official beta-test and can be installed safely. Expect the official release Real Soon Now (TM). Emacs-19.31 seems a little more alert than 19.30. Nice job, Mark! -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: F77 has gone missing?
Dale> I ended up making a link to g77 and things work better. I wouldn't do that. Whenever I compiled octave here, ./configure was perfectly capable of working with either f2c or g77, maybe with the help of a --with-f2c argument, but that's about it. If you make a link f77, one program one day will think it's native Fortran which it is not. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: F77 has gone missing?
Dale Scheetz writes: Dale> I just went back to working on Octave and don't seem to have F77 any Dale> more. I have been keeping pretty up to date on the devel Dale> section. Anyone know what happened to it? There never was one. Take your pick among f2c with fort77 or g77. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: printing in debian/unix is hard...
Carlos> Moral: if you want to have smart printing, use window$ :-( :-( :-( Narrr. I am quite happy with one simple entry in /etc/printcap, plus one gs_filter. All I print is in postscript (generated by genscript or dvipsk) and printed via gs. That I can simply say "print" in any application program. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: reporting in on my new upgraded debian 1.1 beta installation
James> 3) emacs was upgraded to 19.30 (another major enhancement!). when I James> upgraded to 19.29, I had the same problem, namely, emacs looks for James> the file /usr/lib/emacs/19.28/lisp/jka-compr.elc. This problem is James> easily fixed by creating the sym-link 'ln -s 19.30 19.28' in the James> /usr/lib/emacs directory. However, I wish this problem would either James> get fixed, or someone explain to me what I am doing wrong. Emacs sources /usr/lib/emacs/site-lisp/site-start.el, a file which specifies site-specific preferences for modes that should be provided on startup. Following good debian practice, this file is actually softlinked to the file /etc/site-start.el under /etc. Now, the loading path for jka-compr.elc contained, for whatever reason, the absolute path into the emacs lisp directory which depends on the version of emacs. When this was done, it was set to 19.28, and hasn't been fixed since. I had that first changed to (load "jka-compr") which finds it's path automagically, and now, with 19.30, have (auto-compression-mode 1) which provides the same features. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: mailing list as digest?
Matthew> To the point: unfortunately, the volume is very large. I'd much Matthew> prefer to get this mailing list in digest format. Does anyone know Matthew> if this is possible with the list server that debian.org uses? No Matthew> mention is made on their web pages, the place I signed up from. You can use either "procmail" or "mailagent" to filter your incoming mail into different folders. Very, very convenient. Both packages are in the mail directory of the Debian distribution. -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
No more logs since cron.weekly rotation
Files in /var/adm as daemon.log, messages, auth.log don't get updated this morning. Here's the rest of a ls -ltr in /var/adm -rw-r--r-- 1 root adm 0 May 19 06:47 daemon.log -rw-r--r-- 1 root adm 0 May 19 06:47 debug -rw-r--r-- 1 root adm 0 May 19 06:47 daemon.log.0 -rw-r--r-- 1 root adm 0 May 19 06:47 debug.0 -rw-r--r-- 1 root adm 0 May 19 06:47 messages -rw-r--r-- 1 root adm 0 May 19 06:47 messages.0 -rw-r- 1 root adm 0 May 19 06:47 auth.log -rw-r- 1 root adm 0 May 19 06:47 auth.log.0 drwxr-xr-x 2 root root 1024 May 19 06:50 smail -rw-r--r-- 1 root adm 11144 May 19 08:59 wtmp -rw-r--r-- 1 root adm 498 May 19 09:05 diald.log What's even more strange is that the xconsole window has all the relevant information. Here are the last lines: May 19 09:05:33 miles pppd[9109]: pppd 2.2.0 started by root, uid 0 May 19 09:05:33 miles pppd[9109]: Using interface ppp0 May 19 09:05:33 miles pppd[9109]: Connect: ppp0 <--> /dev/ttyS3 May 19 09:05:37 miles pppd[9109]: local IP address 130.15.126.201 May 19 09:05:37 miles pppd[9109]: remote IP address 130.15.126.55 May 19 09:05:37 miles diald[27052]: New addresses: local 130.15.126.201, remote 130.15.126.55. May 19 09:10:07 miles in.comsat[9171]: connect from localhost May 19 09:23:33 miles su: (to root) on /dev/ttyp1 Any idea what's wrong? Is that a sysklogd problem? Or a savelog problem? System info is as this: [EMAIL PROTECTED]:/var/adm> uname -a Linux miles 1.3.98 #1 Tue May 7 16:58:31 EDT 1996 i586 [EMAIL PROTECTED]:/var/adm> w 9:27am up 11 days, 14:59, 1 user, load average: 0.08, 0.05, 0.01 User tty login@ idle JCPU PCPU what edd ttyp0 8:59am w [EMAIL PROTECTED]:/var/adm> dpkg -l syslogd libc5 smail Desired=Unknown/Install/Remove/Purge | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad) ||/ NameVersionDescription +++-===-==- ii syslogd 1.3-2 Kernel and system logging daemons. ii libc5 5.2.18-6 The Linux C library version 5 (run-time libr ii smail 3.1.29.1-21Electronic mail transport system. Thanks for all pointers, Dirk -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Problems with color-ls
Dale> What is in your /usr/local/etc/colour-ls.rc file? That is the usual plain text file that describes which colours are used for which modes, extensions, ... I just didn't like the name DIRCOLORS or whatever is proposed as uppercase is so loud. Dale> I have the eval `dircolors -b` in my /etc/profile my $HOME/.bashrc as Dale> well. It doens't seem to help. It happens when running the Makefile Dale> that changes shells within the Makefile itself. ie It is a /bin/sh Dale> script but calls the /bin/csh in the script. Could that mess things Dale> up? Sure. You can try any of these options (ordered in increasing uglyness): - say "unalias ls" just before the calling make, it will then do a plain old ls without colours (easy) - hardcode calls to ls in the Makefile or script to /bin/ls (ok) - disable the dircolors in /etc/profile and ~/.bashrc (ugly) - rewrite the csh script as sh or bash (very ugly) Hope this helps, -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: Problems with color-ls
Dale Miller writes: Dale> I am working on a package of Lclint which is the free lint for Dale> Linux. Unfortunately I am having problems with color-ls. When I try Dale> to do make test to compile the test cases I get the following error. Dale> dircolors: Unknown shell `csh '. This causes the build to mess Dale> up. Does anyone have any ideas why this happens. If I purge the Dale> color-ls package then everything works fine and the build works. I Dale> like color-ls so I don't want to have to purge it. Any info would be Dale> appreciated. If you need more info answer my question then let me Dale> know. >From dircolors(1): OPTIONS -a, -s Assume the user is using a Bourne-style shell which does not support aliasing. This is default if the base name of the environment variable SHELL is sh or ash. Instead a shell function is used. If the -P option is also used, this function will need to spawn a subshell; this is slow and should be avoided if possible. -b, -k Assume the user is using a Bourne-style shell that supports Korn-style aliasing. This is the default if the base name of the environment variable SHELL is bash or ksh. You may want to specify this option explicitly if your sh is really a more advanced shell, which does support Korn-style aliasing. You have to add the switch to your call of dircolors. I do the following in my /usr/local/etc/profile (which is sourced by /etc/profile): eval `dircolors -b /usr/local/etc/colour-ls.rc` Hope this helps, -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
Re: minicom..
Richard Dansereau writes: Richard> Hi! I'm trying to get minicom properly setup. When I try to run Richard> minicom as a non-priveledged user.. I get the error that /dev/cua1 Richard> permission denied The default permission settings on /dev/cua1 Richard> when you install Debian 0.93R6 is 660. I can get things to work Richard> if I set the permissions to 666 but isn't this somewhat of a Richard> security problem? Is there a secure way of setting this up? Add yourself to the dialout group in /etc/groups. Richard> Also, when I was previously using the slackware distribution Richard> minicom was in color. I currently don't get it in color in an Richard> xterm. Does debian come with a special color xterm? I use "xminicom" which is a script that uses whatever is available on your system. Works with rxvt and the colourised xterm, both are Debian packages. Hope this helps, -- Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd
