Java (aka icedtea) plugin for iceweasel, settings
I can pass options to java programs by setting the environment variable JAVA_ARGS. This works by ultimately passing the contents of the variable on the command line to the java runtime binary, such as /usr/bin/java; well behaved packages do this themselves by using the shell functions in java-wrappers. This is all great, but what about the java web applets, run by the browser plugin (in my case, the icedtea-6-plugin for iceweasel)? As far as I know this doesn't use the runtime binary, and even if it did it wouldn't know to apply the extra options from JAVA_ARGS. So can I control the java runtime options for applets in any way? Googling for the relevant terms tends to bring up something called Java Control Panel, with instructions where to find such a thing on Windoze and Mac. But nothing about Debian, and indeed a search on packages.debian.org for a package file "ControlPanel" shows nothing promising. Right now my JAVA_ARGS is as follows but I expect it to grow. JAVA_ARGS="-Dawt.useSystemAAFontSettings=on -Dswing.aatext=true"
Re: Search engine for documentation indexing?
Olly Automatically uncompressing gzipped files for indexing isn't hard Olly to do, but what can you link to for them in the search results? Olly Of the four web browsers I just tried, only w3m showed the Olly contents of file:///usr/share/doc/coreutils/README.gz rather than Olly downloading it for me. Same for Olly http://localhost/doc/coreutils/README.gz it seems. The dwww cgi uncompresses these files for you, so something like http://localhost/cgi-bin/dwww/usr/share/doc/foo/README.gz works (in any browser). Also, it may be possible to use mod_deflate in Apache to transparently uncompress, though I have never tried that. Olly But as others have said, recoll is probably a better choice for a Olly Xapian-based solution for a desktop situation anyway. Well, but I'm not running a typical desktop, at least not if by that is meant a Gnome or KDE stacked system. The fact that recoll is bound to a particular GUI is definitely a disadvantage. So no orphaning omega please! :-) What I'm ending up doing is a hack: I build a separate tree that is mostly a symlink farm pointing to /usr/share/doc, except that gzipped files are replaced by their uncompressed versions. Then I run omindex on the new tree. I've just tested this and it does the job. Indexing time is about 16 min, which is in between swish-e (9 min) and swish++ (27 min). Not terrible, but maybe there is a way to speed it up by parallelization? The omega docs seem to say nothing about concurrent access to the index. Is it possible to run 2 indexer processes at once, each updating the same index but with different files? -- Ian Zimmerman i...@buug.org gpg public key: 1024D/C6FF61AD fingerprint: 66DC D68F 5C1B 4D71 2EE5 BD03 8A00 786C C6FF 61AD Ham is for reading, not for eating. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Search engine for documentation indexing?
I'd like to build an index of the documentation in /usr/share/doc, but I am quite unhappy with the options I have tried so far: 1. dwww has a built in cgi for searching an index built by swish++. Unfortunately swish++ indexing seems to take forever (it's described as lighting fast on the upstream website, but I can't find the pictures of flying pigs). Also, using the built in dwww integration has the disadvantage that only documents registered in doc-base are indexed, which misses a lot of them. On top of this swish++ shares the main problem of 2. swish-e. This looked very promising for a while, and I even wrote a python module to wrap the API: http://pypi.python.org/pypi?%3Aaction=searchterm=pyswishsubmit=search ... but it can't handle documents encoded other than ASCII and Latin-1 (in particular, it breaks on UTF-8 XHTML documents). This is a show-stopper. 3. xapian-omega. This seems to be the one modern apps are migrating to, I heard of the Gnus mail/newsreader acquiring a xapian based search function. But, out of the box it cannot index gzipped files (and most documents in /usr/share/doc other that HTML pages are gzipped), and there doesn't seem to be a way to add a user-defined filter either to compensate for this (swish-e has user filters). I can't be the only one looking for this, so what do other debianists do? -- Ian Zimmerman i...@buug.org gpg public key: 1024D/C6FF61AD fingerprint: 66DC D68F 5C1B 4D71 2EE5 BD03 8A00 786C C6FF 61AD Ham is for reading, not for eating. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
status of backports?
Is it just me, or is backports.org in trouble? -- Ian Zimmerman i...@buug.org gpg public key: 1024D/C6FF61AD fingerprint: 66DC D68F 5C1B 4D71 2EE5 BD03 8A00 786C C6FF 61AD Ham is for reading, not for eating. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: status of backports?
Jens Did I miss something? Why would it be in trouble? i...@matica:~$ httping -t 5 -h 194.8.57.6 PING 194.8.57.6:80 (http://194.8.57.6:80/): timeout receiving reply from host timeout receiving reply from host timeout receiving reply from host timeout receiving reply from host timeout receiving reply from host timeout receiving reply from host ^C^C^C^Ctimeout receiving reply from host --- http://194.8.57.6:80/ ping statistics --- 7 connects, 0 ok, 100.00% failed -- Ian Zimmerman i...@buug.org gpg public key: 1024D/C6FF61AD fingerprint: 66DC D68F 5C1B 4D71 2EE5 BD03 8A00 786C C6FF 61AD Ham is for reading, not for eating. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
change tk fonts, how?
Tk (as of 8.4) uses a different (smaller, lighter) font for message boxes --- including error messages --- than for the rest of the widgets, at least by default. I have always thought this was a cosmetic glitch, but now I have a huge screen at work and the message font has become simply unreadable. How can I reconfigure tk to use the ubiquitous bold Helvetica for message boxes too? I owned the Osterhout book at one time but IIRC it didn't answer this either. TIA for help, Ian -- His stride expressed the man's pride and self-confidence and his eyes revealed his honest simplicity. His appearance indicated that his belly was of far greater importance than his head. Naguib Mahfouz, _Midaq Alley_ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
export Mozilla Calendar data
Hi, the Mozilla Calendar looks and works really nice, but is there a way
to get the data to any other program, in particular a command line
program? I figured that the calendar data is stored in
~/.mozilla/defualt/${obfuscated_profile_name}/Calendar/CalendarDataFile.ics.
Is there another program that understands data in this format, or do I
have to write it myself? It doesn't look very difficult, still apt-get
would be quicker ;-) A Perl or python library certainly qualifies for an
answer.
TIA, Ian
--
I wonder which is the best virus for unix and if I can write
a better one in Microsoft BASIC ?
Hans-Marc Olsen in comp.unix.programmer
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
cached tex fonts
Is there a simpler way to purge the cached font metrics in /var/cache/fonts than something like find /var/cache/fonts -type f -name *.pk -o -name *.tfm | xargs rm ? In particular, I find that upgrading tetex packages doesn't do this automatically. Should it? Is this a bug? -- Nothing can be explained to a stone. Or to a stoned person, either. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
sawfish and mozilla
Does anybody else notice that sawfish (1.0.1-6) keeps resizing the Mozilla window immediately after Mozilla (2:0.9.8-2) is started? It doesn't happen with other window managers, so I am fairly sure it is sawfish' fault not Mozilla's. If anybody has a clue what causes this, please share. A carbon copy to my address is appreciated but I'll check the list archive too. -- Ian Zimmerman, Oakland, California, U.S.A. GPG: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087 The world has taken on a thickness of vulgarity that raises a spiritual man's contempt to the violence of a passion. Baudelaire
Mozilla SSL root certificates
Hi, this is something that has bothered me for a while. How can the root CA certificates that come with mozilla PSM be converted/exported to a format that is usable with openssl? Background. I want to use fetchmail-ssl with SSL encrypted IMAP (port 993) and verify the server certificate against man-in-the-middle. But the server certificate is chained to the Verisign/RSA root, and Verisign doesn't seem to provide any way to download the root (at least I haven't found any during my interminable journey through their $$$ oriented Web maze). Mozilla has all the roots built in, but not in the DER or PEM formats that openssl groks; in fact the mozilla roots are linked into libnssckbi.so as C structures, it seems. I downloaded the mozilla source (really!) and determined that the bits of the roots originate in the file certdata.txt. But even this file's format is a far cry from something that can be fed into openssl. Can somebody give me a hint how to convert it? I am not familiar with DER; I guess a reference to the definition of DER would be enough to make me grateful (though not yet happy :) Thanks, -- Ian Zimmerman, Oakland, California, U.S.A. Hypocrisy, arrogance and manipulation: sure-fire ways to earn hatred. GPG pub key: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087
svgatextmode - fbset: not so smooth
Hello everyone. I just switched to a more modern graphics board with VESA 2.0 support. This board is not supported by svgatextmode, and won't be as there's no development of svgatextmode anymore. So I have to use the framebuffer console, instead of svgatextmode, to provide high-resolution text display (which is quite important to me). The modes I can select for the framebuffer at boot are limited to the VESA modes, for example to get 128x48 characters I need vga=0x305 which is 1024x768 pixels at 60Hz refresh rate. But the hardware can do much better: in X I can have this resolution with 95Hz! Now it seems I should be able to set the desired mode with fbset after booting, just like I did with svgatextmode. Unfortunately, there's a catch: fbset only changes the mode _for the current virtual terminal_. As soon as I switch to another vt, I am back to the default mode from boot. So I tried something like this #! /bin/sh for i in 1 2 3 4 5 6 7 8 ; do chvt $i ; fbset 1024x768-95 done chvt 1 but not even this seems to work! The mode is changed - for vt1 only, just as before. Somehow the kernel seems to tightly associate the framebuffer settings with vt1 until I switch to another vt _interactively_. So the only way to change the mode for all vts is to log in on each one and run fbset manually. On top of that fbset can only be run as root, so this must be done with su/sudo :-( All this almost makes me regret the upgrade. Is there a way around this difficulty? Thanks, -- Ian Zimmerman, Oakland, California, U.S.A. The easiest way to win an argument: ridicule your opponent's basic assumptions by stating their negation and postfixing it with , right? GPG pub key: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087
policy/philosophy: bugs that depend on input files
I recently filed a bug with, in my opinion, quite sufficient documentation. At least the level of documentation was such that had been routinely accepted with other bugs I'd filed in the past. In a subsequent update, I even enclosed all the input files that caused the erroneous behaviour, verbatim. Nonetheless, the maintainer closed the bug with a dismissive and supercilious comment (on record in the BTS), citing as reason that the input files are not part of Debian. Now, the package in question is a development package, a kind of compiler, one might say. Imagine hitting a bug in gcc, like the one that made all Linux kernels redundantly include -fno-strength-reduce in CFLAGS for eternity, documenting it with the C source and the erroneus assembly output, and being told that it is not sufficient, that you should actually debug the monster of a program that gcc is with gdb. To me this would seem absurd, and so does the present case. What do you think? Please reply to me even if you choose to copy to the list (feel free to do that). -- Ian Zimmerman, Oakland, California, U.S.A. EngSoc adopts market economy: cheap is wasteful, efficient is expensive. GPG pub key: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087
Re: Emacs: Memory exhausted
zolia I am getting message (Memory exhausted--M-x..for recovery..) when zolia starting emacs (20.7-3) on woody. colin Are you sure you aren't actually out of memory? How much memory do colin you have on that machine? colin colin Also, do you have any limits set on memory usage for that user? What colin does 'ulimit -m' say (in bash)? zolia Hello, zolia lydys:~$ ulimit -m zolia unlimited zolia lydys:~$ free zolia total used free sharedbuffers cached zolia Mem:190940 185936 5004 0 6304 119860 zolia -/+ buffers/cache: 59772 131168 zolia Swap: 124956 23244 101712 zolia when i used potato, there weren't any problems. Problem persist zolia even on free system :/ Maybe there is a way to trace the zolia problem? Check out recent archives for gnu.emacs.bug, I remember there was somebody reporting a stack overflow (due to infinite recursion, of course) in y-or-n-p. -- Ian Zimmerman, Oakland, California, U.S.A. EngSoc adopts market economy: cheap is wasteful, efficient is expensive. GPG pub key: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087
Re: True Type fonts
Section Files RgbPath /usr/X11R6/lib/X11/rgb FontPath unix/:7101 FontPath /usr/share/fonts/truetype FontPath /usr/lib/X11/fonts/misc/:unscaled FontPath /usr/lib/X11/fonts/cyrillic/:unscaled FontPath /usr/lib/X11/fonts/100dpi/:unscaled FontPath /usr/lib/X11/fonts/75dpi/:unscaled FontPath /usr/lib/X11/fonts/Type1/ FontPath /usr/lib/X11/fonts/Speedo/ FontPath /usr/lib/X11/fonts/misc/ FontPath /usr/lib/X11/fonts/cyrillic/ FontPath /usr/lib/X11/fonts/100dpi/ FontPath /usr/lib/X11/fonts/75dpi/ ModulePath /usr/X11R6/lib/modules EndSection Stewart Ok. I can leave the unix/:7101 uncommented and it starts x Stewart just fine. The truetype path is what is causing the can't Stewart find default font 'fixed' error. Of course. The X server itself doesn't know iota about True Type. Only xfstt understands them. So X gets some kind of error when processing the truetype line (which shouldn't be there), and never reads the rest of the paths, including the misc one (which is where fixed resides). -- Ian Zimmerman, Oakland, California, U.S.A. EngSoc adopts market economy: cheap is wasteful, efficient is expensive.
Re: problem with new (unstable) xlibs
John I just did an update yesterday, and xlibs won't upgrade. John Basically it fails when it tries to update app-default. I John included my log of what happens if it helps. [...] Preparing to replace xlibs 4.0.1-1 (using .../xlibs_4.0.2-1_i386.deb) ... Unpacking replacement xlibs ... dpkg: error processing /var/cache/apt/archives/xlibs_4.0.2-1_i386.deb (--unpack): trying to overwrite `/usr/X11R6/lib/X11/app-defaults', which is also in package ddd dpkg-deb: subprocess paste killed by signal (Broken pipe) ... Errors were encountered while processing: /var/cache/apt/archives/xlibs_4.0.2-1_i386.deb E: Sub-process /usr/bin/dpkg returned an error code (1) [EMAIL PROTECTED]:~$ exit exit Looks familiar :( AFAIK this is really a problem with all the other packages (like ddd, also groff and a zillion others) that put stuff into /usr/X11R6/lib/X11/app-defaults. They need to to changed to put the stuff into the real location /etc/X11/app-defaults instead. What I ended up doing: a) download the .deb without installing it (if you use apt, it's already in your /var/cache/apt/archives, probably) b) use dpkg-deb --extract and dpkg-deb --control to unpack the package in a private tree (be sure to man dpkg-deb 1st) c) delete the stupid ./usr/X11R6/lib/X11/app-defaults symlink from the private tree d) recreate the package with dpkg-deb --build e) install the recreated package locally with dpkg --install HTH, -- Ian Zimmerman, Oakland, California, U.S.A. EngSoc adopts market economy: cheap is wasteful, efficient is expensive.
Re: Ummm...Losing disk space. HELP.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Rob I installed Debian 2.2 on a Gateway PC (PII-233) with no
Rob prbolems, and it's been running fine. However, it just recently
Rob ran out of disk space for some reason. Uppon investigation, I
Rob found that the /var/log dir was HUGE, specifically the kern.log
Rob file. I removed all the files and rebooted the machine. The
Rob kern.log file is continually growing in size. When i look at it,
Rob I see this over and over again:
Feb 7 20:22:54 Dale kernel: VFS: Disk change detected on device
ide1(22,0)
Feb 7 20:22:54 Dale kernel: hdc: packet command error: status=0x51 {
DriveReady SeekComplete Error }
Feb 7 20:22:54 Dale kernel: hdc: packet command error: error=0x50
Feb 7 20:22:54 Dale kernel: ATAPI device hdc:
Feb 7 20:22:54 Dale kernel: Unknown Error Type: (reserved) -- (Sense
key=0x0d)
Feb 7 20:22:54 Dale kernel: (reserved error code) -- (asc=0x6a,
ascq=0x58)
Feb 7 20:22:54 Dale kernel: The failed Read Subchannel packet
command was:
Feb 7 20:22:54 Dale kernel: 42 02 40 01 00 00 00 00 10 00 00 00
Etcetera, etcetera.
Bingo - exact match to my own problem, which I mentioned here several
months ago, without getting a reply, and which I developed a
workaround - not a solution - just a few days ago.
/dev/hdc is your CDROM, and it is not completely conforming with the
IDE/ATAPI standard. This incompatibility only manifests itself when
there is an error in the Read Subchannel command, which almost all
CDROM audio playing programs (starting with workman, I think actually
many programs simply copied workman's code for the hardware access)
use to poll the drive every 5 seconds or so. And no disc in the drive
qualifies as an error :(
You probably run GNOME or something and have a CDROM playing program
in the panel. Guessing right?
My workaround was to find a CDROM player that doesn't use the polling
design (which I think is terrible anyway even if it works). XFreeCD
is one such. Of course I use no desktop so I am not bound by any
panel compatibility constraints.
Please copy any followups to me; I once again got tired of the volume
on this list.
- --
Ian Zimmerman, Oakland, California, U.S.A.
EngSoc adopts market economy: cheap is wasteful, efficient is expensive.
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.5 and Gnu Privacy Guard
http://www.gnupg.org/
iD8DBQE6gjQMjKMn20M7oIcRApIpAKCAR1Bd0Z1cVPZz0B+RzekLG8V4rQCdFPq0
f3Scj+QZl16N9YRZfDuSz7U=
=5gUB
-END PGP SIGNATURE-
Re: debian init.d scripts
Karsten S is single user. Same as '1'. itz This is not quite right. itz itz S is the real single-user level. 1 OTOH is used just as a itz way-station Ethan not quite, S is for script that need to be run once per boot. Ethan its run on boot regardless of what runlevel will be entered. Ethan things like reading the hardware clock, mounting filesystems Ethan and such need not be done on runlevel changes (even multi - Ethan single and back). itz:~$ ls /etc/rc1.d/ K01wdm K19bindK20ippl K20mysql K20xfs-xtt K90sysklogd K01xdm K20apache K20logoutd K20queue K21autofs S20single K11cron K20eximK20lpd K20snort K25ipfw.sh K12kerneld K20gpm K20makedev K20ssh K83chrony K14ppp K20inetd K20modclean K20swapd.sh K89atd itz:~$ cat /etc/rc1.d/S20single #! /bin/sh # # singleexecuted by init(8) upon entering runlevel 1 (single). # # Version: @(#)single 1.10 22-Jun-1998 [EMAIL PROTECTED] # PATH=/sbin:/bin:/usr/sbin:/usr/bin # Kill all processes. echo -n Sending all processes the TERM signal... killall5 -15 echo done. sleep 5 echo -n Sending all processes the KILL signal... killall5 -9 echo done. # We start update here, since we just killed it. update echo Entering single-user mode... exec init -t1 S itz:~$ itz from multiuser to kill all the multiuser services before entering itz S with clean slate. If you say init S in multiuser you will itz end up with a root password prompt because all the getties will itz be killed (being under direct control of init), but the services itz will stay up. Ethan init S == init 1 This is in contradiction to what you say above, apart from also being factually wrong: itz:~$ cat /etc/inittab # /etc/inittab: init(8) configuration. # $Id: inittab,v 1.8 1998/05/10 10:37:50 miquels Exp $ # The default runlevel. id:2:initdefault: # Boot-time system configuration/initialization script. # This is run first except when booting in emergency (-b) mode. si::sysinit:/etc/init.d/rcS # What to do in single-user mode. ~~:S:wait:/sbin/sulogin # /etc/init.d executes the S and K scripts upon change # of runlevel. # # Runlevel 0 is halt. # Runlevel 1 is single-user. # Runlevels 2-5 are multi-user. # Runlevel 6 is reboot. l0:0:wait:/etc/init.d/rc 0 l1:1:wait:/etc/init.d/rc 1 l2:2:wait:/etc/init.d/rc 2 l3:3:wait:/etc/init.d/rc 3 l4:4:wait:/etc/init.d/rc 4 l5:5:wait:/etc/init.d/rc 5 l6:6:wait:/etc/init.d/rc 6 # Normally not reached, but fallthrough in case of emergency. z6:6:respawn:/sbin/sulogin # What to do when CTRL-ALT-DEL is pressed. ca:12345:ctrlaltdel:/sbin/shutdown -t1 -a -r now # Action on special keypress (ALT-UpArrow). kb::kbrequest:/bin/echo Keyboard Request--edit /etc/inittab to let this work. # What to do when the power fails/returns. pf::powerwait:/etc/init.d/powerfail start pn::powerfailnow:/etc/init.d/powerfail now po::powerokwait:/etc/init.d/powerfail stop # /sbin/getty invocations for the runlevels. # # The id field MUST be the same as the last # characters of the device (after tty). # # Format: # id:runlevels:action:process #1:2345:respawn:/sbin/getty 38400 tty1 2:2345:respawn:/sbin/mingetty tty2 3:2345:respawn:/sbin/mingetty tty3 4:2345:respawn:/sbin/mingetty tty4 5:2345:respawn:/sbin/mingetty tty5 6:2345:respawn:/sbin/mingetty tty6 #8:2345:respawn:/usr/local/sbin/toproot # Example how to put a getty on a serial line (for a terminal) # #T0:23:respawn:/sbin/getty -L ttyS0 9600 vt100 #T1:23:respawn:/sbin/getty -L ttyS1 9600 vt100 # Example how to put a getty on a modem line. # #T3:23:respawn:/sbin/mgetty -x0 -s 57600 ttyS3 itz:~$ itz If you look at /etc/init.d/rcS (the script that init executes itz when entering S) you'll see that _all_ the sub-scripts in itz /etc/rcS.d/ are run with start as a parameter - which means rcS itz _never_ kills anything. If you do init 1 /etc/init.d/rc is run itz instead, which knows how to kill the services by calling itz sub-scripts with a stop parameter. The last sub-script then itz does init S. itz itz I don't know why it is done this way; it looks like an ugly hack itz to me. I'd just have rcS work the same way as rc (indeed they itz should be the same script). Ask me when I am a developer :) Ethan its done this way so that its simple to add something to the Ethan `must be run at boot once' list of things to do. without it Ethan you would have to edit the main initscript /etc/init.d/rc. Ethan (which a debian package can't do without violating policy.) Ethan the rcS.d is basically the equivilent of redhat's rc.local. Ethan redhat has all the `once at boot' stuff crammed into a much Ethan more bloated /etc/rc script, which runs rc.local for anything Ethan packages or the admin needs added. very messy. rcS.d is far Ethan cleaner and simpler. You didn't get my meaning. I don't object to there being a small rcS script which run-parts a directory; I object to there being rc _and_ rcS which do the same thing, but are distinct. -- Ian
Re: Control CPU fan's speed
Nate my p3-800 still runs at 50C(AFAIK when its idle, as i can't Nate query the cpu from linux yet lm_sensors doesn't support the Nate chip..so 50C is from the bios) I labored under the same misconception until Frodo (author of lmsensors) corrected me. The CPU is very far from idle when you're in the BIOS: in fact it is in a tight loop waiting for keyboard input. The reading you get from BIOS is likely _higher_ than one you will/would get from lmsensors on a normally loaded Linux system. -- Ian Zimmerman, Oakland, California, U.S.A. EngSoc adopts market economy: cheap is wasteful, efficient is expensive.
Re: debian init.d scripts
Karsten S is single user. Same as '1'. This is not quite right. S is the real single-user level. 1 OTOH is used just as a way-station from multiuser to kill all the multiuser services before entering S with clean slate. If you say init S in multiuser you will end up with a root password prompt because all the getties will be killed (being under direct control of init), but the services will stay up. If you look at /etc/init.d/rcS (the script that init executes when entering S) you'll see that _all_ the sub-scripts in /etc/rcS.d/ are run with start as a parameter - which means rcS _never_ kills anything. If you do init 1 /etc/init.d/rc is run instead, which knows how to kill the services by calling sub-scripts with a stop parameter. The last sub-script then does init S. I don't know why it is done this way; it looks like an ugly hack to me. I'd just have rcS work the same way as rc (indeed they should be the same script). Ask me when I am a developer :) -- Ian Zimmerman, Oakland, California, U.S.A. EngSoc adopts market economy: cheap is wasteful, efficient is expensive.
Re: upgraded libc6 and libc6 and broke my system (Woody)
Bryan == Bryan Walton [EMAIL PROTECTED] writes: Bryan Hi Andrew, Rather mysterious. Nothing has broken. Restarted Bryan exim (in a deliberate attempt to break it - if the problem was Bryan really there), and is continued to work. Even rebooted (I Bryan would much rather have it break when I am expecting it). Bryan Everything still works. And /sbin/ldconfig is still there. I Bryan don't understand, but, nevertheless, I have downloaded the Bryan potato versions of the the .debs needed to fix things should Bryan the need arise. Is it possible that the problem is not Bryan universal? Since the problem seems to be vaguely located in the db part of libc, it is possible that it only affects exim installations that actually do db lookups (which I think is rare on typical desktops, most people just use flat text file maps). -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: TOT: Virus reports to the list and x-envelope-to:
Ethan == Ethan Benson [EMAIL PROTECTED] writes: George If you are going to bounce a message, send it to the envelope George sender or the Errors-To: address. Ethan looks to me like the only header that seems appropriate is Ethan Return-Path: Ethan that is the only real header with the Ethan bounce-debian-user-blahlbahlbah, but that head appears in From Ethan (NOT From: with the colon) and in an X-From_: The From (sans colon) line is NOT part of the message, it is an artifact of the mailbox file format. By sheer accident it is sometimes set to the envelope sender address :-) -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: sawfish doesn't work
Michael == Michael Smith [EMAIL PROTECTED] writes: Michael I think the default sweetpill theme for sawfish is missing Michael from the sources. This means that you have no raise-lower Michael buttons or even frames on the windows. Try using sawfish and Michael then running the configuration tool. Change the appearance Michael to microgui or anything else, and then close out all your Michael current windows. The new ones should be OK. I've been Michael installing about 5 workstations this week, and that's what I Michael had to do with all of them. The missing theme is in another package, helix-sweetpill or so. The helix people made it the default but omitted adding a dependency to the sawfish package. So if this is the whole problem, just apt-get install helix-sweetpill first and it should work. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
[OT] History: GNUStep vs. Gnome
Hello. This crossed my mind more than once: why was the Gnome project started, when there was already something called GNUStep? Even if not a full desktop, one can certainly see the beginnings of desktop-like functionality in WindowMaker or AfterStep. Was it just not cool-looking enough? :( -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
woody: ssh and rxvt
Last night I did apt-get upgrade to woody. It works well so far, but I came across two small problems: 1/ Something seems to be wrong with rxvt. Full screen programs (jed, nano) suddenly don't understand keypad keys, and I can't use the Compose key (which is correctly defined in my .Xmodmap). This could be a problem with rxvt itself, ncurses, or the rxvt terminfo entry. I worked around this by switching to aterm (I never used enough rxvt features to know if this was an upgrade or downgrade :) 2/ When I ssh into my system from outside, I get the last login notification line _twice_: a correct one as the very first line of output, and a bogus one just before normal shell output (from .profile and such). Here's a sample: Last login: Mon Sep 25 12:38:33 2000 from some.other.host on pts/0 Debian GNU/Linux woody (kernel 2.2.16). You have mail. Last login: Mon Sep 25 12:39:16 2000 from some.other.host Tuesday September 26, 2000 * 7:30pm to 10:30pm Cal Bridge Club at 123 Wheeler Do you want to see the TODO list? n kronstadt:~$ -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Making chrony work
Kendall == Kendall Shaw [EMAIL PROTECTED] writes: Kendall Hi, I've installed chrony, but it doesn't go. Can you make it Kendall go? Kendall More specifically, chronyd is running, and Kendall /etc/chrony/chrony.conf points to a stratum 2 server in santa Kendall cruz and another in quincy, CA. There's an entry in syslog Kendall which reads: Kendall chronyd[4028]: Selected source 165.227.1.1 Kendall But, still the time is wrong. Is there a large difference between the correct time and your system time? chrony (and other NTP clients, as far as I know) never changes the time all at once, because that would hurt cron and other time sensitive things. Instead, it uses the adjtime system call to make the system clock run just a little bit faster (or slower as the case may be) until it's right. Cheers, -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Debian Menu with Sawfish (Helix)
Rino == Rino Mardo [EMAIL PROTECTED] writes: Rino On Wed, Sep 06, 2000 at 04:01:46PM +0200 or thereabouts, Kai Rino Weber wrote: Kai has anyone the same experience with (all packages up-to-date) Kai HelixGnome and Sawfish: Kai Kai The middle mouse button, which brings up sawfish's root menu Kai contained under programs the whole Debian menu with apps Kai ... Since some days I miss it! There are only some entries: Kai xterm, Emacs, Netscape and others. Rino i think you'd have to edit sawmill's config files like in Rino blackbox. in blackbox if i want something in the menu or out of Rino the menu i edit blackbox's config files. it's easy coz but in Rino sawmill you'd have to do it using Lisp. No, this is not the Debian way :) First try running update-menus. (You can run it as a normal user, too). If that doesn't help, it's a bug (in the Debian package, not in sawfish itself). I don't have the potato version of sawfish - the last one I saw was still sawmill. In sawmill the important file which made the menu to be read was /usr/share/sawmill/site-init.jl (or maybe it was in a subdirectory of that called something like lisp or site-lisp, I don't remember exactly). Where that file reads the menu from must agree with the location update-menu creates it (and that is specified in /etc/menu-methods/sawmill). -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: remote backspace/delete problems
loren == loren jan wilson [EMAIL PROTECTED] writes: loren i have debian 2.2 installed to my computer (a very recent loren install via ftp). when i make a remote connection via ssh or loren telnet to a solaris box or a redhat box, my backspace/delete loren keys don't work in nvi/vim. this is ONLY in X...everything loren works fine at the console. loren attempting to backspace this word---hello attempting loren to move to the beginning and delete this word---HELLO (turns loren it to caps--it beeps when i do that, too) loren the keys work fine in bash. in emacs, the backspace key works, loren but the delete key doesn't. everything on the local machine loren works fine. i've looked through many archives and web search loren results trying to fix this problem, and i've spent /way/ too loren long without finding a fix. perhaps this is evidence of loren inferior mental capacity on my part? doesn't this seem a little loren weird to anybody else? maybe i messed up during my x loren configuration? This is a knotty area, but my first line of attack would be termcap/terminfo, and the TERM variable. Basically, you want the terminal database for whatever emulator you're using (xterm, xterm-debian, rxvt?) to be available on the remote machine. If the systemwide files don't have that entry (and it is unlikely that an xterm-debian entry, for instance, is installed on a Solaris), create a private terminal database under your home directory, upload the entry there, and set the corresponding environment variable (TERMCAP or TERMINFO) to point to your private tree. For example: on the remote machine, create a directory ~/terminfo/x upload /usr/share/terminfo/x/xterm-debian to that directory add this to ~/.ssh/environment on remote machine: TERMINFO=~/terminfo:/usr/share/terminfo -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: kernel hangs after trying to free unused memory.
Jon i am running a potato version of debian. I have just compiled a new 2.2.17 Jon kernel with kernel-package and make-kpkg. i executed the following commands Jon as per the README file. Jon make config Jon make-kpkg clean Jon make-kpkg --revision=cutom.1 kernel_image Jon then Jon dpkg -i kernel_image.deb (or something) using the defaults. Jon rebooting starts and then hangs here... Jon Partition check: Jon hda: hda1 hda2 Jon VFS: Mounted root (ext2 filesystem) readonly. Jon Freeing unused kernel memory: 60k freed Jon then it just hangs with a solid cursor. A _lot_ of things happen at that point. It is just before init is executed (during a successful boot, the next message would be INIT v.2.78 booting or something like that). I got hung at this point when my SCSI BIOS setup was wrong for my disk. But obviously you don't have SCSI so that doesn't apply to you. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: kde or gnome?
Joachim == Joachim Trinkwitz [EMAIL PROTECTED] writes: itz window manager, and at present the only window manager with full itz support for Gnome seems to be Enlightenment, AKA `E'. Joachim That's nonsense, many window managers have Gnome support now, Joachim e.g. icewm-gnome, wmaker, sawfish-gnome. Joachim Most integrated in Gnome is sawfish, which is configurable Joachim through the Gnome control center. Sawfish seems to be quite Joachim so lightweight; please try it out. Read the quote again. I wrote the only wm with _full_ support etc. If sawfish is Most integrated, the others are necessarily less integrated, or not? I don't want the wm and the desktop to step over each other, and that's precisely what happens with wmaker when the wmaker frills and the desktop frills overlap. I did omit to mention sawfish, for which I apologize. Another point is that I want to use the debianized versions of things where possible, and I think that's what the original question was about, too. I am sure many of the wm's have better Gnome support now in their bleeding edge versions, but I will not use them. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
gcd (playing audio CDs)
Hi, I installed the gcd package and I was able to play music CDs that way. However, when gcd is running and there's no disc in the drive, I get a constant stream of the following in the logs: Sep 4 02:20:08 kronstadt kernel: ATAPI device hdc: Sep 4 02:20:08 kronstadt kernel: Unknown Error Type: (reserved) -- (Sense key=0x08) Sep 4 02:20:08 kronstadt kernel: (reserved error code) -- (asc=0x08, ascq=0x08) Sep 4 02:20:08 kronstadt kernel: The failed Read Subchannel packet command was: Sep 4 02:20:08 kronstadt kernel: 42 02 40 01 00 00 00 00 10 00 00 00 This is repeated every second, exactly the same values :-( I am now scared to play even though is seems to work, because I remember that my previous CD-ROM drive (also IDE/ATAPI) stopped working (for music _and_ data) shortly after I installed another audio disc player (a text/curses based one) with similar symptoms. The drive works perfectly for reading filesystem data. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: ippl-listfiles vs logrotate
Thomas == Thomas Guettler [EMAIL PROTECTED] writes: Thomas It think the way ippl handles logfiles is confusing. I Thomas searched very long to find out why my logfiles get rotated Thomas although they are not mentioned in logrotate.conf. (The get Thomas rotated by script which calls ippl-listfiles) Thomas What do you think? Should I do send a bug with category Thomas request for feature? AFAIK the logrotate package is mostly there to handle logs of programs invented in Red Hat land, which depend on it. The default Debian way is for each package to handle its own log rotation, like ippl does. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: kde or gnome?
Felix == Felix Natter [EMAIL PROTECTED] writes: Helgi Of course it depends on who you ask.:-) Generally I like Helgi KDE because it is locical in so many ways, but I don't like how Helgi big it is and therefore takes a lot of resources, and it's not Helgi 'free'. Gnome is faster (for me) but I got problems with using Helgi different Felix kde 2.0 might well be faster than GNOME 1.2, because GNOME Felix still uses CORBA for object embedding, while KDE uses DCOP, Felix which builds on inter- process communication. Thus, DCOP is Felix more lightweight, but not fully network-aware (although network Felix communcations will be possible by connecting the DCOP-servers). This is from someone who just half a year ago wouldn't leave text mode, so take it with a grain of salt. First, KDE _is_ totally free. The only problem with it, license-wise, is that it uses Qt, whose license, although _also_ free, is incompatible with GPL. I tried both KDE and Gnome; in fact I tried installing Gnome multiple times, and always came to the same conclusion (which is: I run KDE :). The problem I have with Gnome is a bit subtle, and not obvious to someone who's installing Linux or even a Linux GUI for the first time and deciding between what's out there. Namely, Gnome does not include its own window manager; KDE does. Gnome depends on hooks for Gnome support compiled into an external window manager, and at present the only window manager with full support for Gnome seems to be Enlightenment, AKA `E'. And E is a _HOG_. I mean, it's a ho. There seems to be no easy way to make it not use bitmap textures for everything imaginable on the screen, including caption bars and even menus. The results are predictable. With KDE, the entire Linux boot sequence is still much faster than Windoze; with Gnome and E, it's a toss-up :-( That's on a 32M/P160 machine, which doesn't strike me as minimalistic. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Debian packages..
Michael == Michael Fox [EMAIL PROTECTED] writes: Michael Hi, Problem I am having is with packages. I have an old Michael machine that is still running Slink, and at this time I do Michael not wish to upgrade it. Michael With potato changing to stable, I have since modified my Michael sources.list to reflect the fact that stable is not what I Michael want, but rather slink Michael deb http://http.us.debian.org/debian slink main etc etc Michael I then do a apt-get update and to update my package lists. Michael This works somewhat fine, the problem happens when I wish to Michael install a package for slink, Michael apt-get install package Michael At this point it fails badly, because the stuff that I Michael downloaded with apt-get update, all the paths in this file Michael contain the path Michael blah/dists/stable/binary-i386-package/net/ntop-3-4.deb Michael Which means it will fail, as stable of course is now the Michael potato tree, shouldn't the package list files change to Michael reflect the codename of the release, rather then stable. As I Michael wouldn't be having this problem. Michael Can anyone tell me if the maintainers expect to update the Michael package lists files for slink, so that the paths are right, Michael as since potato is out, it FAILS big time. Michael Please email me anyone if you can help me out. Me too :-( I'm doing this just to make whoever replies post the reply to the list, and not send to Michael alone. There must be more of us in this situation. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: KDE2 debs from TDYC
Michael == Michael Epting [EMAIL PROTECTED] writes: Michael There have been some new .debs made available from tdyc the Michael last couple of days. These *almost* work for me, unlike the Michael ones available before yesterday. If anybody else is Michael successfully using these, could you please tell me how you Michael got a panel to appear? Michael By the way, I had to modify /usr/bin/startkde to get much of Michael anything to work. I added: Michael kdeinit +dcopserver kdeinit +klauncher Michael and, later on Michael kdesktop Michael but no joy on the panel. Hmmm. I installed KDE for the first time this last weekend, and it pretty much works outta the box (including the panel, which I have immediately auto-hidden :) What I found interesting after reading your post: kronstadt:~$ file /usr/bin/startkde /usr/bin/startkde: broken symbolic link to /kde kronstadt:~$ I use kdm (of course), do you? And this is on potato, with the potato debs from tdyc. These shouldn't really change from now on, should they? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Mapping keys to console
Dave == Dave A [EMAIL PROTECTED] writes: Dave Hi, Can anyone help with this problem. I need to map keys to Dave mean different things depending on which console I am on. For Dave example, I need to have a menu screen on one session and an POS Dave order taker on the other. Dave Each has to have keys setup to mean different things. So far it Dave appears that linux only supports a global change. If I map to Dave one console it applies to all sessions. I need them to act Dave independently. If you're a C programmer, Linux has an ioctl that allows a program to set up notifications (with a selectable signal) whenever the virtual terminals are switched. So you can write something that runs in the background, traps these switches and reloads the keymap based on a config file. Start by looking at /usr/include/linux/vt.h if you decide to take the challenge. :) -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
weird ssh message
Starting today, for no obvious reason, when I ssh out from my (one and only remaining) slink box, I get this message twice: Warning: cannot stat authentication directory /tmp/ssh-itz after that, the connection proceeds successfully. I notice there's a directory /tmp/ssh-blahblah (or similar:) on the destination box, owned by me and permed 700, but of course no /tmp/ssh-itz. The destination machines are potatoes :-) This does not happen for connections in the other direction, nor did I ever notice it before in this direction. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: xconsole
john == john gennard [EMAIL PROTECTED] writes: john KDE is installed and seems to behave as before except that on john booting xconsole screen appears with no contents and a new john screen starts each time - I now have 8, one on top of the john other. Whilst these can be minimized or maximized, they cannot john be closed from the 'x' button nor the menu. john I am unable to find out exactly what xconsole is, how it works john and where the configuration details are. I've looked at the man john pages, at /usr/X11R6/bin/xconsole, and john /usr/lib/X11/app-defaults/XConsole, but am too inexperienced to john make any sense of things. Xconsole is supposed to show messages coming from syslogd(8). For it to work under linux, an entry has to be made in /etc/syslog.conf looking like this: # The named pipe /dev/xconsole is for the `xconsole' utility. To use it, # you must invoke `xconsole' with the `-file' option: # #$ xconsole -file /dev/xconsole [...] # # NOTE: adjust the list below, or you'll go crazy if you have a reasonably # busy site.. # daemon.info;mail.info;\ auth.info;*.notice;cron.* |/dev/xconsole and of course /dev/xconsole must be an existing named pipe. What is a bit unusual about xconsole is that it is the only X program that runs while xdm (or its replacements kdm or gdm) waits for user login (though this can be disabled in /etc/X11/xdm/xdm-options). I've observed that attempts to start an additional copy of xconsole after logging in, either from a script like ~/.xsession or from an xterm, don't result in the creation of a new window: instead, the command line parameters (window geometry, font etc.) seem to be applied to the existing window. I don't know myself how this works, and I'd be grateful if someone can explain; but my point is that it seems there is a problem with this mechanism on your system. KDE probably runs xconsole after you log in just because it remembers it had one when you logged out, and instead of reconnecting to the running copy (passed in from kdm) it somehow ends up creating a new window. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Setting proper From: lines
Spinfire == Spinfire Magenta [EMAIL PROTECTED] writes: Spinfire I'm not quite sure if this is an Exim issue or a MUA (Mutt) Spinfire issue, but... here goes: Spinfire I have a machine that is both my workstation and the Mail Spinfire Exchanger for my domain (isomerica.net). Mail sent to Spinfire [EMAIL PROTECTED] works fine, however, mail sent from the Spinfire host is instead tagged with the full hostname (From: Spinfire [EMAIL PROTECTED]) instead of the preferred Spinfire '[EMAIL PROTECTED]' form. Spinfire Is this an Exim issue (i did use isomerica.net as the Spinfire 'outgoing mail name' during the Exim configuration) or if Spinfire this is a Mutt issue. It can be solved with Exim (look for Address rewriting in the docs) but that is more like a work-around. Better tell Mutt to set the From header correctly in the first place. Sorry, no Mutt experience here. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: [ykarant@csci.csusb.edu: xdm potato wm no go]
Yasha == Branden Robinson [EMAIL PROTECTED] writes: Yasha xdm *REFUSES* to exec my wm (fvwm which was formerly known as Yasha fvwm2). When I login via a regular scrolling terminal screen Yasha (e.g., screen on F5), and issue a startx command (the is Yasha background mode in tcsh), everything works. (That is, Yasha everything works after I login as root and issue a Yasha /etc/init.d/xdm stop command so that startx can grab the normal Yasha :0 screen on F7). I remember having a similar problem (not with a Debian packaged X, but with a home-compiled one), and tracing it to the GiveConsole and TakeConsole scripts somewhere under /usr/X11R6/lib/X11. Basically the scripts were broken by a subtle change in semantics in GNU fileutils (having to do with chmod or chown following symlinks). -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
netatalk nits
Hi, I installed netatalk on a potato system (which is also a Samba server). When I try to copy a folder from a share to a Mac, the folder itself is created OK but for every file foo in the folder the Mac displays the following message: The file foo couldn't be read, because it is in use. Do you want to continue copying? The folder is on one of the Samba shares (actually, I set the netatalk shares to be the same as the Samba ones), and Windows machines can read and write it fine. What does this mean? -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: update-alternatives -- changing preferences
Karsten == kmself kmself@ix.netcom.com writes: Dave The easy way would be Dave Dave update-alternatives --config x-window-manager Dave Dave and select wmaker from the list to be your default. Karsten For those who are interested, the following script (run in Karsten /etc/alternatives) will list all manually configured Karsten alternative links: Beware! This does _not_ quite work as expected. It will update the 'master' link but not the 'slave' links (e. g. I bet man editor on your system will still display the nvi manpage). This is just a bug in update-alternatives IMHO, and I reported it as such after discussing these issues on the list some months ago, and I enclosed an ugly perl hack that serves as a temporary replacement for me. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Favorite getty for serial consoles?
Simon == Simon Tennant [EMAIL PROTECTED] writes: Simon I've been tearing my hair out trying to get an old dec VT320 to Simon correctly talk to my computer. I can login but every 2 or 3 Simon lines get a wierd control character and lots of backward ?s. Simon I've tried agetty and gettyps. Simon I'd like to be able to specify parity, hardware flow control Simon etc on the server end. Does anyone have a favorite getty that Simon permits this? mgetty is the normal choice, but I think gettyps should work as well; if it doesn't something is misconfigured. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: char-major-6
Erik == Erik Mathisen [EMAIL PROTECTED] writes: Erik I keep getting this stupid error in my syslog: modprobe: can't Erik locate module char-major-6 Erik now i searched my system, I dont have that module, how do I get Erik this error to stop? It puts 2 or 3 entries in the log a minute. 6 is the major device number for the parallel ports (ie. printer). You probably don't have char-major-6.o :-), but do you have lp.o? If so add this line to /etc/modutils/aliases: alias char-major-6 lp then run /sbin/update-modules. That should stop the annoyance. However, maybe a more pertinent question would be, what/who is trying to access the port? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Unthinkable routing problem
Nagarjuna == Nagarjuna G [EMAIL PROTECTED] writes: Nagarjuna It is unthinkable for us that a machine can forward packets Nagarjuna without itself able to approach the router. !!! Pl mail if Nagarjuna you require more information on this. Post the output of netstat -rn on the gateway. You can alter it to protect the innocent, but take care to substitute equals for equals, and only. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Tape backup software?
Kelly == Kelly Corbin [EMAIL PROTECTED] writes: Kelly Has anyone had much experience w/ tape backup in Linux? I am Kelly looking for tape backup software and was wondering if anyone Kelly knew which was the best. Any input would be appreciated. It Kelly would be for an ATAPI tape backup drive. THANKS! Lots of other people already wrote about the software. I'll only warn about ATAPI. I could not get my drive (HP/Colorado 5000) to work reliably. It would seem to work for a while but then start failing mysteriously in the middle of a tape, always nearly at the same spot (and the tapes were OK, I'm sure of it). I got the source for the last version of the driver and compiled it, it didn't help. Then I bought the cheapest SCSI drive I could find and it works perfectly. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Cannot find map file
chris What does this mean? -chris chris chris Jul 25 08:42:33 cr275960-a kernel: Cannot find map file. chris Jul 25 08:42:33 cr275960-a kernel: No module symbols loaded. Marcin I think You didn't copy System.map from /usr/src/linux (or anywhere you have Marcin kernel sources) to /boot or changed line: map=/boot/map in lilo.conf to Marcin something else. Some serious confusion here! lilo (and /boot/map and the lilo.conf line mentioning it) has absolutely nothing to do with this. And if you install kernels the Debian way (via make-kpkg) it creates a /boot/System.map file for you. So really the correct answer is: man make-kpkg. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Amanda backup: Labeling Tapes
Erik Dear Group, I feel that I have come quite far in setting up Erik amanda on my Debian/GNU Linux 2.2 system. I seem to be able to Erik connect from client to server, access the tape drive (SCSI Erik /dev/nst0) and such. Erik However, since I am still testing I have not yet set up the Erik crontab entry but would like to supply some commands manually. Erik However, as soon as I perform a dump, I get an email asking for: Erik a new tape. I have tried amlabel to write all kinds of Erik labels. This seems to work but I have not found a way to get Erik amanda to accept it. I have not been able to find in the Erik documentation how to go about this, other than (from the Howto): Erik Run amlabel to label all the tapes needed according to your Erik needs. Have you set the labelstr parameter in amanda.conf? Mine says labelstr ^DailySet1-[0-9][0-9]*$ # label constraint regex: all tapes match and my tapes are labelled DailySet1-01, DailySet1-02, etc. .. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: does 'apt-get remove' nuke packages recursively?
chris == Krzys Majewski [EMAIL PROTECTED] writes: chris Suppose package C depends on package B depends on package A. chris Now, it just so happens that no other packages depend on B, but chris package D also depends on A. If I do a apt-get remove C is it chris supposed to do the right thing and remove B as well (but not chris remove A)? I don't think so. How should it know what the right thing is? You're probably thinking of B as some library package, but it can be a full blown application package in its own right, one that you want to keep. OTOH I believe that apt-get remove B will do what you want _in this case_. Of course, that requires you to first check the back dependencies of A, then the forward dependencies of B manually. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
__STDC__
Why does gcc not predefine the __STDC__ macro now? Almost anything I try to compile is broken by this. kronstadt:/usr/include$ gcc -E -dM -xc /dev/null #define __linux__ 1 #define linux 1 #define __i386__ 1 #define __i386 1 #define __GNUC_MINOR__ 95 #define i386 1 #define __unix 1 #define __unix__ 1 #define __GNUC__ 2 #define __linux 1 #define __ELF__ 1 #define unix 1 kronstadt:/usr/include$ -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: __STDC__
kronstadt:/usr/include$ gcc -E -dM -xc /dev/null #define __linux__ 1 #define linux 1 #define __i386__ 1 #define __i386 1 #define __GNUC_MINOR__ 95 #define i386 1 #define __unix 1 #define __unix__ 1 #define __GNUC__ 2 #define __linux 1 #define __ELF__ 1 #define unix 1 Mark That doesn't define __STDC__ on Slink either. Hmmm. I have actually solved the compilation problem, and it was unrelated; in other words, __STDC__ _is_ defined during normal compiles. Any gcc guru here to explain why it doesn't show in the output of the above command? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: __STDC__
Note I've sent a carbon to the list as I think this is still relevant. itz Hmmm. I have actually solved the compilation problem, and it was itz unrelated; in other words, __STDC__ _is_ defined during normal itz compiles. Any gcc guru here to explain why it doesn't show in itz the output of the above command? John I'm not a guru, but... I can only guess that this is defined John internally by the compiler. Only things that may change between John machines and compiler versions would be defined externally John like this. These defines are normally in a spec file John somewhere. gcc -v will tell you where the spec file is, under John /usr/lib/gcc-lib somewhere IIRC. The useful ones are the John architecture/CPU/binary formats, although we don't normally need John to worry about these unless you're embedding assembly code or John doing some cunning tweak. I know where my specs it, I checked it and this symbol is not mentioned. John I initially thought that maybe the -ansi flag was needed, but John this defines __STRICT_ANSI__. I travelled the same road .. John so gcc's pre-processor defines a whole lot of macros. Yes, but according to the info file spitting _all of them_ out is exactly what the -dM flag should do. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
snort
Does anybody have a solution or workaround for Bug #66057 (snort: 5snort from cron.daily exited with status 1)? Happens here too :-( -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Compiling BIND 8.2.2 P5
Derek == Derek Wueppelmann [EMAIL PROTECTED] writes: Derek I keep getting an error with a library. It's looking for -lfl Derek and it can't seem to find it. I'm running Debian 2.1 with some Derek packages from 2.2 Any thoughts on this would be helpful You have installed package flex, haven't you? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: telnet vs ^Z
Will i tried % bash $ vi and then '^Z', and still get nothing but a Will screen-flash (as if doing a quick redraw). That's not good enough, it has to be your login shell. See Stevens, A.P.U.E. pp. 246 and forward, why it makes a difference. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: telnet vs ^Z
Will == Will Trillich [EMAIL PROTECTED] writes: Will i want the remote (linux server) full-screen process (mutt, vi, Will etc) to suspend when i send ^Z, despite the fact that i'm using Will telnet, and not sitting at the console. Will hell, i just tried it from the perl debugger, which doesn't even Will use curses/ncurses for its dialogue, and ^Z blinks the screen Will there, too, with no suspend signal being sent to the job. Will i want my control-Z! (waaah!) % bindkey | grep Z ^Z - tty-sigtsusp ^[^Z First try a working shell like bash, then we'll talk :-) Seriously, I have had problems of this kind with csh type shells. Bash gets most of it right. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: gnuclient vs emacsclient
Timothy == Timothy C Phan [EMAIL PROTECTED] writes: Timothy Hi, What is the diff between the two? gnuserv/gnuclient is a bit more sophisticated. The main difference is that it allows you to execute any editor command remotely, not just opening a file. Timothy I could not get gnuclient to work! More information, please? Do you (gnuserv-start) in your .emacs? Timothy Secondly, How can I bring a file in as readonly from Timothy emacsclient or gnuclient? With emacsclient, AFAIK there's no way but of course you can make your server buffer read-only after it is opened just like any other buffer, with C-x C-q. With gnuclient (actually gnudoit), gnudoit '(find-file-read-only foo.txt)' -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: logrotate
itz Doesn't the logrotate package install a daily cron by itself ??!
itz It certainly did for me (in potato).
Sven In that case, how do I find out all user's crontabs? Are they
Sven all under /var/spool/cron/...?
Read the debian policy document. Debian has additional crontabs
(apart from user crontabs and the root crontab) in
/etc/cron.{d,daily,weekly,monthly}
The logrotate thing is (not surprising) in /etc/cron.daily/logrotate.
--
Ian Zimmerman, Oakland, California, U.S.A.
In his own soul a man bears the source
from which he draws all his sorrows and his joys.
Sophocles.
Re: safe to use woody packages in potato?
John == John Anthony Kazos [EMAIL PROTECTED] writes:
John And for another thing, I looked at ftp.debian.org and the
John disks-i386 directory for woody is empty. What then is the
John upgrade process? Using the dist-upgrade method of apt-get?
Yes.
John If so, I still need to know the answer to my previous question:
John Will system files, like /etc/{fstab,profile}, be overwritten,
John possibly irrevocably, by the upgrade process? The documentation
John on the site is nonexistant.
Files that can be manually edited per debian policy will not be
replaced without warning, I believe. But I haven't done this myself,
yet, so you want a 2nd opinion.
--
Ian Zimmerman, Oakland, California, U.S.A.
In his own soul a man bears the source
from which he draws all his sorrows and his joys.
Sophocles.
Re: logrotate
Sven == Sven Burgener [EMAIL PROTECTED] writes: Sven Hi debs On the subject of logrotate, how do you guys put that to Sven use? For me, root has a crontab entry for this as follows: Doesn't the logrotate package install a daily cron by itself ??! It certainly did for me (in potato). -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: package versions in dselect, and Release file
Jason == Jason Gunthorpe [EMAIL PROTECTED] writes:
Jason On 25 Jun 2000, Ian Zimmerman wrote:
deb file:/usr/local/src/debs localdebs main non-free
Then apt-get update looks for
/usr/local/src/debs/dists/localdebs/{main,non-free}/binary-i386/Packages
but apt-get install pysol looks for
/usr/local/src/debs/main/binary-i386/games/pysol_*.deb.
Jason Which is /usr/local/src/debs+[whatever is in Packages for the
Jason Filename field]
Jason Call dpkg-scanpackages correctly and this will go away.
You're quite right, ... except that man dpkg-scanpackages says
dpkg-scanpackages binarydir overridefile [pathprefix]
Packages
binarydir is the name of the binary tree to process (for
example, contrib/binary-i386). It is best to make this
relative to the root of the Debian archive, because every
Filename field in the new Packages file will start with
this string.
At the very least, this needs a little clarification. What is the
root of the Debian archive? I assumed it was
foobar/dists/{frozen,stable,whatever}, and it was a natural assumption
because it corresponded to the example given. Now we know I was
wrong, foobar is the root. But I would never know just from the
manpage.
Thanks for the help,
--
Ian Zimmerman, Oakland, California, U.S.A.
In his own soul a man bears the source
from which he draws all his sorrows and his joys.
Sophocles.
gap (maths package) versions
The Debian version for the GAP (Groups/Algorithms/Proofs) package is 3.4.4, and yet there are two documentation packages which document GAP4. This situation persists even in woody, which even seems to contain library add-ons for GAP4, but the base GAP package is still at 3.4.4. What's going on? I've installed GAP4 in /usr/local, but I'd prefer a deb .. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
package versions in dselect, and Release file
What are the packaging frontends (dselect in particular) supposed to do when 2 sources from sources.list provide different version numbers of the same package? I have a potato system, but I downloaded a couple of upgraded packages from woody and placed them in a local mirror directory. I generated the Packages files with dpkg-scanpackages. That went fine, so I added a deb file: line for the local mirror to sources.list and run dselect. In the Select phase I could see the newer versions as available. I selected them. But the Install phase ignored my local mirror, tried to download the updates from the potato archive, and failed. I noted while dselect was hitting the sources it said something like deb file:/foo/bar Release Ignored Of course, I don't have a Release file in the mirror directory. Is that necessary for it to be recognized as a worthy source? If so, how do I generate one; dpkg-scanpackages doesn't. Or is there something else wrong with what I'm trying? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: question on sysvinit
Sven == Sven Burgener [EMAIL PROTECTED] writes: itz Why is it this way? Well, there's really no way to tell init to itz execute programs upon _exiting_ a runlevel, or upon a transition itz from one level to another; the /etc/inittab that defines itz runlevels is one-dimensional. So the transition must be defined itz independently of the old level. Sven In their manual, SuSE document that upon changing from say Sven runlevel 2 to 3, first all links matching /sbin/init.d/rc2.d/K* Sven get executed and then the links matching Sven /sbin/init.d/rc3.d/S*. (Funnily they have init.d under /sbin and Sven not /etc) Maybe that's how SuSE does it, but not Debian. Sven specification for this? Or does it simply not matter? :) AFAIK there's no written standard for this yet, but soon there will be (the Linux Standard Base). You can check their drafts. The original System V init did it the Debian way, I believe. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: safe to use woody packages in potato?
John == John Anthony Kazos [EMAIL PROTECTED] writes: John I'm running potato. Is it safe (read, won't screw up the John installation) to change the lines in sources.list to refer to John woody as well as potato, so that I use a newer version if it's John in woody and an older one if it's only in potato? I'm not John significantly concerned about problems with the packages John themselves (I'm running 2.4.0-test2, why not unstable Debian); I John just wanted to make sure that the differences between potato and John woody are just package differences and don't require a special John process to use the newer packages. Are there any caveats with John doing this? (It's difficult for me to recover from an unbootable John Linux on this box.) What will that do that simply switching to woody wouldn't? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: package versions in dselect, and Release file
Olaf == Olaf Meeuwissen [EMAIL PROTECTED] writes:
itz I have a potato system, but I downloaded a couple of upgraded
itz packages from woody and placed them in a local mirror directory.
itz I generated the Packages files with dpkg-scanpackages. That went
itz fine, so I added a deb file: line for the local mirror to
itz sources.list and run dselect. In the Select phase I could see
itz the newer versions as available. I selected them. But the
itz Install phase ignored my local mirror, tried to download the
itz updates from the potato archive, and failed.
Olaf Make sure your local archive is mentioned before the rest. I
Olaf maintain a local mirror but it is usually lagging a bit behind
Olaf (especially with unstable). If your sources.list says something
Olaf like:
Olaf deb file:/pub/debian unstable main deb
Olaf http://ftp.debian.org/debian unstable main
It is not this simple; in fact, I think there's an out-and-out bug.
I'll report it when I have some time to waste.
The problem is that apt-get update and apt-get install (also
upgrade, etc.) make mutually inconsistent assumptions about the
structure of the mirror directory. Suppose the entry in sources.list
reads
deb file:/usr/local/src/debs localdebs main non-free
Then apt-get update looks for
/usr/local/src/debs/dists/localdebs/{main,non-free}/binary-i386/Packages
but apt-get install pysol looks for
/usr/local/src/debs/main/binary-i386/games/pysol_*.deb.
I was only able to work around this by setting up the mirror as
follows.
/usr/local/src/debs---main
^ |
| --non-free
| |
| --dists---localdebs--
||
-
This is related to the thread earlier this week about the semantics of
a trailing slash on the deb URL. Apparently the thorough explanation
given still haven't exhausted the topic :-(
--
Ian Zimmerman, Oakland, California, U.S.A.
In his own soul a man bears the source
from which he draws all his sorrows and his joys.
Sophocles.
Re: question on sysvinit
Sven Hi all I wanted to know why that when changing runlevels from Sven say 2 to 3, the KILL links of _3_ and not 2 get executed before Sven starting the START links of 3...? Marshal I think because K comes before S. So switching run levels Marshal only runs the scripts in the runlevel dircetory (in your case Marshal rc3.d). The kill scripts of 2 have already been run, Marshal presumably when you first boot up, or when you switched to Marshal runlevel 2. Sven Hmm, my idea of the K links have been to clean up the system Sven so as to be able to go into a different state / runlevel... but Sven I am unsure now. Sven I mean, at that point you're leaving runlevel 2 and it would Sven only make sense to stop those runlevel's services and not the Sven new runlevel's, right? Marshal Maybe, but I think to have what you want, you'll need start Marshal up scripts and kill scripts for both going into the runlevel Marshal and leaving the Sven But at first, the system isn't in any runlevel and hasn't any Sven services up, and so it won't need to stop anything and thus K Sven links for runlevel S (or whatever the first actual runlevel is) Sven don't make sense to me. Sven Someone please shed some light on this for me as I want to Sven understand this. What you're missing (and it's easily missed) is that the runlevel changes are _stateless_. It doesn't matter what the _old_ level is, the system will always execute the same scripts upon entering the same (new) level. So, for each level, the K scripts clean up everything that is not needed _on the new level_, potentially doing some redundant work (ie. killing servers that are not running because they are not part of the old level). Then the S scripts start everything that is needed, again potentially restarting things that were already active on the old level (although Debian avoids that as a matter of optimization). Why is it this way? Well, there's really no way to tell init to execute programs upon _exiting_ a runlevel, or upon a transition from one level to another; the /etc/inittab that defines runlevels is one-dimensional. So the transition must be defined independently of the old level. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
HELP! Netscape completely broken by tonight's upgrade!
Navigator-smotif-473 and friends stopped working completely after I unpacked today's versions. The first problem was a syntax error in the shell script /usr/X11R6/bin/netscape, but after I fixed that by hand all hell breaks loose. I had to remove the packages. Please help! -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: HELP! Netscape completely broken by tonight's upgrade!
itz == Ian Zimmerman [EMAIL PROTECTED] writes: itz Navigator-smotif-473 and friends stopped working completely after itz I unpacked today's versions. The first problem was a syntax itz error in the shell script /usr/X11R6/bin/netscape, but after I itz fixed that by hand all hell breaks loose. I had to remove the itz packages. Please help! Here's my temporary fix. I can't imagine what the maintainer was thinking here :-( *** /root/wrapper Fri Jun 23 00:09:38 2000 --- /usr/lib/netscape/base-4/wrapperFri Jun 23 00:11:05 2000 *** *** 271,280 MOZILLA_HOME=$(dirname $netscape) for d in \ ! /usr/lib/netscape/base-4/wrapper.d \ ! /usr/lib/netscape/$VER \ ! /usr/lib/netscape/$VER/$BIN ;do ! for f in (cd $d;ls -1 . | sort); do . $d/$f done done --- 271,278 MOZILLA_HOME=$(dirname $netscape) for d in \ ! /usr/lib/netscape/base-4/wrapper.d ;do ! for f in $(cd $d;ls -1 . | sort); do . $d/$f done done -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: modprobe: Can't locate module char-major-6 ?
Shaul Isn't the following the root for my problem: # # Character devices # CONFIG_VT=y CONFIG_VT_CONSOLE=y CONFIG_SERIAL=m # CONFIG_SERIAL_EXTENDED is not set # CONFIG_SERIAL_NONSTANDARD is not set CONFIG_UNIX98_PTYS=y CONFIG_UNIX98_PTY_COUNT=256 # CONFIG_PRINTER is not set CONFIG_MOUSE=y # # Mice # [13:22:10 /tmp]$ Shaul Why was it made that way? Is it a bug? Depends where the .config (and your kernel) comes from. Is it a custom compiled one or one of the original installation ones? If it is an original, the .config file probably means nothing; you should have had a chance to select the lp module during the installation (it's in the Miscellaneous group, I think). If you compiled the kernel yourself, you should have selected parallel printer support during make config (or make menuconfig or whatever). If you did, and you got the .config file you quote, it is a bug. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: modprobe: Can't locate module char-major-6 ?
Colin == Colin Watson [EMAIL PROTECTED] writes: [23:32:47 /tmp]$ wc /var/log/syslog -l 1316 /var/log/syslog [23:32:50 /tmp]$ grep modprobe: Can't locate module char-major-6 /var/log/syslog | wc -l 214 [23:32:53 /tmp]$ Shaul Why 15% of my syslog is filled with the error message about Shaul that module? What does it do? Colin char-major-6 is an alias for the set of kernel character Colin devices on major number 6; Colin /usr/src/linux/Documentation/devices.txt says that this is Colin parallel printer support, so presumably you've got a printer Colin daemon installed or something. Shaul I am using a kernel-image as is and have not compiled Shaul anything. Why is it all of a sudden gives me messages about it? Shaul Is it a bug in some package? How can I fix it? Colin That kernel-image package *does* have parallel printer support Colin enabled as a module. Have you modified anything in Colin /etc/modutils/ in any way? I don't think so. Rather, the lp module is loaded at boot time, as it is listed in /etc/modules. That would be the place to check for modifications, and the boot script that reads it which is /etc/init.d/modutils. If Shaul wants the driver loaded the non-Debian way, on demand, he needs to add this line to /etc/modutils/aliases: alias char-major-6 lp then run update-modules. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: files/dirs under /var/www/
Olaf I did that. Right now it looks quite safe to add directories and Olaf files below /var/www, but what if a new package comes along and Olaf decides that it needs to install all over the web pages in Olaf /var/www/foo? Ethan hmm /me thinks debian policy should treat /var/www like Ethan /usr/local packages should be forbidden from touching it (at Ethan all, even no directories) the only exception should be apache Ethan when first installed should be allowed to put a example Ethan index.html file only if /var/www did not exist upon install (as Ethan it works now). Ethan policy may already state this i don't recall (and don't have Ethan time to look it up atm) I had this same dilemma and couldn't find anything on it in policy, so I just changed DocumentRoot to /usr/local/share/www. (Web files are mostly static anyway, why under /var?) -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Gnus broken by Emacs upgrade!!
After I did the upgrade today to Emacs 20.6-2, this happens whenever I
try to run Gnus:
begin 644 gnus-backtrace
M4VEG;F%L:6YG.B`H:6YV86QI9UF=6YC=EO;B`H;6%C[EMAIL PROTECTED]:6QE
M(9R97-T(9OFUS*2`B@6!QB!Q8''(8'%@@A!D:@[EMAIL PROTECTED]('+8',
[EMAIL PROTECTED]@V!SH'/('.#$)[EMAIL PROTECTED]!T0I$[EMAIL
PROTECTED])@=0)1('50D*!UH'442!
MUX'[EMAIL PROTECTED]@I$1$2!VX':D2!W$)18'[EMAIL PROTECTED])0D((@:\%1D6!UX'A
M[EMAIL PROTECTED]'(B!;=5MUR97-U;'1S('1E;7`M8G5F9F5R('1E;7`M
M9FEL92!F:6QE(9OFUS(UA:V4MWEM8F]L()T96UP+69I;4B()T96UP
M+6)U9F9EB(@(G1E;7`MF5S=6QTR(@V%V92UE-UG-I;VX@;5T*B`H
M95F875L=UM86IOBUM;V1E(AQ=6]T92!F=6YD86UE;G1A;UM;V1E*2D@
M*AS970M8G5F9F5R(AG970M8G5F9F5R+6-R96%T92`H9V5N97)A=4M;F5W
M+6)U9F9EBUN86UE((@*FYN:5A95R('1E;7`J([EMAIL PROTECTED]:6YD+7!R
M;W1E8W0@')O9VX@V5T2!C;VYD(YU;[EMAIL PROTECTED]@*'0I('-E=UB=69F97(@
M*AB=69F97(MW1R:[EMAIL PROTECTED]!W:5N(YO=!F:6QE+61IF5C=]R2UP
M(9I;4M;F%M92UD:7)E8W1OGD@;6%K92UD:7)E8W1O[EMAIL PROTECTED]'0I('=R:71E
M+7)E9VEO;B`H]I;G0M;6EN*2`H]I;G0M;6%X*2`H;FEL(AQ=6]T92!N
M;VUEVI*2!B=69F97(M;F%M92!K:6QL+6)U9F9EET@,3@*(O=7-R+W-H
M87)E+V5M86-S+S(P+C8O;ES]G;G5S+VYN:5A95R+F5L8R(@+B`Q,CP
M,BE=*2D*(!N;FAE861EBUT96UP+7=R:71E*)^+TUA:6PO86-T:79E(B!N
M:6PIB`@;FYM86EL+7-A=F4M86-T:79E*@H(FUA:6PN;[EMAIL PROTECTED]:6YP=70B
M(@Q([EMAIL PROTECTED](@B;6%I;YD96)I86XM86YN;W5N8V4B(@Q([EMAIL
PROTECTED]DI(@B
M;6%I;YD96)I86XMV=M;(@[EMAIL PROTECTED]@*)M86EL+F1E8FEA;B(@*#$V
M,B`N(#0Q.3$I*2`H(FUA:6PN8V]Q(B`H,2`N(#,[EMAIL PROTECTED])M86EL+F%M86YD
M82(@[EMAIL PROTECTED](F%R8VAI=F4B(@Q(X@,S,[EMAIL PROTECTED])M86EL+FAE
M=F5A(B`H-2`N(#([EMAIL PROTECTED])M86EL+F%U=]?W!A;2(@*#,@+B`U.34I*2`H
M(FUA:6PN8W)O;B(@*#,V-2`N(#0P,2DI(@B;6%I;YL:6YUUC(B`H,2`N
M([EMAIL PROTECTED])M86EL+G1EUK([EMAIL PROTECTED](I*2`H(FUA:6PN;[EMAIL
PROTECTED]
MV5C=7)I='DB(@V-`N(#$X.DI(@B;6%I;YL:6YUUD:6%L9(@*#$@
M+B`Q,RDI(@B;6%I;YGTB(@W,`N(#(X,DI(@B;6%I;YC86UL(B`H
M.2`N(#$T,C(I*2`H(FUA:6PN;6ES8R(@[EMAIL PROTECTED],[EMAIL PROTECTED])M86EL+G!R
M;V-M86EL(B`H-#,@+B`R.#,[EMAIL PROTECTED])M86EL+G=OFMI;F=?87-S971S(B`H
M-`N(#,[EMAIL PROTECTED])M86EL+F5S8V%P961?W!A;2(@[EMAIL
PROTECTED],DI(@B;6%I
M;YU=8B(@[EMAIL PROTECTED],#(I*2`H(FUA:6PN')OV$B(@Q([EMAIL PROTECTED](@B
M;6%I;YD96)I86XM=V5E:VQY(B`H,2`N(#I*2`H(FUA:6PN95B:6%N+65M
M86-S96XB(@Q([EMAIL PROTECTED]DI*2`B?B]-86EL+V%C=EV92(IB`@;FYM86EL+6=E
M=UN97M;6%I;AN;FUL(YN;6PMV%V92UN;W8@(GXO36%I;\B()M86EL
M+F1E8FEA;B(IB`@;FYM;UR97%U97-T+7-C86XH(FUA:6PN95B:6%N(B`B
M(BD*(!G;G5S+7)E75EW0MV-A;[EMAIL PROTECTED];FYM;#IM86EL+F1E8FEA;B(@*YN
M;6P@(B(I*0H@(=N=7,M86-T:79A=4M9W)O=7`H(FYN;6PZ;6%I;YD96)I
M86XB('-C86XI[EMAIL PROTECTED]RUG970M=6YR96%D+6%R=EC;5S*#(I[EMAIL
PROTECTED]
MRUS971UUN97=S*YI;`R('0I[EMAIL PROTECTED];V1E*((A`X`80.`('
M((B3P!QX'((8B!R5QN(1-9H0``V#(0!RB(#AF#+@!RR(@R!S8'.
M7*(@^!T`X:.(@!T2(@=(@B('3#AHAB('4((B!U8'6(8B!UR(@!
MV(@='(B!;9]N=UC;VYN96-T(1I9UC;VYN96-T(=N=7,MW1AG1U
MUF:6QE(=N=7,M8W5RF5N=US=%R='5P+69I;[EMAIL PROTECTED]RUS;%V92!G
M;G5S+75S92UDFEB8FQE+69I;[EMAIL PROTECTED]RUGF]UUQ=6ET(=N=7,MG5N
M+6AO;VMS(=N=7,MW1AG1UUH;V]K(=N=7,M;6%K92UN97=SF,M9FEL
M92!G;G5S+61R:6)B;4MF5A9UF:6QE()B8BUL;V=I;B!A90M:]O:R!G
M;G5S+7-U;6UAGDM;6]D92UH;V]K(=N=7,M9W)O=7!L96YS+6UO9[EMAIL PROTECTED]
MRUS971UUN97=S(YI;!G;G5S+7-E='5P+6YE=W,M:]O:R!G;G5S+7-T
M87)T+61R869T+7-E='5P(=N=7,M9W)O=7`M;ES=UGF]U',@9VYURUG
MF]UUF:7)S=UU;G)E860M9W)[EMAIL PROTECTED]RUC;VYF:6=UF4M=VEN9]W
MR!GF]U!G;G5S+6=R;W5P+7-E=UM;V1E+6QI;[EMAIL PROTECTED]RUS=%R=5D
M+6AO;[EMAIL PROTECTED]RUUV4M9W)O=7!L96YS(QE=F5L72`T*0H@(=N=7,M,[EMAIL
PROTECTED]
M('0@;FEL*0H@(=N=7,H,B!T(YI;D*(!G;G5S+6YO+7-EG9EBTQ*YI
M;!N:6PI[EMAIL PROTECTED]RUN;RUS97)V97(H;FEL*0H@(-A;PM:6YT97)A8W1I
M=F5L2AG;G5S+6YO+7-EG9EBD*(!E5C=71E+65X=5N95D+6-O;6UA
M;F0H;FEL*0H@(-A;PM:6YT97)A8W1I=F5L2AE5C=71E+65X=5N95D
*+6-O;6UA;F0I@``
`
end
Please help! I cannot properly read my mail without Gnus!
Please Cc my address on any followups. Thanks.
--
Ian Zimmerman, Oakland, California, U.S.A.
In his own soul a man bears the source
from which he draws all his sorrows and his joys.
Sophocles.
Re: Colormaps in Linux
Eric == Eric Hagglund [EMAIL PROTECTED] writes: Eric The problem I am trying to fix is that fonts and images in Eric Netscape have never been properly displayed in Netscape. Instead Eric of getting fonts, toolbars, scrollbars and images displayed as Eric the application programmers and web designers intended, I get Eric basic Courier fonts, two dimensional images (I especially notice Eric this with tool bars in netscape and within web pages) and 16-bit Eric colors instead. Eric I believe this probably has something to do with the way I have Eric set up my color map in X as I intermittently get the following Eric error: Unable to allocate default colormap. I remember seeing this long ago and concluding some other package (Tk based, I think) caused it by not playing by the book with the color map, basically grabbing all of it or something. Does it happen when Netscape is the very first thing you start in X? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: apache question
Ethan however one thing you should do on a debian system is chown Ethan /var/www to root and make sure its not group writable. also Ethan chown /var/log/apache/* to root.adm and make sure the Ethan permissions are 640 or 644. (you have to fix the apache cron Ethan jobs to not undo this change) Ethan for some insane reason debian leaves the www-root owned by Ethan www-data.www-data (the same user debian runs apache as) along Ethan with the logs. this is totally wrong as the web server user Ethan should NOT own files or have any write permission to anything. Ethan if it does then all it takes is one of those unprivileged child Ethan processes to be exploited and your web site can be replaced and Ethan your logs can be removed. bad bad bad. As for the document tree, I largely agree. But as for the logs, don't the child servers need to write them, almost by definition? -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: man -- info?
Fabrizio == Fabrizio Polacco [EMAIL PROTECTED] writes: Fabrizio info docs are guides, while manpages are pages of the Fabrizio reference manual. Both type of docs are needed, as they Fabrizio serve different purposes, but only the reference is Fabrizio required. That attitude, to abhor reference manuals in Fabrizio favor of on-line guides, is like pretending to conversate Fabrizio only with persons with the same level of knowledge as you. [...] Fabrizio Now they do not have an excuse any-more. FSF has added a Fabrizio info2man program that extract the manpage from the texinfo Fabrizio source! Do I detect just a slight touch of contradiction here? Where does info2man get the reference information for the manpage, if not from the Info manual? Indeed, every FSF-policy conforming Info manual has a section Invoking foo which contains very nearly the same things as a typical manpage does. Of course, the FSF policy itself is an Info document (standards.info). Maybe that is the real bug :-) -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: xterm xterm-debian
Andrew == Pollywog [EMAIL PROTECTED] writes: Andrew On Fri, 19 May 2000, [EMAIL PROTECTED] wrote: tom What's the best the best way to avoid having ot reset my terminal tom variable everytime I telnet to a site that does not know what tom xterm-debian is? Since this happens quite often it is getting a tom little tedious. Andrew In the remote .bashrc, put Andrew TERM=xterm The right way: In the remote .bashrc or .profile or whatever, do export TERMINFO=$HOME/terminfo:$TERMINFO On remote system, create a subdirectory ~/terminfo/x/ Copy local /usr/share/terminfo/x/xterm-debian to remote ~/terminfo/x/xterm-debian (Modification for termcap based remote systems left as exercise for the user) -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
Re: Latin-1 characters in X, how?
Johann == Johann Spies [EMAIL PROTECTED] writes: itz On the Linux virtual consoles I can enter Latin-1 characters into itz any program using the Compose key sequences, provided these are itz defined in the loaded keymap. The question is, how can I do the itz same with X clients? Johann On my system I use Control-ScrolLock as compose key in X. I Johann think it is the standard setup. No, that only turns on ScrollLock for me, as if Control were not involved. I tried both left and right Control. -- Ian Zimmerman, Oakland, California, U.S.A. In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles.
new perl dependencies
When I run dselect last night, there was a bunch of important packages that had been upgraded. One of them was perl-5.005-doc. I accepted all the suggestions, but when I was ready to exit selection mode and start downloading, I was thrown into a conflict/dependency screen with just perl-5.005 and perl-5.005-doc. And no matter what I did, dselect threw me into that screen again and again. I had to force my way out with an `x'. Everything seems to run for the moment, but I am shaken. Is this some kind of circular dependency among the packages involved? Can it be fixed? -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
more real runlevels?
When will Debian follow the Linux Standard Base with respect to init runlevels? The current scheme (still in potato, I haven't checked woody) basically has just 2 distinct levels - single user and fully operational. This throws away most of the power of sysvinit (which comes at the cost of considerable complexity). I would like to run the system without xdm most of the time, but there are times when I need it. Also sometimes I want to switch off network services but still allow local logins. The LSB allows for such variations with more finely tuned runlevels. But meanwhile I am stuck with Debian. I can make my own runlevels from among 3-5 by removing some of the links from /etc/rc[345].d, but will the next upgrade of the relevant packages re-create the links? If so, what _is_ the proper way to do this with Debian? Thanks for your answers. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: update-alternatives: bug or misunderstanding?
Eric == Eric G Miller egm2@jps.net writes: itz However, that's not how it seems to work on potato. I noticed itz that after the installation I got /etc/alternatives/awk pointing itz to /usr/bin/mawk and /etc/alternatives/awk.1.gz pointing to itz /usr/man/man1/mawk.1.gz. Since I prefer gawk, I did (as root) itz update-alternatives --config /usr/bin/awk awk /usr/bin/gawk Eric You left out the slave link for the man pages, hence they Eric weren't updated. Reread that alternatives man page (yes, it's Eric confusing). I was under the impression that it was the whole point of this feature that it would figure that out automatically. Certainly the system has enough information to do that, looking at the files in /var/lib/dpkg/alternatives. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
update-alternatives: bug or misunderstanding?
man update-alternatives says: It is often useful for a number of alternatives to be synchronised, so that they are changed as a group; for example, when several versions of the vi(1) editor are installed, the man page referenced by /usr/man/man1/vi.1 should correspond to the executable referenced by /usr/bin/vi. update-alternatives handles this by means of master and slave links; when the master is changed, any associated slaves are changed too. A master link and its associated slaves make up a link group. However, that's not how it seems to work on potato. I noticed that after the installation I got /etc/alternatives/awk pointing to /usr/bin/mawk and /etc/alternatives/awk.1.gz pointing to /usr/man/man1/mawk.1.gz. Since I prefer gawk, I did (as root) update-alternatives --config /usr/bin/awk awk /usr/bin/gawk That in fact changed the /etc/alternatives/awk link to point to gawk, but it left the manpage link pointing to mawk.1! The same thing happened again when I wanted to change the etags alternative (to point to the Exuberant version). I ended up changing the manpage links by hand. Is that safe or will it confuse the system even more? -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: SVGA Text Mode ATI Charger video card
John == John Anderson [EMAIL PROTECTED] writes: John Hello, Has anyone been successful at configuring an ATI Charger John video card to use SVGA text mode? I can get it to work at a John generic VGA, but only at 80x25. I have tried ATI mach64, but John It doesn't recognize the card as being that. The resolution I John want is 132 columns by 40 rows. which debian/which kernel? If you use potato with its default installation kernel, chances are you're in framebuffer mode - basically text mode is faked by drawing glyphs in graphics mode. (Does it display the penguin logo when it boots?) To get my favorite text mode resolution I had to recompile my own kernel for potato. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
potato growing pains
The installation went smoothly except for 1 glitch - I was unable to create a boot floppy. When I got to that stage it formatted the blank floppy I provided (and it seemed to use a superformat), then bailed out. When I switched to console 2 where the log messages go during installation, I saw a gripe about invalid argument to umount. Have you seen that? (Yes I tried another floppy, same result). More importantly though, do you know how I can resize the newfangled framebuffer console like I could with SVGATextMode? stm itself doesn't work at all, predictably; but neither does fbset which seems to be the closest thing to what I want. (It complains about a failed ioctl). Even if fbset worked, it's a horrific amount of work to figure out the parameters which are all in terms of pixels, clock rates etc. With stm I could just say stm 112x40 and bingo. For text users like me the framebuffer is definitely a giant step backward unless I'm missing something. I had to install X just to read some basic documents. If resizing is as hard as it seems or even impossible, is there a way to disable the framebuffer and use the old char-cell console? Something like a kernel command line parameter perhaps? I read kernel/Documentation/fb/* but all I can see is a mention of a video= parameter with no explanation what should go on the right side. :-( Thanks for help, -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: ssh loading at startup
Ron == Ron Rademaker [EMAIL PROTECTED] writes: Ron /etc/init.d/ is where you have to put the script Ron /etc/rc[RUNLEVEL].d is where you have to put symlinks to this Ron script The usual way to run sshd is via inetd, though. So maybe Beavis is asking the wrong question, maybe ssh into his box already works without any additional effort. If not, look in /etc/inetd.conf for a line mentioning ssh, perhaps there's something wrong with it. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: ATAPI tape drives.
Edward == Edward Mulholland [EMAIL PROTECTED] writes: Edward The linux hardward compatability HOWTO at Edward http://www.linuxdoc.org/HOWTO/Hardware-HOWTO.html says the Edward following about ATAPI tape drives: ATAPI tape drives For these an alpha driver (ide-tape.c) is available in the kernel. ATAPI tape drives supported are Seagate TapeStor 8000 Conner CTMA 4000 IDE ATAPI Streaming tape drive Edward Does anyone know whether these are the only supported ATAPI Edward tape drives, or just examples? I received another model of Edward ATAPI tape drive (Sony SuperStation) as a gift and I want to Edward be sure that it is totally useless before I return it. It is probably not TOTALLY useless .. but .. I had insuperable problems trying to use the driver with a HP/Colorado 5000 ATAPI drive. Gave up and shelled out $500 for a SCSI drive. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
smbfs - remote directories truncated ?!
Slink + smbfs packaged with slink. smbmounting a share from an NT server works fine, but in large directories I only seem to see a portion of the files! Note that this is NOT Samba FAQ 2.2 - the missing files have names that are as valid as those I can see. In fact, after deleting or moving some visible files I get to see the rest. Has anyone else seen this and have a suggestion? -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: Soft ejects
Ethan == Ethan Benson [EMAIL PROTECTED] writes: Ethan On 9/1/2000 Brian May wrote: Brian [1] Dos/windows copes with this problem in a different (IMHO Brian broken) way - it keeps track of which disk is inserted, and if Brian it needs to read/write to another disk, it complains to the Brian user to reinsert the original disk. Why is this mechanism Brian broken? For starters: some games will automatically eject a Brian CD-ROM and ask you to insert the next CD-ROM. For some reason, Brian windows will often decide that it still needed the original Brian CD-ROM, and ask you to reinsert it!!! It even goes as far as to Brian suggest that the CD-ROM might be dirty. Now thats what I call Brian machine is smarter!!! Ethan I never noticed that back when i briefly tinkered with win95, Ethan one thing I find interesting is windoze does NOT lock the cd Ethan drawer closed when a CD is in use like GNU/Linux and MacOS do, Ethan for example i insert a CD and run some program on it under Ethan win95 then press the eject button and it spits out the CD and Ethan windoze blue screened shortly thereafter. Ethan I have never seen windoze ask for a device back again, i didn't Ethan know it had such a function win* does not appear to really have Ethan a concept of `mounted' filesystems as far as i could tell. Back in the sad, sad days when I was a Windoze programmer, we had to make the APPLICATION PROGRAM do that. I.e. check for the existence of the file D:\flagfile.000 and if not found, bitch at the user to insert the correct CDROM again. Really high tech. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: MTA
peter == peter karlsson [EMAIL PROTECTED] writes: peter Hi! I'll be moving and will lose my direct Internet peter connection, and will have to resort to dial-up. To prepare for peter this, I am switching over to doing mail and news offline peter (slrnpull, fetchmail), but I need some ideas on what to use for peter outgoing mail. I've had sendmail die on me when I'm not peter connected to a network (not in Debian, though, haven't tried peter sendmail in Debian), so I wonder what the best setup is for peter outgoing mail when it is only to send mail when I connect to peter the ISP (and directly when I do that, preferrably without peter manual intervention). With sendmail, use default delivery mode = deferred. Of course there's a funny syntax you need to learn to put that in the config file; you didn't really expect better, did you? :-) To really send the mail when you're online do sendmail -q in a cron job or similar. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: Modem does not work
Pedro == Pedro Quaresma de Almeida [EMAIL PROTECTED] writes: Pedro It is an internal PCI modem, Windows says that it is on COM3, Pedro IRQ10, so it should be in /dev/ttyS3, but I only have two Pedro serial connections working, No, COM3 == ttyS2 ! -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: Stupid question regarding foreign characters and how to output them?
Nathan == Nathan E Norman [EMAIL PROTECTED] writes: Nathan Hi, For some reason I decided I needed to compose a document Nathan with an `n' with a tilde over it - reading through the kbd Nathan package docs it seems I can do this using the compose key. Nathan So, what is the compose key? Usually Ctrl-. , so to get the character you want you type the sequence Ctrl-. ~ n. But it depends on your keytable, I am not sure where Debian puts keytables - anyone? -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: Why
Nate == aphro [EMAIL PROTECTED] writes: Nate i believe the key to linux expansion is pre installed Nate machines. users will be very hesitant to change their OS from Nate something that works for them, even if it means upgrading from Nate win95 to win98 or nt or win3 to win9x. this isnt something that Nate can be blamed on linux but on installing any other OS. Exactly, I agree completely. It would be interesting to see statistics on what percentage of new Windoze users actually install the OS themselves, but the figure I'd bet on is quite low. :-) As for the Great Universal Windows Installer, there are non-SVGA cards out there, like the Hercules and clones. Does Windows 9[0-9] work at all on those? -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: xemacs text-mode and gpm mouse issues
Brian == Brian May [EMAIL PROTECTED] writes: Ian I am not sure how xemacs handles the gpm input. GNU Emacs does Ian it through a subprocess and an associated Elisp module Ian (t-mouse.el). If this is the case for xemacs too, the right Alt Ian key (AltGr) should work. Brian Yes, it does. Thanks for that bit of information. Having it the Brian right TAB key seems a little bit inconvenient though, as I use Brian the mouse with my right hand... In the latest gpm releases, there's an emacs variable (I don't recall the exact name, sorry; I read the debian list at work where I don't have the gpm sources ready) that swaps the right and left Alts as far as Emacs is concerned. Look for it in t-mouse.el and set it in your .emacs. Ian It is also possible, though, that xemacs already has this Ian functionality compiled in; that was another thing I remember Ian hearing from xemacs people. In that case xemacs itself fully Ian controls which mouse events it grabs, and you'd have to deal with Ian this as with an xemacs bug. Brian I doubt xemacs would have it compiled it, otherwise, it would Brian need to know the configuration for my mouse. by compiled in i meant linking Emacs with the gpm library (which in turn talks to the gpm daemon), as opposed to running a separate subprocess (mev) linked with libgpm and talking to the subprocess with t-mouse. GNU Emacs does the latter; I had the vague feeling that (recent) XEmacs does the former. But Emacs (or mev) doesn't need any mouse configuration in either case, only the gpm daemon does. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: X, middle mouse,Artec, do see output in /dev/mouse
Joost == Joost Witteveen [EMAIL PROTECTED] writes: Joost But I've already tried all possible X mice. And everything Joost works with the Microsoft setting, except for the middle Joost button. Strange thing is that I've now got two such mice, Joost bought at different times, and from different suppliers. FWIW (and OT :-) the middle button on an Artec doesn't work in Windoze NT either. It's just a strange mouse. I'd buy a Logitech Mouseman instead. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: xemacs text-mode and gpm mouse issues
Brian == Brian May [EMAIL PROTECTED] writes: Brian This has the following problems: - it is impossible to see what Brian text I have highlighted until I release the mouse button. This is unfortunately impossible without modifications to x?emacs. There was noise from xemacs team about fixing this (in xemacs) but I don't know what became of it. Myself I use GNU Emacs. Brian - more importantly, it means I can't copy and paste between Brian console windows, as everything goes to/from xemacs private Brian buffer. Brian - I have tried combinations like shift+mouse which would work Brian in an xterm, but not here :-( I am not sure how xemacs handles the gpm input. GNU Emacs does it through a subprocess and an associated Elisp module (t-mouse.el). If this is the case for xemacs too, the right Alt key (AltGr) should work. It is also possible, though, that xemacs already has this functionality compiled in; that was another thing I remember hearing from xemacs people. In that case xemacs itself fully controls which mouse events it grabs, and you'd have to deal with this as with an xemacs bug. Brian Is it possible to use normal cutpaste with xemacs? If not, Brian then perhaps this is a bug in gpm for not allowing it? Is it Brian possible to disable mouse support in xemacs on a console? If Brian so how? Similar answer as above. If it's done with t-mouse, disable it in your .xemacs or site-lisp/default.el, wherever it is. If it's compiled in, it's probably a command line option or it can't be disabled at all. Ian, who was upstream gpm maintainer for a while. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: cron messages
Ron What does the following mean and how do I fix it?? Ron /etc/cron.daily/cfengine: Ron cf:mustang:/etc/cfengine/cfengine.conf:26: parse error Ron cfengine:mustang::26: Warning: actionsequence is empty Ron cfengine:mustang::26: Warning: perhaps cfengine.conf has not yet been Ron set up? Ron cfengine:mustang::Execution terminated after parsing due to errors in Ron program Edit /etc/cfengine/cfengine.conf. The syntax is a bit braindamaged and doesn't tolerate empty strings as action lists, so you have to have at least one harmless action in each section (like the one that checks symlinks). Somebody else on the list already helped you with the suidmanager part, I think. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: umount
Dave == Dave Wiard [EMAIL PROTECTED] writes: Dave is there any way to force umount to unconditionally unmount a drive? one Dave of my cd drives is constantly touted as 'busy' when i know there's zero Dave activity, so it refuses to unmount and hence i cannot eject. any help Dave would be apreciated. Well, what do you mean by activity? There's a couple of reasons why the kernel can consider a mount point busy. An obvious one is that a process has a directory on the CD as its cwd. That could be a daemon or a zombie, so you need not know about it. Maybe you should look in the /proc hierarchy for clue the next time it happens. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: I killed my monitor!!
Brad == Brad [EMAIL PROTECTED] writes: Salman Is there any way to check a running X server to find out Salman exactly which modeline is being used at a given time ?? Brad Start xvidtune from a console/xterm window. Press the show Brad button. It will output the current modeline to the Brad console/xterm. Another way is the grabmode program from SVGATextmode (when run from an xterm or rxvt window of course). -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: Problem with named
Nate == aphro [EMAIL PROTECTED] writes: Brian name. How can I fix this problem? Also, how can I add some machine Brian records to the files. I have tried following the howto's and the readmes Brian but I have had no luck yet. Nate best way to do that is to add it to /etc/hosts, and change your Nate /etc/resolv.conf so it has this in it: Nate order local,bind Nate so it will search hosts before it asks the DNS for host info. Hmm, doesn't that defeat the purpose of running named in the first place? I'd add a new zone file and reference it from /etc/named.conf. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: fetchmail vs. ?
Jim == Jim [EMAIL PROTECTED] writes: Jim I'm looking for something similar to fetchmail which will allow Jim me to download msgs via POP3 and deliver them to a local mailbox Jim file, such as /var/spool/mail/user or whatever I want, whereas Jim AFAICT (please correct me if I am wrong) fetchmail will only Jim deliver msgs to an smtp dameon or to a local MDA (procmail, Jim deliver, etc.). Jim I've seen fetchpop and it looks good, and am wondering if there's Jim anything else I might find out there that will do this. Any of the Emacs mail reading packages (Gnus, rmail, vm) will do this; they all rely internally on a helper program called movemail, which can be instructed to do exactly what you want. Even if you don't want to use Emacs (but why shouldn't you? :) you can consider installing it just for movemail. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
Re: html, and a file manager.
tf == tf [EMAIL PROTECTED] writes: tf Also, I want to try out some x file managers. I have mc, of tf course, and file runner. I just got offix-files... what do I type tf to call that up? Can anyone recommend one that I can play with? I have been severely disappointed by all of them (including the ones you mention, plus Linux explorer, Tkdesk, and probably others I forget) with the one and only exception of kfm. It isn't a killer app, but it's the only one that doesn't frustrate my very modest configuration wishes every minute. -- Ian Zimmerman Lightbinders, Inc. 2325 3rd Street #324, San Francisco, California 94107
