Re: Bug#1859: Start/stop scripts do not work when not invoked by init
Ian Jackson writes: > > Heh :) Remember we're being nice to each other now Ian :) > > I didn't mean I didn't like his patch. I'm sure it does the right > thing as far as it goes. It's just that /etc/init.d/functions is > obsolete, and should be replaced with an empty file so that people > stop thinking it's useful :-). > > It shouldn't be removed until all the packages that source it have > been updated not to do so, or they'll all fall over. (None of them > actually *use* any of the functions in it.) Yeah I agree with you, was just trying to find some humour in a not very amusing day. I'm onto my 3rd day of trying to upload a package, my net link is pathetic at the moment :( Andrew -- Dehydration - 34%, Recollection of previous evening - 2%, embarrassment factor - 91%. Advise repair schedule:- off line for 36 hours, re-boot startup disk, and replace head - wow, what a night! -- Kryten in Red Dwarf `The Last Day' Andrew Howell [EMAIL PROTECTED] Perth, Western Australia [EMAIL PROTECTED]
New packaging guidelines - draft
Below is a draft copy of a revised set of packaging guidelines. Please comment if you feel appropriate. In a week or so, barring serious controversy, I'll post a definitive version. I've made various typo fixes, &c. to the draft Ian M. gave me, and made the descriptions of the *.txt files available match reality. I think this is the list of the substantive changes I've made: * Some packages can have a null `build' target, because `build' doesn't make sense. * Control file field names Mixed Case (and NAME changed to Package). * Recommended -> Recommends; Optional -> Suggests. * Source field; Priority, Section, Essential fields. * Control file fields delimited as per RFC822. * Comment about Maintainer field and dots in maintainer name portion. * Revision number folded into Version. * Syntax of package names defined. * Added more stuff about conffiles and what to do about them. * More information about maintainer scripts, including I/O redirection, idempotency, avoidance of prompting, shared files (eg /etc/papersize), &c. * Corrected minor infelicities in syntax descriptions for related-package fields. * New info about Provides. * Requirement to list dependency on libc. * CFLAGS=-O3 (GCC 2.6 supports this and it is supposed to work) The document is Texinfo'd now (that's Ian M.'s work), and the .texi &c files will be available somewhere soon :-). Ian. This is Info file guidelines.info, produced by Makeinfo-1.63 from the input file guidelines.texi. File: guidelines.info, Node: Top, Next: Additional Information, Prev: (dir), Up: (dir) Debian GNU/Linux Packaging Guidelines * (This file was last updated on 6th November 1995. Please check the directory `project/standards' at any Debian GNU/Linux archive for a potentially more up to date copy.) This file documents the steps that must be taken in the preparation of a Debian GNU/Linux package. All submissions to be included in the distribution proper and all packages to be considered for `contrib' or `non-free' availability *must* conform to the guidelines and standards described in this document or they cannot be included or made available at the archive with the distribution. Please read the Guidelines carefully. If you have comments or questions, please contact [EMAIL PROTECTED]'. If you are planning on going further than just contributing a package (i.e., if you plan to maintain it for an extended period of time or if you are generally interested in becoming more involved in the Project), you should join the `debian-devel' mailing list. For more details, read `info/mailing-lists.txt', available at any Debian GNU/Linux archive. * Menu: * Additional Information:: * Package Copyright:: A few words about the importance of understanding the package copyright. * Package Content:: Requirements for the package content. * Source Package:: Creating the source package. * Binary Package:: Creating the binary package. * Control Files:: The binary package control files. File: guidelines.info, Node: Additional Information, Next: Package Copyright, Prev: Top, Up: Top Additional Information ** These Guidelines are intended to be fairly general. More specific information is available about certain aspects of building packages, such as how to use the features of Init under Debian GNU/Linux. This information can be found in the directory `project/standards' at any Debian GNU/Linux archive. At the time of this writing, the following documents are available: `README.etc-skel' A description of `/etc/skel' and `/usr/doc/examples'. `descriptions.txt' How to write an extended (and more useful) `DESCRIPTION' field. `README.init' How to use the features of Init under Debian GNU/Linux in packages. `mailers.txt' How to properly configure packages to use the Debian GNU/Linux mail system. `maintainer-script-args.txt' All the ways that the package maintainer scripts inside a package can be called by dpkg. `dpkg-upgrades+errors.txt' What order things happen in during a package upgrade. `virtual-dependencies.txt' How to use "virtual dependencies" in packages. `virtual-package-names-list.text' The list of virtual package names currently in use, together with the procedure for getting new virtual package names allocated. `dependency-ordering.txt' How to properly order package names in the `DEPENDS' field. There are a number of documents that describe more technical details of dpkg's operation, that will probably only be of minority interest. Please read them if you're doing anything complicated. `auto-deconfiguration.txt' How dpkg can sometimes automatically deconfigure packages in order to do bulk installations smoothly. `dpkg-essential-flag.txt' How to tell dpkg a package is essential and should not be
Re: Bug#1859: Start/stop scripts do not work when not invoked by init
Andrew Howell writes: > Ian Jackson writes: > > Harald Schueler writes: > > > I think "functions" should restore the arguments. The following patch > > > makes works for me: > > > > I prefer the patch below. > > Heh :) Remember we're being nice to each other now Ian :) I didn't mean I didn't like his patch. I'm sure it does the right thing as far as it goes. It's just that /etc/init.d/functions is obsolete, and should be replaced with an empty file so that people stop thinking it's useful :-). It shouldn't be removed until all the packages that source it have been updated not to do so, or they'll all fall over. (None of them actually *use* any of the functions in it.) Ian.
Re: cron mailing root
Alvar Bray writes ("cron mailing root"):
> I have just noticed that at sometime during the slow incremental
> upgrade of my system, cron has stoped mailing root about the jobs it
> runs (mandb updates etc).
>
> Is this correct or have I broken something?
This is correct. It's better to avoid sending users lots of mail that
they don't understand :-), so the cron jobs now just report errors if
there were any.
Ian.
Re: Bug#1834: kubla@goofy.zdv.Uni-Mainz.de: /etc/profile on Debian Linux]
Hi, the file "/etc/profile" should contain only absolut necessary entries. Of course there are thousands of tricks one could do in "/etc/profile", but this stuff (and explanative examples) belong to "/usr/doc/examples". Nobody really wants to mess around with other people taste regarding to enviroment variables and such. Which setting is essential, which not? How do they interact which each other? Here is a very useful setting for you profile. It should be immediatly included in the debian profile. ;-) - snipp - if [ "$TERM" = "xterm" -o "$TERM" = "vs100" ] then PROMPT_COMMAND='echo -ne "\033]2;Machine: $HOST\ User: $LOGNAME\ Directory: $PWD\007\033]1;$LOGNAME:$HOST\007"' export PROMPT_COMMAND fi - snipp - Of course, LOGNAME and HOST must be set ... Have fun Winfried
cron mailing root
Hi, I have just noticed that at sometime during the slow incremental upgrade of my system, cron has stoped mailing root about the jobs it runs (mandb updates etc). Is this correct or have I broken something? alvar -- Alvar Bray Meiko LimitedPhone:+44 1454 616171 650 Aztec West Fax: +44 1454 618188 Bristol BS12 4SD E-Mail: [EMAIL PROTECTED] England WWW: http://www.meiko.com Via: Debian GNU/Linux from home.
Re: Bug#1834: [kubla@goofy.zdv.Uni-Mainz.de: /etc/profile on Debian Linux]
Steve Greenland writes: >[/etc/profile suggestion] >> if [ -f $HOME/$ARCH-linux ]; then >> PATH=$HOME/$ARCH-linux:$PATH >> else >> mkdir $HOME/bin 2>/dev/null >> PATH=$HOME/bin:$PATH >> fi >> > >I don't think /etc/profile should be creating directories in my home >directory. Agree. They should be created when the account is created, or (IMHO better) not at all. >I think if is someone is sufficiently Unix literate to be adding >their own commands, they can probably make the appropriate >modifications to PATH in their own .profile. > >Maybe put "PATH=$HOME/bin:$PATH" in the skel version, commented out? I don't see any harm in including the directory in the PATH even if it's not there. Or even do if test -d $HOME/bin; then PATH=$HOME/bin:$PATH;fi ttfn/rjk
Re: Bug#1834: [kubla@goofy.zdv.Uni-Mainz.de: /etc/profile on Debian Linux]
[/etc/profile suggestion] > if [ -f $HOME/$ARCH-linux ]; then > PATH=$HOME/$ARCH-linux:$PATH > else > mkdir $HOME/bin 2>/dev/null > PATH=$HOME/bin:$PATH > fi > I don't think /etc/profile should be creating directories in my home directory. I think if is someone is sufficiently Unix literate to be adding their own commands, they can probably make the appropriate modifications to PATH in their own .profile. Maybe put "PATH=$HOME/bin:$PATH" in the skel version, commented out? Steve Greenland -- The Mole - I think, therefore I scream "The Fourth Dimension is a shambles?" "Nobody ever empties the ashtrays. People are SO inconsiderate." [Doonsebury]
Re: Bug#1859: Start/stop scripts do not work when not invoked by init
Ian Jackson writes: > > Harald Schueler writes: > > I think "functions" should restore the arguments. The following patch > > makes works for me: > > I prefer the patch below. Heh :) Remember we're being nice to each other now Ian :) Andrew -- Dehydration - 34%, Recollection of previous evening - 2%, embarrassment factor - 91%. Advise repair schedule:- off line for 36 hours, re-boot startup disk, and replace head - wow, what a night! -- Kryten in Red Dwarf `The Last Day' Andrew Howell [EMAIL PROTECTED] Perth, Western Australia [EMAIL PROTECTED]
Bug#1859: Start/stop scripts do not work when not invoked by init
Harald Schueler writes:
> I think "functions" should restore the arguments. The following patch
> makes works for me:
I prefer the patch below.
Ian.
--- /etc/init.d/functions Sat Feb 25 21:20:35 1995
+++ /dev/null Thu Mar 30 21:18:57 1995
@@ -1,76 +0,0 @@
-#! /bin/sh
-#
-# functionsThis file contains functions to be used by most or all
-# shell scripts in the /etc/init.d directory.
-#
-# Version: @(#) /etc/init.d/functions 2.00 03-Oct-1994
-#
-# Author: Miquel van Smoorenburg, <[EMAIL PROTECTED]>
-# Riku Meskanen, <[EMAIL PROTECTED]>
-#
-
- # First set up a default search path.
- export PATH="/sbin:/usr/sbin:/bin:/usr/bin"
-
- # Set RUNLEVEL and PREVLEVEL
- if [ "$RUNLEVEL" = "" ]
- then
- levels=`runlevel`
- if [ $? = 0 ]
- then
- eval set $levels
- PREVLEVEL=$1
- RUNLEVEL=$2
- fi
- fi
-
- # A function to start a program.
- daemon() {
- Usage="Usage: daemon [-n] {program}"
- # Test syntax.
- if [ $# = 0 ]
- then
- echo $Usage
- return 1
- fi
- case "$1" in
- "-n") noecho=1
- shift
- if [ -z "$1" ]
- then
- echo $Usage
- return 1
- fi
- ;;
- *) noecho=0
- ;;
- esac
-
- # See if it already runs (must exclude ourselves!)
- [ ! -z "`pidof -o $$ -o %PPID $1`" ] && return
-
- # echo basename of the program.
- [ $noecho = 0 ] && echo -n "${1##*/} "
-
- # And start it up.
- $*
- }
-
- # A function to stop a program.
- killproc() {
- # Test syntax.
- if [ $# = 0 ]
- then
- echo "Usage: killprog {program}"
- return 1
- fi
- pid=`pidof -o $$ -o %PPID $1`
- if [ "$pid" != "" ]
- then
- echo -n "Killing $pid "
- kill -9 "$pid"
- else
- echo "$1 already down"
- fi
- }
-
Bug#1858: kpathsea package does not install or maintain an ls-R database
Matthew Swift writes: > I recommend that a script like the following, called something like > 'maintain-texmf', be included with kpathsea.deb, run during its > postinst, and that a link be placed to the script in cron.daily or > cron.weekly. (I do not feel strongly about the (new) user group > "texmf"; it's only what I prefer.) The convention in Debian seems to be that these kind of things are all called update-, so update-texmf would probably be better ... > # ensure proper permissions and ownership > chown --recursive root:texmf $DIRS > chmod --recursive ug=rwX,o=rX $DIRS Our packaging guidelines say that the things should be owned by root.root in general. Is there a good reason to use a new group ? Ian.
Bug#1834: kubla@goofy.zdv.Uni-Mainz.de: /etc/profile on Debian Linux]
Ian Murdock writes: > A little complicated, perhaps, but he does have good suggestions. Try. I'd like to take issue with a few of the bits, though: > # /etc/profile - sh(1) initialization file for Linux-based systems. > # Copyright (C) 1993-1995 by Dominik Kubla, ... > # Set the umask: > # umask 000 # -rwxrwxrwxInsane - "Calm down, the doctor's here." > # umask 002 # -rwxrwxr-xInnocent - "Gee, i hear Santa coming!" > umask 022 # -rwxr-xr-xStandard - "Just plain old number 22." > # umask 027 # -rwxr-x---Cautious - "Who are you and can you prove it?" > # umask 077 # -rwx--Paranoid - "The martians are after me ..." This would interwork badly with user private groups and our scheme for ensuring that people's umask is right. The default default umask, so to speak, is 002 and isn't "innocent" :-). > # Set core file size to 0, if a user wants cores, he will have to enable them. > ulimit -c 0 This makes handling bug reports a lot harder. > # Some login programs export LOGNAME, some USER. Make that consistent. > if [ -z "$USER" ]; then > USER=$LOGNAME > export USER > elif [ -z "$LOGNAME" ]; then > LOGNAME=$USER > export LOGNAME > fi This is the business of login, which we control, not of the /etc/profile. > [ environment settings for ARCH, TZ, HOSTNAME, DOMAIN, > MAIL, PATH, WWW_HOME, NNTPSERVER, NLSPATH &c ] Yuk. An ARCH variable is a site-specific thing, and the FSSTND people are thinking of withdrawing arch (which isn't available everywhere) in favour of uname (which is). TZ, MAIL, WWW_HOME, NNTPSERVER, NLSPATH should not be set. If they need to be then the program that is trying to use them is broken. HOSTNAME and DOMAIN are local policy and not particularly helpful. The usual way to do this is to use `hostname' in whatever script in question. Setting the PATH is plausible, but we should set it to one specific thing, rather than having lots of conditionals. > [ fortune, calendar, &c ] This doesn't belong in the global /etc/profile, because then users can't disable it. Most of this belongs in /etc/skel/.profile if anywhere. Note that the more stuff we put in /etc/skel/.profile or /etc/profile the more we will have to edit it, and the more work sysadmins (who will have to edit them too) will have to do because of conffiles prompts relating to these files. Ian.
Bug#1860: fdisk bugs
Observe the session transcripts below. The first bug is the coredump. The second bug is where it won't let me recreate hda3 after having deleted it, claiming that I'm only allowed to go up to cylinder 2047 for some reason. If I delete hda4 too it will let me do it, and then I can recreate hda4 so that it looks the same as it was. When I did this for real I gave up on fdisk at that point, not trusting it to have done hda3 and hda4 correctly (hda4 has a filesystem that I don't want to disturb). cfdisk was perfectly happy to delete hda3 and recreate it. More information on request. Ian. -chiark:~> fdisk /dev/hda You will not be able to write the partition table. The number of cylinders for this disk is set to 2477. This is larger than 1024, and may cause problems with: 1) software that runs at boot time (e.g., LILO) 2) booting and partitioning software form other OSs (e.g., DOS FDISK, OS/2 FDISK) Command (m for help): p Disk /dev/hda: 16 heads, 63 sectors, 2477 cylinders Units = cylinders of 1008 * 512 bytes Device Boot Begin Start End Blocks Id System /dev/hda1 1 11679 846184+ 83 Linux native /dev/hda2102416801718 19656 82 Linux swap /dev/hda3102417192227 2565365 Extended /dev/hda4204822282477 126000 83 Linux native /dev/hda5102417192227 256504+ 83 Linux native Command (m for help): x Expert command (m for help): p Disk /dev/hda: 16 heads, 63 sectors, 2477 cylinders Nr AF Hd Sec Cyl Hd Sec Cyl StartSize ID 1 00 1 10 15 63 1023 63 1692369 83 2 00 15 63 1023 15 63 1023 1692432 39312 82 3 00 15 63 1023 15 63 1023 1731744 513072 05 4 00 15 63 1023 15 63 1023 2244816 252000 83 5 00 15 63 1023 15 63 1023 63 513009 83 Expert command (m for help): r Command (m for help): d Partition number (1-5): 5 Segmentation fault (core dumped) -chiark:~> fdisk /dev/hda You will not be able to write the partition table. The number of cylinders for this disk is set to 2477. This is larger than 1024, and may cause problems with: 1) software that runs at boot time (e.g., LILO) 2) booting and partitioning software form other OSs (e.g., DOS FDISK, OS/2 FDISK) Command (m for help): d Partition number (1-5): 3 Command (m for help): n Command action e extended p primary partition (1-4) p Partition number (1-4): 3 First cylinder (1719-2477): 1719 Last cylinder or +size or +sizeM or +sizeK ([1719]-2047): 2227 Value out of range. Last cylinder or +size or +sizeM or +sizeK ([1719]-2047): 2047 Command (m for help): d Partition number (1-4): 3 Command (m for help): d Partition number (1-4): 4 Command (m for help): n Command action e extended p primary partition (1-4) p Partition number (1-4): 3 First cylinder (1719-2477): 1719 Last cylinder or +size or +sizeM or +sizeK ([1719]-2477): 2227 Command (m for help): n Command action e extended p primary partition (1-4) p Partition number (1-4): 4 First cylinder (2228-2477): 2228 Last cylinder or +size or +sizeM or +sizeK ([2228]-2477): 2477 Command (m for help): p Disk /dev/hda: 16 heads, 63 sectors, 2477 cylinders Units = cylinders of 1008 * 512 bytes Device Boot Begin Start End Blocks Id System /dev/hda1 1 11679 846184+ 83 Linux native /dev/hda2102416801718 19656 82 Linux swap /dev/hda3102417192227 256536 83 Linux native /dev/hda4204822282477 126000 83 Linux native Command (m for help): q -chiark:~> dd if=/dev/hda bs=512 count=1 | uuencode hda.bootsector begin 664 hda.bootsector 1+0 records in 1+0 records out M M M M M M M M M M`0$` [EMAIL PROTECTED](/__\0TQD`D)D/__\%#___H&P:`##4!P`` [EMAIL PROTECTED](@[EMAIL PROTECTED],`5:H/ ` end -chiark:~> dd if=/dev/hda3 bs=512 count=1 | uuencode hda3.bootsector begin 664 hda3.bootsector 1+0 records in 1+0 records out M M M M M M M```
Bug#1841: man_db problems
> "Chris" == Chris Fearnley <[EMAIL PROTECTED]> writes:
Chris> Package: man
Chris> Version: 2.3.10
Chris> 1) man -k pattern gives error messages:
Chris> [EMAIL PROTECTED]:~$ man -k ftp
Chris> apropos: warning: can't read the fallback whatis text database.
Chris> apropos: /usr/local/man/whatis: No such file or directory
Chris> ...
I would strongly suspect that you have added the new man hierachy
/usr/local/man without rerunning the mandb program to record its
contents.
How should you know you should do this? The error message is correct
but doesnt mention mandb. Running mandb was done quietly for you when
you installed the package. I suspect an extra line like in the error
output something like below is necessary:
apropos: if /usr/local/man is new you should rerun mandb.
Maybe the postinstall scripts should install /usr/local/man for you so
everything is setup correctly? - I can imagine many people falling over
this. Maybe the base package should do this?
Chris> 2) man -l to display the man page in the current directory will
Chris> overwrite the page in /var/catman/... --- NOT useful for comparing an
Chris> uninstalled man page with one already installed!
Hmm, it doesnt for me - below you can see I format a local copy of
pppd.8 and it doesnt touch the existing catman page. The man -d shows
it does the expected thing:
# cd /tmp
# cp /usr/man/man8/pppd.8 .
# ls
pppd.8
# ls -l /var/catman/cat8/pppd.8.gz
-rw-r--r-- 1 man meiko 12091 Nov 13 11:27
/var/catman/cat8/pppd.8.gz
# man -l pppd.8
Reformatting pppd.8, please wait...
# ls -l /var/catman/cat8/pppd.8.gz
-rw-r--r-- 1 man meiko 12091 Nov 13 11:27
/var/catman/cat8/pppd.8.gz
# man -ld pppd.8
using more -s as pager
pre-processors `' from default
format: 1, save_cat: 0, found: 1
Reformatting pppd.8, please wait...
trying command: /usr/bin/zsoelim 'pppd.8' | /usr/bin/nroff -mandoc | { export
MAN_PN LESS; MAN_PN='pppd\.8'; LESS="$LESS\$-Pm\:\$ix8mPm Manual page $MAN_PN
?ltline %lt?L/%L.:byte %bB?s/%s..?e (END):?pB %pB\\%.."; more -s; }
close_catalogue()
#
Chris> 3) conflicts with man pages provided in package minicom 1.71-2. When
Chris> I type man sz I get garbage:
Chris> [EMAIL PROTECTED]:~$ man -d sz >& junk
Chris> [EMAIL PROTECTED]:~$ cat junk
< path stuff deleted>
Chris> ult_src: File /usr/man/man1/sz.1
Chris> found ultimate source file /usr/man/man1/sz.1
Chris> chdir /usr/man
Chris> pre-processors `Revision Level ' from file
Chris> man: ignoring unknown preprocessor `R'
Chris> man: ignoring unknown preprocessor `v'
Chris> man: ignoring unknown preprocessor `i'
Chris> man: ignoring unknown preprocessor `s'
Chris> man: ignoring unknown preprocessor `i'
Chris> man: ignoring unknown preprocessor `o'
Chris> man: ignoring unknown preprocessor `n'
Chris> man: ignoring unknown preprocessor ` '
Chris> man: ignoring unknown preprocessor `L'
Chris> man: ignoring unknown preprocessor `v'
Chris> man: ignoring unknown preprocessor `l'
Chris> man: ignoring unknown preprocessor ` '
Here it is trying to use the string `Revision Level' as a set of
pre-processor commands - it only understands the e for run eqn.
It does this 3 times and generates the extra garbage
If the catman page doesnt exist then you can see:
popen ("/usr/bin/zsoelim '/usr/man/man1/sz.1' | /usr/bin/neqn |
/usr/bin/neqn | /usr/bin/neqn | /usr/bin/nroff -mandoc", "r")
The problem is that the sz.1/rz.1/... man pages have the following at
the start:
'\" Revision Level
'\" Last Delta 04-21-88
.TH SZ 1 OMEN
.SH NAME
I suspect that '\" should be .\" (comment)
There is still a bug here to do with pre-processing so I will forward
this on to the man_db maintainers.
Chris> I copied the sz.1 source file into my Slackware and Red Hat
partitions
Chris> and the man program included with them can read the page with no
Chris> trouble at all. Implies that man_db isn't set up robustly enough.
Probably true. :(
Chris> This is under debian 0.93R6, kernel 1.2.13, and libc 4.6.27.
Chris> BTW, I like Debian. Thanks for all the hard work!
Good. :)
--
Alvar Bray
Meiko LimitedPhone:+44 1454 616171
650 Aztec West Fax: +44 1454 618188
Bristol BS12 4SD E-Mail: [EMAIL PROTECTED]
England WWW:
Bug#1853: top fails after 130 processes?
Hello Karl Ferguson! }Package: procps }Version: 0.97-4 }I was running top on my system tonight and it showed the stats of 124 }processes etc etc. Then on the next "view" 5 seconds later it spits out: } }top: Help! Too many processes } }Then it exits, and wont let me run it until it's below 130ish - is this what }it's meant to do? Otherwise killing processes that run away when using over }120 odd processes would be murder going through a normal ps list. As I remember this is a strange behaviour. It is based in two things: a) newer kernel have a process limit of 256, if I'm correct. (I don't remember where this is defined) b) Older kernel only have a limit of 130 or so. A `top' that was compiled on such a system will only allow you to have that much processes. In former times this was a trick to test the distribution. You can easily check wether the maintainer made the work to compile this himself or to copy it from whereever. I'm a bit confused if this beat Debian, too. Regards, Joey -- / Martin Schulze * [EMAIL PROTECTED] * 26129 Oldenburg / / +49-441-777884 * Login&Passwd: nuucp * Index: ~/ls-lR.gz / / http://home.pages.de/~joey/ / The only stupid question is the unasked one /
Bug#1859: Start/stop scripts do not work when not invoked by init
Package: sysvinit Version: 2.57b-1 I want to use the scripts in /etc/init.d to manually start/stop daemons, e.g. do a "xdm stop", then change the config, and finally "xdm start". I noticed that "xdm start/stop" does not work, when called manually. This is because it (and many other scripts) include /etc/init.d/functions, which destroys $1 (start/stop), if $RUNLEVEL is not set (which it is not, if I call the script by hand). I think "functions" should restore the arguments. The following patch makes works for me: --- functions.hsorigSat Nov 4 21:12:25 1995 +++ functions Sun Nov 12 23:16:49 1995 @@ -15,12 +15,14 @@ # Set RUNLEVEL and PREVLEVEL if [ "$RUNLEVEL" = "" ] then + args="$*" levels=`runlevel` if [ $? = 0 ] then - eval set $levels + set $levels PREVLEVEL=$1 RUNLEVEL=$2 + set $args fi fi Harald Schueler Universitaet EssenTel +49-201-1832456 Fachbereich 7 Fax +49-201-1832120 45117 Essen Email [EMAIL PROTECTED]
