fglrx not working for resolution above 1760 (width)
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=450426 am I the only one seeing that bug? fglrx seems to be broken for resolution where width is above 1760 pixels, for several releases (I filed the bug Wed, 07 Nov 2007 02:14:30 -0800), just tried latest version fglrx-driver 8.45.4-1 and it still has the same problem, there's also http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=457298 which was just closed, saying 8.45.4-1 fixes the widescreen problem (not sure if it's the same problem I see, it does not specify what exactly is the problem). any ideas? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
fglrx: openGL broken for screen width greater than 1760 pixels (bug #450426)
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=450426 Since fglrx-driver 8.42.3-2 most of the openGL screen (either full screen or windowed) is corrupted when the width of the openGL window is more than 1760 pixels. The way it looks: top of the window (1/4 or so) looks OK, the rest of the screen is mostly obscured, i.e. some polygons are visible but the screen is mostly black. system: ATI Technologies Inc Radeon R350 [Radeon 9800 Pro] (using the secondary monitor in two monitor setup, primary is off) fglrx-driver-8.42.3-2, fglrx-driver-8.43.2-2 (both have same problem) any ideas? didn't find exactly the same problem on the net (found the one about the corruption of lower right corner which is much different symptom than the problem I have) given that this is openGL problem don't think that open drivers would help, they seem to not offer much 3D acceleration (and radeonhd says it works with R500/R600 only) any ideas? TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Preview of cursors for X
Is there any way to preview the cursor themes for X? I have several cursor packages installed (e.g. chameleon-cursor-theme comixcursors crystalcursors dmz-cursor-theme etc.) and can change from one cursor theme to another using sudo update-alternatives --config x-cursor-theme however I cannot figure out how to preview the cursors. TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: DPMS stops working (ATI, openGl issue?)
jason.public wrote: On 4/9/07, Sven Arvidsson [EMAIL PROTECTED] wrote: On Mon, 2007-04-09 at 01:28 -0700, Erik Steffl wrote: When I start X DPMS is working (monitor is turned off), however after some time it stops working (i.e. monitor is always on). [...] fglrx-driver 8.31.5-1 (proprietary driver) The first thing I would try to eliminate would be the proprietary drivers. Try without them, and see if things improve. Worked for me, going on 4 days. Thanks. hmmm... as far as I can tell the problem is in SDL, I tested it some (openGL programs do not disable DPMS) more plus asked Gish support, found this: http://www.libsdl.org/release/changes-1.2.html The screensaver and DPMS monitor blanking are disabled while SDL games are running under the X11 and DGA video drivers. This behavior will be formalized and selectable in SDL 1.3. can't find more info on this... (I didn't try the free driver yet) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: DPMS stops working (ATI, openGl issue?)
Greg Folkert wrote: On Mon, 2007-04-09 at 20:13 -0700, Erik Steffl wrote: Sven Arvidsson wrote: On Mon, 2007-04-09 at 01:28 -0700, Erik Steffl wrote: When I start X DPMS is working (monitor is turned off), however after some time it stops working (i.e. monitor is always on). [...] fglrx-driver 8.31.5-1 (proprietary driver) The first thing I would try to eliminate would be the proprietary drivers. Try without them, and see if things improve. but free drivers have no openGL so that's not going to help me, as far as I know they are not similar at all so if I figure out that free drivers work it won't shed any more light on proprietary drivers. I checked http://dri.freedesktop.org/wiki/R300 for the status of free driver, seems like the work is progressing but it's not ready yet (as far as I can tell). any other ideas? did anybody else notice that openGL (or somnething) disables DPMS? He said, TRY IT FOR A SHORT TIME. Not forever. It is a tried and true troubleshooting method. Chnage one thing at a time until you find the issue. It might not be the problem. let's say I change it and it works (or doesn't) - what will that help me? I am guessing that the REAL culprit is an OpenGL screensaver. Change your screensaver to just blank the screen and then see if things pan out or not. Also, if you are using the GNOME Screensaver remove it and install the xscreensaver and xscreensavergl packages. There are some inconsistent things that happen with the GNOME version. I use xscreensaver, just tried xset dpms 10, works nicely, I ran openGL screensaver (using xscreensaver-demo, then preview), it still works. So it's not just simple openGL (might be a specific openGL screensaver) tried the same with game of Gish - that seems to disable dpms even though xset q still says: DPMS (Energy Star): Standby: 10Suspend: 700Off: 800 DPMS is Enabled Monitor is On However DPMS is also disabled in similar way even when I do not play Gish (I also sent support request to Gish distributor, they might know something). anyway, Sven posted another replying suggesting that r300 might work pretty well with recent free driver so I guess I'll give it a chance, might be good enough for Gish :-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
DPMS stops working (ATI, openGl issue?)
When I start X DPMS is working (monitor is turned off), however after some time it stops working (i.e. monitor is always on). When I run e.g. xset dpms 60 65 70 it works again (for some time). Only idea I have is that openGL unsets/reset/break dpms (that's about the only graphics weirdness I can think of that happens sort of regularly) Any ideas how to troubleshoot/fix this? system: ATI Technologies Inc Radeon R350 [Radeon 9800 Pro] Apple Cinema connected via DVI kernel 2.6.18 fglrx-driver 8.31.5-1 (proprietary driver) Debian bugs don't mention anything relevant, net search revealed similar (same) problems but no solutions. TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: DPMS stops working (ATI, openGl issue?)
Sven Arvidsson wrote: On Mon, 2007-04-09 at 01:28 -0700, Erik Steffl wrote: When I start X DPMS is working (monitor is turned off), however after some time it stops working (i.e. monitor is always on). [...] fglrx-driver 8.31.5-1 (proprietary driver) The first thing I would try to eliminate would be the proprietary drivers. Try without them, and see if things improve. but free drivers have no openGL so that's not going to help me, as far as I know they are not similar at all so if I figure out that free drivers work it won't shed any more light on proprietary drivers. I checked http://dri.freedesktop.org/wiki/R300 for the status of free driver, seems like the work is progressing but it's not ready yet (as far as I can tell). any other ideas? did anybody else notice that openGL (or somnething) disables DPMS? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [OT/FLAME] Horrible GNOME File Picker (Was: Open (helper application chooser) for iceweasel/icedove is too simple)
Sven Arvidsson wrote: On Wed, 2007-01-10 at 00:27 -0800, Erik Steffl wrote: it takes few minutes to open /usr/bin here (almost no load on machine), next time (I assume cache helps a lot) it takes 10-20 seconds. system: debian unstable icedove 1.5.0.9.dfsg1-1 pentium 2.4 GHz 1GB RAM do you think I should file a bug? against what? where? I found a bug dealing directly with opening /usr/bin, http://bugzilla.gnome.org/show_bug.cgi?id=322314 There is also a few links to performance related bugs here, http://live.gnome.org/GtkFileChooser yeah, this one in particular seems relevant: http://bugzilla.gnome.org/show_bug.cgi?id=310642 However they all seem to think that it's all because the directory is read (to be able to autocomplete etc.) which is not entirely true cause bash autocompletes as well yet I cannot seem to be able to make it stall on directories with few thousand files (local disk). Seems like the problem is getting mime type (as mentioned here before) or something of that nature... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [OT/FLAME] Horrible GNOME File Picker (Was: Open (helper application chooser) for iceweasel/icedove is too simple)
Wim De Smet wrote: ... gnome file chooser discussion snipped ... indeed very annoying) but I don't think it's open()ing every file in those directories. This would require an ordinate amount of processing power not to mention disk I/O which I'm just not seeing. it takes few minutes to open /usr/bin here (almost no load on machine), next time (I assume cache helps a lot) it takes 10-20 seconds. system: debian unstable icedove 1.5.0.9.dfsg1-1 pentium 2.4 GHz 1GB RAM do you think I should file a bug? against what? where? Glad I've found out about ctrl-l and ability to just start typing (thanks to all who responded) but the performance is incredibly bad (compare to bash - there is no noticeable delay between me hitting tab twice and bash asking Display all 3758 possibilities? and it's same for other directories which are not as likely to be cached). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Perl Programming within Debian
David Baron wrote: ... perl discussion snipped ... With Windows, we bought a nice debugger--any one know of such a beast GPLed under Linux? Being able to watch those vairables was a life-saver. A decent IDE would be nice as well (For using .net, one has Microsoft's Visual Studion in Windows--result can run in linux using mono). there's ddd (apt-get install ddd) which is a GUI frontend for other debuggers, for perl it uses perl itself (perl has build-in debugger), it's fairly intuitive (in a sense that it has menus and dialogs for all common debugger functionality) and also gives you direct access to perl debugger. as far as IDE goes - xterms, vim or gvim (emacs?), etags etc. (as far as I can tell IDEs are mostly for environments where you don't have tools that play well together and no _good_ window managers) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Open (helper application chooser) for iceweasel/icedove is too simple
Not sure if it's standard gtk/gnome file open dialog, but it's the one used to pick application when opening an attachement and downloading files (it's probably used in other places as well). The dialog has two parts - left part has shortcut to user home dir, desktop and filesystem, right side has list of files/directories. it is not possible to just write in the file (like it is common with other file open/save dialogs) which is VERY annoying, e.g. if I want to open a text file using gvim I have to figure out where gvim is and click all the way there, instead of just typing gvim (and assume it's in path). is there any way to configure these? are these standard gnome file open dialogs? TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
SOLVED: Re: how to convert DVD to something uploadable to youtube?
thanks to all who responded, in the end here's what I used: dvdrip to rip dvd (got *vob files) mencoder (see the command line quoted below, I just changed the scale=352:288 to be as close as possible to 320x240 which youtube.com wants, keeping the same x to y ratio as original vob file) avidemux to cut the avi (there were quite a few failed attempts, using different codes/programs etc. so hope this will help...) erik Carl Fink wrote: On Wed, Nov 29, 2006 at 11:42:51AM -0800, Erik Steffl wrote: any pointers to solutions? Programs to use? Settings? Howtos? Well, mplayer's mencoder works on VOBs if you're willing to dig through the options. Something like: mencoder -ovc lavc -lavcopts vcodec=msmpeg4 -lavcopts vbitrate=1000 -oac mp3lame -lameopts q=3 -lameopts mode=1 -vf scale=352:288 -o mymoviename.avi VTS_01_01.VOB would produce a good file for YouTube. They recommend DivX video compression and mp3 audio compression and a frame size close to 320x240 because they're going to resize to that, anyway. Maximum length is 100 megabytes, and maximum duration is 10 minutes. Once the file is an AVI, you could use Avidemux to chop it up. Of course, converting a protected DVD is a violation of the US Digital Millennium Copyright Act if done in the USA. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
how to convert DVD to something uploadable to youtube?
I have a DVD (PAL system, not sure if it matters, no copy protection) and I'd like to convert it to something that I can upload to youtube (or just send to friends). I am using debian linux system (unstable). I tried dvdrip and cinelerra (from http://www.kiberpipa.org/~minmax/cinelerra/builds/pentium4/) however got nowhere: - DVD rip by dvdrip ok, I get *vob files I can view with e.g. xine (same quality as DVD so I guess that part worked) - dvdrip transcode - I tried several options but avis are of very low quality - cinelerra - tried to export to avi, the video is all blue with broken timing (slows down at a time) any pointers to solutions? Programs to use? Settings? Howtos? TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
pavuk: does it work for anybody? (segfaults here)
Just tried to use pavuk, it seems pretty unstable just trying to configure the download scenario and then reliably segfaults when starting the download (both with X frontend and commandline). There are bugs already filed (200+ days old) for this (as far as I can tell, they also mention segfaults when beginning the download) with no response from maintainer. Anybody knows more? Anybody was able to run pavuk? Are there any similar spiders (with javascript support, auth support, flexible download options etc.) available that you would recommend? TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: ATI Driver Installer ATI Installer Version: 8.27.10
Mitch wrote: On 08/09/06 00:12 AM, Mathieu Malaterre wrote: Hello, Has anyone tried the new ati driver installer. Here is what I did: Well now all I get is: Mesa GLX Indirect (*). Is there a step that I am missing ? I use those drivers on two machines. I build my modules with make-kpkg, but that's not important. The problem is that fglrx looks for dri drivers in /usr/X11R6/lib/modules/dri/fglrx_dri.so, but it doesn't get installed there. You'll find it in /usr/lib/dri. I'm pretty sure you'll need to mkdir that modules dir and then I just ln -s the dri dir. You can confirm this with 'setenv LIBGL_DEBUG verbose' (tcsh, translate to yours) before running glxinfo. BTW you can also strace fglrxinfo which will reveal what it's trying to open you can specify where openGL programs look for the driver (sh and derivates syntax): LIBGL_DRIVERS_PATH=/usr/lib/dri export LIBGL_DRIVERS_PATH for csh and derivates: setenv LIBGL_DRIVERS_PATH /usr/lib/dri erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SOLVED Re: xorg 7 and ATI proprietary driver (fglrx)
Liam O'Toole wrote: On Tue, 20 Jun 2006 07:20:02 -0500 Ron Johnson [EMAIL PROTECTED] wrote: [SNIP] Is there an ATI Debian package? Yes, but only in unstable. The packages of interest are fglrx-driver and fglrx-kernel-src. The latter can conveniently be built using module-assistant. I use these on my Thinkpad, although they can be a bit flaky. that's what I'm using but I think the ATI change in /etc/profile was from before, back from times when I used their rpm package (converted to deb) overall the quality of the driver and packaging is pretty sad, like freezing if you have more than one X server or not being able to get console back after X server exits, nasty behavior like writing into /etc/profile, aticonfig overwriting (and rearranging in weird ways) your xorg.conf (at least they create a backup) etc. Considering they could get people doing the stuff for free (there is an ongoing effort to create free drivers) it is kinda funny... ventlook how far we got since the days of free and pretty much perfect drivers for voodoo with openGL and everything/vent erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Xorg 7.0 and ATI working!
Bill Wohler wrote: In case there are still folks out there that like me have xserver-xorg 6.8 on hold because of the reports that X would freeze up if you had an ATI card, I'd like to report that all is well. I share the sentiments that the upgrade was a non-event for such a major upgrade (at least now that the problems exposed with the early reports have been fixed). I went into aptitude, unheld xorg-xserver, hit U, and then g. Although I was expecting to have to apt-get install -f a few times, I didn't have to run it once! I have: IBM ThinkPad t40p ATI Technologies Inc Radeon R250 Lf [FireGL 9000] (rev 02) Debian etch I haven't seen any strange behavior in a few days now. I believe I read that the ATI fixes were applied to Xorg 7.1, but it appears that the these fixes were also applied to xserver-xorg-video-ati 6.5.8.0-1 which is available to Debian users of Xorg 7.0. My thanks and appreciation go out to the X Strike Force. A couple of notes: 1. glxgears missing The fix was to install mesa-utils. do you have 3D hw acceleration working? I thought it's not working or VERY experimental in free ati driver 2. The courier fonts were ugly. The fix was to rebuild xorg.conf (I'm assuming with a proper font path). I did not have to resort to the workaround presented in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=367593 some of my fonts look prety ugly too, how did you figure out what the proper font path is? Whatever was there after dpkg-reconfigure xserver-xorg? TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Xorg 7.0 and ATI working!
Liam O'Toole wrote: On Tue, 20 Jun 2006 23:39:52 -0700 Erik Steffl [EMAIL PROTECTED] wrote: Bill Wohler wrote: [SNIP] 2. The courier fonts were ugly. The fix was to rebuild xorg.conf (I'm assuming with a proper font path). I did not have to resort to the workaround presented in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=367593 some of my fonts look prety ugly too, how did you figure out what the proper font path is? Whatever was there after dpkg-reconfigure xserver-xorg? TIA, erik Hi again, Erik. Fonts are looking pretty good here (deja-vu and msttcorefonts, with bitmap fonts enabled too). What problems are you having? pretty much everything looks ok except of fonts in fvwm - window titles, menus, fonts in pager. The menu and window titles look like really badly resized bitmap fonts (before xorg 7 upgrade they looked ok) this is the font specified in fvwm config: default: -adobe-helvetica-bold-r-*-*-12-* menu:-adobe-helvetica-bold-r-*-*-12-* the font in pager is too small for any font to appear good but I think it's somewhat worse than before pager font: -*-times-medium-r-*-*-*-80-*-*-*-*-*-* tried these in xfontsel and they still look pretty bad even though they don't have exactly same look as in fvwm (and there's no 80 size for times, only 120) these look like pretty common fonts, not sure what happened, they are not missing, just ugly. btw what's deja-vu? I assume it's a font or set of fonts but apt-cache found nothing thanks, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
SOLVED Re: xorg 7 and ATI proprietary driver (fglrx)
Liam O'Toole wrote: On Mon, 19 Jun 2006 12:04:58 -0700 Erik Steffl [EMAIL PROTECTED] wrote: [SNIP] ... The next thing I would check is that various symlinks in /usr/lib/ point to the fglrx libraries rather than the mesa ones. You should at least look at /usr/lib/libGL.so* and /usr/lib/libGLcore.so*. The command 'strace fglrx' will help you to identify any others. I checked that before and it looks ok (diversion by fglrx-driver to: /usr/lib/fglrx/diversions/libGL.so.1.2) however I did strace fglrxinfo and here's what it revealed: open(/usr/X11R6/lib/modules/dri//fglrx_dri.so, O_RDONLY) = -1 ENOENT (No such file or directory) it indeed is not in that directory (it's in /usr/lib/dri/fglrx_dri.so). now imagine this, it turns out that ATI tools modified my /etc/profile (2005/07/13 or before): ### START ATI FGLRX ### ### Automatically modified by ATI Proprietary driver scripts ### Please do not modify between START ATI FGLRX and END ATI FGLRX ... doing nasty things with LD_LIBRARY_PATH, LIBGL_DRIVERS_PATH which is why it didn't work. One simple command (unset LIBGL_DRIVERS_PATH) and it works again. thanks a lot for the pointers! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: xorg 7 and ATI proprietary driver (fglrx)
thanks, seems like changing kernel config helped, still no cigar though, xorg log says: (II) fglrx(0): DRI initialization successfull! and lsmod confirms that fglrx is loaded but then fglrxinfo says: display: :0.0 screen: 0 OpenGL vendor string: Mesa project: www.mesa3d.org OpenGL renderer string: Mesa GLX Indirect OpenGL version string: 1.2 (1.5 Mesa 6.4.1) despite the fact the ldd confirms that /usr/lib/libGL.so.1 is used which is a diversion by fglrx-driver to: /usr/lib/fglrx/diversions/libGL.so.1 and the openGL programs are slow so I'm pretty sure software rendering is used. any ideas where to go from here? Would you (or somebody else) mind posting relevant parts of xorg.conf (and perhaps kernel but I guess I have that part working since the kernel module is loaded) TIA erik Richard wrote: On 19/06/06, Erik Steffl [EMAIL PROTECTED] wrote: Is it possible to make ATI proprietary driver fglrx work with the latest xorg packages? I installed fglrx packages (build the kernel module etc.) but it doesn't seem to accept xorg 7: (II) fglrx(0): UMM Bus area: 0xd0acb000 (size=0x07535000) (II) fglrx(0): UMM area: 0xd0acb000 (size=0x07535000) (II) fglrx(0): driver needs X.org 6.8.x.y with x.y = 99.8 (II) fglrx(0): detected X.org 7.0.0.0 (II) Loading extension ATIFGLRXDRI (II) fglrx(0): doing DRIScreenInit drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is -1, (No such device) drmOpenDevice: open result is -1, (No such device) drmOpenDevice: Open failed drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is -1, (No such device) drmOpenDevice: open result is -1, (No such device) drmOpenDevice: Open failed [drm] failed to load kernel module fglrx (II) fglrx(0): [drm] drmOpen failed (EE) fglrx(0): DRIScreenInit failed! (WW) fglrx(0): *** (WW) fglrx(0): * DRI initialization failed! * (WW) fglrx(0): * (maybe driver kernel module missing or bad) * (WW) fglrx(0): * 2D acceleraton available (MMIO) * (WW) fglrx(0): * no 3D acceleration available* (WW) fglrx(0): * * does it mean it's not going to work until ATI releases a new driver with support for xorg 7? Anybody had any success with the free driver (it has at least some 3D experimental features). relevant packages: ii fglrx-control 8.24.8-1 ii fglrx-driver8.24.8-1 ii fglrx-driver-dev8.24.8-1 ii fglrx-kernel-2.6.15jojda8.24.8-1+jojda.0 ii fglrx-kernel-src8.24.8-1 ii xserver-xorg7.0.22 TIA, erik Hi Eric I also had a lot of trouble to get this going on debian ... it seems the fglrx driver doesn't like it when the kernel was compiled with including CONFIG_DRM=y and CONFIG_DRM_RADEON=y ... When you compiled this kernel-module you also have to modprobe this into the kernel modprobe fglrx ... when DRM was already compiled into the kernel this modprobe will fail. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Determinate the directory whithin a script
Paolo Pantaleo wrote: 2006/6/18, Paolo Pantaleo [EMAIL PROTECTED]: I have a script in something like /media/sda1/backup/script.sh since sda1 could be also sda2 or anything, I want to determintate at run-time what is the directory in which the script is located, how can I do? pwd doesn't work, since I cuold call the script from some other direcotry Thnx PAolo -- if you have a minute to spend please visit my photogrphy site: http://mypic.co.nr $0 is just the command line... maybe `pwd` + $0 could do the work... but not if the scipt is in the PATH... I can't use udev... but maybe I'll read something about it Well for now I will assume that the script is launched only from the directory in which it is located... it seems the easyer way. I don't think you can reliably figure out where the script was loaded from, after all it didn't even had to be loaded from file system. guess you know it but just in case: general practice is to put script in some bin dir (~/bin or /usr/local/bin or /opt/myVeryOwnBackupSystem-version/bin or something along those lines) and use a command line argument to specify directory to process (script.sh --dir /media/sda1/backup). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: xorg 7 and ATI proprietary driver (fglrx)
Liam O'Toole wrote: On Sun, 18 Jun 2006 19:16:15 -0700 Erik Steffl [EMAIL PROTECTED] wrote: Is it possible to make ATI proprietary driver fglrx work with the latest xorg packages? Yes. I have fglrx working with xorg and the stock Debian 686 kernel in an up-to-date sid installation. ... The first thing you should do is ignore xorg and check whether the fglrx kernel module can be loaded. What does 'modprobe fglrx' tell you? I removed drm from kernel and now the module is loaded: jojda:/home/erik# lsmod | grep fglrx fglrx 462048 - and even xorg log indicates that everything works, at least I think that's what this message means: (II) fglrx(0): DRI initialization successfull! but it's still not working, fglrxinfo says: display: :0.0 screen: 0 OpenGL vendor string: Mesa project: www.mesa3d.org OpenGL renderer string: Mesa GLX Indirect OpenGL version string: 1.2 (1.5 Mesa 6.4.1) any ideas? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: xorg 7 and ATI proprietary driver (fglrx)
Liam O'Toole wrote: On Mon, 19 Jun 2006 06:19:38 -0700 Erik Steffl [EMAIL PROTECTED] wrote: Liam O'Toole wrote: On Sun, 18 Jun 2006 19:16:15 -0700 Erik Steffl [EMAIL PROTECTED] wrote: Is it possible to make ATI proprietary driver fglrx work with the latest xorg packages? Yes. I have fglrx working with xorg and the stock Debian 686 kernel in an up-to-date sid installation. ... The first thing you should do is ignore xorg and check whether the fglrx kernel module can be loaded. What does 'modprobe fglrx' tell you? I removed drm from kernel and now the module is loaded: jojda:/home/erik# lsmod | grep fglrx fglrx 462048 - and even xorg log indicates that everything works, at least I think that's what this message means: (II) fglrx(0): DRI initialization successfull! but it's still not working, fglrxinfo says: display: :0.0 screen: 0 OpenGL vendor string: Mesa project: www.mesa3d.org OpenGL renderer string: Mesa GLX Indirect OpenGL version string: 1.2 (1.5 Mesa 6.4.1) any ideas? erik Do you load the GLcore module explicitly in the Module section of xorg.conf? If so then remove that line. no I don't, here's the module section of my xorg.conf: Section Module # erik: does not exist: LoadGLcore Loadbitmap Loaddbe Loadddc # erik: nvidia doesn't use it: Load dri Loadextmod Loadfreetype # Load glx Loadint10 # erik: does not exist: Loadpex5 Loadrecord # erik: no symbols found: Load speedo Loadtype1 Loadvbe # erik: does not exist: Loadxie # erik: for ati radeon 9800 # This loads the miscellaneous extensions module, and disables # initialisation of the XFree86-DGA extension within that module. SubSection extmod Optionomit xfree86-dga # don't initialise the DGA extension EndSubSection # This loads the GLX module Loadglx # libglx.a Loaddri # libdri.a EndSection see anything suspicious? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
xorg 7 and ATI proprietary driver (fglrx)
Is it possible to make ATI proprietary driver fglrx work with the latest xorg packages? I installed fglrx packages (build the kernel module etc.) but it doesn't seem to accept xorg 7: (II) fglrx(0): UMM Bus area: 0xd0acb000 (size=0x07535000) (II) fglrx(0): UMM area: 0xd0acb000 (size=0x07535000) (II) fglrx(0): driver needs X.org 6.8.x.y with x.y = 99.8 (II) fglrx(0): detected X.org 7.0.0.0 (II) Loading extension ATIFGLRXDRI (II) fglrx(0): doing DRIScreenInit drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is -1, (No such device) drmOpenDevice: open result is -1, (No such device) drmOpenDevice: Open failed drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is -1, (No such device) drmOpenDevice: open result is -1, (No such device) drmOpenDevice: Open failed [drm] failed to load kernel module fglrx (II) fglrx(0): [drm] drmOpen failed (EE) fglrx(0): DRIScreenInit failed! (WW) fglrx(0): *** (WW) fglrx(0): * DRI initialization failed! * (WW) fglrx(0): * (maybe driver kernel module missing or bad) * (WW) fglrx(0): * 2D acceleraton available (MMIO) * (WW) fglrx(0): * no 3D acceleration available* (WW) fglrx(0): * * does it mean it's not going to work until ATI releases a new driver with support for xorg 7? Anybody had any success with the free driver (it has at least some 3D experimental features). relevant packages: ii fglrx-control 8.24.8-1 ii fglrx-driver8.24.8-1 ii fglrx-driver-dev8.24.8-1 ii fglrx-kernel-2.6.15jojda8.24.8-1+jojda.0 ii fglrx-kernel-src8.24.8-1 ii xserver-xorg7.0.22 TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
drupal package status?
Any ideas what's the status of drupal package in debian? drupal 4.6.x is tagged pending (for about a year), maintainer said he has no time, drupal version 4.7.0 is out now but if there's no update it's going to be hard to update (drupal site recommends to upgrade to 4.7 only from 4.6) ? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Otazka
[EMAIL PROTECTED] wrote: Dobry den. Mam zaujem si nainstalovat Debian do pocitaca ale chcem tam mat aj windows xp profesional ktory mam taktiez uz zakupeny. Akym stylom to treba instalovat tak aby tam boli obidva systemy? ktory prvy a ktory druhy? ja mam uz v notebooku nainstalovanu verziu debian gnu/linux a potrebujem tam dostat este ten windows xp profesional. Co mam teraz spravit? riaditel asks how to install both win xp pro and debian on same computer riaditel: this is english language mailing list so you'd get a lot more responses if you'd write in english (toto je anglicky hovoriaci mailing list, dostal by si o dost viac odpovedi keby si komunikoval po anglicky) not sure how to make win wp stop fighting, I resolved the issue by installing linux on hdc (c:) and set lilo so that it tricks win xp into thinking d: is c: so it happily sits on D: and think it has disk C: available (so whatever it does to MDR etc. is not a problem for LILO). this is LILO entry for windows: other = /dev/hdb2 label = win boot-as = 0x80 # pretend to be C: you can do similar trick using grub, I just don't know how:-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: mouse
sela wrote: I am going to sound like a goof ball but, my mouse roller working in the wrong direction mean scrolling up will take down and vice verse. here is my debian XF86Config-4 , which work wrong: Section InputDevice Identifier Configured Mouse Driver mouse Option CorePointer Option Device/dev/psaux Option Protocol ImPS/2 Option ZAxisMapping 4 5 EndSection and here is my RH XF86config which works just fine with the same mouse Section InputDevice Identifier Mouse0 Driver mouse Option Protocol IMPS/2 Option Device /dev/psaux Option ZAxisMapping 4 5 Option Emulate3Buttons no BTW i am not emulating 3 buttons on the debian.. anyone ? anyone ? :) did you try to replace non-working debian config by working redhat config? (just the mouse section) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: weirdest network problems of my ilfe
[EMAIL PROTECTED] wrote: Hi, Ok after installing the new router that came w/ my dsl i'm having the most mind-boggling internet problems of my life. Here's some background on my network: +Linux box |Linux laptop (wireless) Wireless/Wired Router/Modem+Windows box #1 |Windows laptop +Windows laptop (wireless) Ok, so first off... all of my family's windows boxes are working fine. Everything is just normal. The linux boxen however, have problems. Internet generally works (although some sites (slashdot, gmail, etc) won't load in firefox (for some reason they will in links). I can't connect to yahoo/jabber/msn/etc w/ gaim, but i can w/ centericq. I've been trying to figure out all this for forever and so i tried to tcptraceroute my way to messenger.hotmail.com as an experiment (on port 1863 or whatever it was). It took a while, but the route was traced. The funny thing is that if i start tracing the route, and then start up gaim, i connect just fine. same goes for the sites that didn't load up in my browser... I can't figure out why everything works on the windows, but not on my linux boxen. the network stuff is all assigned by the router (dhcp). What should i do? you might want to try few liveCD distros to see if some of them works differently, if some of them works better see what settings it uses you can use some network monitoring tool to see what's being sent (e.g. tcpdump) so that you can see what's going on, it might give you soem ideas... use ping with different packet sizes to different destinations (your ISP gateway etc.), see response times, packet loss etc. and compare to windows boxes (the ones that work OK). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
SOLVED: Re: cyrus21: Address family not supported by protocol
Henrique de Moraes Holschuh wrote: On Tue, 13 Sep 2005, Erik Steffl wrote: Sep 13 03:06:40 jojda cyrus/master[9483]: unable to create imaps listener socket: Address family not supported by protocol Something gone wrong in your IPV6 (or lack thereof) setup? Tell cyrus to bind explicitly to a ipv4 socket, and the error might go away. Also, make sure you are not refering by name to a service that is not defined in /etc/services. I don't recall the exact error message Cyrus outputs (or, actually, glibc does) in that case. Refer to the service by port number, instead. thanks, further investigation revealed that imap (143) is working, imaps(993) is not. What happened: famd was listening on imaps, after adding proto=tcp4 to imap and imaps in cyrus.conf cyrus complained that address is already in use (too bad it didn't say which one and even worse that it wasn't complaining before I added proto=tcp4). anyway, fam NEWS/Debian.gz mentions the problem and refers to bug http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=277528 hmmm... seems like a pretty bad idea to use well known ports for random daemons... Also seems like it would be really nice for cyrus to actually complain about the problem in the first place (would have saved me few hours). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
cyrus21: Address family not supported by protocol
got the following error message from cyrus after reboot: Sep 13 03:06:40 jojda cyrus/master[9483]: unable to create imaps listener socket: Address family not supported by protocol and cyrus imapd does not listen on 993 (MUAs cannot connect to it). It was working fine before reboot. Other than not being able to listen cyrus seems to be working, the emails are received and stored as usual. system: debian unstable kernel 2.6.11 cyrus21-admin 2.1.18-1 cyrus21-clients 2.1.18-1 cyrus21-common 2.1.18-1 cyrus21-docs 2.1.12-2 cyrus21-imapd 2.1.18-1 shorewall 2.4.3-2 (is this relevant?) any ideas what this means? google didn't find any answers (found same Qs), didn't find any relevant bugs either. TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: installation problems with SATA drives
SALAH NOURI wrote: Hi, did anybody figure out how to install debian on a machine with 2 or more SATA drives? i get the error no partitionable media when booting the installer with the default parameters. and booting with expert26 gives me no common cd-rom drive was detected. - make sure you have recent install, I think you need at least kernel 2.6.9 (or maybe higher, the early 2.6.x kernels didn't work) - you can choose SATA disks to appear as regular IDE disks (/dev/hd*) or SCSI disks (in kernel config: CONFIG_SCSI_SATA=y, plus module for your chipset (lspci should tell you what you have) and SCSI disk support) then check your boot messages, the disks should be recognized... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: gaim/gtk themes
Wayne Sitton wrote: I'm running etch,KDE, and Gaim. Problem is the incomming messages window, the text is too small to read. I've changed the type of gtk theme, and even increased the font size. But it increases the font size for everything but the incoming messages window. does anyone know how to increase the font size on the incoming messages. I looked at Gaim's website and followed their direction on how to change it, but once again it increased the size for everything but the incoming message window. which messages appear small? yours: you can set the font etc. in Tools | Preferences | Intrerface | Conversations | Message text, check the 'Send default formatting with outgoing messages' and use the dialog below it to set the font style/size etc. your buddy: not sure how to set default for that but you can ask it to ignore font sizes (which might or might not help) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
drupal error: cannot instantiate xmlrpc_client (cron)
since latest upgrade of drupal I get the following error from Cron [EMAIL PROTECTED] [ -x /usr/share/drupal/scripts/cron.sh ] /usr/share/drupal/scripts/cron.sh: Fatal error: Cannot instantiate non-existent class: xmlrpc_client in /usr/share/drupal/modules/drupal.module on line 142 I suspect it has something to do with the xmlrpc security fix, anybody else seen it? Just asking before filing a bug report (don't want to file a bug report if it's something wrong on my end). BTW the error is in function that tries to ping the drupal directory server. TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: xorg in sid
Paul Scott wrote: I see that some xorg packages are now in sid. Are there enough packages to switch from xfree86? Are there any problems? the new open gl packages remove the old ones and thus whole bunch of programs: apt-get install libglu1-xorg (or x-window-system-core wich installs libglu1-xorg) and the following packages are removed (on my system, there are probably whole lot more): 3dwm-server amarok-engines amarok-xine armagetron audacity billard-gl bittornado-gui blender bzflag celestia-glut chromium chromium-data csound freeglut3 freeglut3-dev frozen-bubble gliv glutg3 glutg3-dev gtkglarea5 junior-games-gl libclan2-gl libfltk1 libfltk1.1c102 libgle3 libglut3 libglut3-dev libgtkglext1 libpolhem libqt3-mt-dev libsdl-perl libsdl1.2-dev libwxgtk2.4 libwxgtk2.4-python libwxgtk2.5.3 libwxgtk2.5.3-python libxine1 planetpenguin-racer plib1.8.4 pornview qt3-apps-dev qt3-examples race rss-glx snd snd-gtk space-orbit ssystem tipptrainer tipptrainer-data-de tuxkart tuxracer vreng wmanager wxpython2.5.3 xine-ui xlibmesa-dev xlibmesa-glu xlibmesa-glu-dev xlibmesa3 xlockmore-gl xpp xracer xscreensaver-gl xt I guess that's going to be fixed over time... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is 'rename' missing or is it just me?
Paul Johnson wrote: On Friday June 17 2005 12:19 pm, Aarre Laakso wrote: Hello, I run Debian unstable. I noticed today that I don't have a 'rename' command in /usr/bin anymore, I can't recall 'rename' being a command anywhere but MS-DOS. You probably meant 'mv'. there's a fancy rename program that comes with perl, it's managed by alternatives so I guess there are other rename programs but I only have one (how does one find about all alternatives? --list lists only the installed ones) on my machine (unstable): [EMAIL PROTECTED]:~$ which rename /usr/bin/rename [EMAIL PROTECTED]:~$ ls -l /usr/bin/rename lrwxrwxrwx 1 root root 24 Jun 10 01:47 /usr/bin/rename - /etc/alternatives/rename [EMAIL PROTECTED]:~$ ls -l /etc/alternatives/rename lrwxrwxrwx 1 root root 16 Jun 10 01:47 /etc/alternatives/rename - /usr/bin/prename [EMAIL PROTECTED]:~$ dlocate -S /usr/bin/prename perl: /usr/bin/prename [EMAIL PROTECTED]:~$ erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: On IMAP servers
Steve Lamb wrote:
Rogério Brito wrote:
Not only Thunderbird, but other MUAs, independently of what platform you're
confined to use. That's the beauty of IMAP, IMVHO.
That's the beauty of properly implemented IMAP on the client side. I
remember back in my PMMail/2 beta test days ('94-'95?) trying to explain what
a proper implementation of IMAP was. At the time and for a long time
afterwards IMAP was viewed as a glorified POP server and not a means to access
remote mail folders. This is evident in the fact that until Thunderbird no
email client, and I do mean none, did it right. They all failed on the
simplest of tests. Were they able to be configured so their sent-mail folder
was IMAP? Drafts? In every case it wasn't possible In fact many people
never even considered that and often when IMAP was brought up in the context
of what on Linux does IMAP people would point to fetchmail. Fetchmail!
*sigh*
Thunderbird has warts, to be sure, the most notable being no
reply-to-list. But at least it keeps the accounts separate and allows those
accounts to place all mail, even what is traditionally local only on the
IMAP server.
when I started using IMAP (2001/10/27) I tried number of IMAP capable
MUAs and all of them were kinda OK, mutt is the only one that I cannot
make save sent messages in imap folder but I guess I just need to read
the docs (I didn't try much). I don't remember all the email clients I
tried, I used mozilla client mostly, also tried kmail, evolution,
sylpheed etc. and they were able to use imap folder for all special
operations (maybe not all of them, I don't remeber). I remember that in
mozilla the folders for templates, drafts and sent mail had to have
different name (i.e. I still save sent mail in SentMail, drafts in
Unfinished and templates in Sablony in my IMAP account).
are you talking about pre-2k times only? I mean during last four
years imap support seems to be pretty good (and improving). Thunderbird
definitely isn't the first usable MUA, as far as imap support goes.
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: On IMAP servers
Steve Lamb wrote: Erik Steffl wrote: are you talking about pre-2k times only? I mean during last four years imap support seems to be pretty good (and improving). Thunderbird definitely isn't the first usable MUA, as far as imap support goes. Nope. In the past few years I've tried Netscape, TheBat!, Sylpheed-Claws, ok, so how come I was using mozilla email client since 2001/10/27 and saving emails, moving to trash, saving drafts etc.? that qualifies as past few years, right? BTW that's only with cyrus server (my current setup), I was briefly using uw-imap before that and it worked even then (otherwise I'd never switch to imap). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
ATI driver errors after last dist-upgrade ([fglrx:firegl_rmmap] *ERROR* map 0xde28ff10 still in use)
after last aptitude ugg (upgrade of everything) I suddenly see huge number of error message like this in /var/log/syslog: May 17 22:02:37 jojda kernel: [fglrx:firegl_rmmap] *ERROR* map 0xde28ff10 still in use (map_count=1) there is huge number of them (436 between May 18 07:45:31 and May 18 11:55:01) and thery seem to be printed even if there's no openGL app running. relevant system info: debian unstable kernel 2.6.9 Radeon 9800 Pro ati driver: fglrx_4_3_0-8.10.19-1.i386.rpm (via alien) more details: openGL used to work before last upgrade, no error messages. It still works but it freezes for short periods of time (when the messages are written to syslog) so it's not very usable anymore. haven't found much info on google, except of somebody suggesting setting UseFastTLS 2 (instead of default 0) in X config, haven't tried that yet but does anybody know more details? Apparently it's some change in X server that triggers this problem (don't know whether it's ati driver problem or X server problem). any details? TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Windows vs Linux Functionality?
Hugo Vanwoerkom wrote: ... I know nothing about digital cameras. My daughter suddenly shows up with one and I google for the brand (I forget it now) and the site shows only XP info, I tell her: no support for your camera. Was I wrong? You mean I plug it in anyway and it might work? it's fairly likely it would work (as a storage device, i.e. you can get photos from it, some photo cameras also have a PC cam/webcam mode, that probably wouldn't work, as far as ic an tell). of course, you need the USB storage support in the kernel. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: iTunes Replacement for Linux
Mike wrote: I use my ipod frequently. What program would you guys recommend I use in a linux world. apt-get install gtkpod works fairly well, it's only for managing your iPod, not a general music/media management program. you also need kernel support for mac style paritions and hfs+ (if you have mac style iPod) and usb or firewire mass media storage (via scsi emulation). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: aptitude keeps trying to replace my vim-gtk and ftpd
Wayne Topa wrote: Jules Dubois([EMAIL PROTECTED]) is reported to have said: On Wed, 03 Nov 2004 10:54:56 -0500, Wayne Topa wrote: I just ran aptitude and it got The following packages are unused and will be REMOVED. There are 8 packages it wants to remove, one of which is bluefish, which I am using as I run aptitude.(?) ... If you want to keep bluefish, tell aptitude you installed it manually. I didn't install it manually. I had a total system wipeout about a AFAIK aptitude considers the package intentionally installed if you do installation using aptitude. Otherwise it will try to remove the package (if nothing else depends on it). to fix it just hit '+' (ask aptitude to install it) when it tries to remove it. From then on it will know you want the package. I guess the underlying problem is that apt-get doesn't differentiate between packages installed as a dependency and packages installed explicitly by user so aptitude has nothing to decide whether you installed the package. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: esound and alsa not compatible?
Eric Gaumer wrote: On Sun, 2004-10-24 at 22:10 -0700, John L Fjellstad wrote: Eric Gaumer [EMAIL PROTECTED] writes: I think you are confusing the two. Alsa is a sound architecture but esound is a sound daemon. Alsa makes sounds where as esound plays more of a traffic cop role. Bottom line is that they serve two different purposes when dealing with sound. Alsa plays the driver or module role. Someday it might be able to do the job of a sound daemon as well but I don't think this is the intent of the project. I thought the never ALSA could play the role of 'traffic cop'? As in, it can get input from different streams and then merge them before sending it to the soundcard. Or does sound daemons do more? Alsa cannot play multiple audio streams simultaneously. From what I understand, this is more of a hardware limitation than an alsa limitation. They claim that some sound cards can do automatic hardware mixing. If your card can't do this then there is a plugin called dmix that does software mixing (i.e. allow sounds to play simultaneously) . I've never tried it. Just search alsa dmix for plenty of how-to's. I would imagine just using esd or arts would be easier and work flawlessly at the moment. I think a lot of Gnome apps are probably programmed to use esound. Give it a try... it can't hurt. I'm no sound expert by any means so take this with a grain of salt... there's dmix plugin in alsa (for software mixing), see: http://alsa.opensrc.org/index.php?page=DmixPlugin erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: fvwm question..
bing yu wrote: I can not get message from fvwm list, Please forgive me posting here. in my .fvwmrc Key rA MExec exec crxvt-gb so Alt+r can open a crxvt terminal After I open crxvt, I usually do this : $su password:(enter my pwd) # cd /mnt/media/ my question is straitforward(let alone the security problem), Can I write above in the .fvwmrc ? so after I Alt+r, the crxvt can go to [EMAIL PROTECTED] I read the man page Key rA MExec exec crxvt-gb -e su;pwd still , It ask me to imput my password can't you setup your /mnt/media so that you don't need root? something along the lines (one of the following): - make it owned by you - create media group, make /mnt/media/ writeable by group, you member of the group (IIRC you need to relogin after adding yourself to the group) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: usb storage kernel module with 2.4 kernel
Nori Heikkinen wrote:
on Wed, 29 Sep 2004 10:00:19PM -0400, Nori Heikkinen insinuated:
i'm trying to mount my new iPod on my home computer now. at work,
running the 2.6.6 kernel, the only kernel module i needed to
modprobe was usb-storage. here, running the 2.4.26 kernel, i don't
seem to have a module called anything like that, nor will the
computer recognize the iPod when it's plugged in (nothing in
/var/log/{syslog,messages}).
amended: i'm a big dork, and had it plugged into the wrong cable. it
DOES see it:
Sep 29 22:04:42 orange kernel: hub.c: new USB device 00:07.2-2,
assigned address 19
but doesn't know what to do with it:
Sep 29 22:04:42 orange kernel: usb.c: USB device 19 (vend/prod
0x5ac/0x1203) is not claimed by any active driver.
and nothing shows up in /proc/scsi/scsi.
i'm using USB (1.1), fwiw. (but i don't think that matters for this).
yet in my kernel config, i have:
CONFIG_USB_STORAGE=m
so it seems like it should be modprobe-able ...
what am i missing, here?
you also need scsi, not sure what exactly, here's what I have:
CONFIG_SCSI=y
CONFIG_SCSI_PROC_FS=y
CONFIG_BLK_DEV_SD=m
CONFIG_BLK_DEV_SR=m
CONFIG_CHR_DEV_SG=m
CONFIG_SCSI_REPORT_LUNS=y
CONFIG_SCSI_CONSTANTS=y
CONFIG_SCSI_LOGGING=y
CONFIG_SCSI_SPI_ATTRS=m
CONFIG_SCSI_SATA=y
CONFIG_SCSI_ATA_PIIX=y
not all of that is relevant (i.e. SATA is for SATA support) but you
need at least some of them, take a look at kernel config help to figure
out which ones are relevant for you.
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [solved] Re: mounting iPod with USB
Nori Heikkinen wrote: ... /dev/sda /mnt/ipod hfsplus rw,user,noauto,noatime 0 0 ), /var/log/syslog gives the error: Sep 27 16:12:48 homeruns kernel: HFS+-fs warning: Filesystem was not cleanly unmounted, running fsck.hfsplus is recommended. mounting read-only. and then it mounts it read-only, like it says. can't you force it to remount rw? That can often be one with regular HDs. Why don't you run fsck on it? is there no fsck.hfsplus? apt-file didn't find it... there is a way to set the number of mounts after which it requires fsck, can't you set it to higher number? for regular HDs I see different message if they exceed the mount count (it says something to that effect), maybe your iPod really was unmounted improperly. ... homeruns:~# hpmount /dev/sda /mnt/ipod hpmount: /dev/sda: Neither Wrapper nor native HFS+ volume header found (Unknown error 4294967295) which is just wack, afaict, seeing as it was mounting happily with `mount -t hfsplus` up until 32 mounts (and still does, just r-o). why are you mounting sda? generally you don't mount the disk but a partition (that has a filesystem), on my iPod (bought august 2003) the sda3 is the one to mount. I use firewire but that shouldn't make any difference. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: vfat iPod suddenly fails to write [was: Re: [solved] Re: mounting iPod with USB]
Nori Heikkinen wrote: on Tue, 28 Sep 2004 09:02:31AM -0700, Erik Steffl insinuated: Nori Heikkinen wrote: ... /dev/sda /mnt/ipod hfsplus rw,user,noauto,noatime 0 0 ), /var/log/syslog gives the error: Sep 27 16:12:48 homeruns kernel: HFS+-fs warning: Filesystem was not cleanly unmounted, running fsck.hfsplus is recommended. mounting read-only. and then it mounts it read-only, like it says. can't you force it to remount rw? you mean, like, remount -o rw,remount /dev/sda2 (or whatever device)? -- no, i can't. tried that; should have said. That can often be one with regular HDs. Why don't you run fsck on it? is there no fsck.hfsplus? apt-file didn't find it... yeah, i couldn't find one. there is a way to set the number of mounts after which it requires fsck, can't you set it to higher number? hm, probably ... but now i converted it to vfat. for regular HDs I see different message if they exceed the mount count (it says something to that effect), maybe your iPod really was unmounted improperly. it's entirely possible. ... homeruns:~# hpmount /dev/sda /mnt/ipod hpmount: /dev/sda: Neither Wrapper nor native HFS+ volume header found (Unknown error 4294967295) which is just wack, afaict, seeing as it was mounting happily with `mount -t hfsplus` up until 32 mounts (and still does, just r-o). why are you mounting sda? generally you don't mount the disk but a partition (that has a filesystem), on my iPod (bought august 2003) the sda3 is the one to mount. on mine, it's /dev/sda2 ... I use firewire but that shouldn't make any difference. apparently the hfsplus tools figure out which partition on the device given has the data you want for you -- magic! so, that's legit when i mount with -t hfsplus, but not when i do it with -t vfat. now, i'm getting a weird problem, however. after dumping about 200 songs on my newly-vfat'd iPod, i tried to add another batch, and gtkpod crashed, and my kernel told me: Sep 28 12:14:44 homeruns kernel: FAT: unable to read inode block for updating (i_pos 1219121)3FAT: Directory bread(block 76088) failed Sep 28 12:14:44 homeruns kernel: FAT: Directory bread(block 76089) failed Sep 28 12:14:44 homeruns kernel: FAT: Directory bread(block 76090) failed Sep 28 12:14:44 homeruns kernel: FAT: Directory bread(block 76091) failed ... ... and so on, for about 100 lines. now i can't mount it. i get the error: mount: /dev/sda2 is not a valid block device when i try. when this happened the first time, i tried rebooting (i know, bad linux user! shouldn't need to reboot), and that solved the problem until it happened agin. i'm going to try this on a friend's mac tonight, but what's up? do i have a lemon? why is the filesystem suddenly ceasing to be unwritable? you didn't mention it yet: sometime the two finger salute helps, I had few occurences of iPod being unresponsive and sometime it helps (other times reboot of linux machine helped, I think there's something wrong with (at least my setup of) firewire (or somewhere between firewire and scsi). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: mousewheels in 2.4.26 vs 2.6.7
John Summerfield wrote: Hugo Vanwoerkom wrote: Hi Debian! I use the same X Debian package + XF86Config-4 in the 2 kernel versions and the same Sarge release. But... in 2.6.7 my 2 mousewheel A4Tech optical mouse has only the upper wheel active and in 2.4.26 both. So since the only difference is the kernel version, that has to be the difference. [EMAIL PROTECTED]:~$ uname -r 2.6.7-1-686 [EMAIL PROTECTED]:~$ Bus 001 Device 004: ID 09da:0006 A4 Tech Co., Ltd This optical mouse has two wheels. Both work. X configuration might need change, I had to change mine: 2.4.x Option Protocol MouseManPlusPS/2 2.6.x Option Protocol ImPS/2 however my side button (button 6) does not work. The other options is to use evdev driver (I haven't tried it yet): Option Protocol evdev Option Dev Name dev name not sure what exactly would work for you, the way mouse works is a lot different in 2.6.x (instead of raw device kernel recognizes mouse itself, however the device reacts differently depending on which protocol you use to communicate with it). search debian-devel for 'logitech mouse [was: Lost Trust]', messages from 2004/5/31 - 2004/6/8 (read the whole thread, some messages are not relevant to your problem but some of them are quite informative, like the one that mentions evdev). there are also quite a few discussions on kernel mailing list (mostly from around the time 2.6 was released) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: xcdroast
Micha Feigin wrote: On Fri, Sep 10, 2004 at 07:24:09PM -0400, Tom Allison wrote: I have the 2.6 kernel. using this I've assumed I can run ATAPI burners, but XCDroast has heart burn itself... I have 2.6.5 kernel, debian unstable xcdroast warns me tht I should be using scsi devices but other than that it seems to be working (only burned few CDs, didn't compare times). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Executing script on X login
Wim De Smet wrote: On Thu, 9 Sep 2004 15:18:46 -0700 (PDT), Matt Perry [EMAIL PROTECTED] wrote: Is there a file where I can put programs to be run when I log into X via XDM? I tried .Xsession but that just dumps me back out to XDM. That's logical. The .xsession is run as a script. So it starts all programs and if none of these tie it up the script ends and so does your xsession. What I want is for several programs to be run when I log in such as xterm and xmodmap. I *don't* want to start a window manager from this. I want to use the window manager and whatever startup that's provided by the system. That way if the sysadmin changes them I'll see the changes. I think you can use the generic 'x-window-manager' which will point to the selected default. So you could end your .xsession with either 'exec x-window-manager' (though the exec is probably not necessary). it would be nice if it really worked, unfortunately it does not, because alternatives system is, well, not keeping promises or something. problem 1 (that's just something to be careful about): if you just have the default x-window-manager it can be replaced during update and when you restart WM the X session will crash/end. problem 2 (IMO serious alternative issue): even if you set up the x-window-manager to whatever you want manually and alternatives promises to not touch it (see the docs) it actually does change it (it switches the x-window-manager to another WM), I am not sure why because I always found out about it too late. You restart your WM and the X session is gone. I filed the bug, there was a discussion on mailing list (iirc debian-devel) etc. no result (other than nothing:-). not sure if the problem still persists (I am not using x-window-manager anymore). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Executing script on X login
Stefan O'Rear wrote: On Thu, Sep 09, 2004 at 04:47:38PM -0700, Matt Perry wrote: On Thu, 9 Sep 2004, Erik Steffl wrote: problem 1 (that's just something to be careful about): if you just have the default x-window-manager it can be replaced during update and when you restart WM the X session will crash/end. problem 2 (IMO serious alternative issue): even if you set up the x-window-manager to whatever you want manually and alternatives promises to not touch it (see the docs) it actually does change it (it switches the x-window-manager to another WM), I am not sure why because I always found out about it too late. You restart your WM and the X session is gone. Did you use sudo update-alternatives --config or did you manually fiddle the symlinks / priorities? of course I used update-alternatives. it happened few times so I read the docs and all. plus there was a discussion about it on debian-user (or user), so it's not like I was making links myself and blaming alternatives. it might have been solved because I just checked and fvwm2 is still the selected one. It used to be overriden (few times) when some of the other WMs were updated. here's the bug I filed: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=178333archive=yes here's a what google found (search for steffl): http://www.faqchest.com/linux/DEBIAN/debian-03/debian-0303/debian-030301/debian03030504_08492.html http://lists.debian.org/debian-user/2002/12/threads.html erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
famd eating 100% CPU?
I just noticed that famd is eating (almost) 100% CPU, before that I didn't even know such a thing existed. Anybody has any ideas why would it go crazy? top says: PID USER PR NI VIRT RES SHR S %CPU %MEMTIME+ COMMAND 19629 erik 25 0 3616 1364 2572 R 87.4 0.2 510:36.14 famd 2653 root 6 -10 97.7m 52m 20m S 5.6 6.0 660:17.24 XFree86 ... strace says (this is repeated over and over...): --- SIGIO (I/O possible) @ 0 (0) --- write(2, *** overflow sigqueu..., 58) = 58 rt_sigreturn(0x2) = 51 TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: copying text from console to X with gpmdata?
Carl Fink wrote: On Wed, Aug 25, 2004 at 12:20:30PM -0700, Stefan O'Rear wrote: gpmdata only carries mouse clicks, it doesn't do anything with selections. My advice: Use XTerm. I like my 212x78 console. (And I don't need GPM!) I use xterm, but I can't copy from an xterm into a fully graphical application (e.g. OpenOffice.org, FireFox). Am I missing a trick? hey, how is xterm not fully graphical! :-) apart from other solutions offered in the thread sometime the xcutsel helps. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: So you think you are (or wanna be) a hacker
John Summerfield wrote: Jason Rennie wrote: On Thu, Aug 12, 2004 at 10:59:36AM +0800, John Summerfield wrote: So set the ball rolling, here is a snippet from a program I found via freshmeat the other day: configfile = malloc(strlen(getenv(HOME)) + 20); sprintf(configfile,%s/%s,getenv(HOME), cfgfile); Something a bit safer... char *home = getenv(HOME); if (home == NULL || cfgfile == NULL) hittheuseronthehead(); ;-) I like this because it's conservative and clearly coded. There are many short cuts in C which make more concise code. However, the most important code readers aren't compilers, they're people. When the code doesn't work as expected, it's people and not compilers who have to figure out what's wrong, and it's then that clear code becomes more important. Often too, clear code won't have as many bugs in the first place, but then even bug-free code needs changes when requirements change. int sz = strlen(home) + strlen(cfgfile) + 2; char *configfile = malloc(sizeof(char)*sz); sprintf(configfile,%s/%s,home,cfgfile); I'm sure someone can do better (and be more creative :) On Wed, Aug 11, 2004 at 08:31:49PM -0700, Stefan Nicolai O'Rear wrote: * It's very ugly (atleast to me.) Is there such a thing as pretty C code? Or C++ code for that matter? :-) I have some aprehensions about use of extensions as in Stefan's reply because they're non-standard and reduce portability. Is there interest here in more of this kind of thing? IMO for stuff like the above the c++ strings should be used (unless there is some reason to use pure C or you really need every bit of performance boost you can get), even in otherwise C source. There's number of C++ fueatures/libs that can be used withotu going fully OO and IMO using containers (STL) at least somewhat is a lot better that juggling pointers (except of special cases maybe). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote: Erik Steffl [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]... [. . .] My point is that it's really is a choice. The fact is that in majority of cases the heap is the better choise. But, I don't see any inherent reason why you shouldn't use the stack for large data. I said the what? I just provided you with bunch of reasons. And you provided another one (you can get around the problem you described by having the variable declared in highest level function where it's used, in your example above the workspace would be allocated in whoever calls f and would be passed as an argument to f). In a particular situation these might not outweight whatever advantage you get for using stack but that does not mean there are no reasons. stack-is-not-for-large-data is inertia, but it may be better called a myth. By the way, I don't know how hard to keep track of stack use. But, the stack is released when the function is finished. Isn't it, then, harder to keep track of heap use? I've heard a lot about memory leak, which refers to forgetting to 'delete' allocated heap memory. I'm yet to hear stack leak :) Of course, I'm joking. Please don't take this seriously. of course there are reasons to use one or another, I am not advocating heap over stack, you asked for sme reasons why one should be careful about using stack (you wrote: ...shouldn't use much stack ... nobody seems to be able to give a convincing argument ...) so I provided you with some reasons. Obviously, a general rule like that does not apply in all situations... Then we agree! I may have misunderstood what you said. I agree that one should be careful about using stack (your words) but I don't agree that one shouldn't use much stack (mine). The reasons you gave are reasons for being careful about stack, not ones that one should't use it in any case. Why am I arguing this? Because I once used a Unix system where the system administrator imposed a low stacklimit. I wanted to use large stack in a situation I described earlier. Before asking the system administrator to raise the limit, I asked knowledgeable people if there's some inherent reasons against using large stack. The answers I got were that you should use heap for large data, without reference to reasons. That was a long time ago. But even today I sometimes see people saying something like: if your program uses more than 2MB of stack, your program is broken. well, it was a lot worse before, compilers were poor, sometime you could overwrite stack or heap, there was a lot less memory overall etc. So I guess part of the stack-is-bad mindset is just inertia... I forgot to add that Stroustrap says that lot lot of memory management problems can be avoided by simply using variables on the stack and not dynamically allocate it using new (so you're not alone:-), I don't have reference but I guess you'll be able to find it if you want to... If you use STL (or something similar) you can have the containers in the stack and let the containers (better debugged than your code) handle the allocation of the data (that way you get advantage of using heap for large data plus the advantage of stack - code is cleaner/shorter, data are automatically freed without you having to take care of it and you can handle the out of memory error). Not saying that you should use that, just that there are various combinations that can be used, depending on what one needs... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote: Roel Schroeven [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]... [...] It's interesting that some of the new features are (or seem to me to be) mainly for numerical computations, such as the restrict keyword and the builtin complex numbers. Possibly, but isn't this feature's usefullness in numerical computations quite limited if you're restricted to the size of the stack? That's a good question. In fact, most modern Unix/Linux systems allow you to use as large stacksize as you like, roughly speaking. Although the default stakesize limit is often set to a very low value, say, 2MB, each user can raise the limit, unless the system administrator (root) prohibits it. So, no. That feature in question is very useful. Some people believe that the stacksize limit should be kept low (in other words, that you shouldn't use much stack), but, as far as I know, nobody seems to be able to give a convincing argument for that position. I suspect that the posision is just an inertia, a remnant of an old habit. But, I'm ready to be convinced otherwise if somebody has a good argument. stack is something you do not have control over and you have no information about, it's not good to use it for large data - you cannot figure out how big it is, how much you're using, there's nothing you can do if space cannot be allocated etc. (maybe you can figure some of it out but that would be compiler specific, I think) on the other hand when using malloc/free or new/delete you can do _something_ if there's not enough memory. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Postfix-Cyrus-Web-cyradm-HOWTO for Debian
Jacob Friis Larsen wrote: Those of you who installed Postfix and Cyrus. Did you follow this howto? http://www.delouw.ch/linux/Postfix-Cyrus-Web-cyradm-HOWTO/html/index.html Can I see your config files? I am using debian packages (unstable) for both cyrus2 and postfix, here are the relevant parts (note lmtp transport instead of cyrus): /etc/postfix/main.cf mailbox_transport = lmtp /etc/postfix/transport: # using sockets because they do not require auth, I # couldn't figure out how to do the authentification:-) localhost lmtp:unix:/var/run/cyrus/socket/lmtp zasran.com lmtp:unix:/var/run/cyrus/socket/lmtp jojda.zasran.comlmtp:unix:/var/run/cyrus/socket/lmtp localhost.zasran.comlmtp:unix:/var/run/cyrus/socket/lmtp /etc/postfix/master.cf # the following line should be there already (IIRC) lmtp unix - - n - - lmtp /etc/cyrus.conf: # IIRC this line is already there, you might need # to uncomment it lmtpunixcmd=lmtpd listen=/var/run/cyrus/socket/lmtp prefork=0 maxchild=20 /etc/imapd.conf: # IIRC already there, must match the listen argument # in /etc/cyrus.conf lmtpsocket: /var/run/cyrus/socket/lmtp end of config quotes --- I think the aliases don't work if you use lmtp (I am just figuring it out), you need to use virtual aliases (defined in /etc/postfix/virtual?), there is some info about aliases and virtual aliases on postfix mailing lists (see postfix.org) as far as I can tell that's it, let me know if you need more info erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote:
Erik Steffl [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]...
Ryo Furue wrote:
[. . .]
That's a good question. In fact, most modern Unix/Linux systems allow
you to use as large stacksize as you like, roughly speaking. Although
the default stakesize limit is often set to a very low value, say, 2MB,
each user can raise the limit, unless the system administrator (root)
prohibits it. So, no. That feature in question is very useful.
Some people believe that the stacksize limit should be kept low
(in other words, that you shouldn't use much stack), but, as far as
I know, nobody seems to be able to give a convincing argument for that
position. I suspect that the posision is just an inertia, a remnant
of an old habit. But, I'm ready to be convinced otherwise if somebody
has a good argument.
stack is something you do not have control over and you have no
information about, it's not good to use it for large data - you cannot
figure out how big it is, how much you're using, there's nothing you can
do if space cannot be allocated etc. (maybe you can figure some of it
out but that would be compiler specific, I think)
on the other hand when using malloc/free or new/delete you can do
_something_ if there's not enough memory.
Right. But, that depends on your design goal. There are many applications
where the heap is the right choise for large data. But, there are others
where the stack is better.
First of all, there are many cases where the heap is the only choise, as
when your function needs to return the pointer to the space allocated
within it. So, the real choise arises in cases like this:
int f() {
//. . .
double workspace0[N]; // stack
double* workspace1 = new double[N]; // heap; Which is better?
//. . .
delete[] workspace1;
}
As you say, if there's _somthing_ to do when there is not enough memory,
the heap is the right choice. That something may be, and in many cases,
is, just to print an error message and quit.
If, on the other hand, you don't care if your program crashes in case
of memory shortage, the stack is the better choice. One obvious reason
is that the stack is the faster. Another is that you can forget about
'delete'ing the allocated space; deallocation is automatic for the stack.
Many scientific applications are like this: If it doesn't run, it doesn't.
That's fine. Use a larger computer.
My point is that it's really is a choice. The fact is that in majority
of cases the heap is the better choise. But, I don't see any inherent
reason why you shouldn't use the stack for large data. I said the
what? I just provided you with bunch of reasons. And you provided
another one (you can get around the problem you described by having the
variable declared in highest level function where it's used, in your
example above the workspace would be allocated in whoever calls f and
would be passed as an argument to f). In a particular situation these
might not outweight whatever advantage you get for using stack but that
does not mean there are no reasons.
stack-is-not-for-large-data is inertia, but it may be better called
a myth.
By the way, I don't know how hard to keep track of stack use.
But, the stack is released when the function is finished. Isn't
it, then, harder to keep track of heap use? I've heard a lot about
memory leak, which refers to forgetting to 'delete' allocated
heap memory. I'm yet to hear stack leak :)
Of course, I'm joking. Please don't take this seriously.
of course there are reasons to use one or another, I am not
advocating heap over stack, you asked for sme reasons why one should be
careful about using stack (you wrote: ...shouldn't use much stack ...
nobody seems to be able to give a convincing argument ...) so I provided
you with some reasons. Obviously, a general rule like that does not
apply in all situations...
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Migration: Mozilla Mail to Mozilla Thunderbird
Aaron B wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Wednesday 04 August 2004 11:16 am, Jerome BENOIT wrote: Hello List, is there a secure (and easy) to migrate from Mozilla Mail to Mozilla Thunderbird ? Thanks in advance, Jerome Yes, in fact, Mozilla Mail and Mozilla Firefox use the same mail format, so you can simply copy your mail folder from one profile to another. First you need to find where your mozilla profile was stored (probably something like ~/.mozilla/foo/foo/mail) and then copy that folder into your thunderbid profile's mail folder (probably something like ~/.thunderbird/foo/foo/mail) If that doesn't work, do a google search- it's a pretty common operation that's probably covered in detail on mozilla forums. or use IMAP and don't worry about email format for various mail clients (and use number of mail clients at the same time, access mail remotely etc.) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Postfix-Cyrus-Web-cyradm-HOWTO for Debian
Jacob Friis Larsen wrote: Those of you who installed Postfix and Cyrus. Did you follow this howto? http://www.delouw.ch/linux/Postfix-Cyrus-Web-cyradm-HOWTO/html/index.html Any tips? I am using lmtp for delivery, I think it is required if you want to use sieve. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
postfix aliases ignored
it looks like postfix (for a long time, it's not a new problem) aliases are not working on my system, all aliases are ingored, any pointers appreciated, here are the details: debian unstable postfix 2.1.4-4 relevant lines in /etc/postfix/main.cf: alias_maps = hash:/etc/aliases alias_database = hash:/etc/aliases relevant line in /etc/aliases (erik is a user on the system, can receive email from both local machine and from outside, mail erik works): www-data: erik I ran newaliases which according to man 5 aliases should rebuild the /etc/aliases.db and it looks like it did (I mv'd the old one of the way and the new one was rebuilt (see Aug 2 date): jojda:/etc# ls -l aliases* -rw-r--r-- 1 root root 1192 Jun 15 01:39 aliases -rw-r--r-- 1 root root 12288 Aug 2 12:55 aliases.db -rw-r--r-- 1 root root 12288 Jul 3 06:04 aliases.db.20040802 -rw-r--r-- 1 root root 12288 Apr 20 2003 aliases.db.db3.orig -rw-r--r-- 1 root root 12288 Jun 15 01:40 aliases.db.org I restarted both postfix and cyrus (I don't know when the alias/db is read). looks like postfix is still trying to deliver to www-data (after Aug 2 12:55), cyrus complains: Aug 2 13:00:01 jojda postfix/lmtp[6880]: 224D2C80D0: to=[EMAIL PROTECTED], relay=/var/run/cyrus/socket/lmtp[/var/run/cyrus/socket/lmtp], delay=0, status=bounced (host /var/run/cyrus/socket/lmtp [/var/run/cyrus/socket/lmtp] said: 550-Mailbox unknown. Either there is no mailbox associated with this 550-name or you do not have authorization to see it. 550 5.1.1 User unknown (in reply to RCPT TO command)) any ideas? TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote: Erik Steffl [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]... [. . .] it's [1]. debian has very slow release cycles, but unstable is much better quality than you'd expect from name. In general I see unstable being equal to latest releases of other distros (both features and stability) while stable is extremely (the kind you need for servers) but does not have all the latest features. Thanks for the answer. Does that mean that *ordinary* users should use the unstable and that the stable should be used only by those who want extremely stable mission-critical servers? By an ordianry user, I mean somebody like this: She can become a root and shut down or restart her Debian work station; she can use apt-cache and apt-get to install packages but is not sure how to configure /etc/apt/source.list; When she wants an application which needs to be build from source, she asks thesysadmin to do that; she's never looked into /var/log; etc. I think an ordianry Windows user is like this. Most of the Linux/Unix users around me are like this. (Please don't say they should use Windows if they can't admin their Unixes :-) They seem to be happy with the situaltion and they are certainly productive in their own field.) that's really hard to tell. it's up to individual user, there are very few serious bugs but sometime the dependency and/or install scripts break which apt (dpkg) have hard time to deal with so it might be hard to upgrade the system (if you now the package is broken it's easy to put it on hold but if you don't know it it's not completely trivial to clean up). during times of transition of some fundamental pieces (gcc, libc, new major version of gnome, kde) things are more shaky than usual. it's kinda sad but I don't see (in general) how an ordinary user (as defined above) could use debian as a desktop - for desktop use (as a home copmputer etc.) you need the latest software (because desktop software is only now becoming mature enough), support for latest HW (USB and firewire devices, video card etc.) which is only available for unstable. And unstable is just about two inches beyond the reach of an ordinary user (as described above). you don't need to be a real sysadmin to use unstable but you need to know more about packaging system (at least one tool - apt-get, aptitude, dselect etc.) than you wanna know:-) then again, it's easy to try it (I've heard that new installer is really a lot better than the old one that scared at least some people) so any ordinary user can judge by herself... I don't see how this is different in windows or redhat or anything else. [. . .] But, vendors don't sell Windows software that doesn't work on the current version of Windows XP. Perhaps, in your perception, of course they do, do you remember example of game I bought and couldn't make it work on win xp pro while it was running fine on win98 (that was few month ago, win xp was out for quite some time) [the game is RC Cars published by Whiptail Interactive, developed by Creat Studio, relese date 12/11/2003] Um, I didn't think of that possibility :-) Weren't we talking about situations where the target OS of the software is too new? The Intel compiler doesn't run correctly on my workstation because the stable Debian is too old. That was the begining of this subthread. I don't think Vendors sell software which runs only on a preview (beta) version of Windows. And I *thought* it's not _exactly_ the same but my point was that windows is a family of OSs that are very similar, just like linux distros (and you have a LOT more control over linux, e.g. you can make debian look a lot like redhat for an application). (but you pointed out that I was wrong) that the unstable Debian roughly corresponded to a beta version of the OS. well, opinions differ but I'd say that unstable is about the same as latest release, stable is about the same as previous release of most other OSs/distros (e.g. no software is added to stable). Maybe it would be more descriptive of it's role to call it 'current'. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How do I read a Mac disk on Linux.
John Summerfield wrote: Erik Steffl wrote: The 2.6 kernel I'm running here looks like it will work. THere's a 2.6 kernel there too, but it's in a different box. If I have to go onsite, I might as well use OSX on the powerbook. However, I _can_ build a Sarge 2.4 kernel. Maybe. My efforts at building Sarge packages on Woody haven't met universal success. stable has kernel source up to 2.4.19 - isn't it enough? Myabe you just need unstable... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How do I read a Mac disk on Linux.
John Summerfield wrote: Erik Steffl wrote: John Summerfield wrote: Erik Steffl wrote: The 2.6 kernel I'm running here looks like it will work. THere's a 2.6 kernel there too, but it's in a different box. If I have to go onsite, I might as well use OSX on the powerbook. However, I _can_ build a Sarge 2.4 kernel. Maybe. My efforts at building Sarge packages on Woody haven't met universal success. stable has kernel source up to 2.4.19 - isn't it enough? Myabe you just need unstable... _just_ unstable?? I tried the three most recent Debian kernels to no avail. I guess I _should_ have just coaxed a prebuilt kernel into place. well, woody might have some tools that are too old... not sure what's the problem... unstable should be current enough for current kernel (I have 2.6.5) and it's just need, not just unstable:-) I don't know just when the hfsplus driver first appeared: I just assumed it's not in the Woody kernel coz it's too new and so I just went for the newest kernel I thought would be easy, without going to backports. I used it somewhere around 2.4.20, don't remember exactly which one (2.4.23?) so 2.4.19 might be good enough... Anyway, the time for talking is past and I'll be attaching it to my Powerbook later today, and after that the drive will be reformatted for Linux. come on, there's no need for such give-up attitude! :-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How do I read a Mac disk on Linux.
John Summerfield wrote: I have an external (USB) disk that has Linux files iin a Mac HFS+ disk. It's connected to my IA32 peecee: Parted sees it thus: Echidna:~# parted /dev/sda GNU Parted 1.4.24 Copyright (C) 1998, 1999, 2000, 2001, 2002 Free Software Foundation, Inc. This program is free software, covered by the GNU General Public License. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. Using /dev/sda Information: The operating system thinks the geometry on /dev/sda is 30401/255/63. Therefore, cylinder 1024 ends at 8032.499M. (parted) print Disk geometry for /dev/sda: 0.000-238475.179 megabytes Disk label type: mac MinorStart End Filesystem Name Flags 1 0.000 0.031 Apple 3128.031 238475.171 Apple_HFS_Untitled_2 (parted) I've installed hfsplus and hfsutils but I don't see how to access it. The peecee is running Woody. Is there a way to access the data on the disk w/o attaching it to my powerbook? (using the powerbook entails a visit to the site). did you try to mount the disk? from partition table it looks like this might work: mount -t hfsplus /dev/sda3 /mnt/macDisk you need mac partition support in kernel (looks like you have that otherwise parted would not recognize partitions, I guess) and hfsplus filesystem support in kernel. You also need support for USB mass storage (these appear as scsi disks, e.g. /dev/sda) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How do I read a Mac disk on Linux.
John Summerfield wrote: Erik Steffl wrote: John Summerfield wrote: I have an external (USB) disk that has Linux files iin a Mac HFS+ disk. It's connected to my IA32 peecee: Parted sees it thus: Echidna:~# parted /dev/sda GNU Parted 1.4.24 Copyright (C) 1998, 1999, 2000, 2001, 2002 Free Software Foundation, Inc. This program is free software, covered by the GNU General Public License. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. Using /dev/sda Information: The operating system thinks the geometry on /dev/sda is 30401/255/63. Therefore, cylinder 1024 ends at 8032.499M. (parted) print Disk geometry for /dev/sda: 0.000-238475.179 megabytes Disk label type: mac MinorStart End Filesystem Name Flags 1 0.000 0.031 Apple 3128.031 238475.171 Apple_HFS_Untitled_2 (parted) I've installed hfsplus and hfsutils but I don't see how to access it. The peecee is running Woody. Is there a way to access the data on the disk w/o attaching it to my powerbook? (using the powerbook entails a visit to the site). did you try to mount the disk? from partition table it looks like this might work: mount -t hfsplus /dev/sda3 /mnt/macDisk you need mac partition support in kernel (looks like you have that otherwise parted would not recognize partitions, I guess) and hfsplus filesystem support in kernel. You also need support for USB mass storage (these appear as scsi disks, e.g. /dev/sda) USB is fine. I think parted interprets the data it reads. Kernel sucks: Echidna:~# mount -t hfsplus /dev/sda3 /mnt/cdrom/ mount: fs type hfsplus not supported by kernel Echidna:~# ls -l /lib/modules/2.4.18-1-686/kernel/fs/ total 156 drwxr-xr-x2 root root 4096 Apr 27 14:13 adfs drwxr-xr-x2 root root 4096 Apr 27 14:13 affs drwxr-xr-x2 root root 4096 Apr 27 14:13 autofs drwxr-xr-x2 root root 4096 Apr 27 14:13 autofs4 drwxr-xr-x2 root root 4096 Apr 27 14:13 bfs -rw-r--r--1 root root 6788 Apr 15 04:02 binfmt_aout.o -rw-r--r--1 root root10008 Apr 15 04:02 binfmt_misc.o drwxr-xr-x2 root root 4096 Apr 27 14:13 coda drwxr-xr-x2 root root 4096 Apr 27 14:13 efs drwxr-xr-x2 root root 4096 Apr 27 14:13 ext2 drwxr-xr-x2 root root 4096 Apr 27 14:13 ext3 drwxr-xr-x2 root root 4096 Apr 27 14:13 fat drwxr-xr-x2 root root 4096 Apr 27 14:13 freevxfs drwxr-xr-x2 root root 4096 Apr 27 14:13 hfs drwxr-xr-x2 root root 4096 Apr 27 14:13 hpfs drwxr-xr-x2 root root 4096 Apr 27 14:13 intermezzo drwxr-xr-x2 root root 4096 Apr 27 14:13 isofs drwxr-xr-x2 root root 4096 Apr 27 14:13 jbd drwxr-xr-x2 root root 4096 Apr 27 14:13 lockd drwxr-xr-x2 root root 4096 Apr 27 14:13 minix drwxr-xr-x2 root root 4096 Apr 27 14:13 msdos drwxr-xr-x2 root root 4096 Apr 27 14:13 ncpfs drwxr-xr-x2 root root 4096 Apr 27 14:13 nfs drwxr-xr-x2 root root 4096 Apr 27 14:13 nfsd drwxr-xr-x2 root root 4096 Apr 27 14:13 nls drwxr-xr-x2 root root 4096 Apr 27 14:13 ntfs drwxr-xr-x2 root root 4096 Apr 27 14:13 qnx4 drwxr-xr-x2 root root 4096 Apr 27 14:13 ramfs drwxr-xr-x2 root root 4096 Apr 27 14:13 reiserfs drwxr-xr-x2 root root 4096 Apr 27 14:13 romfs drwxr-xr-x2 root root 4096 Apr 27 14:13 smbfs drwxr-xr-x2 root root 4096 Apr 27 14:13 sysv drwxr-xr-x2 root root 4096 Apr 27 14:13 udf drwxr-xr-x2 root root 4096 Apr 27 14:13 ufs drwxr-xr-x2 root root 4096 Apr 27 14:13 umsdos drwxr-xr-x2 root root 4096 Apr 27 14:13 vfat Echidna:~# The 2.6 kernel I'm running here looks like it will work. THere's a 2.6 kernel there too, but it's in a different box. If I have to go onsite, I might as well use OSX on the powerbook. However, I _can_ build a Sarge 2.4 kernel. Maybe. why wouldn't you be able to build the kernel? use the existing kernel config and make oldconfig, just add the mac partition support and hfsplus (you need BOTH for the disk to work) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote: Erik Steffl [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]... Ryo Furue wrote: [...] In fact I looked at the homepage of nptl. That was kind of scary to me :) The page says that you need to use a rather new kernel and libc. [...] I don't think you can expect a really old version of debian to support new software. If you want to use new software use new software. Excuse me, but I interpret what you say in two ways: 1) The current stable version (which I use) is old; 2) You thought I was using an obsolete version (potate or earlier). I don't know which is the correct interpretation, although your reference to unstable seems to suggest (1). I'm using the newest stable version. :) it's [1]. debian has very slow release cycles, but unstable is much better quality than you'd expect from name. In general I see unstable being equal to latest releases of other distros (both features and stability) while stable is extremely (the kind you need for servers) but does not have all the latest features. unstable/stable terms don't only mean how many bugs and how robust the distros are, it also refers to the feature set - stable does not change anymore (bugfixes, including security fixes, are released but that's it), unstable changes all the time - new software and new releases of existing software. in lot of cases debian stable is simply too old, sometime you can install whatever it's missing yourself, sometime it's better to go to unstable. There are also few options in between (there are new non-standard debs available (I forgot where, you'd have to ask/search), you can install _some_ packages from unstable distro etc.) I don't see how this is different in windows or redhat or anything else. But, vendors don't sell Windows software that doesn't work on the current version of Windows XP. Perhaps, in your perception, of course they do, do you remember example of game I bought and couldn't make it work on win xp pro while it was running fine on win98 (that was few month ago, win xp was out for quite some time) [the game is RC Cars published by Whiptail Interactive, developed by Creat Studio, relese date 12/11/2003] the unstable version of Debian is the current version? :) If that be so, the name unstable is misleading, which has scared me off thus far. well, it's a topic of long discussions:-) debian is well known for releasing new stable versions fairly infrequently so for many people the stable version is just not that useful - no support for latest HW, old versions of software... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote: Steven Jones [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]... Working in a MS, Solaris, Linux, Tru64 shop, I find that for the vast majority of our servers the usability of Linux is as good as Unix if not better. While Unix might have high end bits Linux lacks for 95% of the world's servers that small missing % I suspect is not an issue. This is not a Linux-vs-Unix issue, but I've recently been experiencing a downside of Linux. I think one of the biggest problems for developers of commercial software for Linux is that there's no such thing as the Linux OS. There are simply too many combinations of the kernel version, libc version, pthreads version, etc. to support all. The consequence is usually the vendor supports only the RedHat Linux. I'm using the Intel Fortran Compiler (IFC). Its version 7 runs on Debian without any problem whatsoever, although Intel doesn't support Debian. But, last year Intel released a total rewrite of the compiler, version 8, with which my Fortran programs don't work at all (*). Since Debian isn't supported, about particular software (nptl thread library) not being available for woody: why not install it yourself? just because it's not available as debian package does not mean you cannot install it. I can imagine that this might not be possible in all cases but in general this approach solves lot of portability problems (from one linux distro/version to another linux or even unix distro/version). even if I paid (which I don't), Intel wouldn't fix my problem. (If paying would fix it, I would pay.) This is a big headache. Uniformity is sometimes good. I also heard from a programmer that her company develops software only for Windows because it's so uniform and ubiguitous. Her company, being small, wouldn't be able to support Linux. If a costomer doesn't have a Windows machine, the company makes the costomer buy one. (The sofwares so expensive that the cost of a lowly Windows machine is nothing.) Unfortunately, uniformity and community efforts don't come together. I don't understand it. if you (they) think that it is acceptable to buy computer with windows just to make sure that the client has OS that software company supports why wouldn't it be acceptable to buy computer with redhat linux? if you are willing to have a dedicated machine then you can have a dedicated machine that plays nicely with other linux machines... depending on the circumstances (not all solutions are possible/desirable in all situations): - statically link everything - provide your own shared libs (do not install in system dirs!) - have an experienced sysadmin set up the machine - develop portable software, do not depend on random quirks of kernels/libs uniformity and community don't come together??? compare linux distros to unix versions! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Kent West wrote: Erik Steffl wrote: Ryo Furue wrote: I think one of the biggest problems for developers of commercial software for Linux is that there's no such thing as the Linux OS. There are simply too many combinations of the kernel version, libc version, pthreads version, etc. to support all. The consequence is usually the vendor supports only the RedHat Linux. about particular software (nptl thread library) not being available for woody: why not install it yourself? just because it's not available as debian package does not mean you cannot install it Sure, that works for sysadmins who know what they're doing. But that's not going to work for the masses. Which means the vendor must put in extra effort to get things to work. The result is that the vendor chooses not to go down that path, and Linux remains a niche product rather than moving into the mainstream, due to lack of developers developers developers developers . . . (courtesy of Steve Monkey Boy Balmer). let's see another example. I bought a game for windows (few month ago, newly released game). Mainstream one, not a relatively obscure intel fortran compiler. it didn't work on my game machine (win xp pro). I sent an email to support, they said I need to run as non admin, I tried that, did work even less, in the end it would not even uninstall. I wish there was a happy ending to this, in the spirit of 'sure, that works for sysadmins'. BTW the game ran fine on another computer that ran win 98 (IIRC). so there you go, windows will remain a niche product blah blah blah... I also heard from a programmer that her company develops software only for Windows because it's so uniform and ubiguitous. I don't understand it. if you (they) think that it is acceptable to buy computer with windows just to make sure that the client has OS that software company supports why wouldn't it be acceptable to buy computer with redhat linux? You keep looking at this from the customer's point of view; the customer do you work for MS? why did you snipped my suggestion on how to deal with the issue FROM VENDOR SIDE and then you complain I don't look at it from vendor side? quote from my email you responded to: - statically link everything - provide your own shared libs (do not install in system dirs!) - have an experienced sysadmin set up the machine - develop portable software, do not depend on random quirks of kernels/libs 3 out of 4 are advices for a vendor (because the guy asked from the vendor perspective). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Ryo Furue wrote: Hi, Since Kent West has kindly clarified what I wanted to say, I'm not going to repeat my main point. Only the following: Erik Steffl [EMAIL PROTECTED] wrote in message news:[EMAIL PROTECTED]... [...] about particular software (nptl thread library) not being available for woody: why not install it yourself? just because it's not available as debian package does not mean you cannot install it. I can imagine that this might not be possible in all cases but in general this approach solves lot of portability problems (from one linux distro/version to another linux or even unix distro/version). In fact I looked at the homepage of nptl. That was kind of scary to me :) The page says that you need to use a rather new kernel and libc. So, the installation of nptl would mean: Back up everything, learn how to build kernel since I've never done that before (my machine is dual-processor, and my kernel was built by another guy), learn how large an impact upgrading libc will mean, and so on. If I had a leisure few days, I might try. But, if I had a few days off, I'd rather spend the time away from the computer screen! I've had too much these days in front of my workstation. I don't think you can expect a really old version of debian to support new software. If you want to use new software use new software. It looks like libc6-i686 in unstable supports nptl: jojda:~ apt-cache show libc6-i686 ... Version: 2.3.2.ds1-13 ... This package includes support for NPTL. ... I don't see how this is different in windows or redhat or anything else. Yes it would be nice if everything worked with everything but that's not possible... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Is Linux Unix?
Kent West wrote: Erik Steffl wrote: Kent West wrote: The result is that the vendor chooses not to go down that path, and Linux remains a niche product rather let's see another example. snip so there you go, windows will remain a niche product blah blah blah... Calling a product that has 3% of the market a niche product seems to make sense to me. Calling a product that has 90% of the market a niche product doesn't make sense to me. it's not about the percentage but about the line of reasoning plus a little bit of kidding on the side. anectodal evidence against anecdotal evidence - not much reasoned in the end, what my example was meant to imply is that you cannot generalize from one example (intel fortran does not determine fate of linux) not very strong argument but I also wanted to suggest that version incompatibility _exists_ in other OSs (that are not really niche, as you point out) also note that linux is niche on desktops but not on servers, the original poster didn't talk specifically about desktops so one could also argue that linux already is not niche product. You keep looking at this from the customer's point of view; the customer do you work for MS? Nope. why did you snipped my suggestion on how to deal with the issue FROM VENDOR SIDE and then you complain I don't look at it from vendor side? Because I mistakenly was overly aggressive in my snipping, and because I misunderstood what you were saying. Up until this point, it seemed to me that you were looking at things from the customer viewpoint rather than the vendor viewpoint Here's your paragraph just before your bullet points: I don't understand it. if you (they) think that it is acceptable to buy computer with windows just to make sure that the client has OS that software company supports why wouldn't it be acceptable to buy computer with redhat linux? if you are willing to have a dedicated machine then you can have a dedicated machine that plays nicely with other linux machines... The paragraph seems to start out from the customer's viewpoint; I assumed that it ended from the customer's viewpoint. After re-reading the paragraph in light of your response, I'm no longer sure that it starts from the customer's viewpoint (it was the if you ... think that it is acceptable to buy computer... that made me think that). oh, that was a response to particular case that original poster mentioned, about the comapny that makes customer buy a machine for their software. Immediately after this paragraph was: depending on the circumstances (not all solutions are possible/desirable in all situations): - statically link everything - provide your own shared libs (do not install in system dirs!) - have an experienced sysadmin set up the machine - develop portable software, do not depend on random quirks of kernels/libs Not being a developer/programmer, I did not pay enough attention to these bullet points to recognize that three of these things are vendor-oriented. There was no intended slight on what you had to say; merely ignorance/misunderstanding on my part. I apologize. oh ok, it wasn't connected to the previous, it was my take on what vendor could do, I guess I could have titled it appropriately so that it's clear. - Karsten made the very interesting point in one of his posts that the entire Oracle enterprise database software was ported to Linux with a single make command (although Paul indicates that the reference is not easily found, and I haven't bothered searching for it). I too remembered oracle but can't find it, however I have found (Informix and linux port: we just typed make): http://slashdot.org/articles/980801/0948200.shtml ... Part of this perception stems from earlier attempts with installing VMWare on a Debian box. The installer had to compile kernel modules, which involved me getting the headers (nope, that wasn't enough; get the full source), etc, and even then, the version of VMWare I last tinkered with wouldn't work with the newer kernel (2.6 maybe? it's been a while) that I had recently installed. So in that experience, VMWare wouldn't work on my particular version of Linux. 1) Could the vendor have made the installation more bullet-proof? 2) Without lots of resources to do so? well, WMWare is not a good example (to generalize from, it's fairly unique) because it is very tricky to emulate a processor, it is much closer to kernel than most other apps. There are few classes of problems that you can run into and there are different solutions for each: - specific kernel features: most apps should not care about kernel but some do - e.g. real time (or almost real time) apps (multimedia, games), apps using unstable (I mean changing API, not crashing) kernel interfaces (like alsa was until fairly recently etc.). As linux matures this is getting better. This is hard for developer (vendor) to fix, what they can do is clearly state what
Re: Is Linux Unix?
Ryo Furue wrote: Paul E Condon [EMAIL PROTECTED] wrote in message news: [...] I'm using the Intel Fortran Compiler (IFC). Its version 7 runs on Debian without any problem whatsoever, although Intel doesn't support Debian. But, last year Intel released a total rewrite of the compiler, version 8, with which my Fortran programs don't work at all [...] If your programs work with version 7, why do you feel a need to migrate to version 8? Well maybe because version 7 will become unavailable soon (?). For a couple of reasons. One is that Intel doesn't want to fix bugs in version 7. For example, I discovered a problem in version 7 and reported it to Intel. They aknowledged that it is indeed a problem, but they say they won't fix it because version 8 doesn't have the problem. They strongly urged me to migrate to version 8. I suspect that they wish they hear nothing about version 7 and want to pull out as soon as possible. Another reason is that I'm not sure if you can get a copy of version 7 for free any more. When I installed version 7, I didn't think it will go away so that I didn't store the package. Now, what if my machine crashes and I need to install the compiler again? one or more of the following: - backup your machine - see where the compiler is installed, backup those directories and files (if you can't find all the compiler files do lsof when the compiler is running or run compiler with strace and see which files it opens - that way you will find all the libraries, config files etc. that it uses) - get the package, if intel doesn't offer it anymore I am pretty sure there is some place on the net where users of intel fortran meet/talk so they should be able to help you (I assume it is at least sort of legal since it was free) ... (Or migrating to C/C++?) That's out of question! :) For numerical computation, Fortran 95 is better than C++. (I have plenty of experience in C++ programming.) Sure I greatly miss C++'s objects and templates and other goodies. But codes for numerical computation can be much more easily written in Fortran 95. That's the most important point for a non-proffessional programmer like me. I find that hard to believe. is it because of fortran libs that are not available for c/c++ or is performance much better or does fortran have some features that make it easy to write numerical computation code (I forgot most of what I had to learn about fortran but I don't remember ANYTHING being easy/readable in fortran). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Fri, Jul 16, 2004 at 04:13:51PM -0700, Erik Steffl wrote: Emma Jane Hogbin wrote: Ok, this worked (yay!!) but it is read-only. I found a document which said the partition may need to be cleaned first on a Mac, but I tried this and it didn't help. I didn't have to clean anything on mac (I don't even have access to a Mac) what happens if you explicitly try to mount the partition as rw: mount -t hfsplus -w /dev/sda3 /mnt/ipod Same thing...it's still read-only. any errors? what if you try to remount it rw, does it say anything? Is there anything interesting/related in /var/log/syslog? ... Ths time the iPod ran out of battery while I was trying to work things out and now I can't umount it and I get the following errors when I try to read the files: which universe is this in? iPod battery is _charged_ via firewire connection. how can it run out of battery??? smeagol:~ 09:13:59 $ ll /mnt/ipod/ ls: /mnt/ipod/Calendars: Input/output error ls: /mnt/ipod/Contacts: Input/output error : Input/output error ls: /mnt/ipod/Notes: Input/output error total 0 drwxrwxrwx1 99 99 6 Jul 10 10:35 iPod_Control And I can't properly unmount it anymore. It says: smeagol:~ 09:15:50 $ sudo umount /mnt/ipod umount: /mnt/ipod: device is busy Who knew that winning an iPod would be so difficult. :( sometime I have to reset iPod to make it work again (switch hold button to HOLD, wait 5s, switch hold off, press menu and play buttons for few seconbs). sometime I have to reboot computer, not sure what's going on but even if I unload and reload all the relevant modules it cannot connect again (rescanning scsi doesn't find the disk), when I reboot it works OK... and just now: it found iPod 8 times (scsi 0, channel 0, id 0 - 7, lun 00). Seems like it's somewhat less stable since I upgraded to 2.6.5. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Sun, Jul 18, 2004 at 03:41:13AM -0700, Erik Steffl wrote: Emma Jane Hogbin wrote: On Fri, Jul 16, 2004 at 04:13:51PM -0700, Erik Steffl wrote: Emma Jane Hogbin wrote: mount -t hfsplus -w /dev/sda3 /mnt/ipod Same thing...it's still read-only. any errors? what if you try to remount it rw, does it say anything? Is there anything interesting/related in /var/log/syslog? Other than it saying the disk wasn't properly unmounted ...there's nothing out of the ordinary that I can see. ... HFS+-fs warning: Filesystem was not cleanly unmounted, running fsck.hfsplus is recommended. mounti g read-only. this is why it is mounted read-only (filesystem marked dirty). You said that resetting the iPod fixed the file system problem - is it still mounted read-only? What messages you get with clean filesystem? What messages you get when you try to force it to be mounted rw (messages from mount and in /var/log/syslog)? ... Ths time the iPod ran out of battery while I was trying to work things out and now I can't umount it and I get the following errors when I try to read the files: which universe is this in? iPod battery is _charged_ via firewire connection. how can it run out of battery??? No...not the new ones. The new ones are both PC/Mac. If you're using a Mac it is charged through the firmware but if you're using a PC you have to charge it separately. /me shrugs. At least that's what the information booklet says and that's what my experience tells me. oh, now I vaguely remember reading something about this, I think on PC it depends on firewire port, I use the one on Intel D865PERL motherboard, it does charge the iPod. sometime I have to reset iPod to make it work again (switch hold button to HOLD, wait 5s, switch hold off, press menu and play buttons for few seconbs). This fixed the file system mix up, but not the read-write problem. what messages do you get after the filesystem problem was fixed (filesystem problem was the reason why it was mounted read-only so once the filesystem problem is gone it should be mounted rw)? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Mon, Jun 21, 2004 at 08:06:41PM -0700, Erik Steffl wrote: what about sda3? This is what works for me (that's without relying on /etc/fstab): mount -t hfsplus /dev/sda3 /mnt/ipod Ok, this worked (yay!!) but it is read-only. I found a document which said the partition may need to be cleaned first on a Mac, but I tried this and it didn't help. I didn't have to clean anything on mac (I don't even have access to a Mac) what happens if you explicitly try to mount the partition as rw: mount -t hfsplus -w /dev/sda3 /mnt/ipod do you have a line in /etc/fstab for ipod? (if yes, does it say ro?) this is what I have in /etc/fstab for ipod: /dev/sdb3 /mnt/ipod hfsplus rw,user,noauto 0 0 erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Copy Paste into vi??
Bijan Soleymani wrote: [EMAIL PROTECTED] writes: Installed woody. My system boots to a prompt. I then do startx to run xserver. I then run mozilla to surf. What I would like to do is copy some text from the browser and paste it into the vi editor. How can this be done? Select the text in the browser (highlight it) and then switch to vi and middle click (of course have to be in insert mode). If your mouse doesn't have a middle button you might be able to click both buttons at the same time. ctrl-insert often works too (depends on terminal) note that gvim can recognize mouse input and doesn't do autoindent when you paste using mouse (it does not (cannot?) recognize pasting using keyboard) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Recent alsa midi problem
Silvan wrote: On Wednesday 07 July 2004 12:47 pm, Roy Pluschke wrote: After a recent dist-upgrade on my sid system midi has stopped working. Which leads me to believe that snd-emu10k1 and snd-emu10k1-synth have been loaded correctly. is it anything similar to my problem? my email to debian-user is not in archives (http://lists.debian.org/debian-user/2004/07/threads.html) yet (archives are up to 2004/07/04, my email is from 2004/07/05 :-) so here the quote. Subject: modprobe -v snd-seq-oss freezes (sb live! kernel 2.6.5) alsa used to work on my system, I didn't change anything I think might be related and suddenly the modprobe -v snd-seq-oss freezes. system: debian unstable kernel 2.6.5 (with alsa) I did not change kernel or modules (I checked and there are no new/changed files under /lib/modules/2.6.5) yet suddenly (after reboot) the oss sequencer cannot be loaded. There are no (related) messages in /var/log/syslog. jojda:/home/erik# modprobe -v snd-seq-oss insmod /lib/modules/2.6.5/kernel/sound/core/seq/snd-seq.ko insmod /lib/modules/2.6.5/kernel/sound/core/seq/snd-seq-midi-event.ko insmod /lib/modules/2.6.5/kernel/sound/core/seq/oss/snd-seq-oss.ko and the modprobe does not exit, it just sits there. udev creates devices (see below) but there are not usable, e.g.: jojda:/var/log# sfxload /data/music/soundfont/creative/8MBGMSFX.SF2 /dev/sequencer: No such device udev says (in daemon.log): Jul 5 01:18:24 jojda udev[2821]: configured rule in '/etc/udev/rules.d/udev.rules' at line 38 applied, 'midiC0D2' becomes 'snd/%k' Jul 5 01:18:24 jojda udev[2821]: creating device node '/dev/snd/midiC0D2' Jul 5 01:18:25 jojda udev[2791]: configured rule in '/etc/udev/rules.d/udev.rules' at line 36 applied, 'hwC0D2' becomes 'snd/%k' Jul 5 01:18:25 jojda udev[2791]: creating device node '/dev/snd/hwC0D2' Jul 5 01:18:25 jojda udev[2798]: configured rule in '/etc/udev/rules.d/udev.rules' at line 38 applied, 'midiC0D1' becomes 'snd/%k' Jul 5 01:18:25 jojda udev[2798]: creating device node '/dev/snd/midiC0D1' Jul 5 01:18:25 jojda udev[2768]: configured rule in '/etc/udev/rules.d/udev.rules' at line 40 applied, 'seq' becomes 'snd/%k' Jul 5 01:18:25 jojda udev[2768]: creating device node '/dev/snd/seq' Jul 5 01:18:25 jojda udev[2775]: creating device node '/dev/sequencer' Jul 5 01:18:25 jojda udev[2782]: creating device node '/dev/sequencer2' Jul 5 01:18:25 jojda udev[2805]: creating device node '/dev/amidi' Jul 5 01:18:25 jojda udev[2812]: creating device node '/dev/admmidi' Jul 5 01:20:22 jojda ntpd[2392]: kernel time sync enabled 0001 any ideas? TIA erik erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
modprobe -v snd-seq-oss freezes (sb live! kernel 2.6.5)
alsa used to work on my system, I didn't change anything I think might be related and suddenly the modprobe -v snd-seq-oss freezes. system: debian unstable kernel 2.6.5 (with alsa) I did not change kernel or modules (I checked and there are no new/changed files under /lib/modules/2.6.5) yet suddenly (after reboot) the oss sequencer cannot be loaded. There are no (related) messages in /var/log/syslog. jojda:/home/erik# modprobe -v snd-seq-oss insmod /lib/modules/2.6.5/kernel/sound/core/seq/snd-seq.ko insmod /lib/modules/2.6.5/kernel/sound/core/seq/snd-seq-midi-event.ko insmod /lib/modules/2.6.5/kernel/sound/core/seq/oss/snd-seq-oss.ko and the modprobe does not exit, it just sits there. udev creates devices (see below) but there are not usable, e.g.: jojda:/var/log# sfxload /data/music/soundfont/creative/8MBGMSFX.SF2 /dev/sequencer: No such device udev says (in daemon.log): Jul 5 01:18:24 jojda udev[2821]: configured rule in '/etc/udev/rules.d/udev.rules' at line 38 applied, 'midiC0D2' becomes 'snd/%k' Jul 5 01:18:24 jojda udev[2821]: creating device node '/dev/snd/midiC0D2' Jul 5 01:18:25 jojda udev[2791]: configured rule in '/etc/udev/rules.d/udev.rules' at line 36 applied, 'hwC0D2' becomes 'snd/%k' Jul 5 01:18:25 jojda udev[2791]: creating device node '/dev/snd/hwC0D2' Jul 5 01:18:25 jojda udev[2798]: configured rule in '/etc/udev/rules.d/udev.rules' at line 38 applied, 'midiC0D1' becomes 'snd/%k' Jul 5 01:18:25 jojda udev[2798]: creating device node '/dev/snd/midiC0D1' Jul 5 01:18:25 jojda udev[2768]: configured rule in '/etc/udev/rules.d/udev.rules' at line 40 applied, 'seq' becomes 'snd/%k' Jul 5 01:18:25 jojda udev[2768]: creating device node '/dev/snd/seq' Jul 5 01:18:25 jojda udev[2775]: creating device node '/dev/sequencer' Jul 5 01:18:25 jojda udev[2782]: creating device node '/dev/sequencer2' Jul 5 01:18:25 jojda udev[2805]: creating device node '/dev/amidi' Jul 5 01:18:25 jojda udev[2812]: creating device node '/dev/admmidi' Jul 5 01:20:22 jojda ntpd[2392]: kernel time sync enabled 0001 any ideas? TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [OT] yahoo protocol switching
Nori Heikkinen wrote: noticed today that yahoo has again switched protocols, making it impossible to connect from gaim. the quick fix suggested on /. [1] -- ... has anyone else been bitten by this, and found a workaround? usually the yahoo web messenger works, good as an emergency until the other ones catch up erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Kevin Mark wrote: On Tue, Jun 22, 2004 at 01:21:42AM -0700, Erik Steffl wrote: Erik Steffl wrote: Emma Jane Hogbin wrote: ... smeagol:/mnt 21:28:04 $ mount ipod2 mount: wrong fs type, bad option, bad superblock on /dev/sda2, or too many mounted file systems what about sda3? This is what works for me (that's without relying on /etc/fstab): mount -t hfsplus /dev/sda3 /mnt/ipod what does fsdisk (or other partitioning tools) say about /dev/sda? so I looked at it on my computer, the fdisk, cfdisk, sfdisk are useless, they claim there is not partition information. IIRC, all these do not handle hfs?! it's not about filesystems, it's about partitions. I don't think these could handle it even if there was no hfs+ partition on iPod. parted (apt-get install parted, I have unstable parted 1.6.9-3.1) works, it displays partition table which looks reasonable: (parted) print Disk geometry for /dev/sdb: 0.000-28610.859 megabytes Disk label type: mac MinorStart End Filesystem Name Flags 1 0.000 0.030 partition map 2 0.031 32.030 firmware 3 32.031 28610.858 hfs+disk This is interseting to me. I will have to try this on my mac! -Kev erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Erik Steffl wrote: Emma Jane Hogbin wrote: ... smeagol:/mnt 21:28:04 $ mount ipod2 mount: wrong fs type, bad option, bad superblock on /dev/sda2, or too many mounted file systems what about sda3? This is what works for me (that's without relying on /etc/fstab): mount -t hfsplus /dev/sda3 /mnt/ipod what does fsdisk (or other partitioning tools) say about /dev/sda? so I looked at it on my computer, the fdisk, cfdisk, sfdisk are useless, they claim there is not partition information. parted (apt-get install parted, I have unstable parted 1.6.9-3.1) works, it displays partition table which looks reasonable: (parted) print Disk geometry for /dev/sdb: 0.000-28610.859 megabytes Disk label type: mac MinorStart End Filesystem Name Flags 1 0.000 0.030 partition map 2 0.031 32.030 firmware 3 32.031 28610.858 hfs+disk from that you can see that mount -t hfsplus /dev/sdb3 /mnt/ipod (sda3 in your case) should work, it works on my system. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Sun, Jun 20, 2004 at 06:05:39PM -0700, Erik Steffl wrote: note that you need both HFS+ filesytem AND mac style partitions kernel support to be able to work with the iPod. I have both... From File Systems - Miscellaneous filesystems: x x* Apple Macintosh file system support (EXPERIMENTAL) x x* Apple Extended HFS file system support that's not it, those are both file systems (in .config: # Miscellaneous filesystems), you need support for mac style partitions: # Partition Types ... CONFIG_PARTITION_ADVANCED=y # not sure what this does ... CONFIG_MAC_PARTITION=y # this is what you need ... I had the same problem... once you have the mac partition support you should be able to use iPod with e.g. gtkpod erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Sun, Jun 20, 2004 at 11:23:18PM -0700, Erik Steffl wrote: that's not it, those are both file systems (in .config: # Miscellaneous filesystems), you need support for mac style partitions: # Partition Types Ahh! Recompiling... Does it really matter for the 2.6.x kernels if things are compiled as modules versus compiled directly into the kernel? I've put everything directly into the kernel... (maybe one day when I understand all this stuff a little better I'll go back to fighting with modules). I don't think it matters, stuff you need for boot (e.g. drivers for the disk you boot off) has to be compiled in. Generally I prefer modules because it makes troubleshooting easier: you see which ones are loaded, can remove and load them with different options, see what exactly happens when you load them (messages in /var/log/syslog) etc. Another plus of modules is that you can add modules and don't have to reboot - e.g. in your case you can (could have) just change kernel config, build and install modules and modprobe the mac partition module (if it can be compiled as a module, I don't remember), no need to reboot or new kernel installation. with udev you have to specify the modules you want to load in /etc/modules (or load them manually when you know you need them), there's no auto-load of modules on demand (which IMO is a step backward). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Sun, Jun 20, 2004 at 11:23:18PM -0700, Erik Steffl wrote: that's not it, those are both file systems (in .config: # Miscellaneous filesystems), you need support for mac style partitions: # Partition Types Ouch. Bad idea. I get kernel panics when I try to boot with these two new options enabled. It dies around the VFS section and fails to mount hda1. are you sure you didn't do any other changes? it looks like you don't have support for PC partitions. I just checked, the partition support cannot be compiled as a module (I use 2.6.5), here's what I have in that section: # # Partition Types # CONFIG_PARTITION_ADVANCED=y # CONFIG_ACORN_PARTITION is not set # CONFIG_OSF_PARTITION is not set # CONFIG_AMIGA_PARTITION is not set # CONFIG_ATARI_PARTITION is not set CONFIG_MAC_PARTITION=y CONFIG_MSDOS_PARTITION=y CONFIG_BSD_DISKLABEL=y CONFIG_MINIX_SUBPARTITION=y CONFIG_SOLARIS_X86_PARTITION=y CONFIG_UNIXWARE_DISKLABEL=y # CONFIG_LDM_PARTITION is not set # CONFIG_NEC98_PARTITION is not set # CONFIG_SGI_PARTITION is not set # CONFIG_ULTRIX_PARTITION is not set # CONFIG_SUN_PARTITION is not set # CONFIG_EFI_PARTITION is not set erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Mon, Jun 21, 2004 at 11:20:45AM -0700, Erik Steffl wrote: are you sure you didn't do any other changes? it looks like you don't have support for PC partitions. I just checked, the partition support cannot be compiled as a module (I use 2.6.5), here's what I have in that section: Ahh. This was my first time turning on the Advanced Partition information. I hadn't turned on enough of the options. I now have the following and the iPod is STILL showing up as having no partitions on it... CONFIG_PARTITION_ADVANCED=y # CONFIG_ACORN_PARTITION is not set # CONFIG_OSF_PARTITION is not set # CONFIG_AMIGA_PARTITION is not set # CONFIG_ATARI_PARTITION is not set CONFIG_MAC_PARTITION=y CONFIG_MSDOS_PARTITION=y CONFIG_BSD_DISKLABEL=y CONFIG_MINIX_SUBPARTITION=y CONFIG_SOLARIS_X86_PARTITION=y CONFIG_UNIXWARE_DISKLABEL=y # CONFIG_LDM_PARTITION is not set # CONFIG_NEC98_PARTITION is not set # CONFIG_SGI_PARTITION is not set # CONFIG_ULTRIX_PARTITION is not set # CONFIG_SUN_PARTITION is not set # CONFIG_EFI_PARTITION is not set that's weird, I'd check: (are you really running the kernel you think you are running?) (is the iPod OK?) if you're using firewire (I don't remember the beginning of the thread): these are the modules I load for iPod (I don't have any other firewire devices): # firewire modprobe ieee1394 modprobe ohci1394 # firewire disk support (looks like scsi) modprobe sbp2 plus I run the rescan-scsi-bus.sh that I got somewhere (IIRC one of the iPod and linux sites), see attachement. The iPod is not usable (not visible) until you load the modules above and run the script. this is my /etc/fstab line to mount iPod: /dev/sdb3 /mnt/ipod hfsplus rw,user,noauto 0 0 don't remember why it's sdb3 (it's sdb because I have another disk on sda but I don't remember why it's sdb3, it's probably how the iPod disk is partitioned, there might be something else on sdb1 and sdb2) BTW I had some problems with fdisk, I don't remember whether it was with iPod or SATA but if you can't see aprtition info using fdisk try to use cfdisk or other partitioning programs or just blindly mount /dev/sdb3 (or whatever the disk is on your system). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: ... smeagol:/mnt 21:28:04 $ mount ipod2 mount: wrong fs type, bad option, bad superblock on /dev/sda2, or too many mounted file systems what about sda3? This is what works for me (that's without relying on /etc/fstab): mount -t hfsplus /dev/sda3 /mnt/ipod what does fsdisk (or other partitioning tools) say about /dev/sda? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Most Polite Apps for Window Managers?
Esteban Martinez wrote: Ed Sutherland wrote: I'm considering moving from one of the giant desktop environments (Gnome) to a svelter window manager (blackbox or windowmaker.) I no need to consider:-) you can just try different WMs, most of them even without restarting X (last time I tried it KDE didn't alow this), just go to the main (debian) menu and pick a WM from Window Manager menu. understand some apps require only the gnome or kde toolkits, while others require the whole kit-and-kaboodle to operate. I'm wondering if these app categories can politely run in a window manager (that is, using just the gnome toolkit.) E-mail (Thunderbird) Web (Firefox) Office (OpenOffice) these run ok with different WMs in general there are (almost) no problems related to running apps in different WMs Contacts (Rubrica) never tried this one, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firewire, iPod and Linux
Emma Jane Hogbin wrote: On Sun, Jun 20, 2004 at 04:44:58PM -0500, James Abella wrote: When sbp2 can log in, use fdisk to get basic partition info of /dev/sda. If there are only sda1 and sda2, it's Win mode. If not, the easiest way to convert it to Win mode is to install iTune on one Windows box. Ok, I'm back to the laptop seeing the iPod. I upgraded udev and re-wrote the rules. Then I made a new symbolic link. I also have leftover in /etc/fstab the following (I'm not sure if it makes a difference or not): /dev/sda2 /mnt/ipod2 auto ro,user 0 0 dmesg says: ieee1394: Node added: ID:BUS[0-00:1023] GUID[000a27000266158c] ieee1394: Node changed: 0-00:1023 - 0-01:1023 scsi0 : SCSI emulation for IEEE-1394 SBP-2 Devices ieee1394: sbp2: Logged into SBP-2 device ieee1394: Node 0-00:1023: Max speed [S400] - Max payload [2048] Vendor: Apple Model: iPod Rev: 1.50 Type: Direct-Access ANSI SCSI revision: 02 sda: Spinning up diskready SCSI device sda: 39063024 512-byte hdwr sectors (2 MB) sda: test WP failed, assume Write Enabled sda: asking for cache data failed sda: assuming drive cache: write through sda: unknown partition table Attached scsi removable disk sda at scsi0, channel 0, id 0, lun 0 And finally fdisk smeagol:~ 19:22:11 $ sudo fdisk /dev/sda Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel Building a new DOS disklabel. Changes will remain in memory only, until you decide to write them. After that, of course, the previous content won't be recoverable. The number of cylinders for this disk is set to 19073. There is nothing wrong with that, but this is larger than 1024, and could in certain setups cause problems with: 1) software that runs at boot time (e.g., old versions of LILO) 2) booting and partitioning software from other OSs (e.g., DOS FDISK, OS/2 FDISK) Warning: invalid flag 0x of partition table 4 will be corrected by w(rite) Disk /dev/sda: 20.0 GB, 2268288 bytes 64 heads, 32 sectors/track, 19073 cylinders Units = cylinders of 2048 * 512 = 1048576 bytes Device BootStart EndBlocks Id System (and then nothing) My laptop appears to think it's an empty iPod. I guess that means it's HFS? Unfortunately I do not have a Windows machine with firewire that I can try it with (and I am not willing to install Windows on my laptop). I do have HFS support enabled in the kernel, I guess I'll try updating the automount scripts to try reading the device as HFS? note that you need both HFS+ filesytem AND mac style partitions kernel support to be able to work with the iPod. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
postfix: /etc/aliases ignored?
system: debian unstable, postfix 2.1.1-8 I have /etc/aliases, recreated /etc/aliases.db using newaliases even though I am not sure it matter because here's what I have in main.cf: alias_maps = hash:/etc/aliases alias_database = hash:/etc/aliases the problemn is that aliases are ignored, e.g. this one (from /etc/aliases): root: erik here's what I get in syslog after sending email to root (it tries to pass [EMAIL PROTECTED] to lmtp (cyrus delivery mechanism). Is the problem in adding the hostname to the local address, i.e. address is changed from root to [EMAIL PROTECTED] and then not aliased (because it's not just root anymore)? Jun 15 02:15:52 jojda postfix/pickup[31601]: 39CABC814E: uid=1000 from=erik Jun 15 02:15:52 jojda postfix/cleanup[31624]: 39CABC814E: message-id=[EMAIL PROTECTED] Jun 15 02:15:52 jojda postfix/qmgr[31602]: 39CABC814E: from=[EMAIL PROTECTED], size=334, nrcpt=1 (queue active) Jun 15 02:15:52 jojda postfix/lmtp[31625]: 39CABC814E: to=[EMAIL PROTECTED], relay=/var/run/cyrus/socket/lmtp[/var/run/cyrus/socket/lmtp], delay=0, status=bounced (host /var/run/cyrus/socket/lmtp[/var/run/cyrus/socket/lmtp] said: 550-Mailbox unknown. Either there is no mailbox associated with this 550-name or you do not have authorization to see it. 550 5.1.1 User unknown (in reply to RCPT TO command)) Jun 15 02:15:52 jojda postfix/cleanup[31624]: 3C67EC814F: message-id=[EMAIL PROTECTED] Jun 15 02:15:52 jojda postfix/qmgr[31602]: 3C67EC814F: from=, size=2384, nrcpt=1 (queue active) Jun 15 02:15:52 jojda postfix/qmgr[31602]: 39CABC814E: removed Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: duplicate_check: [EMAIL PROTECTED] user.erik0 Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: mystore: starting txn 2147492688 Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: mystore: committing txn 2147492688 Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: duplicate_mark: [EMAIL PROTECTED] user.erik1087290952 Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: mystore: starting txn 2147492689 Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: mystore: committing txn 2147492689 Jun 15 02:15:52 jojda cyrus/lmtpd[31490]: duplicate_mark: [EMAIL PROTECTED] .erik+.sieve.1087290952 Jun 15 02:15:52 jojda postfix/lmtp[31625]: 3C67EC814F: to=[EMAIL PROTECTED], relay=/var/run/cyrus/socket/lmtp[/var/run/cyrus/socket/lmtp], delay=0, status=sent (250 2.1.5 Ok) Jun 15 02:15:52 jojda postfix/qmgr[31602]: 3C67EC814F: removed TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: accessing a fuji digital camera
LeVA wrote: Hi! I want to download the pictures from a fuji digital camera. ... there was a lot of discussion about using the camera in usb mass storage mode. Did anybody try to use it in PC cam mode? I googled for it (even read the webcam howto) but it's hard to figure out what does it mean - which apps and what drivers are needed etc. There are two drivers for fuji on sourceforge but they are not very active... any pointers? (I already signed up for video4linux mailing list) TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Solved with a modeline: Re: NEC2080UX+ - Supported Future Video Modes
William Ballard wrote: On Wed, Jun 09, 2004 at 04:26:25PM -0700, Erik Steffl wrote: if you set the frequencies right it should work without edid (and without modelines). it's possible you will need modeline though, I don't know the details of your situation... You sir are my hero! I didn't need the Modeline. All I needed was IgnoreEDID=True and in the monitor section: HorizSync 64 - 64 VertRefresh 60 - 60 which are the numbers the monitor reports in Windows. It's a flat-panel monitor; 60hz is the natural frequency. It only really matters when you are watching Mpegs2; at the artificial 75 hz you can see tearing that goes away when you run it at 60hz. I didn't need to add UseEdidFreqs=False, only the IgnoreEDID=true. I tried changing the HorizSync and VertRefresh without using IgnoreEDID, but it did nothing. Thank you very much for helping me solve a tough problem. glad it helped, I wonder why would edid info (that comes directly from monitor, as far as I can tell) result in worse settings then what X does based on vert/horiz frequencies. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Solved with a modeline: Re: NEC2080UX+ - Supported Future Video Modes
Zenaan Harkness wrote: On Wed, 2004-06-09 at 17:09, William Ballard wrote: From this snippet of XFree86.0.log, can anyone construct a modeline which will come up at [EMAIL PROTECTED] The monitor is NEC LCD1850E with a Radeon 9800XT. (II) fglrx(0): Supported additional Video Mode: (II) fglrx(0): clock: 135.0 MHz Image Size: 359 x 287 mm (II) fglrx(0): h_active: 1280 h_sync: 1296 h_sync_end 1440 h_blank_end 1688 h_border: 0 (II) fglrx(0): v_active: 1024 v_sync: 1025 v_sync_end 1028 v_blanking: 1066 v_border: 0 (II) fglrx(0): Ranges: V min: 55 V max: 85 Hz, H min: 31 H max: 82 kHz, PixClock max 140 MHz Well, if the same technique works as did for me: Modeline 1280x1024 135 1280 1296 1440 1688 1024 1025 1028 1066 Which actually looks the same as a line further down in the log: (II) fglrx(0): Monitor name: NEC LCD1850E (II) fglrx(0): Serial No: 204011324 (II) fglrx(0): (II) fglrx(0): DesktopSetup 0x (**) fglrx(0): PseudoColor visuals disabled (**) fglrx(0): Overlay disabled (**) fglrx(0): Overlay disabled (II) fglrx(0): PLL parameters: rf=2700 rd=12 min=2 max=4; xclk=36500 (==) fglrx(0): Using gamma correction (1.0, 1.0, 1.0) (**) fglrx(0): Center Mode is disabled (==) fglrx(0): TMDS coherent mode is enabled (II) fglrx(0): Total 5 valid mode(s) found. (--) fglrx(0): Virtual size is 1280x1024 (pitch 1280) (**) fglrx(0): *Default mode 1280x1024: 135.0 MHz (scaled from 0.0 MHz), 80.0 kHz, 75.0 Hz (II) fglrx(0): Modeline 1280x1024 135.00 1280 1296 1440 1688 1024 1025 1028 1066 If it doesn't work, we'll have to find someone more knowledgeable than us. Let us know how you go. didn't catch the beginning but if you are sure that monitor supports it you can include the following in Section Device: Option UseEdidFreqs false Option IgnoreEdid true I had to do the same for Sony monitor that was reporting wrong values (Edid said it couldn't do 1280x1024) but given the HorizSync and VertRefresh it should have supported it and indeed it worked fine (once edid was disabled). I did set the HorizSync and VertRefresh according to monitor docs so I wasn't 'overclocking' it, I ony wanted resolution that wasn't in edid (slightly higher) guess it can damage your monitor so you'd better be carefull... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Solved with a modeline: Re: NEC2080UX+ - Supported Future Video Modes
William Ballard wrote: On Wed, Jun 09, 2004 at 03:32:35PM -0700, Erik Steffl wrote: didn't catch the beginning but if you are sure that monitor supports it you can include the following in Section Device: Option UseEdidFreqs false Option IgnoreEdid true When I put in IgnoreEdid the monitor just comes up with VGA out of sync UseEdidFreqs doesn't appear to have an effect. When you simply place a Modeline in the monitor section, how do you make X use it? Maybe I'm IgnoringEdid but it's not picking up the modeline. are you using proper frequencies? X tries to set freq as high as possible for given resolution, I didn't read the beginning of the thread but it seems like you want it to use 60 instead of 75? if so make sure HorizSync has 60 as upper bound. if you set the frequencies right it should work without edid (and without modelines). it's possible you will need modeline though, I don't know the details of your situation... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: X Server Crashing on DPMS Event
Barry Skidmore wrote: I believe I am having a problem with the xserver failing whenever my monitor's DPMS kicks in. After either windowmaker or enlightenment have run for 15 minutes, the window manager quits and I am put back into the console with the following errors: Fatal server error: Caught signal 4. Server aborting. xinit: connection to X server lost. I have all power management turned off in the BIOS, and do not have apm or acpi installed. I am using a ViewSonic VG150 LCD monitor, and according to the manual and web site, there is no documented way of turning the monitor's DPMS off. you can use xset to control DPMS (when, what), see man xset for more info: To control Energy Star (DPMS) features: -dpms Energy Star features off +dpms Energy Star features on dpms [standby [suspend [off]]] force standby force suspend force off force on (also implicitly enables DPMS features) a timeout value of zero disables the mode are you sure it's dpms that causes the problem? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: X Server Crashing on DPMS Event
Barry Skidmore wrote: When I did a: # xset -dpms xserver immediately crashed with the following error: xterm: fatal IO server error 32 (Broken pipe) or KillClient on X Server :0.0 /usr/bin/Windowmaker warning.got signal 1 (Hangup1-exiting... looks like it's confirmed that it has something to do with dpms. however the errors above are not from X server, these are just errors from X clients (xterm and Windowmaker) essentially complaining that X server is gone. check the /var/log/XFree86.0.log and /var/log/xdm.log for X server errors. Try to search google for the error, that helps sometime. Also try to search for X config info specific to your video card (don't think the monitor matters). erik Barry On Sun, Jun 06, 2004 at 08:30:15PM -0500, Barry Skidmore wrote: No, I am not sure it is dpms that is causing the problem, but I will see what is the effect of the -dpms option. I did not see that option documented. Barry On Sun, Jun 06, 2004 at 04:30:22PM -0700, Erik Steffl wrote: Barry Skidmore wrote: I believe I am having a problem with the xserver failing whenever my monitor's DPMS kicks in. After either windowmaker or enlightenment have run for 15 minutes, the window manager quits and I am put back into the console with the following errors: are you sure it's dpms that causes the problem? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Multimedia Optimisation
Simon L wrote: Do you know where I could find informations on how to make my Debian computer go faster for all the media things? I know that my computer can go much more faster than now, because it does on Windows and I would like to do the same on linux. depends on what you want and what is the current problem, here are some pointers: X config - make sure you use correct driver (for normal work it should be enough, for games and other 3d apps you might need openGL but I guess that's not what you're asking about). use 2.6.x kernel and enable preemptive kernel (in kernel config) do not run unneccessary services that might eat lot of CPU (most of them don't so you problably do not have to bother) use hdparm to check whether your disks have DMA enabled (big difference) use e.g. gkrellm (or other tools like top, etherape) to monitor your system and see what resources are used, how, and which programs use them. If your system simply feels unresponsive it's most probably problem with X server (you are using generic driver instead of driver for your specific card) or not enough memory. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Need to prevent X from restarting
Aldous Huxley wrote: Yes, I already knew that trick. It's just that sometimes I need to back out of X completely to free up system resources to allow me to play mp3s using console-based players like splay. So I'd like to be able to boot into X login, then back out of X when I'm through working, free up that precious memory (64mb total) and play some music and use other console-based progs too. if you have enough swap then lot of X (and X clients, if any) will be swapped out, that might be enough, just make sure you don't have X programs running that do something on their own (like browser with flash plugin animation etc.) So is there a way to prevent X from always restarting? as others pointed out it is restarted by display manager (probably gdm, kdm or xdm), apart from stopping the display manager you can also: - remove it completely, start X on demand by startx, when you exit X it won't be automatically restarted - use update-rc.d to not automatically start [xkg]dm (and use startx to start X only when you want X) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: desktop Linux for a debian user?
Ignatz Sol wrote: I'm fairly new at Linux, but I really enjoy using Debian. I'm looking to install a desktop Linux at home in a dual-boot mode. As the Debian desktop still has a ways to go, can ya'll recommend a distro for me? It needs to be slick to please my wife, but I love the Debian spirit. A friend recommended Suse, which looks great, but I'm put off by the idea of paying for it. (Not that it's out of the question.) distros mostly default in installation, package management, default themes etc. So if you are happy with debian just configure it the way you (and your wife) like it and that's it. There's really no much difference in desktop suitability between distros, as long as you are happy with system administration. when you read about how good a distro is for desktop it mostly means how good is distro for _novice_ user after _fresh_ install. Which does not apply to you (you already said you like debian) so you can safely ignore all the desktop advices. Pretty much all the distros can be downloaded for free (sometime the paid for version has more, usually non-free, stuff) so don't hesitate to try few of them and see how you like them (just remember, installation is done only once (or at least not very often), maintenance is done all the time (=a lot more often)) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Strange lockups in X
Nate Bargmann wrote: Hi. About two weeks ago I updated both of my machines to the then latest versions in Testing. At the same time the xfree packages of version 4.3.0.dfsg.1-1. Shortly after that I began to experience instant freezes on both the laptop and the desktop which are two very different machines, but i386 architecture. I can't pin down any common action that I might be doing. I've had the desktop run for nearly 24 hours and often times it will run less than six hours before locking up. It has always locked up when not in use so xscreensaver has been active. how exactly it freezes? - can you ping the machine? - does it freeze when you do not use X? - does it freeze when you do kernel compile (or memory test)? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: please help- I can't even figure my mailer
chuck boothe wrote: ... Could you, would you plleeeaaae look down from your mount of knowledge just long enough to give me- line by line (VERBATIM)- the directions that I need to access root through a terminal (yes, I do you need to provide more info about what problem you have. Generally on login prompt you type root as user name and root password as password. Do you get errors after you do that? Are you talking about X login (via xdm, gdm, kdm etc.) or terminal (text mode) login? know the password) and then give me the line by line (VERBATIM) that I need to allow myself and my family access to my two cdr/rw s and to allow us to hear the beautiful music which once so freely flowed from our speakers (yes, we can't even afford a stereo). I am begging you! My soundcard is an Allegro ESS Maestro (1988?), my video driver is an Intel !810, and my two cdr?rw s are LG's. I would try knoppix or mepis, there is a high probablity that it will recognize your HW and you will get a usable configuration (working X windows, working soundcard etc.) it's really hard to help you if you don't say what you did and what errors you see (or what are the symptoms of your problems). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SATA hard drives?
Colin Watson wrote: On Mon, Apr 19, 2004 at 09:07:27AM +0100, [EMAIL PROTECTED] wrote: I have a new computer that has a Soyo Dragon motherboard, with Onboard SATA/RAID: 2x Serial ATA, RAID 0/1, .. and a IBM 160 GB Hard drive 160 GB 7200RPM SATA Hard Drive. Can Debian be installed on this setup? Serial ATA still requires a 2.6 kernel; support for it is scheduled for inclusion in 2.4.27. There are 2.6 versions of sarge's debian-installer in preparation, but they're still somewhat experimental ... ac kernels also work (I used 2.4.21 ac4, with libata patches for support for disks over 137GB), I just tried 2.6.5 and it seems to work fine (matrox 250GB SATA drive, intel D865PERL motherboard), no patches needed. as far as the installation goes: next knoppix will have optional 2.6 kernel so I guess that might be an installation option... (the CD was already available somewhere, but it's not available for download, see knoppix page for more details, it might be possible to get it somehow, I guess, if there's no other way to install). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
