Netconfig tool or do I have to edit file manually?
I configured my Debian box so it uses dhcp, but now I want to change it to a static IP. Is there a netconfig utility for Debian, or do I just have to edit /etc/network/interfaces manually? brian -- Brian Lavender http://www.brie.com/brian/
wu-ftpd exploit and patch I got, debs too
A friend of mine emailed me this glob.c patch for the recent wu-ftpd
exploit. I don't understand how the exploit works, but I am sure someone
will tell me if this patch gaurds against it. The only other patch out
there seems to be the Dead Rat src rpm, so I'll post what I found. I also
used the patch build Debian packages, so if you are running Debian, you
can use my Debian packages. Use them at your own risk though. I don't
have extended experience at building Debian packages, and basically I
took the source from the previous package, upped the rev on the changelog,
and did a
$ fakeroot debian/rules binary
and voila, I had new deb packages. You can get those at:
ftp://brie.com/pub/debian/potato
Below is the patch a friend emailed me.
brian
Generic patch against globc.c for:
Subject: Wu-Ftpd File Globbing Heap Corruption Vulnerability
-- SNIP --
--- glob.c.orig Sat Jul 1 14:17:39 2000
+++ glob.c Wed Nov 28 00:43:38 2001
@@ -298,7 +298,7 @@
for (lm = restbuf; *p != '{'; *lm++ = *p++)
continue;
-for (pe = ++p; *pe; pe++)
+for (pe = ++p; *pe; pe++) {
switch (*pe) {
case '{':
@@ -314,11 +314,19 @@
case '[':
for (pe++; *pe *pe != ']'; pe++)
continue;
+ if (!*pe) {
+ globerr = Missing ];
+ return (0);
+ }
continue;
}
+}
pend:
-brclev = 0;
-for (pl = pm = p; pm = pe; pm++)
+if (brclev || !*pe) {
+ globerr = Missing };
+ return (0);
+}
+for (pl = pm = p; pm = pe; pm++) {
switch (*pm (QUOTE | TRIM)) {
case '{':
@@ -352,19 +360,18 @@
return (1);
sort();
pl = pm + 1;
- if (brclev)
- return (0);
continue;
case '[':
for (pm++; *pm *pm != ']'; pm++)
continue;
- if (!*pm)
- pm--;
+ if (!*pm) {
+ globerr = Missing ];
+ return (0);
+ }
continue;
}
-if (brclev)
- goto doit;
+}
return (0);
}
@@ -416,11 +423,10 @@
else if (scc == (lc = cc))
ok++;
}
- if (cc == 0)
- if (ok)
- p--;
- else
- return 0;
+ if (cc == 0) {
+ globerr = Missing ];
+ return (0);
+ }
continue;
case '*':
--
Brian Lavender
http://www.brie.com/brian/
--
Brian Lavender
http://www.brie.com/brian/
fakeroot on debian/rules binary?
How do you do a fakeroot when you build a package from source? $debian/rules binary brian
Samba 2.2.1a on potato?
Ok, I am trying to build a Samba 2.2.1a deb package for potato, but when I did a $ debian/rules binary it complained it could not find dh_installlogrotate So, I am wondering, what do I need to do to create my Samba package? I already tried grabbing the latest debhelper, but that has a number of other dependencies. Do I just go in and modify the rules file? brian -- Brian Lavender http://www.brie.com/brian/
Re: creating debian rescue disk
I don't know if you noticed, but that disk is mountable. You can actually copy a new kernel onto it which has reiserfs support. Compile your new kernel, mount the disk, and copy the new kernel over the old one. It's in msdos file format, so you can even copy over in Windows. The disk is a syslinux disk. For details, check out the SYSLINUS web page. http://syslinux.zytor.com/ brian On Tue, Aug 21, 2001 at 10:03:01PM +0200, Timo BlazkoBoewing wrote: Hello, I have quite trouble with my laptop: it will only boot with the ide rescue disks (e.g. http://ftp.de.debian.org/debian/dists/ woody/main/disks-i386/3.0.9-2001-08-11/images-1.44/ide/), neither any other CD image neither floppy image works (not even idepci). The thing is that it locks the machine when probing for any SCSI devices, I presume. My problem is that I *strongly* intend to use reiserfs. But being derived from idepci, the reiser floppies wont boot. Cos the ide flavour does not support reiserfs, can anyone give me advice how to create a debian rescue disk for woody??? Of course I am not afraid of compiling my own kernel, but I dunno how to bind it with the bootstrap menu things. Any ideas??? Greetings thanx, Timo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- Brian Lavender http://www.brie.com/brian/
slrn newsreader, How do I specify my from address?
In slrn when I post to a newsgroup, it puts my from email address as [EMAIL PROTECTED] I would like it to put it as my real email address which is: [EMAIL PROTECTED] Is there a config file I need to modify? brian -- Brian Lavender http://www.brie.com/brian/
Sendmail work as smtp server for internal private subnet
I have sendmail on my gateway box which has a dynamic IP. How do I make it so it will allow the internal hosts which are on a 192.168.1.x subnet use it as their smtp server? brian -- Brian Lavender http://www.brie.com/brian/
Potato install termwrap problem
I just installed the base system of potato from the binary-i386 iso image disk 1. Once it goes through the install of the base system, I get the following error on boot. /bin/sh: /sbin/termwrap: No such file or directory /bin/sh: exec: /sbin/termwrap: cannot execute: No such file or directory INIT: Id 1 respawning too fast: disabled for 5 minutes What is wrong here? I looked in /sbin and there is no termwrap program there. brian -- Brian Lavender http://www.brie.com/brian/
dump package list to file?
I want to dump a list of the packages I have installed on a current potatoe installto a file, so I can later build a machine with the same packages? How do I do this? brian -- Brian Lavender http://www.brie.com/brian/
mimencode, which package?
In which package do I find mimencode brian -- Brian Lavender http://www.brie.com/brian/
Debian talk, Sacramento, CA, May 10
I don't know if there is an appropriate Debian mailing list for Debian events, but I figure this one ought to be good. I also posted this to debian-devel. I figure that should cover it. I am proud to announce that we will have Joey Hess and Sean 'Shaleh' Perry will speak at the upcoming sacLUG meeting, Wednesday, May 10, 2000 7-9pm sacLUG Sacramento, CA. The public is welcome to attend. Joey will speak about debconf, and Sean will speak about a topic which we have yet to decide upon. I am sure that this will be an exciting talk. For meeting details and directions see http://www.saclug.org If you happen to fly to Sacramento to attend the meeting, the Sacramento airport is extremely easy to get into and out of. It is about twenty five minutes from the meeting location. brian -- Brian Lavender http://www.brie.com/brian/
Re: removing hard drives
Remove or modify the mount point entries from your /etc/fstab In addition, compare the devices your /etc/fstab has listed for your mount points compared to the devices that are coming up in bootup. You can always do a # dmesg | less to see how your devices (drives) are coming up. For example, /home /dev/hdc1 may now be /home /dev/hdb1 you may very well have to use a rescue disk to do this. On Fri, Apr 21, 2000 at 08:21:19AM -0400, Chris Mason wrote: I had three hard drives, one boot drive and one partitioned into two. I physically removed the second hard drive ( who needs 20GB on a linux system). Now I get errors on boot(fsck). How do I removed the dives from the boot sequence? I, for one, could use an extra 20 Gigs. -- Brian Lavender http://www.brie.com/brian/
Re: OT: Video cards? Creating Athlon Debian box..
It's not Debian you are matching your card up to, it's xfree86. In fact, you will want to upgrade xfree86 to 3.3.6, so it will support your card. I just built a couple machines with Matrox G200 16 MB cards and they work well. I am not a gamer, so I couldn't give you much info as far as the 3D cards. check http://www.xfree86.org/3.3.6/ for supported cards. Also, you can get the latest xfree86 deb packages from the developer's site: http://www.debian.org/~branden/ brian On Sun, Apr 09, 2000 at 10:38:28PM -0700, [EMAIL PROTECTED] wrote: I'm putting together my first machine (upgrading from a 486, so it should be interesting :) and have trouble deciding on a video card. I thought 3dfx was supposed to be a good choice (looking at the 3000), as was Matrox (the G400). But older posts on the archives seem to indicate trouble with both (or at least _references_ to trouble). NVidia (sp?) was mentioned as a possibility, but I didn't know it was supported well. I hope people don't mind the poll. I honestly am at a bit of a loss with all the new stuff I'm discovering out there, and want to make a good match with Debian. Thanks for any input! Kenward Vaughan ps. My work will involve molecular modeling imagery, with VRML quite likely a part of that. Certainly 2D, but I don't know about 3D. Any games are likely under Winblows for my kids (until I find out about Descent under Linux.. ;-) -- Brian Lavender http://www.brie.com/brian/
Re: raid patches for 2.2.14
For the latest linux raid patches and howto visit: http://www.redhat.com/~mingo/ I just built a couple of raid 1 2.2.14 debian boxes. This is exceptional code and is by far superior to the old way of doing raid. The raid tools dangerous tar ball also contains the howto and information about the linux-raid mailing list. The dangerous portion of the name means, make sure you are not operating on mission critical data which you don't have backed up or can afford to lose when configuring raid. linux-raid@vger.rutgers.edu and if you are interested in the archives check: http://mail-archive.com/ brian On Tue, Apr 11, 2000 at 03:22:54PM +0200, Robert Varga wrote: Are there raid0.90 patches for 2.2.14 kernel? The last patch that I can find on the kernel mirror is 19990824 and it is against 2.2.11, however it cannot be applied to the kernel found in kernel-source-2.2.14 debian package. The last raid kernel-patch debian package is 19990724 and is against 2.2.10 kernel. However the raidtools2 package is version 19990824. What to do now? Robert Varga -- Brian Lavender http://www.brie.com/brian/
CVS question
I have been using cvs to track development of the my latest tk/perl application. I have been coding along starting with version 1.1 cruising along and making commits along the way. I got up to version 1.7 when I realized that I wanted to back up to version 1.6 make some changes, and continue from there. Here is how I backed up. I was on version 1.7, so I deleted the file I had. I then did a cvs checkout -r 1.6 which warned me that the current file had been deleted, but gave me 1.6. I made my changes, but now and can't seem to reintegrate them. Here is a what happens when I try to do a cvs commit [EMAIL PROTECTED]:~/src/scripts cvs commit tkcorrespond.pl cvs commit: sticky tag `1.6' for file `tkcorrespond.pl' is not a branch cvs [commit aborted]: correct above errors first! How do I commit this? I suppose I would want it to be version 1.8 brian -- Brian Lavender http://www.brie.com/brian/
Re: contents-i386.gz
I believe it is the contents of the i386 set. I found it on one of Debian's mirrors. ftp://ftp.freesoftware.com/pub/linux/debian/dists/slink/ freesoftware:/pub/linux/debian/dists/slink ls ChangeLog Contents-m68k.gzcontrib/ Contents-alpha.gz Contents-sparc.gz main/ Contents-i386.gzDebian2.1r5@non-free/ brian On Mon, Apr 03, 2000 at 06:06:57AM -0700, pplaw wrote: debs, where's the contents-i386.gz file? (i.e., where does one get it?) ia, t. -- Bentley Taylor __ [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: what is the num value of s in -rwsr-x--- (chmod)?
That means it is suid. It runs under the effective user id of the owner of the file. To make it this way you do a chmod 4755 foo.pl I am sure there are more details to this, so consult the man pages for more info. (Maybe I should say consult the info pages :) brian On Mon, Apr 03, 2000 at 06:50:43AM +, john smith wrote: what is the s letter stand for? r stands for read access,w write, and x access files and subdirectories, I can even c a d i.e. drwxr-xr-x. what does those stand for and what are their numerical access mode values? to further add to my question, how do I give a user access so that he/she will be able to use the pppd/pon command? __ Get Your Private, Free Email at http://www.hotmail.com -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
which mailing list to follow potatoe
I want to follow the development and evaluation of potato. Which list should I subscribe to. brian -- Brian Lavender http://www.brie.com/brian/
cdimage problem with rsync
I am trying to create my own CD Image. http://cdimage.debian.org/ I am trying to run the following rsync command to update my pseudo image but I get the following error. $ rsync --verbose --progress --stats --block-size=8192 \ --temp-dir=/home/brian/tmp \ rsync.kernel.org::mirrors/debian-cd/2.1/i386/binary-i386-1.iso . link_stat /debian/2.1/i386/binary-i386-1.iso : No such file or directory client: nothing to do -- Brian Lavender http://www.brie.com/brian/
Re: (eh?) Perl library shifted...?
use lib /path/to/your/lib; This will append to perl's include path. Otherwise, I think it's hard coded. brian On Wed, Mar 29, 2000 at 12:06:37PM -0500, Jeff Gordon wrote: Not sure, but it looks like a recent Potato update must've rewritten a variable holding the location of Perl libraries -- at least, I've got a little program running here, that can no longer locate a needed module. Anybody know where to reach in and change the values Perl sees as '@INC' ...? -- -- Jeff -- http://www.wellnow.com There's nothing left in the world to prove. All that's worth doing is to love one another, using whatever means are available to serve. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: Debian Perl Modules
I still have a question regarding the debian perl modules. It appears that dpkg and apt have perl modules. If I went in and replace perl with a newer version from source, I would have to rebuild those modules. If I look at my libraries of my Potatoe installation I see some Debian perl modules. You see there is a DebianNet.pm and a Debian directory. Did these perl modules come out of the dpkg and apt debian packages? I am thinking about replacing the perl with my source version. If I want to do that and have it replace the debian version, I am going to have to make sure that those modules exist. Where would I get these debian perl modules? I don't see them on CPAN or in the debian source tree. Of course maybe I did not look thoroughly. I think I do see a point where I could install a new version of perl in /usr/local and have that version co-exist with the perl supporting my Debian system. $ ls -F /usr/lib/perl5 5.004/ 5.005/ 5.00503/ Bundle/ Data/ Date/ Debian/ DebianNet.pm Dpkg/ File/ HTML/ HTTP/ I18N/ LWP/ LWP.pm Mail/ Net/ SGMLS/ SGMLS.pm Text/ Time/ URI/ URI.pm WWW/ auto/ dialog.pl i386-linux/ lwpcook.pod sgmlspl-specs/ site_perl/ $ ls -F /usr/lib/perl5/Debian DebConf/ DpkgFtp.pm $ ls -F /usr/lib/perl5/Dpkg Archive/ Package/ On Fri, Mar 24, 2000 at 01:40:36PM +1030, John Pearson wrote: On Thu, Mar 23, 2000 at 10:10:00AM -0800, Brian Lavender wrote I was taking a look at the perl modules on my debian system, and I noticed that there are a couple of Debian perl modules. In fact I once built a newer version of perl on my debian system, and I noticed that apt, and I believe dpkg stopped working. My Question: Is it possible to build a new version of perl on my debian system and install these perl modules? Say I have slink I want to upgrade perl to the latest. Can I add in these perl modules, and where would I get the source to add them? Slink perl uses a different layout for /usr/lib/perl5 to that used for later perls; it changed for Potato, I'm guessing to allow multiple perl versions to co-exist; slink perl has directories like /usr/lib/perl5/i386-linux/5.004/auto/ /usr/lib/perl5/i386-linux/5.004/CORE/ and so on, whereas perl5 from potato uses /usr/lib/perl5/5.005/i386-linux/auto/ I seem to recall that both layouts differ from the default specified in the perl distribution. As a result, slink packages that insert files in /usr/lib/perl5 are not compatible with later perl packages. Your options appear to be: - Stick to slink perl related packages; - Build the version of perl that you want, but ensure that it uses the 'slink' directory conventions; you may also have to rebuild perl-dependent packages if the changes to perl break them in other ways; - Upgrade to potato (or at least, upgrade perl related packages) and be happy with the perl versions that provides; - Upgrade at least perl related packages to potato and then build the perl of your choosing using the potato conventions or installing into /usr/local/, without replacing the potato perl that other packages rely on. If you haven't actually replaced the slink version of perl (i.e., you simply installed the new perl alongside of the old one, either in /usr/local/ or replacing /usr/bin/perl) then you may be able to have the two coexist peacefully by renaming your new perl binary to (e.g.) 'perl-local' and (if necessary) replacing /usr/bin/perl by hand from the slink package; it depends if your new perl clobbered files from the slink perl. HTH, John P. -- [EMAIL PROTECTED] [EMAIL PROTECTED] Oh - I - you know - my job is to fear everything. - Bill Gates in Denmark -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: Debian Perl Modules
I just downloaded perl 5.6.0 which was released yesterday, and I did a $ ./configure.gnu --prefix=/usr/local and it looks if all is well. I ran dselect, and it looks as if it still works. I also checked to make sure that my original perl 5.005_03 is still in /usr/bin and it looks as if it is there too. So, all looks good, and it appears that I can successfully maintain two different versions of perl, on that supports my debian, and one that I can use to implement bleeding edge features. brian On Fri, Mar 24, 2000 at 08:51:53AM -0800, Brian Lavender wrote: I still have a question regarding the debian perl modules. It appears that dpkg and apt have perl modules. If I went in and replace perl with a newer version from source, I would have to rebuild those modules. If I look at my libraries of my Potatoe installation I see some Debian perl modules. You see there is a DebianNet.pm and a Debian directory. Did these perl modules come out of the dpkg and apt debian packages? I am thinking about replacing the perl with my source version. If I want to do that and have it replace the debian version, I am going to have to make sure that those modules exist. Where would I get these debian perl modules? I don't see them on CPAN or in the debian source tree. Of course maybe I did not look thoroughly. I think I do see a point where I could install a new version of perl in /usr/local and have that version co-exist with the perl supporting my Debian system. $ ls -F /usr/lib/perl5 5.004/ 5.005/ 5.00503/ Bundle/ Data/ Date/ Debian/ DebianNet.pm Dpkg/ File/ HTML/ HTTP/ I18N/ LWP/ LWP.pm Mail/ Net/ SGMLS/ SGMLS.pm Text/ Time/ URI/ URI.pm WWW/ auto/ dialog.pl i386-linux/ lwpcook.pod sgmlspl-specs/ site_perl/ $ ls -F /usr/lib/perl5/Debian DebConf/ DpkgFtp.pm $ ls -F /usr/lib/perl5/Dpkg Archive/ Package/ On Fri, Mar 24, 2000 at 01:40:36PM +1030, John Pearson wrote: On Thu, Mar 23, 2000 at 10:10:00AM -0800, Brian Lavender wrote I was taking a look at the perl modules on my debian system, and I noticed that there are a couple of Debian perl modules. In fact I once built a newer version of perl on my debian system, and I noticed that apt, and I believe dpkg stopped working. My Question: Is it possible to build a new version of perl on my debian system and install these perl modules? Say I have slink I want to upgrade perl to the latest. Can I add in these perl modules, and where would I get the source to add them? Slink perl uses a different layout for /usr/lib/perl5 to that used for later perls; it changed for Potato, I'm guessing to allow multiple perl versions to co-exist; slink perl has directories like /usr/lib/perl5/i386-linux/5.004/auto/ /usr/lib/perl5/i386-linux/5.004/CORE/ and so on, whereas perl5 from potato uses /usr/lib/perl5/5.005/i386-linux/auto/ I seem to recall that both layouts differ from the default specified in the perl distribution. As a result, slink packages that insert files in /usr/lib/perl5 are not compatible with later perl packages. Your options appear to be: - Stick to slink perl related packages; - Build the version of perl that you want, but ensure that it uses the 'slink' directory conventions; you may also have to rebuild perl-dependent packages if the changes to perl break them in other ways; - Upgrade to potato (or at least, upgrade perl related packages) and be happy with the perl versions that provides; - Upgrade at least perl related packages to potato and then build the perl of your choosing using the potato conventions or installing into /usr/local/, without replacing the potato perl that other packages rely on. If you haven't actually replaced the slink version of perl (i.e., you simply installed the new perl alongside of the old one, either in /usr/local/ or replacing /usr/bin/perl) then you may be able to have the two coexist peacefully by renaming your new perl binary to (e.g.) 'perl-local' and (if necessary) replacing /usr/bin/perl by hand from the slink package; it depends if your new perl clobbered files from the slink perl. HTH, John P. -- [EMAIL PROTECTED] [EMAIL PROTECTED] Oh - I - you know - my job is to fear everything. - Bill Gates in Denmark -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/ -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: Moving /home to new Parition
First mount your new home partition under a different mount point. # mkdir /mnt/home # mount /dev/hda12 /mnt/home # cd /home # find . -xdev | cpio -pm /mnt/home # mv /home /oldhome Edit your fstab so that it mounts hda12 as home ie /dev/hda12 /home ext2 defaults 0 2 # init 6 This will reboot, and if all works, your new home should be under /dev/hda12 Now you can safely remove /oldhome # rm -rf /oldhome brian On Fri, Mar 24, 2000 at 11:51:25AM -0800, Dzuy M. Nguyen wrote: I mounted a new partition /hda12 and use the /home as the mountpoint. When I go: mount /dev/hda12 /home I end up with an empty directory. How do I move the current contents of my /home directory into the new mountpoint /home on /dev/hda12? thanks. -- Brian Lavender http://www.brie.com/brian/
Debian Perl Modules
I was taking a look at the perl modules on my debian system, and I noticed that there are a couple of Debian perl modules. In fact I once built a newer version of perl on my debian system, and I noticed that apt, and I believe dpkg stopped working. My Question: Is it possible to build a new version of perl on my debian system and install these perl modules? Say I have slink I want to upgrade perl to the latest. Can I add in these perl modules, and where would I get the source to add them? brian -- Brian Lavender http://www.brie.com/brian/
Re: Remote Printing
I believe the answer lies in the /etc/lpd.perms That is the print server that will recieve the jobs. It is commented well, but the time I tried to make it so I could print from one Linux box to the other, I could not print from the print server to itsself. Then of course you have to configure the remote machine so it will print to the print server. This is a printcap entry I have for printing to a remote machine. lp|hplj5si|HP LaserJet 5si:\ :lp=/dev/null:sd=/var/spool/lpd/hplj5si:rm=imhplj5si:rp=hplj5si:\ :sh:pw#80:pl#66:px#1440:mx#0:\ :if=/etc/magicfilter/ljet4m-filter :af=/var/log/lp-acct:lf=/var/log/lp-errs: If you get the permissions set on the machine with the physically connected printer so that a remote printer can print to it, along with a user on that machine being able to print to it, let me know. I have yet to get this configured. Of course check /usr/doc/lprng for info too. brian On Mon, Mar 06, 2000 at 10:54:56PM -, [EMAIL PROTECTED] wrote: Hi, I have the lpd running on both the machines. In fact, after the new entries in my printcap, I have restarted lpd. The results were posted only after these things failed. The HOWTO was not much of a help. sridhar --- Sridhar M.A.Powered by: University of Mysore, Manasagangotri Debian GNU/Linux 2.1 Mysore 570 006, INDIA -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: dc390 what does it mean?
Do you have a dc390 Tekram SCSI controller installed? What sort of controller are you using? brian On Tue, Mar 07, 2000 at 06:08:50PM +1100, Pikuan wrote: i just bought a debian 2.1 r 4 CD i was installing from cd, pressed enter at the main menu. it does all the initialisation when it reach this line and stop: DC390: 0 adapters found anyone know what is wrong? thank you, pikuan === Pikuan Susanto We love u. There's nothing to worry Just Rest Thank you for the unforgettable memories -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: Time taken by a script
I have yet to use this module, but I was just reading about it in the Camel book. $ perldoc Benchmark brian On Mon, Apr 17, 2000 at 04:44:17PM +0200, Ron Rademaker wrote: I've written a cgi script in perl an I would like to know how long (exactly, 10th of second) it takes to execute the script, how can I do this? Ron -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: MS Frontpage
You can install libc5 runtime libraries. That is what I did with the Excite for Web Servers search engine, that is still available, but no longer maintained. brian On Sun, Mar 19, 2000 at 03:59:37PM +0100, Werner Reisberger wrote: Does anybody know how to install Frontpage Server Extension on a Debian system with libc6? I was able to compile Apache with the frontpage module but the installation of the extension failed becaused the binary therein are compiled against libc5 and there aren't any sources @= I hate this stuff but some customers wants it. -- Werner -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: Promise Ultra 66
On Sat, Mar 18, 2000 at 02:34:16PM +0100, J.H.M. Dassen Ray wrote: On Fri, Mar 17, 2000 at 21:32:35 -0800, Brian Lavender wrote: but it just flashes the characters 2FA: That's from the Master Boot Record program. I guess you are correct about this, because I took the drive and reformated it with dos, and it was able to boot. Does this mean I am going to have to create a dos partition on the disk, and put loadlin in it with the kernel, and have an autoexec which will in turn boot linux? Now, I did find the IDE patches, and I was able to create a boot disk that will boot my system and find the drive on the controller. It would just be nice if that MBR with lilo could start up my disk. I will have to take a closer look at the UDMA howto. In the meantime, I am having trouble with fmt seeing the full size of the disk. I have the Large Drive HOWTO, but that does not seem to tell me exactly what I need to type in, so fmt will use the proper parameters and see the full 20 Gigs. Side note, my scsi system sure does seem a whole lot easier. Perhaps twice the price for the same drive space is not so bad after all. brian I also tried booting with a floppy with IDE support, but it won't detect it either. The bootfloppies for Slink don't handle UDMA66. You need to use bootfloppies with a kernel on them that's patched for UDMA66 support (using the IDE patches from ftp.kernel.org:/pub/linux/kernel/people/hedrick/). I'm not sure if potato's bootfloppies (will) support UDMA66. HTH, Ray -- UNFAIR Term applied to advantages enjoyed by other people which we tried to cheat them out of and didn't manage. See also DISHONESTY, SNEAKY, UNDERHAND and JUST LUCKY I GUESS. - The Hipcrime Vocab by Chad C. Mulligan -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
pass boot disk kernel parameters?
How do you create a boot disk where you can pass it kernel parameters such as disk=linear append = hd=683,16,38 hd=64,32,202 I just grabbed the append specification from LILO stuff in the Running Linux book [129]. The book says how to add it in lilo, but gives no indication for putting the information onto a kernel boot disk. Is there a way that you can specify these parameters with rdev? I mean can you do # rdev ./vmlinuz /dev/hda6 hda=16383,16,63 brain -- Brian Lavender http://www.brie.com/brian/
Re: XEmacs use 'Alt' as 'Meta'
I just use xkeycaps to redefine the keys. Check dselect for it. Am I following this thread correctly? brian On Sat, Mar 18, 2000 at 04:55:23PM -0800, [EMAIL PROTECTED] wrote: I've been rather fighting the inconsistancies myself. I might get to fixing up console someday. Oh yeah, there's something called ICCCM which xemacs follows under X which among other things describes a method (IMO the Right Way) to consistantly read and interpret keypresses. Basically (going from memory) Apps like Enlightenment and gnome seem to assume that any key that generates mod1 modifier is ALT ICCCM apps like xemacs assume that the shift keys that generate ALT_L and ALT_R with the same modifier bit, said bit not shared with any other shift keys, is ALT. Also assumes that the shift keys that generate META_L and META_R with a dictinct modifier bit (doesn't matter exactly WHICH modifier bit) is META. I mean, you can map SHIFT_L and SHIFT_R to use mod1 and Eterm will suddenly decide that pressing SHIFT+R will be ALT-R. People in general don't seem to care, at least not the last time I brought up the subject. See the list archives and the bug system. Heh, maybe I'll make a proposal in my spare time, if I could get at least one official developer to back me. This is on my list of peeves. -- Beiad Ian Q. Dalton On 18 Mar 2000, Brian May wrote: Sean == Sean 'Shaleh' Perry [EMAIL PROTECTED] writes: Sean On 16-Mar-2000 Joseph A. Martin wrote: Hello, I feel like this should be a simple thing to do, but I haven't been able to figure it out so far. I want XEmacs to my Left Alt key as the Meta key. It works like that when I am running XEmacs from the console. However, if I run XEmacs from XWindows ESC is my meta key and Left-Alt is not recognized by XEmacs. What do I need to change to have XEmacs in XWindows use Left-Alt as the Meta key? Sean if you have one, the windows key is bound as meta. Is it possible to use the windows key as meta in XEmacs for text mode sessions? I don't like this inconsistency... Some combinations work, others don't. Tested: text mode console: alt -- meta gnome-terminal:alt -- meta xterm: windows -- meta -- Brian May [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: swat for samba???
I just searched for swat in dselect and found it. I found it in usual debian sources. Are you using apt sources? I have Samba 2.0.5 listed in dselect. brian On Fri, Mar 17, 2000 at 05:19:45PM +0100, FreeMan wrote: Hi, I would like to use swat on my Linux-box for easier configuration of samba. I thought it would be a seperate package, but couldn't find it in dselect. Do I just have to enable swat by a parameter in a configuration file or is it really a seperate package and I have to get it from somewhere? thanks for help!ks for help! -- Brian Lavender http://www.brie.com/brian/
Re: Promise Ultra 66
Once again, how do you specify the CHS specs for cfdisk or fdisk? brian On Sun, Mar 19, 2000 at 01:37:17PM -0800, Percival wrote: It really sucks, basically. I found out that I had to track down the manufacturer of my large hard drive, get the specs from them, and then pass them on to cfdisk/fdisk and ignore the warnings that those programs gave me. I also remember that one of them sinply would not work (fdisk, I think), so I had to use cfdisk and specify the CHS specs. But then it worked great! I was installing SuSE at the time, which has cfdisk in the install ramdisk. Beware of strangeness - some drives have multiple CHS specs, as they respond to different 'virtual' setting. -Percival On Sun, Mar 19, 2000 at 12:51:59PM -0800, Brian Lavender wrote: How do you specify the CHS specs? THis has been a big delima. The Large Hard Drive HOW-TO talks about it, and boom, doesn't say anything as far as how-to do it. brian On Sun, Mar 19, 2000 at 12:18:15PM -0800, Percival wrote: I have found that with large drives (8G+) you have to manually specify the CHS specs to fdisk (or cfdisk) when creating partitions, or you end up with non-reliable partitions (I suspect that this might be true with fdisk under DOS as well). I didn't think that ATA/66 specs added anything other than more data lines for faster data transfer. Are there new specs for controll as well? -Percival On Sat, Mar 18, 2000 at 09:46:10PM -0800, Brian Lavender wrote: On Sat, Mar 18, 2000 at 09:41:43PM -0800, Percival wrote: Do you have more than one drive in the computer? Yes, in fact I am doing raid1. I built one half of the raid, transferred the files to the raid from regular partitions. changed the /etc/fstab to boot using the /dev/mdx devices, rebooted, did a hotraidadd on the original devices, and there I was. I have been able to totally boot from a floppy. I have the kernel rdev to /dev/hdc6 which is one half of the raid. Now I have a little more to this story. I booted with a dos disk. I formatted the first disk with fat32, and the controller will totally boot to it. Now I am trying to go back and reformat that disk, and hotraidadd that half to it again. I also have the patched lilo, so it can boot a raid device. My current problem is I can't get fmt to recognize the whole disk. When I start format, it sees the 20 Gigs, but it says the existing partition is exceeding the physical limits of the disk. In thought maybe I could just start adding partitions, but it seems that once I reach 2 Gigs, that is the limit. So, partitioning is first. Then back to making the Ultra 66 boot. Like I believe I said, I could create a dos partition on the disk. That would be a yucky kludge though. brian The reason I ask is that the error you are getting is far before linux gets involved. It seems to me that you have two partitions set active (DOS terms) that is, two partitions set bootable (normal term). Also, from what I understand, you do not need any special support or anything to use UDMA/66 drives - they are backward compatible. You should be able to boot just fine from stock boot disks. Get the system running first, then work on building a special kernel with full support for UDMA/66 to see the speed benefit. -Percival On Sat, Mar 18, 2000 at 09:32:29PM -0800, Brian Lavender wrote: On Sat, Mar 18, 2000 at 02:34:16PM +0100, J.H.M. Dassen Ray wrote: On Fri, Mar 17, 2000 at 21:32:35 -0800, Brian Lavender wrote: but it just flashes the characters 2FA: That's from the Master Boot Record program. I guess you are correct about this, because I took the drive and reformated it with dos, and it was able to boot. Does this mean I am going to have to create a dos partition on the disk, and put loadlin in it with the kernel, and have an autoexec which will in turn boot linux? Now, I did find the IDE patches, and I was able to create a boot disk that will boot my system and find the drive on the controller. It would just be nice if that MBR with lilo could start up my disk. I will have to take a closer look at the UDMA howto. In the meantime, I am having trouble with fmt seeing the full size of the disk. I have the Large Drive HOWTO, but that does not seem to tell me exactly what I need to type in, so fmt will use the proper parameters and see the full 20 Gigs. Side note, my scsi system sure does seem a whole lot easier. Perhaps twice the price for the same drive space is not so bad after all
Large Hard Drive Problems, HELP!
I built a system in which I put two 20 Gig Maxtor Hard Drives. When I do # cfdisk /dev/hda I can see the full 20 Gigs on the first drive, but when I do # cfdisk /dev/hdc I only see 8 Gigs. I know there is/was a problem with large hard drives. I am using the Asus P5A motherboard, with its on board IDE controllers. I have each drive as master on the two controllers. Is the second controller, not as good as the first? Can I issue a parameter to the kernel, so it will see the full 20 Gigs? Should I get a Promise Ultra ATA 66 controller, or some other IDE controller? brian -- Brian Lavender http://www.brie.com/brian/
Promise Ultra 66
Is anyone using the Promise Ultra66 controller with Linux? I put my drive on it, and I get nothing when I boot the machine. The card shows on the screen with the devices, but it just flashes the characters 2FA: on the screen. If I press return, it just repeats the characters. I disabled my IDE controller on the MB, so these are the only disks. I also tried booting with a floppy with IDE support, but it won't detect it either. brian -- Brian Lavender http://www.brie.com/brian/
Dial In problems, HELP!
I am trying to create a dial in connection where I can dial into my computer using dialup networking on win 95 and create a ppp connection. I installed mgetty and ppp. I am able to dial into my machine and it will automatically answer, but it will not authenticate the user with PAP. I check the syslog and messages and they both indicate a login failure. Below the log output are my configurations. What is wrong with my configuration? brian /var/log/syslog Mar 10 17:04:21 mammoth pppd[1654]: pppd 2.3.5 started by a_ppp, uid 0 Mar 10 17:04:21 mammoth pppd[1654]: Using interface ppp0 Mar 10 17:04:21 mammoth pppd[1654]: Connect: ppp0 -- /dev/ttyS1 Mar 10 17:04:21 mammoth pppd[1654]: sent [LCP ConfReq id=0x1 asyncmap 0x0 auth pap ma gic 0x53eaa09c pcomp accomp] Mar 10 17:04:24 mammoth pppd[1654]: rcvd [LCP ConfReq id=0x3 asyncmap 0xa magic 0x3 60f90 pcomp accomp callback 0x600] Mar 10 17:04:24 mammoth pppd[1654]: sent [LCP ConfRej id=0x3 callback 0x600] Mar 10 17:04:24 mammoth pppd[1654]: sent [LCP ConfReq id=0x1 asyncmap 0x0 auth pap ma gic 0x53eaa09c pcomp accomp] Mar 10 17:04:24 mammoth pppd[1654]: rcvd [LCP ConfReq id=0x4 asyncmap 0xa magic 0x3 60f90 pcomp accomp] Mar 10 17:04:24 mammoth pppd[1654]: sent [LCP ConfAck id=0x4 asyncmap 0xa magic 0x3 60f90 pcomp accomp] Mar 10 17:04:24 mammoth pppd[1654]: rcvd [LCP ConfAck id=0x1 asyncmap 0x0 auth pap ma gic 0x53eaa09c pcomp accomp] Mar 10 17:04:24 mammoth pppd[1654]: sent [LCP EchoReq id=0x0 magic=0x53eaa09c] Mar 10 17:04:24 mammoth pppd[1654]: rcvd [PAP AuthReq id=0x1 user=roger password=testin g] Mar 10 17:04:24 mammoth pppd[1654]: PAP authentication failure for roger Mar 10 17:04:24 mammoth pppd[1654]: sent [PAP AuthNak id=0x1 Login incorrect] Mar 10 17:04:24 mammoth pppd[1654]: sent [LCP TermReq id=0x2 Authentication failed] Mar 10 17:04:24 mammoth pppd[1654]: rcvd [LCP EchoRep id=0x0 magic=0x360f90] Mar 10 17:04:24 mammoth pppd[1654]: rcvd [LCP TermAck id=0x2] Mar 10 17:04:24 mammoth pppd[1654]: Connection terminated. Mar 10 17:05:09 mammoth chat[1655]: Failed Mar 10 17:05:09 mammoth pppd[1654]: disconnect script failed Mar 10 17:05:10 mammoth pppd[1654]: Exit. /var/log/messages Mar 10 17:04:21 mammoth pppd[1654]: pppd 2.3.5 started by a_ppp, uid 0 Mar 10 17:04:21 mammoth pppd[1654]: Using interface ppp0 Mar 10 17:04:21 mammoth pppd[1654]: Connect: ppp0 -- /dev/ttyS1 Mar 10 17:04:24 mammoth pppd[1654]: PAP authentication failure for roger Mar 10 17:04:24 mammoth pppd[1654]: Connection terminated. Mar 10 17:05:09 mammoth chat[1655]: Failed Mar 10 17:05:09 mammoth pppd[1654]: disconnect script failed Mar 10 17:05:10 mammoth pppd[1654]: Exit. $ cat /etc/mgetty/ debug 4 fax-id speed 38400 $ cat /etc/mgetty/login.config /AutoPPP/ - a_ppp /usr/sbin/pppd auth -chap +pap login debug * - - /bin/login @ $ cat /etc/ppp/options.ttyS1 mammoth:mammoth-s1 $ cat /etc/ppp/options ms-dns 192.168.1.100 disconnect chat -- \d+++\d\c OK ath0 OK asyncmap 0 auth crtscts lock modem netmask 255.255.255.0 +pap proxyarp lcp-echo-interval 30 lcp-echo-failure 4 pap-restart 3 pap-max-authreq 10 noipx -- Brian Lavender http://www.brie.com/brian/
Re: Dial In problems, HELP!
That totally worked with the pap-secrets. Now my client windows machine will authenticate when I connect to the ppp server. But now I have a new problem. I can only send packets between the ppp client and the ppp server. Do I need to configure some route commands, so the packets can go elsewhere? So far all I have is $ cat /etc/ppp/options.ttyS1 mammoth:mammoth-s1 where mammoth is the host name, and mammoth-s1 is a name I have resolved to another IP address for the serial port. Do I need to add addition routing information, so the modem user can connect elsewhere on the network? brian On Sat, Mar 11, 2000 at 02:26:04PM +1030, John Pearson wrote: On Fri, Mar 10, 2000 at 07:15:49PM -0800, Brian Lavender wrote I am trying to create a dial in connection where I can dial into my computer using dialup networking on win 95 and create a ppp connection. I installed mgetty and ppp. I am able to dial into my machine and it will automatically answer, but it will not authenticate the user with PAP. I check the syslog and messages and they both indicate a login failure. Below the log output are my configurations. What is wrong with my configuration? brian [snip] Most likely (and assuming that the password you're using is correct), the problem is in your /etc/pap-secrets file. Depending on the actual version of pppd you're using, it should contain an entry like * mammoth * or perhaps * mammoth If you want to make the entry specific to your client, replace the first * with the client's account name. John P. -- [EMAIL PROTECTED] [EMAIL PROTECTED] Oh - I - you know - my job is to fear everything. - Bill Gates in Denmark -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: HOW DO I Apply debian patch to original source?
Ok, I got the debhelper package so that when I do $ apt-get source wu-ftpd $ cd wu-ftpd-2.60 $ debian/rules binary It will build. Now here is the thing. When I look at the rules file It appears that it has a logrotate configuration on the install. So, I am thinking that I can comment out that logrotate, and use the way that slink rotates the logs to have them rotated. Would this be logical? I commented out the following line binary-arch: install-stamp dh_testdir dh_testroot dh_installdocs ERRATA README doc/HOWTO/VIRTUAL.FTP.SUPPORT doc/TODO \ doc/HOWTO/upload.configuration.HOWTO debian/FAQ install -m 644 debian/README.config $(tmp)/etc/wu-ftpd/README ln -s /etc/wu-ftpd/README $(tmp)/usr/share/doc/wu-ftpd/README.config install -m 644 debian/*.8 $(tmp)/usr/share/man/man8/ cd $(tmp)/usr/share/man/man8 ln -s addftpuser.8.gz rmftpuser.8.gz \ ln -s xferstats.8.gz xferstats.wu-ftpd.8.gz dh_installexamples `find doc/examples -type f` debian/*.es dh_installchangelogs CHANGES #install -m 644 debian/logrotate $(tmp)/etc/logrotate.d/wu-ftpd dh_installinit -r -udefaults 50 dh_installpam dh_strip dh_compress dh_fixperms dh_installdeb dh_shlibdeps dh_gencontrol dh_md5sums dh_builddeb brian On Tue, Mar 07, 2000 at 02:22:20PM -0700, Bob Nielsen wrote: On Tue, Mar 07, 2000 at 11:07:04AM -0800, Brian Lavender wrote: I am having problems installing the wu-ftpd package from source. The whole reason I am grabbing the source is that I want to use a more recent version of wu-ftpd than what is available with the slink binary packages. If I can build the package from the source files to give me a slink compatible package, it is all the better. I update my /etc/apt/sources.list with http://http.us.debian.org/debian unstable main contrib non-free and then I did a # apt-get update $ apt-get source wu-ftpd and it retrieved the 2.6.0 wu-ftpd. It looks as if it applied the patch for me. I got an erro with the debian/rules binary though $ cd wu-ftpd-2.6.0 $ debian/rules binary dh_testdir make: dh_testdir: Command not found make: *** [build-stamp] Error 127 Do you know what this error indicates? It means that the command dh_testdir is not available. $ dpkg -S dh_testdir debhelper: /usr/share/man/man1/dh_testdir.1.gz debhelper: /usr/bin/dh_testdir You need to install the debhelper package. -- Bob Nielsen, N7XY (RN2)[EMAIL PROTECTED] Tucson, AZ DM42nh QRP-L #1985 SOC #77http://www.primenet.com/~nielsen -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Apply debian patch to original source?
I am trying to compile a debian package from source. I can go to the package's web page and get the original source along with the diff. How do I apply the patch? If I do a $ tar zxf wu-ftpd_2.6.0.orig.tar.gz $ zcat wu-ftpd_2.6.0-4.diff.gz | patch -p1 can't find file to patch at input line 3 Perhaps you used the wrong -p or --strip option? The text leading up to this was: -- |--- wu-ftpd-2.6.0.orig/src/ftpcount.c |+++ wu-ftpd-2.6.0/src/ftpcount.c -- File to patch: I am interested in the wu-ftpd package. http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0.orig.tar.gz http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0-4.diff.gz brian -- Brian Lavender http://www.brie.com/brian/
HOW DO I Apply debian patch to original source?
I posted this to the list once already, but I don't know if it made it. I am trying to compile a debian package from source. I can go to the package's web page and get the original source along with the diff. How do I apply the patch? If I do a $ tar zxf wu-ftpd_2.6.0.orig.tar.gz $ zcat wu-ftpd_2.6.0-4.diff.gz | patch -p1 can't find file to patch at input line 3 Perhaps you used the wrong -p or --strip option? The text leading up to this was: -- |--- wu-ftpd-2.6.0.orig/src/ftpcount.c |+++ wu-ftpd-2.6.0/src/ftpcount.c -- File to patch: I am interested in the wu-ftpd package. http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0.orig.tar.gz http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0-4.diff.gz brian -- Brian Lavender http://www.brie.com/brian/
Re: HOW DO I Apply debian patch to original source?
I am having problems installing the wu-ftpd package from source. The whole reason I am grabbing the source is that I want to use a more recent version of wu-ftpd than what is available with the slink binary packages. If I can build the package from the source files to give me a slink compatible package, it is all the better. I update my /etc/apt/sources.list with http://http.us.debian.org/debian unstable main contrib non-free and then I did a # apt-get update $ apt-get source wu-ftpd and it retrieved the 2.6.0 wu-ftpd. It looks as if it applied the patch for me. I got an erro with the debian/rules binary though $ cd wu-ftpd-2.6.0 $ debian/rules binary dh_testdir make: dh_testdir: Command not found make: *** [build-stamp] Error 127 Do you know what this error indicates? On Mon, Mar 06, 2000 at 11:37:32PM -0700, Bob Nielsen wrote: On Mon, Mar 06, 2000 at 06:24:50PM -0800, Brian Lavender wrote: I posted this to the list once already, but I don't know if it made it. I am trying to compile a debian package from source. I can go to the package's web page and get the original source along with the diff. How do I apply the patch? If I do a $ tar zxf wu-ftpd_2.6.0.orig.tar.gz $ zcat wu-ftpd_2.6.0-4.diff.gz | patch -p1 can't find file to patch at input line 3 Perhaps you used the wrong -p or --strip option? The text leading up to this was: -- |--- wu-ftpd-2.6.0.orig/src/ftpcount.c |+++ wu-ftpd-2.6.0/src/ftpcount.c -- File to patch: I am interested in the wu-ftpd package. http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0.orig.tar.gz http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0-4.diff.gz To create a package from the Debian source, you should also install dpkg-dev and download http://ftp.debian.org/debian/dists/frozen/main/source/net/wu-ftpd_2.6.0-4.dsc Then type dpkg-source -x wu-ftpd_2.6.0-4.dsc (as root). This will unpack the source and apply the patch. You should then cd to the wu-ftpd-2.6.0 directory which was created and type debian/rules binary. This will compile the source and create the Debian package for you. -- Brian Lavender http://www.brie.com/brian/
packages in different selections on install
on install, there are certain preselected categories of packages that you can pick from, and get a set of preselected packages. Where can I find that list? How would I incorporate those packages into an existing installation? brian -- Brian Lavender http://www.brie.com/brian/
build raid on install
Is it possible to build RAID-1 on install before installing any packages. I was looking at the software-RAID howto, and it says to install linux on a single partition, then create your RAID devices, and then move the data over to the raid devices. Here is my setup: /dev/sda1 swap 64 Megs /dev/sda5 / 100 Megs /dev/sda6 /boot 10 megs /dev/sda7 /usr 700 Megs /dev/sda8 /tmp 250 Megs /dev/sda9 /var 350 Megs /dev/sda10 /home 7000 Megs /dev/sdb164 Megs /dev/sdb5100 Megs /dev/sdb610 megs /dev/sdb7700 Megs /dev/sdb8250 Megs /dev/sdb9350 Megs /dev/sdb10 7000 Megs As you can see, I have configured two disks almost identical. I want to turn /dev/sda7 and /dev/sdb7 into /dev/md0 which in turn will be mounted at /var and then I want to do /dev/sda10 and /dev/sdb10 into /dev/md1 which in turn will be mounted at /home When I read the software raid-HOWTO, it says install Linux on one root partition and then prepare those other partitions, but don't mount them. Then once you have a small running system, create the raid partitions, copy your data over, change you /etc/fstab so it mounts the /dev/mdx and voila, you are off and running. Here's the problem. I only have 100 Megs as root. I suppose I could install a teenie tiny system on this, but it would be nice to create the /dev/mdx devices before installation. Is this possible? brian -- Brian Lavender http://www.brie.com/brian/
Re: TeX
Go to an ftp archive and go into the proposed-updates for slink. There was some sort of y2k problem with tex where the make format or something to that effect failed. Anyway, the updated package will work. In fact here is the url for the cdrom debian archive. ftp://ftp.cdrom.com/pub/linux/debian/dists/slink-proposed-updates/.4/linux/debian/dists/slink-proposed-updates brian On Mon, Jan 03, 2000 at 12:56:30PM -0800, G. Crimp wrote: Hi, I've been using TeX at work for a few weeks now. Last week I got a new computer and installed 2.1 on it (same as was on the old box). As far as I can tell, I have all the appropriate tetex packages in place. TeX no longer works, though. When I try to run tex on a .tex file, I get the following error: This is TeX, Version 3.14159 (Web2C 7.2) I can't find the format file `tex.fmt'! The tetex docs say to run the command `texconfig init' in such a case. When I run it I get tons of output, including the following excerpts: [...] Beginning to dump on file tex.fmt (format=tex 1999.12.30) [...] No pages of output. Transcript written on tex.log [...] The command continues to run producing similar output concerning latex and metafont. However, the file tex.fmt is nowhere to be found on my file system. Maybe the line No pages of output. explains that. I thought I could get some answers from the transcript in tex.log, but that file appears not to have been written anywhere either. Anybody know what I have done wrong on this new computer, or more positively, how I can back to writing documentation with TeX ? Thanks, Gerald Crimp -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Re: installing apache from source
If apache that was installed was compiled with DSO support, then you should be able to add php support by going into dselect and and selecting it. DSO means Dynamic Shared Objects. Search your system for apxs which is the perl script that allows you to dynamically compile in modules. Then find your httpd binary and execute it as follows $ httpd -l It should list all the compiled in modules. If it lists mod_so.c then you have DSO support. If you feel inclined, you can grab the php source, and compile its module. In the PHP source you will find INSTALL.DSO where it will explain the steps to install it. Good Luck, brian On Mon, Jan 03, 2000 at 05:24:59PM +, hypnos wrote: I'm gonna be downloading the sources for apache and php3, so that I can compile apache with support for php3. I currently have apache installed (from slink .deb) on the machine. What's the best way to go about this change? I don't want my new installation of apache to conflict with the current installation. Should I use dselect to remove the current apache installation, and then install from source? I believe that [R]emove (from dselect) will not delete my configuration files, so I should still be able to use those with the new installation, right? Thanks -- hypnos mailto:[EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
Error with LaTeX
I am trying to use Latex and I am getting an error when I try to process the sample2e.tex file. Here is the error I get: [EMAIL PROTECTED]:~/docs latex sample2e.tex This is TeX, Version 3.14159 (Web2C 7.2) I can't find the format file `latex.fmt'! Is there supposed to be a latex.fmt somewhere on my system? brian -- Brian Lavender http://www.brie.com/brian/
Re: Error with LaTeX
On Fri, Dec 24, 1999 at 08:46:56AM -0800, Pann McCuaig wrote: On Fri, Dec 24, 1999 at 09:59, Paul Huygen wrote: Brian Lavender [EMAIL PROTECTED] wrote: Here is the error I get: [..] I can't find the format file `latex.fmt'! Is there supposed to be a latex.fmt somewhere on my system? Yes, there is. You can generate one using either a program initex or to run TeX in a special way. Usually latex.fmt comes with the TeX distribution or it is generated on the fly during the installation of the TeX package. I suggest you to look at the documentation of your TeX package. Alas, this won't work (I'm assuming the package from slink). Some critical date has 'expired'. I saw that an upload to correct this happened a few days ago so you might start looking for it in incoming if you have access or proposed-updates. An alternative (what I did when this happened to me a couple weeks ago) is to go to the teTeX site and get the tarball and do a /usr/local/ install. I am using slink. Obtaining the new package from proposed-updates did the trick. Thanks, brian -- Brian Lavender http://www.brie.com/brian/
glibc's crypt
Does the glibc in Debian slink have the good crypt by default? brian -- Brian Lavender http://www.brie.com/brian/
Re: fetchmail, sendmail, and mutt and sending mail with the correct a ddress
If you installed sendmail as a package from the Debian distribution, then you need to add the below line to your /etc/mail/sendmail.mc file and run the program sendmailconfig as root MASQUERADE_AS(ntrnet.net)dnl This should produce a line in your /etc/mail/sendmail.cf which looks like DMntrnet.net The sendmailconfig should restart sendmail, and everything should work. brian On Thu, Oct 21, 1999 at 06:07:56PM -0400, John Davis wrote: Hello I thought I had my mail system setup correctly, but then I noticed a problem. When I send mail from my linux box (which is connected to the internet via ppp,) it sends mail as if its coming from suit.ntrnet.net which is wrong. Suit is my hosthame and ntrnet.net is my isp domain. Ntrnet doesn't know about my hostname and thus can't forward mail to me. How can I configure my mailsystem so it sends mail as if its from ntrnet.net? ie., [EMAIL PROTECTED] instead of [EMAIL PROTECTED] Do I need to mess with the sendmail configuration, fetchmail, or mutt? Clueless in NC - John -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Brian Lavender http://www.brie.com/brian/
glibc questions in slink: crypt, linux threads
As I understand, Debian slink comes with MD5 crypt because of export restrictions. I want to have a real crypt so I went to a german ftp server, found glibc 2.0.7 source and I downloaded it. Of course the real crypt is in a separate tar file. I downloaded that too and unpacked it in the glibc source tree I read the faq as far as compiling glibc and it said to do a configure like ./configure --enable-add-ons=crypt,linuxthreads What I am wondering is, is glibc on slink compiled with linux threads? It seems as if there are issues either way and that I probably ought to go with it if it was compiled with it or vice a versa if it wasn't. ftp://ftp.gwdg.de/pub/linux/glibc/2.0.7pre6 brian -- Brian Lavender http://www.brie.com/brian/
