Re: xterm defaults
On Thu, Mar 21, 2002 at 07:01:52PM -0300, Marcelo Chiapparini wrote: Hello, how can I change the default colors for xterm?. I need that the command xterm open a xterminal with the desired colors, instead using xterm -bg color, for example. I didn't find this information in the man pages. [...] Quick solution: Edit ~/.Xresources XTerm*background: color_of_your_choice XTerm*foreground: color_of_your_choice Read /usr/X11R6/lib/X11/rgb.txt which contains lots of possible color-names. Kind regards, Andreas
Re: .deb packages from outside debian/dists/ ?
On Sun, Feb 10, 2002 at 12:04:01AM -0500, Ben Collins wrote: On Sat, Feb 09, 2002 at 08:31:50PM -0800, Paul E Condon wrote: I want to install some software that I downloaded as a .deb file, but it did not come from a Debian mirror. Instead it came from a ftp download from a developer's web site. The only way I have ever used dselect, dpkg, or apt-get have been to get stuff from a Debian mirror and then unpack and install it. This happened all automatically. Now I have a file in my home directory. Where do I put it? What commands and options do I use? dpkg -i blah.deb [...] If you decided to uninstall 'blah.deb', wouldn't you run into trouble? One may create a 'Packages' file for some extra deb-package: localhost:~# cd /usr/local localhost:/usr/local# mkdir debs localhost:/usr/local# cd debs localhost:/usr/local/debs# mkdir package_name localhost:/usr/local/debs# cp path/to/debfile package_name localhost:/usr/local/debs# dpkg-scanpackages package_name\ some_empty_override_file\ debs/ Packages Additional entry in /etc/apt/sources.list: deb file:/usr/local debs/ This way installing or uninstalling may be more properly achieved. At least I have been doing it this way, and it has been working fine. Any comments? Kind regards, Andreas.
Re: icewm starts gnome
On Mon, Feb 11, 2002 at 11:16:45PM -0800, Lars Jensen wrote: When have both icewm and gnome on my machine, and want to be able to start up icewm without gnome starting up, so I have used update-alternatives to set my default x-window-manager. However when I start up icewm gnome also starts up, and I get both the icewm menubar, and the gnome menubar on top at the bottom of the window. How do I start icewm without gnome also starting up? [...] Hello Lars, check ~/.xsession for gnome-relational entries and uncomment them eventually. Basically my ~/x.session contains nothing but: exec icewm I hope this helps. Kind regards, Andreas.
Re: icewm starts gnome
On Tue, Feb 12, 2002 at 06:46:45PM +0100, Andreas Hetzmannseder wrote: [...] check ~/.xsession for gnome-relational entries and uncomment them ^ Sorry, of course I mean 'comment out' ...
Re: .deb packages from outside debian/dists/ ?
On Tue, Feb 12, 2002 at 09:32:38AM -0800, Craig Dickson wrote: begin Andreas Hetzmannseder quotation: dpkg -i blah.deb [...] If you decided to uninstall 'blah.deb', wouldn't you run into trouble? One may create a 'Packages' file for some extra deb-package: localhost:~# cd /usr/local localhost:/usr/local# mkdir debs localhost:/usr/local# cd debs localhost:/usr/local/debs# mkdir package_name localhost:/usr/local/debs# cp path/to/debfile package_name localhost:/usr/local/debs# dpkg-scanpackages package_name\ some_empty_override_file\ debs/ Packages Additional entry in /etc/apt/sources.list: deb file:/usr/local debs/ This way installing or uninstalling may be more properly achieved. At least I have been doing it this way, and it has been working fine. Any comments? Just one: it's completely unnecessary. dpkg can uninstall any currently-installed package. It has no need, and probably completely ignores, the original .deb file. [...] Thanks, Craig, I will try out dpkg -i right now... Andreas.
Re: .deb packages from outside debian/dists/ ?
On Tue, Feb 12, 2002 at 01:39:23PM -0500, David Z Maze wrote: Andreas Hetzmannseder [EMAIL PROTECTED] writes: On Sun, Feb 10, 2002 at 12:04:01AM -0500, Ben Collins wrote: dpkg -i blah.deb [...] If you decided to uninstall 'blah.deb', wouldn't you run into trouble? Of what sort? dpkg doesn't need the package file to uninstall an installed package. One may create a 'Packages' file for some extra deb-package: ... Additional entry in /etc/apt/sources.list: The only thing this gains you is having your packages not be advertised as obsolete or locally installed in dselect or aptitude. It also might be a win if you're sharing custom packages between several machines using a local Web server or a shared filesystem. [...] Thanks for your suggestions, Mr. Maze. I just tried out installing an extra deb-package called phoenix via dpkg -i ...much simpler anyway :) Interestingly I found no difference in dselect's hanling: It is listed under 'Up-to-date Extra packages in section alien' as it was before. I have no experience with aptitude yet. Removing/purging the package again was no problem at all. I tend to use dselect all the time. Kind regards, Andreas.
Re: How to add plugins to Mozilla
On Fri, Apr 27, 2001 at 05:36:23PM -0500, Kent West wrote: For months off-and-on I've tried to add Java and/or Flash to Mozilla. Nothing I've done seems to work. I'm currently running Mozilla Build ID: 2001032614. Here's the contents of my /usr/local/mozilla/plugins directory: westek[westk]:/home/westk ls -l /usr/local/mozilla/plugins/ total 976 -rwxrwxr-x1 westkstaff2363 Feb 15 21:39 ShockwaveFlash.class -rwxrwxr-x1 westkstaff 947992 Feb 15 21:39 libflashplayer.so -rwxrwxr-x1 westkstaff 38499 Mar 26 16:05 libnullplugin.so I am logged on and running X Mozilla as westk, and I'm a member of both the westk and staff groups. [...] Concerning the java-plugin - this is how it worked for me: I opened a document with java contents, to make Mozilla download and install the plugin automatically. BTW I needed to do this as root - permissions for my /usr/local/mozilla/plugins are: drwxr-xr-x3 11191uucp 4096 Apr 7 21:00 plugins and I *am* member of the uucp group... Now I realize that I didn't have the permission to create the necessary subdirectories as normal user... You definitely need write permission for /usr/local/mozilla/plugins. After download everything has been installed under the java2/ subdirectory. Additionally one needs to create a symlink like the following: cd /usr/local/mozilla/plugins ln -s java2/plugin/i386/ns600/libjavaplugin_oji.so . Finally you may chown westk.staff libjavaplugin_oji.so to get the permissions properly. Good luck, Andreas.
Re: How to add plugins to Mozilla
On Fri, Apr 27, 2001 at 05:36:23PM -0500, Kent West wrote: For months off-and-on I've tried to add Java and/or Flash to Mozilla. Nothing I've done seems to work. I'm currently running Mozilla Build ID: 2001032614. Here's the contents of my /usr/local/mozilla/plugins directory: westek[westk]:/home/westk ls -l /usr/local/mozilla/plugins/ total 976 -rwxrwxr-x1 westkstaff2363 Feb 15 21:39 ShockwaveFlash.class -rwxrwxr-x1 westkstaff 947992 Feb 15 21:39 libflashplayer.so -rwxrwxr-x1 westkstaff 38499 Mar 26 16:05 libnullplugin.so I am logged on and running X Mozilla as westk, and I'm a member of both the westk and staff groups. [...] Concerning the java-plugin - this is how it worked for me: I opened a document with java contents, to make Mozilla download and install the plugin automatically. BTW I needed to do this as root - permissions for my /usr/local/mozilla/plugins are: drwxr-xr-x3 11191uucp 4096 Apr 7 21:00 plugins and I *am* member of the uucp group... Now I realize that I didn't have the permission to create the necessary subdirectories as normal user... You definitely need write permission for /usr/local/mozilla/plugins. After download everything has been installed under the java2/ subdirectory. Additionally one needs to create a symlink like the following: cd /usr/local/mozilla/plugins ln -s java2/plugin/i386/ns600/libjavaplugin_oji.so . Finally you may chown westk.staff libjavaplugin_oji.so to get the permissions properly. Good luck, Andreas.
Re: permissions and fstab
On Mon, Mar 26, 2001 at 06:04:21PM -0300, Marcelo Chiapparini wrote: Hi! I have a dos partition which belongs to a win98 disk. I need to access this dos partition from linux, in order to write some stuff on it. This partition is mounted automatically at startup trhough the appropriate entry in the fstab file. The problem are the permissions. This dos partition has root as owner, so I don't have permission to write to it from my personal account. How can I fix it at boot time? i.e I need this partition mounted with the correct permissions which allows me to write to it from my personal account (which is not root! :)). Hi, Marcelo I don't know how your fstab looks like but I think you need the user option. The entry for my win95-partition looks like this: /dev/hda1 /mnt vfat defaults,user,noauto 0 0 This is mounted manually by any normal user and does always have the user's permissions. Hope this helps. Kind regards, Andreas.
Re: permissions and fstab
On Mon, Mar 26, 2001 at 06:04:21PM -0300, Marcelo Chiapparini wrote: Hi! I have a dos partition which belongs to a win98 disk. I need to access this dos partition from linux, in order to write some stuff on it. This partition is mounted automatically at startup trhough the appropriate entry in the fstab file. The problem are the permissions. This dos partition has root as owner, so I don't have permission to write to it from my personal account. How can I fix it at boot time? i.e I need this partition mounted with the correct permissions which allows me to write to it from my personal account (which is not root! :)). Hi Marcelo Sorry for the corrupted message... I don't know how your fstab looks like but I think you need the user option. My fstab entry for the windows-partition looks like this: /dev/hda1 /mnt vfat defaults,user,noauto 0 0 /dev/hda1 is mounted manually by any normal user and every user who mounts it gets the right permissions. I hope this helps. Kind regards, Andreas.
Re: permissions and fstab
On Mon, Mar 26, 2001 at 06:04:21PM -0300, Marcelo Chiapparini wrote: Hi! I have a dos partition which belongs to a win98 disk. I need to access this dos partition from linux, in order to write some stuff on it. This partition is mounted automatically at startup trhough the appropriate entry in the fstab file. The problem are the permissions. This dos partition has root as owner, so I don't have permission to write to it from my personal account. How can I fix it at boot time? i.e I need this partition mounted with the correct permissions which allows me to write to it from my personal account (which is not root! :)). Hi Marcelo Sorry... third try... seems to be a bad day for me... I don't know how your fstab looks like but I think you need the user option. My fstab entry for the windows-partition looks like this: /dev/hda1 /mnt vfat defaults,user,noauto 0 0 /dev/hda1 is mounted manually by any normal user and every user who mounts it gets the right permissions. I hope this helps. Kind regards, Andreas.
Re: gzipped dvi's
On Fri, Mar 23, 2001 at 11:23:16AM +0100, Sebastiaan wrote: Hi, I wondered if there is an easier way to view .dvi.gz files than to unpack them first to /tmp. I looked through the optionlist of xdvi, tried pipes, but found nothing usefull. [...] Hi, Sebastiaan, You may want to try `see yourfile.dvi.gz` The xdvi-viewer should pop up automatically. Hope this hels, Andreas.
Re: Log rotation in Debian /var/log
On Wed, Mar 21, 2001 at 12:31:28PM -0500, Paul D. Smith wrote: [...] I mean, I know about logrotate and I ass-u-me that the logs in /var/log are rotated using logrotate... no? When I look in /etc/cron.daily/logrotate I see an invocation of logrotate with the config file /etc/logratate.conf. All well and good. But, there is no entry in /etc/logrotate.conf that pertains to the general contents of /var/log! It does /var/log/wtmp and /var/log/btmp and that's it. How are the rest of the log files getting rotated? Is this built into logrotate somehow so it doesn't need to be configured? Or what? [...] Most of the system files are being rotated by sysklogd. Regards, Andreas.
Why is learning groff being made so difficult?
Hello, debian-users, Well, groff's manpages are ...interesting... right? So I would like to read the documents that came with the package in /usr/share/doc/groff, but I don't know how to display them correctly. The best solution I found was something like: groff -Tascii meintro.me | less which is still terrible to read (missing words, ugly spaces etc.) I feel like being told to learn encrypting by an encrypted manual :( Can someone help, me please? Thank you, Andreas
Re: Why is learning groff being made so difficult?
On Thu, Mar 08, 2001 at 04:50:52PM -0500, Andrew Perrin wrote: Actually, no, try: zcat meintro.me | groff -me -Tascii | less which will actually translate it. If you're trying to learn groff, reading groff source probably won't help :) Yes! This works perfectly. I didn't even know how to invoke the macro... Thank you very much. Andreas
Re: Q: Need cron is anacron loading?
On Sat, Mar 03, 2001 at 01:34:54PM +0100, Jonathan Gift wrote: Hi, I don't have my box on 24hrs a day and cron never worked its magic until I loaded anacron. But I noticed on boot that both cron and anacron are loading. Do I need cron if I have anacron? Or do they complement one another? Hello Jonathan, I just found this in anacron's README: -QUOTE- [...] Anacron is not an attempt to make cron redundant. It cannot currently be used to schedule commands at intervals smaller than days. It also does not guarantee that the commands will be executed at any specific day or hour. It isn't a full-time daemon. It has to be executed from boot scripts, from cron-jobs, or explicitly [...] -END OF QUOTE- Kind regards, Andreas.
Re: Log Rotate on Debian via Cron Not working!
On Thu, Mar 01, 2001 at 07:30:03PM -0300, Martin Marconcini wrote: Hi, I have indicated syslog to log kernel msgs onto /var/log/kernel Now Since i use ipmasq with some loggin the file will is growing fast... 2mb per day. I'd like to set a cron job that keeps a couple of days back logs and starts a new each day. How is cron setted up on Debian? Could not make it work properly. (On redhat i did it some time ago) Thanks in advance. Hello Martin, does your computer run 24 hours a day? If not you should check whether anacron is installed - otherwise cron won't work as expected. hth regards Andreas.
logfiles' rotation doesn't work at all...
Dear debian-users, In spite of the sysklogd scripts in /etc/cron.daily and /etc/cron.weekly my system's log files keep growing and growing. In fact they never have been rotated since potato's installation in October 2000... Here are the relevant entries in my /etc/cron.daily/sysklogd: #! /bin/sh cd /var/log for LOG in `syslogd-listfiles` do if [ -f $LOG ]; then savelog -g adm -m 640 -u root -c 7 $LOG /dev/null fi done for LOG in `syslogd-listfiles --auth` do if [ -f $LOG ]; then chown root.adm $LOG chmod o-rwx $LOG fi done /etc/init.d/sysklogd reload My /etc/cron.weekly/sysklogd looks like this: #! /bin/sh cd /var/log for LOG in `syslogd-listfiles --weekly` do if [ -f $LOG ]; then savelog -g adm -m 640 -u root -c 4 $LOG /dev/null fi done for LOG in `syslogd-listfiles --auth` do if [ -f $LOG ]; then chown root.adm $LOG chmod o-rwx $LOG fi done /etc/init.d/sysklogd reload 'syslogd-listfiles --auth' gives me: /var/log/auth.log 'syslogd-listfiles' lists: /var/log/messages /var/log/kern.log /var/log/syslog /var/log/debug 'syslogd-listfiles --weekly' returns: /var/log/messages /var/log/mail.warn /var/log/mail.err /var/log/mail.info /var/log/uucp.log /var/log/lpr.log /var/log/user.log /var/log/kern.log /var/log/auth.log /var/log/mail.log /var/log/daemon.log /var/log/debug Logfiles must not occur in /etc/cron.weekly and /etc/cron.daily simultaneously. This would affect the following: /var/log/messages /var/log/kern.log /var/log/debug /var/log/auth.log I would like to exclude everything but /var/log/syslog in /etc/cron.daily/sysklogd, like this: # /bin/sh cd /var/log savelog -g adm -m 640 -u root -c 7 /var/log/syslog /dev/null /etc/init.d/sysklogd reload -but it doesn't work :( Would someone please let me know how to fix this? How do I edit my sysklogd scripts properly? How do I determine what is listed in 'syslogd-listfiles' and 'syslogd-listfiles --weekly'? Thank you in advance. Andreas.
Re: Compiling java (gcj, jikes, kaffe...)
Thanks very much, D-Man,
Sorry for my late reply, I was on a holiday.
I have no experience with C/C++ yet, but I will keep your explanations,
as I am willing to learn.
On Tue, Jan 30, 2001 at 04:18:04PM -0500, D-Man wrote:
On Sat, Jan 27, 2001 at 08:43:04PM +0100, Andreas Hetzmannseder wrote:
| Dear debian-users,
|
| This is really stupid, but I don't know any further.
| Every time I try to compile some java source code - actually the HelloWorld
| example from the Debian Java FAQ - I get error messages:
|
| with gcj:
| /usr/lib/crt1.o: In function `_start':
| /usr/lib/crt1.o(.text+0x18): undefined reference to `main'
| collect2: ld returned 1 exit status
gcj outputs C++ object files from the Java source. Then (unless you
gave it the -c option) it links it to form an ELF binary (just like
C/C++). The problem is the linker can't find main.
int main( int argc , char** argv )
{
/* ... */
return 0 ;
}
The solution is to tell gjc where main is:
gcj --main=MyJavaClassDefiningMain *.java
gcj will then run a script (installed with it) to spit out some C/C++
code with the appropriate main() function that will call the main you
defined in your Java class. If you give gcj the -c option it will
stop when it has produced C++ object files (.o). You can then link
them at the end.
gcj also has a -C option will tells it to output Java bytecodes
(.class) instead. These will then need some JVM (jre, kaffe, etc) to
run.
The -C option is exactly what I needed...
|
| with jikes:
| Found 2 system errors:
|
| *** Error: Could not find package named
| /usr/share/java/repository/java/util
|
| *** Error: Could not find package named
| /usr/share/java/repository/java/lang
|
The packages java.util and java.lang can't be found by the compiler.
I haven't used jikes before so I can't be more specific. Maybe a
CLASSPATH problem?
I managed to run the following:
jikes -classpath /usr/share/java/libgcj.zip MyApp.java
I still don't know what to do with /usr/share/java/repository - it's
empty...
| with kaffe's kjc:
| java.lang.NoClassDefFoundError: at/dms/kjc/Main
| at java.lang.Throwable.init(Throwable.java:38)
| at java.lang.Error.init(Error.java:21)
| at java.lang.LinkageError.init(LinkageError.java:21)
| at
java.lang.NoClassDefFoundError.init(NoClassDefFoundError.java:21)
|
I guess that kjc is a shell script that tries to run a java program.
Apparently the main() function here is in the class Main in the
package at.dms.kjc and the class can't be found. This is definitely
a CLASSPATH problem.
When I tried out kaffe, it had the commands javac for compiling and
java for executing bytecodes. I think they were really just
symlinks to the kaffe commands. Perhaps trying these commands will
help?
javac is not available in my version of kaffe (1.0.5)
| BTW: Running ready programs with kaffe works just fine...
|
| I put the following entry into my ~/.profile:
| export CLASSPATH=/usr/share/kaffe
I would recommend against this. At my present job I am using a Win2k
box (with cygwin of course) and Sun's jdk. I found it is much better
to make a whole collection of shell scripts that execute the proper
JVM (I have 1.1.8, 1.2.2, and 1.3 installed) with the proper classpath
stuff on the command line. [...]
For the time being I only have kaffe's JVM installed (Sun's jdk 1.2 is
running with win95...) - I will keep this in mind when it comes to
additional JVMs.
Thank you very much again,
Andreas.
Re: partial success (was: getting started with compiling java...)
On Sun, Jan 28, 2001 at 10:03:14PM +0100, Andreas Hetzmannseder wrote: Hi debian-users, For the time being I would like to compile some kind of HelloWorld application, be it with gcj, jikes or kjc. All of the three give me error messages, and I have no idea how to fix them. Here they are: with gcj: /usr/lib/crt1.o: In function `_start': /usr/lib/crt1.o(.text+0x18): undefined reference to `main' collect2: ld returned 1 exit status with jikes: Found 2 system errors: *** Error: Could not find package named /usr/share/java/repository/java/util *** Error: Could not find package named /usr/share/java/repository/java/lang with kaffe's kjc: java.lang.NoClassDefFoundError: at/dms/kjc/Main at java.lang.Throwable.init(Throwable.java:38) at java.lang.Error.init(Error.java:21) at java.lang.LinkageError.init(LinkageError.java:21) at java.lang.NoClassDefFoundError.init(NoClassDefFoundError.java:21) My CLASSPATH would be /usr/share/kaffe [...] O.K., I will answer to myself... I managed to run jikes -classpath /usr/share/java/libgcj.zip MyApp.java - though I still don't know what to do with /usr/share/java/repository... Still no success with gcj and kjc... Greetings, Andreas.
getting started with compiling java - please help
Hi debian-users, For the time being I would like to compile some kind of HelloWorld application, be it with gcj, jikes or kjc. All of the three give me error messages, and I have no idea how to fix them. Here they are: with gcj: /usr/lib/crt1.o: In function `_start': /usr/lib/crt1.o(.text+0x18): undefined reference to `main' collect2: ld returned 1 exit status with jikes: Found 2 system errors: *** Error: Could not find package named /usr/share/java/repository/java/util *** Error: Could not find package named /usr/share/java/repository/java/lang with kaffe's kjc: java.lang.NoClassDefFoundError: at/dms/kjc/Main at java.lang.Throwable.init(Throwable.java:38) at java.lang.Error.init(Error.java:21) at java.lang.LinkageError.init(LinkageError.java:21) at java.lang.NoClassDefFoundError.init(NoClassDefFoundError.java:21) My CLASSPATH would be /usr/share/kaffe Any suggestion on what is wrong or missing will be highly appreciated. Thank you, Andreas.
Compiling java (gcj, jikes, kaffe...)
Dear debian-users, This is really stupid, but I don't know any further. Every time I try to compile some java source code - actually the HelloWorld example from the Debian Java FAQ - I get error messages: with gcj: /usr/lib/crt1.o: In function `_start': /usr/lib/crt1.o(.text+0x18): undefined reference to `main' collect2: ld returned 1 exit status with jikes: Found 2 system errors: *** Error: Could not find package named /usr/share/java/repository/java/util *** Error: Could not find package named /usr/share/java/repository/java/lang with kaffe's kjc: java.lang.NoClassDefFoundError: at/dms/kjc/Main at java.lang.Throwable.init(Throwable.java:38) at java.lang.Error.init(Error.java:21) at java.lang.LinkageError.init(LinkageError.java:21) at java.lang.NoClassDefFoundError.init(NoClassDefFoundError.java:21) BTW: Running ready programs with kaffe works just fine... I put the following entry into my ~/.profile: export CLASSPATH=/usr/share/kaffe Could someone please tell me what is wrong or missing? Thanks in advance, Andreas.
Re: exim and mtt probs continue
On Sat, Nov 18, 2000 at 03:01:22PM -, John-Mark wrote: [...] what is my visible mail name [...] The visible name should be the name you gave to your computer. I can't fine where to set up the isp smtp info re sending mail. When exim asks for your smarthost, enter the name for your provider's smtp server. Fetchmail seems to be ok but where does it write the emails to. Your incoming mail will go to /var/spool/mail/jmj by default. I have installed mutt but when i run it i get the message that there is no /var/spool/mail/jmj directory jmj being my user name. Apparently exim should set this up for me/ I am truly at a loss. Don't worry, just make sure it already exists. Mutt is also supposed to make a file called .muttrc in my home directory but this does not exist [...] You have to do this yourself. There is a /etc/Muttrc which should already work for you. You may want to copy it into your home directory and rename it as '.muttrc'. If you have any idea how to proceed. I also have a ppp-dialup-connection to my internet provider. Below are my answers during /usr/sbin/eximconfig. Replace them with your's where necessary. - What is the 'visible' name of your system:k-star - Does this system have any other names?:none - Are there any domains you want to relay mail for?:none - Are there any networks of local machines you want to relay mail for?:none - Would you like to use the RBL?:n - Which machine will act as the smarthost?:smtp.netway.at - Which user account should system administrator mail go to?:andy After you have finished your configuration, you should check your /etc/exim.conf for your rewriting rule in the section 'REWRITE CONFIGURATION'. Make sure that it is enabled, i.e. uncommented. Then you have to enter something like the following into your /etc/email-addresses: jmj:[EMAIL PROTECTED] Now you should be able to send and receive mail. If you want to use mutt's mailboxes you have to configure procmail additionally. I hope this helps. By the way: Wrap your lines. Also make question marks when asking questions. You will certainly be rewarded with better responses. Kind regards, Andreas.
Re: XF86Setup
On Wed, Oct 25, 2000 at 03:51:25PM -0500, Paul T. McNally wrote: [...] I suspect XF86Setup and xvidtune will pop up but are not really doing anything. You have to be root to make changes via xvidtune effective. If I select the option to start from scratch I get the errno=11 message as if I had not installed vga 16 server. You need xserver-vga16 for the graphical display of XF86Config. [...] XF86Setup is using the svga server and can't revert back? [...] No. You will be asked whether you want to make your newly installed xserver-vga16 the default. You may want to answer 'no'. Good luck, Andreas.
Re: mouse and keyboard problems in x 3.3.6
On Wed, Oct 25, 2000 at 07:56:48AM -0500, Bob Edwards wrote: [...] The line in the /etc/gpm.conf file reads : type =PS/2, so I changed the same line to type=PS/2 in the XF86Config file. I just did the same thing yesterday, and I also got into trouble... In the /etc/gpm.conf file there is a symbolic link from /dev/mouse to /etc/gpmdata. However, in the /etc/gpm.conf.file, an earlier line reads: device/etc/psaux. There is not a symbolic link between /dev/psaux to /dev/mouse. I took care of most of the keyboard problems, the only thing is that I have a microsoft natural keyboard, and I cannot use the number pad on the far right. Everything else on the keyboard works fine, and the server no longer sends me error messages when I try to start x. For your review, relevant sections of my XF86Config file as it now stands: ** keyboard section Protocol standard AutoRepeat500 5 XkbModel Microsoft EndSection *** Section Pointer ProtocolPS/2 ^^ You may want to try Microsoft instead. That was given as the default protocol during my XF86Config. Well, I thought that couldn't be right (although the mouse was moving perfectly with it) as I also have a PS/2 mouse. So I changed the protocol to PS/2 and the device to /dev/psaux and the mouse was frozen forever... Device /dev/gpmdata ^^ This should be OK. BaudRate 1200 #ZaxisMapping 4 5 Buttons3 EndSection ** X does start, and the pointer goes to the bottom of the screen, but it will only move back and forth across the bottom of the screen, but not up and down. I hope the highlighted settings work for you as they do for me. Kind regards, Andreas.
Re: Problem solved (was: Silly Problem: Unable to access deb-files via CD-ROM)
Markus Fischer schrieb: On Wed, Oct 11, 2000 at 06:08:42PM +0200, Andreas Hetzmannseder wrote : However I want to be able to use /dev/hdb as well, as it is a CD-writer. Do I have to install a specific module to make it work? Depends, what kind of manufactor, model, etc ? You know, without more details its a bit hard ;) Hallo Markus, I reedited /etc/fstab and made two separate symlinks for my two CD-drives: /dev/cdbrenner -- /dev/hdb /dev/cdrom -- /dev/hdc I now also have two separate mount points in my / -partition: /cdrom /cdbrenner according to my entries in 'fstab' /dev/hdb is working fine since then, though I just don't know what was wrong before: It's a mystery... My CD-writer is an ATAPI Mitsumi CR-4804 TE, but now it seems that there is no additional module required. By the way: This is the drive where I can read the one CD which is locked up in the regular drive (/dev/hdc), and - no big surprise - I also burned it in this drive. So I'd better remember to read the CDs in their appropriate drives, and this should keep me from getting into trouble. You see, this is not really a big problem for me. Lately I've just been testing it several times, when I had nothing better to do :) But then again it might become really annoying in special situations. Perhaps I will tell Jens Axboe about this as you suggested. Do you know his address? Best regards, Andreas.
Re: Problem with '/etc/shutdown.allow'
Carel Fellinger schrieb: [...] You see, shutdown really needs root privilege, soit. So either you use the suid trick on /sbin/shutdown (better not, and if not then there is no need to have shutdown (/sbin) in your path either), or signal some process running with root privilege to call shutdown for you. The latter is what happens with Ctrl-Alt-Del, it triggers the keyboardhandler to call shutdown. The keyboardhandler being a kernel process runs with root privilege, so that works. But... ... there is more to this story if shutdown is called with the -a flag and the file /etc/shutdown.allow exists. You see, in that case shutdown *itself* will refuse to do its work if neither root nor one of the users listed in /etc/shutdown.allow are actually logged-on on a virtual console. An xterm or XDM log-on screen won't do, you really have to be logged-on. But even if you're logged-on, shutdown -a still needs to be run with root privilege (see above), so typing it from the command line will only work if you're root or the suid bit is set. [...] Hello Carel, Thank you very much again. Concerning being root or having the suid bit: Mr. Benson showed up a third solution earlier in this thread. With sudo you can specify exactly for which command(s) you want to give root privileges to mere mortals. I just did it yesterday and it works fine - and there is no suid bid set to shutdown... This is great, don't you think? Cheers, Andreas.
Re: CD-drive blocked after successless mounting
Chris Gray wrote: [...] Try # echo 0 /proc/sys/dev/cdrom/lock as root. Cheers, Chris P.S.: This works for me, but I have a 2.4.0-testX kernel, so YMMV. [...] Well, that's funny: CD remains in jail again :) I checked /proc/sys/dev/cdrom/lock: It did contain '0' (zero). I have kernel 2.2.17 (potato). Thanks very much, anyway. Cheers, Andreas.
Re: Problem with '/etc/shutdown.allow'
Ethan Benson wrote: [...] normally anyone whether anyone is logged into the console or not can press control alt delete and reboot the system (if you have that line in inittab) adding shutdown.allow changes this so that control-alt-delete only works when a listed user is logged into any of the console ttys. if none are logged in then control-alt-delete is disabled. [...] Ah, I see: 'shutdown.allow' is used for making rebooting more restrictive - I thought I could use it for enabling mere mortals to execute the halting of the system. Now when I type 'shutdown -a -h now' it still tells me, I have to be root. It looks like I have to set a SUID-flag. But I would prefer a better solution. Otherwise: What would 'shutdown.allow' be good for? you could do two different things: add a group `shutdown' and add any authorised users to this group and do a chgrp shutdown /sbin/shutdown chmod 4754 /sbin/shutdown. or use sudo. i highly reccomend sudo over making shutdown suid. making shutdown suid allows users in that group to call shutdown with whatever arguments they want, which is not a good thing. with sudo you can give specific users the ability to run ONLY shutdown -h now with only those exact arguments and no other. then you would run instead: sudo shutdown -h now here is the relevant lines to put in /etc/sudoers (use visudo): Cmnd_Alias SHUTDOWN=/sbin/shutdown -h now Cmnd_Alias REBOOT=/sbin/shutdown -r now username hostname=SHUTDOWN,REBOOT or if you don't want the user password to be required: username hostname=NOPASSWD: SHUTDOWN,REBOOT Thank you very much, Mr. Benson - 'sudo' works just fine. just be sure to use visudo to edit sudoers it will check your syntax properly. if you don't like the editor it runs (probably vi) then export EDITOR=whatever I am definitely a vi fan and I had no problems with editing 'sudoers'. Best regards, Andreas.
Re: CD-drive blocked after successless mounting
Jason Lunz schrieb: [...] I've seen it happen where the gnome panel CD player somehow locked the tray when you put in a non-audio disc, and only its eject button would work. Try using it, or killing it and using something else. [...] I haven't got X installed yet. Regards, Andreas.
Re: Problem with '/etc/shutdown.allow'
Hello, Carel! You should have received the following message earlier, but somehow it was lost. Here it is again: Carel Fellinger wrote: On Mon, Oct 16, 2000 at 06:26:20PM +0200, Andreas Hetzmannseder wrote: Hello, debian-users! There are two users on my debian system, who I want to be able to reboot and halt without being root. For this purpose I created /etc/shutdown.allow with the corresponding user names. [...} shutdown isn't in your PATH, but even if it were it has no SUID flag so it won't do you no good. You are absolutely right. Well I wouldn't believe it, so I had to see for myself: /sbin is now in my path but what's the point of it? [...] However when I attempt to type 'shutdown -a -r now' or 'shutdown -a -h now' by myself, it says 'shutdown: command not found'. I still have to be root in this case. I've just read the Fine Manual just to please you:) and I think it only effects Ctrl-Alt-Del from the console, not calling shutdown from the command-line unless you make it a SUID program. (for the latter to have a change of working make sure the shutdown command is in your PATH, as it isn't by default for us mortal users) 'shutdown' is now in my path - what else shall I do? What happens is: the keyboard handler (running as root) intercepts the magic Ctrl-Alt-Del and tells init to shut-up. The keyboard handler knows to do this as that is what it was told by the /etc/inittab entry for the Ctrl-Alt-Del key combo. Thank you for your explanation. 'man shutdown' tells me that I should add an entry in /etc/inittab, but I don't know what to enter... Can somebody help me please? Well, you shouldn't add the entry but modify the excisting one and add to that entry the '-a' flag. [...] Now I feel completely stupid. Which existing entry shall I modify? Please have a look at my /etc/inittab: id:2:initdefault si::sysinit:/etc/init.d/rcS ~~:S:wait:/sbin/sulogin l0:wait:/etc/init.d/rc0 l1:wait:/etc/init.d/rc1 l2:wait:/etc/init.d/rc2 l3:wait:/etc/init.d/rc3 l4:wait:/etc/init.d/rc4 l5:wait:/etc/init.d/rc5 l6:wait:/etc/init.d/rc6 z6:6:respawn:/sbin/sulogin ### ca:12345:ctraltdel:/sbin/shutdown -t1 -a -r now ### kb::kbrequest:/bin/echo Keyboard Request -- edit /etc/inittab to let this work. pf::powerwait:/etc/init.d/powerfail start pn::powerfailnow:/etc/init.d/powerfail now po::powerokwait:/etc/init.d/powerfail stop 1:2345:respawn:/sbin/getty 38400 tty1 2:23:respawn:/sbin/getty 38400 tty2 3:23:respawn:/sbin/getty 38400 tty3 4:23:respawn:/sbin/getty 38400 tty4 5:23:respawn:/sbin/getty 38400 tty5 6:23:respawn:/sbin/getty 38400 tty6 Please note the highlighted candidate and that I have *not* changed anything so far. I would be glad if you could give me a clue. Thank you very much so far, Andreas.
Problem with '/etc/shutdown.allow'
Hello, debian-users! There are two users on my debian system, who I want to be able to reboot and halt without being root. For this purpose I created /etc/shutdown.allow with the corresponding user names. Now when I press Ctrl-Alt-Del the system will reboot as expected - although I have to be logged in as a qualified user, according to shutdown.allow. This looks perfectly reasonable to me as this procedure is described in /etc/inittab. However when I attempt to type 'shutdown -a -r now' or 'shutdown -a -h now' by myself, it says 'shutdown: command not found'. I still have to be root in this case. 'man shutdown' tells me that I should add an entry in /etc/inittab, but I don't know what to enter... Can somebody help me please? Or am I simply doing something wrong? Thanks in advance, Andreas.
CD-drive blocked after successless mounting
Hello, debian-users! I have two CD-drives: one for regular use and one for additional writing. I have been experiencing problems, when I tried to read burned CDs in the regular drive - well, that wouldn't matter much to me, but I wanted to try it out anyway. Here is what happened: $ mount /cdrom [... lots of error messages ...] mount: I could not determine the filesystem type, and none was specified So I pressed the eject key but it was blocked. I typed 'umount /cdrom' but it told me that cdrom was not mounted. So I was left with the CD stuck in the wrong drive, which was neither mounted nor unmounted... Of course a reboot would always help me out, but is there another way to unlock a CD-drive? If anyone knows one, please let me know. Thanks in advance, Andreas.
Re: Problem with '/etc/shutdown.allow'
Kasatenko Ivan Alex. wrote: Hello Andreas, Monday, October 16, 2000, 8:26:20 PM, you wrote: AH There are two users on my debian system, who I want to be able to reboot AH and halt without being root. For this purpose I created AH /etc/shutdown.allow with the corresponding user names. Now when I press AH Ctrl-Alt-Del the system will reboot as expected - although I have to be AH logged in as a qualified user, according to shutdown.allow. When you press Ctrl+Alt+Del, init(1) gets it and executes reboot sequence, afaik. So, the only right way in this situation is to disable Ctrl+Alt+Del at all. Sorry, Ivan, I didn't make this very clear. I *do* want to use Ctrl-Alt-Del. I just wanted to point out that it is behaving differently, since I have '/etc/shutdown.allow'. Now I have to be logged in at least as a normal user - that wasn't the case before... AH [...] However when I attempt to type 'shutdown -a -r now' or AH 'shutdown -a -h now' by myself, it says 'shutdown: command not found'. AH I still have to be root in this case. Add /usr/sbin into your path with: export PATH=$PATH:/usr/sbin in BASH, if I'm not mistaken. :) Actually 'shutdown' is in /sbin - so I added /sbin to my path... Everything should work. Now when I type 'shutdown -a -h now' it still tells me, I have to be root. It looks like I have to set a SUID-flag. But I would prefer a better solution. Otherwise: What would 'shutdown.allow' be good for? Thank you very much, anyway. Best regards, Andreas.
Re: CD-drive blocked after successless mounting
mike wrote: Try /usr/bin/eject /cdrom or /dev/hd?. I did so, but with no success :( I think the problem is, that the CD remains in a twilight status: neither mounted nor unmounted. Thank you very much anyway. Andreas. On Mon, 16 Oct 2000 18:47:23 +0200, Andreas Hetzmannseder said: Hello, debian-users! I have two CD-drives: one for regular use and one for additional writing. I have been experiencing problems, when I tried to read burned CDs in the regular drive - well, that wouldn't matter much to me, but I wanted to try it out anyway. Here is what happened: $ mount /cdrom [... lots of error messages ...] mount: I could not determine the filesystem type, and none was specified So I pressed the eject key but it was blocked. I typed 'umount /cdrom' but it told me that cdrom was not mounted. So I was left with the CD stuck in the wrong drive, which was neither mounted nor unmounted... Of course a reboot would always help me out, but is there another way to unlock a CD-drive? If anyone knows one, please let me know. Thanks in advance, Andreas. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- gEEk||dOOd^Deb+ianXFce$everything goesProX-Mozilla-Status: 0009e Oct 17 01:23:47 2000 X-Mozilla-Status: 0801 X-Mozilla-Status2: FCC: /C|/Programme/Netscape/Users/mailempfang/mail/Sent Message-ID: [EMAIL PROTECTED] Date: Tue, 17 Oct 2000 01:23:47 +0200 From: Andreas Hetzmannseder [EMAIL PROTECTED] X-Mozilla-Draft-Info: internal/draft; vcard=0; receipt=0; uuencode=0; html=0; linewidth=0 X-Mailer: Mozilla 4.73 [de]C-CCK-MCD QXW0324c (Win95; U) X-Accept-Language: de,en MIME-Version: 1.0 To: Kasatenko Ivan Alex. [EMAIL PROTECTED] CC: debian-user@lists.debian.org Subject: Re: Problem with '/etc/shutdown.allow' References: [EMAIL PROTECTED] [EMAIL PROTECTED] Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Kasatenko Ivan Alex. wrote: Hello Andreas, Monday, October 16, 2000, 8:26:20 PM, you wrote: AH There are two users on my debian system, who I want to be able to reboot AH and halt without being root. For this purpose I created AH /etc/shutdown.allow with the corresponding user names. Now when I press AH Ctrl-Alt-Del the system will reboot as expected - although I have to be AH logged in as a qualified user, according to shutdown.allow. When you press Ctrl+Alt+Del, init(1) gets it and executes reboot sequence, afaik. So, the only right way in this situation is to disable Ctrl+Alt+Del at all. Sorry, Ivan, I didn't make this very clear. I *do* want to use Ctrl-Alt-Del. I just wanted to point out that it is behaving differently, since I have '/etc/shutdown.allow'. Now I have to be logged in at least as a normal user - that wasn't the case before... AH [...] However when I attempt to type 'shutdown -a -r now' or AH 'shutdown -a -h now' by myself, it says 'shutdown: command not found'. AH I still have to be root in this case. Add /usr/sbin into your path with: export PATH=$PATH:/usr/sbin in BASH, if I'm not mistaken. :) Actually 'shutdown' is in /sbin - so I added /sbin to my path... Everything should work. Now when I type 'shutdown -a -h now' it still tells me, I have to be root. It looks like I have to set a SUID-flag. But I would prefer a better solution. Otherwise: What would 'shutdown.allow' bX-MoX-Mozilla-Status: 0009ery much, anyway. Best regards, Andreas.
Thank you! (was:Symlinking /tmp to /var...)
Andreas Hetzmannseder wrote: [...] Please can you show me a way to make sure that my /tmp-files will be going to /var? [...] Big thanks to all of you: Bruce Sven Ethan Mike ... Sorry for my late reply. I was completely offline yesterday. Cheers, Andreas.
Symlinking /tmp to /var...
Dear debian-users, The disk space for my root-partition is 40 MB, while I supplied 80 MB for my /var-partition. I would like to make a symbolic link from /tmp, which resides in the root partition, to /var. This was my plan: 1. Copying /tmp to /var/tmp.root (with /var/tmp.root being created) 2. Deleting /tmp 3. Doing ln -s /var/tmp.root/ /tmp/ Even the first step didn't succeed - it said: Omitting file ... Nothing happened, but then again I might have done something stupid. Please can you show me a way to make sure that my /tmp-files will be going to /var? - My concern is that /tmp might be getting too big for /, thus I would like to have all temporary files to go into /var. Thanks in advance, Andreas.
Re: Installation problem
Claude Aeschlimann wrote: Hi, I have a problem installing Debian 2.2 on my system [...] : - The first part goes well (both my CD-ROM and CD-Recorder are recognized) and I can install the base system. But after the reboot, the installation procedure loops no ends on the root password setting [...] . I tried going on a another console and set a password there. It worked but then it loops on creating a user. I succeded passing this step, but then it loops on scanning the CDs [...] and I can't get to the selection of packages to install. This is a loop story :-( Did someone have this problem, can someone please help ? Hallo, Claude, I had a similar problem two days ago. I also have got two CD-drives - one for regular use (/dev/hdc) and one for writing (/dev/hdb). Installing the base system worked fine, but after rebooting /dev/hdb wouldn't be recognized any more. It says: /dev/hdb is not a valid block device. That's really strange because before rebooting I was installing via /dev/hdb. Fortunately I was able to use the other drive. I assume this is a problem with missing modules but I don't know yet. I will have to experiment with modconf - you may want to try as well. But as you say that you are captured within loops, you may additionally try to deactivate one of your drives. Well, that is not a big help at all, just a little hint. I hope you can do something with it. Another problem is that when configuring lilo, I don't get the choice to install it on hda (only sda /boot or /). /dev/sda sounds OK to me, as you have a SCSI-HD. Tell me, if I'm wrtong. Ciao, Andreas.
Problem solved (was: Silly Problem: Unable to access deb-files via CD-ROM)
Andreas Hetzmannseder wrote: [...] I am attempting to install Debian from a set of official potato-CDs. Well, the installation process for the base system was rather painless up to the point when it asked me for configuring 'apt': Choose the method apt should use to access the Debian archive... [...] the installer complained, that it could not find the device (which is actually /dev/hdb in my case). After several attempts to edit /etc/fstab for /dev/hdb I realized that the system can't recognize /dev/hdb at all. It says: /dev/hdb is not a valid block device Fortunately I tried out my second CD-ROM drive (/dev/hdc) - and it worked just fine ... apt-cdrom add ... no complications ... simply fantastic! However I want to be able to use /dev/hdb as well, as it is a CD-writer. Do I have to install a specific module to make it work? Thanks for your attention, Andreas.
Silly Problem: Unable to access deb-files via CD-ROM
Dear debian-users, I am attempting to install Debian from a set of official potato-CDs. Well, the installation process for the base system was rather painless up to the point when it asked me for configuring 'apt': Choose the method apt should use to access the Debian archive... Of course I entered cdrom, but the installer complained, that it could not find the device (which is actually /dev/hdb in my case). I inserted every single CD with no success, so I finally cancelled apt-configuration. So far I have got Debian installed just with the most basic files from base2_2.tgz - and it _did_ come from CD-ROM. Up to now everything seems properly installed, but what about all the other packages? Could you please tell me how to configure apt correctly? Any kind of hint will be highly appreciated. Cheers, Andeas.
Re: Internet woes
Michael Tanney wrote: Everytime I try and use wvdial it sits there and does nothing when it reaches the point of launching pppd while the modem shows On Hold. And since wvdial does not require any extra configuration other than that which it asks you to enter at its configuration time, how do I stop pppd messing it up ? I think that the problem might be that pppd is trying to specifiy it's own IP to my ISP, because when pppd is executed without any parameters it shows on the last line none of the secrets (passwords) tried would let it use an IP address or something similar.. Can I pre-load a chatscript into pppd before using wvdial to connect. I don't actually care how I connect to the internet but no matter what I try it never works, always with the exact same problem. This comes rather late, so I don't know if you solved your problem already. If you did, just ignore the following... Does your shell prompt reappear immediately after you typed in 'pon'? If it doesn't, you may look into your /etc/ppp/options whether '-detach' is enabled. If it is, you may disable the corresponding line with a hash mark. Regards, Andreas.
Re: Sending mail via exim: spooky....
Phillip Deackes schrieb: I posted some information on this recently. I am sending it again: I'm sorry. My attention was completely captured by another thread... My machine's name (hostname) is scgf. If I look in /etc/hosts I see: 127.0.0.1 scgf localhost In /etc/hostname: scgf scgf does not exist in the real world, just on my machine. Your Linux machine needs a name. Be kind and give it a name. So my machine's name is woof... To help you understand what I have written below, my real-world email address is [EMAIL PROTECTED] My /etc/exim.conf it set up like this (there is a lot more besides, of course, but these lines are relevant to this thread): qualify_domain = gmx.co.uk For me: qualify_domain = netway.at qualify_recipient = scgf.gmx.co.uk local_domains = localhost:scgf.gmx.co.uk My version of exim (2.05-2) seems to be unfamiliar with your dot notation, so my entries are like this: qualify_recipient = woof local_domains = localhost:woof host_accept_relay = localhost This option doesn't appear in my exim.conf, but it seems that it doesn't matter in my case. The above lines ensure that all outgoing mail appears to come from [EMAIL PROTECTED] Local mail is delivered locally and does not leave my system. All incoming mail from my ISP, collected by fetchmail, is detected by Exim and is delivered according to the filters in my ~/.forward file. If you follow my example, your mail system should work OK. Well, I have to check local deliveries like system messages, but sending remote mail works fine now! Thank you very much. Andreas
Re: Sending mail via exim: spooky....
Lehel Bernadt wrote: On 26-Aug-2000 Andreas Hetzmannseder wrote: Dear debian-users I tried to send a test message to my friend who has an e-mail account on the same PC and from the same provider as me, let's say [EMAIL PROTECTED] This should have gone to my provider's smtp-server named smtp.netway.at. I edited just a little text using mutt, hit the key for 'send' [...] Well, the message hasn't arrived at our mailserver :( [...] Besides it seems, all that exim does in my case is deliver messages locally. /var/log/exim/mainlog includes the following: [...]unknown local-part friendsname in domain netway.at with netway.at being the 'visible mail name of my system' according to the following entry in my exim.conf: qualify_domain = netway.at Now I think that exim tried to deliver an error message to [EMAIL PROTECTED] according to my /etc/email-addresses: andy:[EMAIL PROTECTED] Like before /var/log/exim/mainlog said: unknown local-part hetzmann in domain netway.at [...] I think your problem is with the local_domains setting in exim.conf. Make sure you have qualify_recipient set to your hostname (not netway.at) and local_domains is commented out. According to your suggestion my settings are now as follows: qualify_recipient = woof # ...which is localhost's name... local-domains = localhost:netway.at:localhost # localhost even twice? Well, I don't mind, if exim insists # on that... My problem remains. Two questions: - 'my hostname': Do you mean the name of my PC or the name of my providers smtp-server? - 'commenting out': Do you mean a) disable by taking out of the context through commenting? b) enable by taking out the comment mark You see, I have a special talent of understanding things quite contrary, like: west wind: Does it come from the west or does it go to the west? Thank you and kind regards... Andreas.
Sending mail via exim: spooky....
Dear debian-users I tried to send a test message to my friend who has an e-mail account on the same PC and from the same provider as me, let's say [EMAIL PROTECTED] This should have gone to my provider's smtp-server named smtp.netway.at. I edited just a little text using mutt, hit the key for 'send', the test message has been relaid in my outbox file, which is probably alright. But then I recognized quite a lot of modem activity, which went on for two or three minutes. Now I do have a really slow internet connection, but several minutes for a text of just two or three lines... It semed suspicious to me, so I turned it off. Well, the message hasn't arrived at our mailserver :( But what about the modem activity? This is really spooky Besides it seems, all that exim does in my case is deliver messages locally. /var/log/exim/mainlog includes the following: [...]unknown local-part friendsname in domain netway.at with netway.at being the 'visible mail name of my system' according to the following entry in my exim.conf: qualify_domain = netway.at Now I think that exim tried to deliver an error message to [EMAIL PROTECTED] according to my /etc/email-addresses: andy:[EMAIL PROTECTED] Like before /var/log/exim/mainlog said: unknown local-part hetzmann in domain netway.at Besides, I created this file for the purpose of handling outgoing mail, but now it looks like I can't receive system messages any more which used to go to /var/spool/mail/andy. Is my /etc/email-addresses OK? But the big question is: How do I manage to make exim send remote mail correctly? Hoping for your kind support, Andreas.
Re: Bad exim configuration (was: fetchmail gives me headache)
Morten Liebach wrote: [...] My /etc/inetd.conf looks like this: #:MAIL: Mail, news and uucp services. smtpstream tcp nowait mail/usr/sbin/exim exim -bs nntpstream tcp nowait news/usr/sbin/tcpd /usr/sbin/leafnode #:INFO: Info services finger stream tcp nowait nobody /usr/sbin/tcpd /usr/sbin/in.fingerd ident stream tcp waitidentd /usr/sbin/identdidentd These are the only lines that are not commented out. Try to copy the line for smtp from above, it should work. And so I did... It works! Do you actually need all the things you have running? Certainly not. I barely know what they are good for. You see I have a lot to learn... Is it slink you use? Yes, but I want to upgrade to potato as soon as the CDs are easily available for me. I downloaded quite a big part of slink from the Internet (about 210 packages) and I have a *really* slow connection, so I don't want to go through this again. I will need potato for X. The slink version doesn't work for my video card. [...] This is the output of nmap in my case: 9 tcp discard 13 tcp daytime 21 tcp ftp 23 tcp telnet 37 tcp time 79 tcp finger 111 tcp sunrpc 113 tcp auth 512 tcp exec 513 tcp login 514 tcp shell 515 tcp printer It isn't a very secure setup on anything but a ``trusted network'', whatever that is. I'm not surprised to read that. As you are on a dial-up like me, you probably don't need any of these ports, nmap of my machine: PortState Protocol Service 22 opentcpssh secure shell. 25 opentcpsmtp EXIM!!! 79 opentcpfinger nifty thing, not important. 80 opentcphttp Apache, dwww on-line docs 113 opentcpauth identd, for IRC ... 119 opentcpnntp leafnode newsserver. 515 opentcpprinter ... 1024opentcpunknown What is this?? 6000opentcpX11 ... Ports 79 and 113 could be shut down ... but doing that isn't very inportant for me. nmap now gives me the following list: Open ports on woof (127.0.0.1): Port Number Protocol Service 25 tcp smtp 111 tcp sunrpc 515 tcp printer ...though I still don't know what sunrpc is good for. [...] Please would you be so kind to check them if they are correct? Sure! [...] Q: Which user accounts should system administrator mail go to? A: andy Is ``andy'' your username? Yes. If so, your outgoing mail will claim to come from [EMAIL PROTECTED] with this setup [...] eximconfig says: Mail for the 'postmaster and 'root' accounts is usually redirected to one or more user accounts of the actual system administrator [...who is me...]. By default [...] mail for postmaster and for various system accounts is redirected to root, and mail for 'root' is redirected to a real user [...] Which user account(s) should system administrators mail go to? Enter one or more usernames [...] Enter 'none if you want to leave this mail in root's mailbox - NB this is strongly disencouraged [...] By entering 'andy' I have got messages from my system delivered to /var/spool/mail/andy. I have to explain that this isn't really local delivery as I have a standalone PC with no LAN. It just goes from localhost to localhost. This happened several times and I felt nothing bad about it. Isn't this common practice? If not, what else should I enter? Well, at least I know that exim works for _local deliveries_ [...] So it is just a broken inetd.conf. Shouldn't inetd.conf have been updated during the installation of exim? Perhaps it is a bug in the exim package, but I haven't checked the bug report yet. Fix it as per above, and do a ``killall -HUP inetd''. And so I did. Mail is now delivered to /var/spool/mail/andy just like I thought. Thank you very much again. The next thing will be setting up mutt. [...] Look at my homepage [...] I will have a thorough look at your .muttrc... Cheers, Andreas.
Re: Bad exim configuration (was: fetchmail gives me headache)
Morten Liebach wrote: [...] _IS_ exim running and listening on port 25? If not, ``SMTP Transaction error'' would be the error message, since fetchmail speaks SMTP to exim on port 25 (fetchmail says ``Hi, I've got mail for you'', and exim says nothing, so fetchmail times out and tell you ``SMTP Transaction error''). Look in /etc/inetd.conf for a line that starts with ``smtp'', what does it say? There is no such line. This is how my inetd.conf looks like: discard stream tcp nowait root internal discard dgram udp wait root internal daytime stream tcp nowait root internal daytime dgram udp wait root internal timestream tcp nowait root internal timedgram udp wait root internal ftp stream tcp nowait root /usr/sbin/tcpd /usr/sbin/in.ftpd telnet stream tcp nowait root /usr/sbin/tcpd /usr/sbin/in.telnetd shell stream tcp nowait root /usr/sbin/tcpd /usr/sbin/in.rshd login stream tcp nowait root /usr/sbin/tcpd /usr/sbin/in.rlogind execstream tcp nowait root /usr/sbin/tcpd /usr/sbin/in.rexecd talkdgram udp wait nobody.tty /usr/sbin/tcpd /usr/sbin/in.talkd ntalk dgram udp wait nobody.tty /usr/sbin/tcpd /usr/sbin/in.ntalkd finger stream tcp nowait nobody /usr/sbin/tcpd /usr/sbin/in.fingerd ident stream tcp nowait nobody /usr/sbin/identd identd -i Try a portscanner too (nmap f.ex., do a `nmap localhost', look for this line: 25 opentcpsmtp ) There isn't such a line either. This is the output of nmap in my case: 9 tcp discard 13 tcp daytime 21 tcp ftp 23 tcp telnet 37 tcp time 79 tcp finger 111 tcp sunrpc 113 tcp auth 512 tcp exec 513 tcp login 514 tcp shell 515 tcp printer If exim isn't listenig, that's the real problem, not fetchmail. You were right! I never would have found it out by myself. So it seems that I have to reconfigure inetd.conf as well. Do you know how I can manage this? Or maybe it is just that I gave the wrong answers to eximconfig... Next comes a list of eximconfig's questions and my answers to them. Please would you be so kind to check them if they are correct? Q: What is the visible mail name of your system? This will appear on From: lines of outgoing messages. A: woof Q: Does this system have any other names which may appear on incoming messages, apart from the visible name above (woof) and the system's hostname (woof)? [...] A: localhost Q: [...] Are there any domains you want to relay mail for --- that is you are prepared to accept mail for them from anywhere on the internet, but they are not local domains. [...] A: none The answer above might be wrong in the first place, but I just didn't know what to enter else. Is this the place for my provider's domain? Q: [...] Are there any networks of local machines you want to relay mail for? A: none The answer above looks right to me as I have a standalone PC with dialup connection to my provider. Q: [...] Would you like to use the RBL? A: n (for no) Q: Which machine will act as the smarthost and handle outgoing mail? A: smtp.netway.at Q: Which user accounts should system administrator mail go to? A: andy Well, at least I know that exim works for _local deliveries_, as I get some error messages to /var spool/mail/... every now and then. [...] Look at my homepage [...] I certainly will. I already have installed mutt, it looks really nice. Thank you very much so far, Andreas.
Re: fetchmail gives me headache (was: Strange things like drwx--S--- with elm...)
Morten Liebach wrote: [...] I don't have any mda defined in my ~/fetchmailrc, delivering to port 25 on localhost is the default. Does this mean /var/spool/mail/... ? My fetchmailrc: set postmaster local_username ...OK... set bouncemail set properties Fetchmail complained about these two options (parse error), so I disabled them. Bouncemail doesn't even seem to be available in my fetchmailrc template file. What version of fetchmail do you have? Mine is 4.6.4-1.1. poll my_pop3_server with proto POP3 user userid there with password very_secret is local_username here warnings 3600 antispam 571 550 501 554 It works very well, and has done so for a year now. :-) I envy you... :) Now have a look at my fetchmailrc: set postmaster local_username #set bouncemail #set properties poll my_pop3_server with proto pop3 user userid there with password very_secret is local_username here warnings 3600 antispam 571 550 501 554 This looks like it should work, don't you think? Instead I always get an SMTP Transaction error. It reads the first incoming message for a few seconds, then it exits with connection failed and I tried it over and over again... Do I really need 'set bouncemail' and 'set properties '? Let me get the following clear: I am not having X yet, fetchmailconfig doesn't work for me, so I set up ~/.fetchmailrc by myself. For the time being all I want is to have mail delivered anywhere, so I am not even talking about any MUA, be it mutt, elm, mozilla or whatever. This will be my next big problem ;) Do you invoke fetchmail via your MUA? Sorry for my confused asking, but there are just too many problems all at once. I never thought this would become so complicated... Thanks in advance, Andreas
Re: trouble with fetchmail via exim
Florian Friesdorf wrote: On Tue, Aug 22, 2000 at 02:32:49AM +0200, Andreas Hetzmannseder wrote: [...] My ~/.fetchmailrc looks like this: poll pop.netway.at proto pop3 user my-remote-username password my-password is andy here mda /usr/bin/exim I just switch over from qmail to exim. my fetchmailrc looks very similar, except I'm not using the mda option. Somehow fetchmail knows how to contact exim. But if you use the mda option, I think it should be mda /usr/bin/exim -bs That's the way exim gets called from inetd. [...] fetchmail complains about the -bs switch. It says 'parse error'... problem remains... Regards, Andreas.
Re: trouble with fetchmail via exim
Dave Sherohman wrote: Andreas Hetzmannseder said: The following domain(s) will be recognized as referring to this system: woof Unless your incoming email is addressed to [EMAIL PROTECTED], exim will not recognize it as being for a user on your system. You need to give exim a valid local domain. So I reconfigured exim. Now it says: The following domain(s) will be recognized as referring to this system: woof, netway.at with 'netway.at' being the second part of my e-mail address. The problem remains. Did I make a stupid entry? Is there something else missing? Interestingly the error messages differ in respect to whether the mda option in ~/fetchmailrc is specified or not. If it is it says: '...neither action flags...' etc. if it is not it says: '...SMTP transaction error... Kind regards, Andreas.
Re: Strange things like drwx--S--- with elm...
Eric G . Miller wrote: Your description of fetchmail opening an elm display does not jibe with my experience in using fetchmail. The reason for this was probably the following entry in my ~/.fetchmailrc: mda /usr/bin/elm fetchmail --version gave me the following explanation: Messages will be delivered with /usr/bin/elm. Generally fetchmail polls a POP or IMAP server for mail and delivers it to the local smtp server. I guess you mean the mail transport agent (exim in my case). Please correct me if I'm wrong. I changed the mda option to /usr/sbin/exim anyway. Is this OK? I hope it will work. The mail should be delivered to /var/mail/user if everything is configured properly. This looks like /var/spool/mail/user in my case. Do you have a ~/.forward file that may be causing weird behavior? No. It might have been just the wrong entry for mda. But I will see... Thanks for your help, Andreas
trouble with fetchmail via exim
Dear debian-users, I ran fetchmail, it read the first incoming message, but then it stopped and told me the following: ..exim: neither action flags nor mail addresses given. fetchmail terminated with signal 13 My ~/.fetchmailrc looks like this: poll pop.netway.at proto pop3 user my-remote-username password my-password is andy here mda /usr/bin/exim eximconfig gives me the following summary: Mail generated on this system will have 'woof' used as the domain part (after the @ in the From: field and similar places). The following domain(s) will be recognized as referring to this system: woof Mail for postmaster, root, etc. will be sent to andy. Local mail is delivered. Outbound remote mail is sent via smtp.netway.at. If anybody knows, what might be wrong or missing, please let me know. Thanks in advance, Andreas.
Strange things like drwx--S--- with elm...
Dear debian-user, After running elm (version elm-me+_2.4pl25ME+65-0.slink.0) for the first time yesterday it created the following file in my home directory: drwx--S--- myusername myusername 1024 Aug 19 18:29 Mail/ which is supposed to be for incoming mail. Please note the capital S. Furthermore there was no .elm/elmrc configuration file yesterday, but today there were two of them: elmrc and elmrc.old ... Now when I try to receive e-mail via fetchmail it opens the elm-display, the incoming messages move quickly by - no chance to read them - then it says they are flushed and it seems that they are gone forever... But that's not all: After about a dozen of messages elm asks to create a folder with a funny name which seems to be a substring of some part of the received mail. This looks like: rom: Boy George [EMAIL PROTECTED] or similar in my home directory, which is again drwx--S--- ... And yet some other weird stuff: I checked my elmrc configuration file and found some options being replaced by text which seems to have been taken from the mail. I reedited the file, ran fetchmail again, but with the same flushing, with elm asking me to create a funny folder and with similar funny text entries in elmrc. I would like to know: - What does the capital S mean? - How can I configure elm properly? -In the User's Guide it is said to run Configure which is presumably a shellscript but I haven't found it anywhere. Please let me know, I'm really curious. Thanks in advance, Andreas
Re: dselect shows all packages to be removed?
null void schrieb: Suddenly...dselect now shows all packages pending removal...this is dangerous...Can I make dselect somehow correct itself? Highlight the corresponding header for your packages and press the key of your choice, e.g. '+' for installation. Regards, Andreas
Re: Keyboard troubles...
Christian Pernegger wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I've skimmed the German_HOWTO, played with console-tools and generally familiarized myself a bit with keymaps and the like - here's what I've come up with... The default loaded keymap is de-latin1-nodeadkeys which is the correct one. The relevant entries are: keycode 12 = ssharpquestionbackslash keycode 26 = +udiaeresis +Udiaeresis keycode 39 = +odiaeresis +Odiaeresis keycode 40 = +adiaeresis +Adiaeresis As long as diaeresis means umlaut that should be correct. '/usr/bin/showkey -m' displays the correct characters for all keys, but the results on the console are: szlig; does cursor up auml; Auml; does nothing at all uuml; beeps Uuml; does nothing ouml; Ouml; beeps Help! Christian -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.2 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE5geb8txWmQklOL8URAonBAJ0Y7pLcPKPBbfnN0afsbFVRbxIZkACgjao8 5JvKabu+X9onCeZ4Zhv0zPk= =K9jP -END PGP SIGNATURE- To display sharp S and umlauts with bash you may edit a file named '.inputrc' which contains just three lines: set meta-flag on set convert-meta off set output-meta on Put this file into your home directory. I hope this helps. Andreas
Re: Newbie Install Questions
Dale Morris wrote: [...] If I download a file from the net, where does it go? [...] The best thing is to create a directory tree like the one on the ftp- server: /debian/dists/stable/main/... etc. You just need to build the paths that are important to you. Put the downloaded files along these paths and dselect will find the packages you want to install, automatically. apt-get might work similar, but I haven't used it yet. Andreas
Re: magicfilter trouble (was: apsfilter on Epson Stylus Color... Who can help?)
[EMAIL PROTECTED] wrote: You also want to install everything that magicfilter 'suggests' [...] You can find the names of the programs by grepping through the filter file and checking the against the contents file. Or just run a pass through dselect. ;) Actually I did install magicfilter with dselect, but there seem to be some hidden dependencies, especially on recode, xfig and transfig. I found this out when I checked my filter-file as you suggested. For the moment I'm happy with magicfilter - Thank you very much. Andreas On Sun, 23 Jul 2000, Andreas Hetzmannseder wrote: [...] right now the printer does *nothing at all* - not even error messages. [...]
Re: magicfilter trouble (was: apsfilter on Epson Stylus Color... Who can help?)
I have the SC 660, amd the magicfilter package works well with it. I would like to see your printcap entry. Mine looks like this: lp|esc600|Epson Stylus Color 600:\ :lp=/dev/lp1:sd=/var/spool/lpd/esc600:\ :sh:pw#80:pl#72:px#1440:mx#0:\ :if=/etc/magicfilter/stylus_color_720dpi-filter:\ :af=/var/log/lp-acct:lf=/var/log/lp-errs: But right now the printer does *nothing at all* - not even error messages. I have checked /dev/lp1, so this can't be the problem The SC 600 is well supported. Magicfilter has 3 filters for it: [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] Did you create the filters by yourself? I only found two corresponding driver entries: stylus_color_360dpi and stylus_color_720dpi (with magicfilter version 1.2-28). Please tell me, how you made it work, I'm really curious... Best regards, Andreas
Re: apsfilter on Epson Stylus Color... Who can help?
Attempting to print a simple text file with lpr filename on my Epson Stylus Color 600 produces a single line: Unknown device: escp2 Then it prints the same sentence on new sheets of paper over and over again, so that I have to remove the paper to stop the printer. I also have that printer, and use apsfilter. (Version 5.1.4-1) Do you know from where can I download that version? I can only find version 4.9.7-5 on the ftp-server... The /dev/lp0 thing is presumably a difference in kernel or something (I seem to remember the numbering changed). However, the obvious difference is uniprint. Maybe it's worth you trying that driver, instead? uniprint is not in the drivers list - maybe it's listed in your newer version... Thanks for your advice anyway, Andreas
Re: magicfilter trouble (was: apsfilter on Epson Stylus Color... Who can help?)
I have the SC 660, amd the magicfilter package works well with it. I would like to see your printcap entry. Mine looks like this: lp|esc600|Epson Stylus Color 600:\ :lp=/dev/lp1:sd=/var/spool/lpd/esc600:\ :sh:pw#80:pl#72:px#1440:mx#0:\ :if=/etc/magicfilter/stylus_color_720dpi-filter:\ :af=/var/log/lp-acct:lf=/var/log/lp-errs: But right now the printer does *nothing at all* - not even error messages. I have checked /dev/lp1, so this can't be the problem Are you running a 2.2 or later kernel? If so, use /dev/lp0 instead of /dev/lp1. I have version 2.0.38 Andreas
Re: LILO and Printer
I'm used to RedHat and am having problems with LILO... LILO: No changes allow me to select which OS to enter on bootup. I can change it to Windows or Linux, but not to choose between them. I do this by editing lilo.conf and then running lilo. Concerning your LILO-Problem: If you changed the Master Boot Record, you can do the following: Escape Booting by pressing SHIFT in the right moment. You should see something like 1FA or 3FA for example. If you type the displayed number you choose the harddrive-partition with the default OS. Type F, if you wish to boot from Floppydisk. If you type A, you'll get a list of all hd-partitions plus the F-Option, something like 1234F. You can choose the OS by typing the number of the corresponding partition. So you can switch between the OSs just by typing A1, A2, A3 or whatever. I hope, this helps. Andreas.
apsfilter on Epson Stylus Color... Who can help?
I've grown a long grey beard (no, not really :) ...) about the
following problem: Attempting to print a simple text file with
lpr filename on my Epson Stylus Color 600 produces a single line:
Unknown device: escp2
Then it prints the same sentence on new sheets of paper over and over
again, so that I have to remove the paper to stop the printer.
The standard printcap entry by apsfilter looks like this:
lp|lp2|escp2-a4-auto-mono|escp2 auto mono:\
:lp=/dev/lp1:\
:sd=/var/spool/lpd/escp2-a4-auto-mono:\
:lf=/var/spool/lpd/escp2-a4-auto-mono/log:\
:af=/var/spool/lpd/escp2-a4-auto-mono/acct:\
:if=/usr/lib/apsfilter/filter/aps-escp2-a4-auto-mono:\
:mx#0:\
:sh:
The corresponding logfile tells me something like this:
[Total: 1 page on 1 sheet]
/usr/lib/apsfilter/filter/aps-escp2-a4-auto-mono: line 111:
246 Broken pipe
$DECOMPRESS a2ps ${A2PS_OPTS}
247 Done(1) | eval $PRINT_PS
I made several troubleshooting checks:
1. I created a simple printcap entry, which worked just fine. It looked
like this:
lp:\
:lp=/dev/lp1:\
:sd=/var/spool/lpd/lp:\
:mx#0:\
:sh:
2. a2ps and rewindstdin seem to be correctly installed, although not in
/usr/lib/apsfilter (the base directory) as proposed, but in
/usr/lib/apsfilter/bin. To be sure, i copied the two files into the
base directory, but the problem remained.
3. Ghostscript (gs) is installed (as well as netpbm and two
jpeg-libraries: libjpeg62 and libjpegg6a). gs resides in the /usr/bin
directory, which is in the search path of GLOBAL.sh as required.
4. I even forced some debugging output, but I haven't learnt shell
scripting yet, so I don't know what to make of it ... If someone is
interested in this, I will post it.
Have I forgotten something? Did I choose the wrong printer driver
(escp2)? Do I need to install additional packages (like X, which I don't
have at present)? Who has experience in using apsfilter on an Epson?
Please, answer...
Re: apsfilter on Epson Stylus Color... Who can help?
I have an Epson Stylus Color 740 printer and I could not get it to work with apsfilter. I used Magicfilter instead, and the printtool and both worked for me. Thanks for your recommendation. I would like to give apsfilter a last chance. If I can't find a way to make it work, I will gladly switch to magicfilter.
Re: Man command
Is there somewhere where I can download man (the manual command) from the Internet? You can download man-db, manpages and manpages-dev(as well as any other package) via ftp (e.g. ftp://ftp.xy.debian.org - for xy insert a country-code). You find these files in the ./debian/dists/stable/main directory. The following subdirectory depends on your computer (for me binary-i386). The next subdirectory should be doc which includes the man-files.
