Re: Mount behaving oddly under sarge

[CW Harris]

On Thu, May 05, 2005 at 11:20:37AM -0600, Monique Y. Mudama wrote:
> On 2005-05-05, Roberto C. Sanchez penned:
> >
> > Yes.  The more correct solution is make sure that the vfat, fat and msdos
> > (since floppies are actually msdos, not vfat) modules are loaded into your
> > kernel on boot (this can be done by adding them as lines into /etc/modules).
> > Thing is, if the kernel has no support for a FS, the auto option is pretty
> > well useless.
> >
> > -Roberto
> 
> Wait -- auto only works if the module has already been loaded?  It
> doesn't work if the module is available (but not loaded)?
> 
> I hadn't realized that, but it could explain a number of weirdnesses
> I've had with auto.  Annoying, though.
> 
> What about if you put the type in /etc/filesystems, but don't have the
> module loaded?

Yes this will load the module... I've used it for some time.

I posted a longer reply above... not seeing this thread got broken from
the main one in my mutt reader.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Mount behaving oddly under sarge

[CW Harris]

On Thu, May 05, 2005 at 12:23:57PM +0100, Matthew Roberts wrote:
[...]
> After a reboot, I can't mount a floppy as a normal user:
> 
> $mount /media/floppy
> mount: I could not determine the filesystem type, and none was specified
> 
> Even though I have this line in /etc/fstab:
> 
> /dev/fd0/media/floppy0  autorw,user,noauto  0   0
> 
> However, I can become root using
> 
> $su
> 
> and do
> 
> #mount -t vfat /dev/fd0 /media/floppy
> 
> which works fine.
> 
> After that the original command (as a normal user) works fine:
> 
> $mount /media/floppy
> 
> But the next reboot sets it all off again.

You don't have the vfat module loaded, so as a normal user mount looks
for filesystems of the type specified in /proc/filesystems.  Running
mount -t vfat as root loads the vfat module for you, it now  shows up in
/proc/filesystems so it subsequently will work as a normal user.

Two solutions:
1) If you use vfat a lot, consider loading it at boot time in
/etc/modules or compiling into the kernel.

2) Probably the solution you want:
Add vfat to the file /etc/filesystems (create the file if necessary).
>From man mount:

  The type iso9660 is the default.  If no -t option is  given,  or
  if  the auto type is specified, the superblock is probed for the
  filesystem type (adfs, bfs, cramfs, ext, ext2, ext3, hfs,  hpfs,
  iso9660,  jfs,  minix,  ntfs,  qnx4,  reiserfs, romfs, udf, ufs,
  vxfs, xfs, xiafs are supported).  If  this  probe  fails,  mount
  will try to read the file /etc/filesystems, or, if that does not
  exist, /proc/filesystems.  All of the  filesystem  types  listed
  there  will  be tried, except for those that are labeled "nodev"
  (e.g., devpts, proc and nfs).  If  /etc/filesystems  ends  in  a
  line  with  a  single  * only, mount will read /proc/filesystems
  afterwards.

  The auto type may be useful for user-mounted floppies.  Creating
  a  file /etc/filesystems can be useful to change the probe order
  (e.g., to try vfat before msdos) or if you use a  kernel  module
  autoloader.  Warning: the probing uses a heuristic (the presence
  of appropriate `magic'), and could recognize the wrong  filesys-
  tem  type, possibly with catastrophic consequences. If your data
  is valuable, don't ask mount to guess.


HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: bash, perl, C

[CW Harris]

On Sat, Jan 08, 2005 at 12:18:33AM +0100, Gerard Robin wrote:
[...]
> 
> I think that my question was not very clear.
> 
> timerest3.pl in fact is a subroutine in a little script Perl that I wrote 
> myself ( I am not a student who expect that others do his work :-))
> In that subroutine I wrote
> ..
> while ($i < 6) {
> print "\e[0;46;31m", 5-$i, "\e[0m";
> `sleep 1`; # the sleep of the shell
> print "\r";
> ...
> If i use the sleep of Perl there nothing is displayed at though with the 
> sleep 
> of the shell the numbers 5 4 3 2 1 0 are displayed as I want.
> And, I don't understand why the sleep of Perl doesn't work ?
> I think it is not coherent to use the sleep of the shell whereas Perl has 
> such 
The sleep function is working as you expect, it is the output that is
buffered.  Look at the perl info for the $| variable.  (Hint: set $| to
a non-zero value for unbuffered output.)

(Not a perl guru, but I think this is what you want.)

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Can a particular network card be permenantly bound to an eth'x' number?

[CW Harris]

On Fri, Jan 07, 2005 at 09:21:46PM +, rich wrote:
> I have several network cards in my laptop - wired lan, wireless lan,
> loopback & firewire.  After a recent update (I'm running testing) my
> interface numbers all jumped around so that instead of the wired lan
> being eth0, it's now eth1 & the firewire is eth0.  What defines what
> eth'x' number is given to which network device?  It's a pain having to
> change configuration each time they move numbers (as also happens
> depending on whether I boot with my wireless cardbus in the slot or
> not)!

And I believe the order is reversed between 2.4 kernels and 2.6 kernels.

I believe this is easily done by the new udev (although I haven't used
it).

You might also look at "nameif" in the net-tools package.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Debian on an old PC

[CW Harris]

On Fri, Jan 07, 2005 at 12:53:11PM -0600, Kent West wrote:
> Paul Johnson wrote:
> 
> >On Thursday 06 January 2005 01:44 pm, John Schmidt wrote:
> >
> > 
> >
> >>Old pcs often can't boot from a CD even if they have one.  You might 
> >>be able to flash the BIOS to upgrade it, but that assumes there is an 
> >>update out there (highly unlikely).  
> >>   
> >>
> >
> >If it's too old to boot from a CD, wouldn't it also pre-date flashable 
> >BIOS?
> > 
> >
> No, I remember flashing old AT&T 6300 (4.77MHz 8086 PC) PCs here on 
> campus because they had a Y2K-style glitch in the BIOS; it wouldn't go 
> beyond 1987 or thereabouts (my memory is hazy).

Hey! I'm using one of those as a text terminal (typing this on it
actually).  The display is a bit slow on long listings, but I *really*
like the feel/key placement of the keyboard.

Back more on-topic for this thread, I still use some older PC's.  But
*fast enough* is pretty relative.  One can manage with some pretty slow
hardware until you try something faster.  I used gdm/enlightenment on an
AMD 686 ~150Mhz with 48MB memory for sometime. But after trying faster
hardware it really starts to seem intolerably slow.

What I find is the biggest problem on older hardware is actually getting
large amounts of memory for them.  They don't really have enough memory
slots to add, and by the time you replace them with larger DIMMS the
cost becomes a significant chunk of an all-around better used machine.

(Of course if the machine is new enough to use current memory modules
 then it becomes much easier.)

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: locatedb is world readable

[CW Harris]

On Fri, Jan 07, 2005 at 08:29:55AM -0600, Kent West wrote:
> David Garamond wrote:
> 
> >>Sorry, a followup question. I deleted /etc/cron.daily/find several 
> >>days ago. How do I get it back? Reinstalling the findutils package 
> >>(using synaptic) doesn't bring it back.
> >
> >>Do I have to purge and install?
> >
> [EMAIL PROTECTED]:/home/westk> apt-file search /etc/cron.daily/find
> findutils: etc/cron.daily/find
> findutils: etc/cron.daily/find
> 
> Since you've already reinstalled findutils and that didn't do it, then 
> I'd say, "Yep. Try purging and installing." I guess since this file is 
> in /etc, it's considered a configuration file and therefore doesn't get 
> replaced on a plain reinstall.

You can always extract the file from the .deb with dpkg-deb -x
and then copy it back to etc/cron.daily/


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Multiple installed kernel-image packages?

[CW Harris]

Putting this back to the list...

On Wed, Dec 29, 2004 at 04:33:43PM -0700, Paul E Condon wrote:
> On Wed, Dec 29, 2004 at 12:45:09PM -0700, CW Harris wrote:
>> On Wed, Dec 29, 2004 at 12:31:33PM -0700, CW Harris wrote:
>>> On Wed, Dec 29, 2004 at 12:33:14PM -0500, William Ballard wrote:
>>>> On Wed, Dec 29, 2004 at 10:19:23AM -0700, CW Harris wrote:
>>>>> There are source packages kernel-latest-{version}-{arch}.  Is  this what
>>>>> you are looking for? (E.g. kernel-latest-2.6-i386)
>>>> 
>> 
>> Following up on this a bit more... 
>> 
>> I don't know what kernel-latest-2.6-i386 is (maybe patches?) but it is
>> definitely too small to be the full kernel source.
>> 
>>  $ apt-get -s source kernel-latest-2.6-i386
>>  Reading Package Lists...
>>  Building Dependency Tree...
>>  Need to get 3691B of source archives.
>>  Fetch Source kernel-latest-2.6-i386
>> 
>> Sorry for the trouble. Maybe these are just packages the maintainers use
>> to keep things up to date (since they don't show up properly in
>> searches).
>> 
> 
> For most Debian packages, one can upgrade to the latest version
> without fear of breaking anything. But, installing a new kernel
> requires a reboot of your computer in order to check if it has been
> correctly. You don't have to do the reboot, but if its not correctly
> installed and if there is a power failure and you try to get going
> again after it, you have a problem. So, Debian policy is that kernel
> upgrades require manual intervention. You can build a script that
> finds the latest kernel patches and applies them to the latest
> kernel source, and compiles the kernel, builds the kernel deb, and
> installs it, but the high priests of Debian are too timid to do
> such a thing. (And, IMHO, you should be, also.)

This was in response to automatically installing the latest *source*
package, not automatically installing the latest image.  I agree it
would not be good to automatically install a new kernel.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Multiple installed kernel-image packages?

[CW Harris]

On Wed, Dec 29, 2004 at 12:31:33PM -0700, CW Harris wrote:
> On Wed, Dec 29, 2004 at 12:33:14PM -0500, William Ballard wrote:
> > On Wed, Dec 29, 2004 at 10:19:23AM -0700, CW Harris wrote:
> > > There are source packages kernel-latest-{version}-{arch}.  Is  this what
> > > you are looking for? (E.g. kernel-latest-2.6-i386)
> > 

Following up on this a bit more... 

I don't know what kernel-latest-2.6-i386 is (maybe patches?) but it is
definitely too small to be the full kernel source.

$ apt-get -s source kernel-latest-2.6-i386
Reading Package Lists...
Building Dependency Tree...
Need to get 3691B of source archives.
Fetch Source kernel-latest-2.6-i386

Sorry for the trouble. Maybe these are just packages the maintainers use
to keep things up to date (since they don't show up properly in
searches).


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Multiple installed kernel-image packages?

[CW Harris]

On Wed, Dec 29, 2004 at 12:33:14PM -0500, William Ballard wrote:
> On Wed, Dec 29, 2004 at 10:19:23AM -0700, CW Harris wrote:
> > There are source packages kernel-latest-{version}-{arch}.  Is  this what
> > you are looking for? (E.g. kernel-latest-2.6-i386)
> 
> Do you mean kernel-source-2.6?

No.

> 
> There are no packages named kernel-latest-* in the repository.

Okay. This is what I did. I noticed in the description of
kernel-image-2.6-686:

$ apt-cache show kernel-image-2.6-686
Package: kernel-image-2.6-686
Priority: optional
Section: base
[...]
Source: kernel-latest-2.6-i386
   This.
Version: 100
Depends: kernel-image-2.6.8-1-686
[...]

So I searched package source names on packages.debian.org (since I don't
have any source lines in my apt setup on this machine). And it returned
information on packages like kernel-latest-2.6-i386, as I described in
my email to this list.

In response to your statement I now added a sarge src line to my apt
config, and:

$ apt-cache showsrc kernel-latest-2.6-i386
Package: kernel-latest-2.6-i386
Binary: kernel-headers-2.6-k7, kernel-image-2.6-686,
kernel-image-2.6-k7-smp, kernel-image-2.6-386,
kernel-image-2.6-686-smp, kernel-headers-2.6-686,
kernel-headers-2.6-386, kernel-image-2.6-k7,
kernel-headers-2.6-k7-smp, kernel-headers-2.6-686-smp
Version: 100
Priority: optional
Section: devel
Maintainer: Debian kernel team <[EMAIL PROTECTED]>
Build-Depends: debhelper (>= 4)
Architecture: i386
Standards-Version: 3.6.1
Format: 1.0
Directory: pool/main/k/kernel-latest-2.6-i386
Files:
 375238fac45257cff33b5ef47c1a1fd0 1542 kernel-latest-2.6-i386_100.dsc
 e0eab2db85dee02b8a40c9ae0c1c86c8 2149 kernel-latest-2.6-i386_100.tar.gz
Uploaders: William Lee Irwin III <[EMAIL PROTECTED]>, Andres Salomon
<[EMAIL PROTECTED]>


However the apt-cache search kernel-latest doesn't seem to find any
returns for kernel-latest, so I guess kernel-latest is a strange type of
package.  I guess we need some clues from the repository team?

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Multiple installed kernel-image packages?

[CW Harris]

On Tue, Dec 28, 2004 at 11:30:15PM -0500, William Ballard wrote:
> On Tue, Dec 28, 2004 at 11:08:19PM -0500, Adam Aube wrote:
> > Paul Gear wrote:
> [snip nested attributions, correctly I hope]
> > > Thanks for the detailed response.  Are you saying that once my system is
> > > installed (on 2.6.8, as it happens), it will never get an upgrade to
> > > 2.6.9 (once it is released) unless i explicitly install it?
> > 
> > Correct.
> 
> Why in the heck isn't there a:
> kernel-source-latest, which is upgraded whenever a new kernel source package
> comes out, and depends on that, and a
> kernel-source-latest-only, which depends on kernel-source-latest and
> conflicts with earlier kernel-source packages?

There are source packages kernel-latest-{version}-{arch}.  Is  this what
you are looking for? (E.g. kernel-latest-2.6-i386)

> 
> Doing the same for kernel-images would result in an explosion of
> kernel-image-* packages in the archive, so maybe just do it for the source.
> Only two extra packages.

As another thread pointed out there are meta-packages for images that
will always depend on the latest kernel.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Correct syntax for mkinitrd command?

[CW Harris]

On Tue, Dec 28, 2004 at 12:43:10PM -0500, Paul Tsai wrote:
> Andrew,
> I believe if you compile the kernel "the debian way" as they  like to 
> say on this list. your job is a lot easier.
> 
> after configuring your kernel instead of make bzimage, do
> make-kpg --initrd -rev 1 kernel_image

Typo should be "make-kpkg" (from kernel-package)


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: FAT32 (was: dual-OS system)

[CW Harris]

On Thu, Dec 16, 2004 at 04:12:35PM +, Daniel Goldsmith wrote:
> On Thu, 16 Dec 2004 15:44:40 +, David Dorward <[EMAIL PROTECTED]> wrote:
> > Daniel Goldsmith wrote:
> 
[...]
> > > o Why were the dos/win filesystem supports removed from Sarge's
> > > kernels?
> > 
> > They aren't, as far as I know. Being a module doesn't prevent the kernel
> > accessing them; AFAIK the modules will be loaded automatically when
> > mount attempts to mount a vfat partition.
> 
> Hmmm - an older thread on debian-user mentions /etc/filesystem for the
> auto mounting usage. That may be my problem - the Sarge install
> doesn't appear to have created an /etc/filsesystem.

This is only needed for auto mounting if the kernel needs to load a
module. If the module is compiled in, or *already* loaded, the kernel
should try that fs type.

You say you have loaded vfat and *still* the kernel doesn't recognize
the fat32 partition?  So "$/sbin/lsmod" shows vfat support?  What about
fat support (it should be auto loaded when vfat is loaded)?

If these two are loaded your kernel should mount the partition.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: switching X resolution modes?

[CW Harris]

On Tue, Dec 14, 2004 at 01:18:38PM +0100, Michal R. Hoffmann wrote:
> On 13-12-2004 23:39,Robert Storey wrote:
> > Dear Michal,
> > 
> > I don't know if this is exactly what you want, but here's my idea. If
> > you are starting X with the "startx" command, there is a switch
> > available for color depth:
> > 
> [cut]
> 
> Oh, I see I have asked unclearly. What I want is changing the modes
> (real resolutions, without touching the virtual screen size) from a
> script. So xrandr cannot be used, neither startx. BTW, I use gdm.
> 
> I look for an application/script what would do exactly the same what
> Ctrl-Alt-GrPlus/GrMinus do.

Well, 
$ xvidtune -next
$ xvidtune -prev

will simulate the Ctrl-Alt-KP-{Plus/Minus}.  But I don't know how to
specify a specific resolution.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: paper size

[CW Harris]

On Mon, Dec 06, 2004 at 06:14:30PM -0500, Antonio Rodriguez wrote:
> In a sarge installation, I should have leter as paper size for my
> system, but it seems more like a4. Checking config.ps I see a4 listed
> before letter. A few years ago, it meant that ps docs produced locally
> would be a4 by default. Then I had to manually tweak it to get
> paper size. Afterwards someone wrote in the list that the
> configuration for dvips was done differently than before. Mozilla
> pages printed through postscript default are fine, but docs processed
> through dvis, psresize, pstops, etc are a mess. Would anyone know

To configure dvips use "texconfig"

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: lspci and /sys

[CW Harris]

On Sat, Dec 04, 2004 at 10:22:41AM +0100, Petter Senften wrote:
> Hello!
> 
> I'm new to this list, but after googling for a few hours after a fix to 
> my problem it seems I have no other option than to try to get some 
> attention here :)
> 
> My server is a Debian Sarge-box (dist-upgraded from Woody a few months 
> back). Anyhoo, anytime I use lspci I get the following error: pcilib: 
> Cannot open /sys/bus/pci/devices
> 
> Also, when the PHP-script "PHPsysinfo" runs it generates the same error 
> in the Apache Error-log, as well as taking forever to finish the 
> operation. Both items still list my PCI devices, but after a delay 
> (which to me is annoying).
> 
> If I go to /sys it's empty, so obviously something is missing. I 
> wouldn't say that I'm a complete Linux-bewbie, but this problem has me 
> completely stumped, and as mentioned Google has been of no help this 
> time. So I'm asking all the experts out there - what can I do to correct 
> this?

Did you mount sysfs?

$grep sysfs /etc/fstab
sysfs   /syssysfs   defaults0  0


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Newbie Having Much Trouble Getting Deb 2.4.18-bf2.4 To Work

[CW Harris]

On Fri, Dec 03, 2004 at 01:45:33PM -0600, Leonard Chatagnier wrote:
> Thanks so much, that added enough clarity that I now know what was wrong.
> I'll read the manual.

You are welcome.  Also, please do not cc me. I read the lists. Please
browse through the Debian Mailing lists introduction, note especially
the Code of Conduct indicates some etiquette points, notably do not CC
unless requested. (I am sending you a copy in this particular case. I
hope this doesn't seem hypocritical :-P)

You can browse this here:
http://www.debian.org/MailingLists/#codeofconduct

I notice others have commented on top-posting.  Although I don't recall
anything explicit about top/bottom posting for this list, it is my
observation that most on this list prefer bottom posting. (Not trying
to start a flame war here :-)


Again, if you are new to Debian and don't have a lot of files to save
(or if you have some spare hard drive space for playing around) I would
recommend you try the new installer.  It is not quite released yet, but
is in very good shape, last time I looked (certainly much better shape
than the Woody installer---many improvements have been made [cudos to
the installer team]).

You can download a 100M network install CD image from here: 
http://www.debian.org/devel/debian-installer/

(This is a few links off of the Getting Debian link on the main web
 site. Follow "Getting Debian" -> "downloading as you install"
-> "Minimal CD" )



-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Newbie Having Much Trouble Getting Deb 2.4.18-bf2.4 To Work

[CW Harris]

On Thu, Dec 02, 2004 at 11:12:34PM -0600, Leonard Chatagnier wrote:
> Chris.
> Thanks, I'll try your suggestion and report back on results.  FYI, the url's
> were copied verbatum from the Debian mirror list.  If they are formatted
> incorrectly, I'm not aware, and why would the Debian site misformat there
> own links?.

Maybe I wasn't clear.  The links are okay...but you need information
after them in the sources.list file. Notice all the other lines.
Following the URI is information about the distribution you want.
E.g.
deb http://ftp.us.debian.org/debian/ stable main contrib

This gives the URI and specifies the stable distribution, main and
contrib sections.

>From the sources.list man page:
   The  format  for  a  sources.list entry using the deb and deb-src types
   are:

   deb uri distribution [component1] [component2] [...]

Note that the distribution is *required* information. (man sources.list)


[...]
> > Now my comments:
> >
> > deb ftp://ftp.us.debian.org/debian/
> > deb ftp://ftp.uk.debian.org/debian/
> > deb ftp://ftp.au.debian.org/debian/
> > ^^ these are lines 26-28 they are not properly formatted.
> > These 3 lines all give you the same info, they are just located in
> > different countries.  Pick one mirror close to you and stick with it.
> >
> > deb http://ftp.us.debian.org/debian/ stable main non-free contrib
> > This line will get the US mirror of the stable distribution (I guess
> > that is what you were trying to do above?)
> >
> > So delete lines 26-28 and I think your sources.list is fine.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Woody and kernel 2.6

[CW Harris]

On Thu, Dec 02, 2004 at 04:14:06PM +0100, Andreas Janssen wrote:
> Hello
> 
> Thomas Beresford (<[EMAIL PROTECTED]>) wrote:
> 
> > Is it possible to install a 2.6 kernel on woody?
> 
> Yes, if you install the necessary userspace utilities. This includes
> module-init-tools, and newer versions of a lot of other packages,
> perhaps procps, util-linux, e2fstools, lm-sensors and some more. There
> are 2.6 backports available from backports.org, that packages should
> satisfy the most important dependencies.
> 
> > If so what version is the most recommmended?
> 
> If you compile the source from kernel.org, use the newest version.
> 
Additionally, this info was in the last Debian Weekly News[1].  I haven't
looked at it, but maybe it is helpful?


(From Debian Weekly News - November 30th, 2004)

Installing Linux 2.6 on Debian. Falko Timme wrote a [20]tutorial on
running Linux 2.6 together with Debian GNU/Linux 3.0 (woody). In order
to have module support some backports need to be installed. One of the
final steps is creating a ramdisk so that the kernel will boot
properly.

 20. http://www.falkotimme.com/howtos/debian_kernel2.6_compile/



[1] List-Archive: 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Newbie Having Much Trouble Getting Deb 2.4.18-bf2.4 To Work

[CW Harris]

On Thu, Dec 02, 2004 at 01:51:21AM -0600, Patrick Albuquerque wrote:
> On Wed, Dec 01, 2004 at 11:55:39PM -0600, Steve Block wrote:
> > Patrick Albuquerque wrote:
> > >On Wed, Dec 01, 2004 at 04:38:19PM -0600, Leonard Chatagnier wrote:
> > >
> > >You have encrypted your /etc/apt/sources.list file?  Also, what version
> > >of Debian are you installing?  
> > >
> > >Patrick,
> > 
> > The file is plain text, just typed as binary in the attachment for some 
> > reason.
> > -- 
> 
> Strange,  I see no attachments in mutt, just see line after line like this...
> 
> begin 666 sources.list
> M"B-D96(@8V1R;VTZ6T1E8FEA;B!'3E4O3&EN=7@@,RXP('(R(%]7;V]D>5\@
> M+2!/9F9I8VEA;"!I,[EMAIL PROTECTED]($)I;F%R>2TW("@R,# S,3(P,2E=+R!U;G-T86)L
> 
> Is my system misconfigured?  I can't see anything obvious in .muttrc
> 
It's uuencoded and just in-line rather than mime-attached.  In the
future he should just include the relevent portion of the file in the
email, rather than attaching it, and especially not including it in this
strange way. The decoded sources.list file is:

#deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-7 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
[skipping similar commented lines...]

# deb http://security.debian.org/ stable/updates main

deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-7 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-6 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-4 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-5 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-3 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-2 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb cdrom:[Debian GNU/Linux 3.0 r2 _Woody_ - Official i386 Binary-1 
(20031201)]/ unstable contrib main non-US/contrib non-US/main
deb http://security.debian.org/ stable/updates main
deb ftp://ftp.us.debian.org/debian/ 
deb ftp://ftp.uk.debian.org/debian/
deb ftp://ftp.au.debian.org/debian/

deb http://ftp.us.debian.org/debian/ stable main non-free contrib
deb-src http://ftp.us.debian.org/debian/ stable main non-free contrib
deb http://non-us.debian.org/debian-non-US stable/non-US main contrib non-free
deb-src http://non-us.debian.org/debian-non-US stable/non-US main contrib 
non-free

Now my comments:

deb ftp://ftp.us.debian.org/debian/ 
deb ftp://ftp.uk.debian.org/debian/
deb ftp://ftp.au.debian.org/debian/
^^ these are lines 26-28 they are not properly formatted.
These 3 lines all give you the same info, they are just located in
different countries.  Pick one mirror close to you and stick with it.

deb http://ftp.us.debian.org/debian/ stable main non-free contrib
This line will get the US mirror of the stable distribution (I guess
that is what you were trying to do above?)

So delete lines 26-28 and I think your sources.list is fine.

You might want to browse http://www.debian.org/doc/  Debian
documentation.  There are also many other sites, newbiedoc.sourceforge.net
is one other I have seen mentioned.  Google this list for newbie docs
and you'll find dozens of suggestions, I'm sure.

As an aside, if you are really new to Debian you might want to jump in
with the Sarge distribution (currently testing, soon to be the new
stable).  The current stable (Woody) is quite old.  It is still going
through changes though, that might be intimidating to a newbie.

Browse here if you want some history of Debian:
http://www.debian.org/doc/manuals/project-history/ch-intro.en.html

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: remote build of Debian router

[CW Harris]

Ray you sent this to me privately...

Fwd this one back to the list...


CW Harris wrote on 2004-11-30 02:33:
> You might google, or ask, over on debian-firewall... I seem to recall
> someone mentioning a package or setup they used that allowed one to
> restore the old ruleset after a defined time period---just for these
> cases where you shoot yourself in the foot and lock yourself out.
> 
> Unfortunately, I can't seem to find it now. :<

fwbuilder has that feature --a recent version is needed, I think.

After a pre-specifiable amount of time, unless you upload a non-expiring
version of your policies, fwbuilder will reboot the firewall and hence
restore the prev settings.

This assumes, of course, you have an initial firewall on the box.

HTH,

Cheers,
Ray


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: remote build of Debian router

[CW Harris]

On Mon, Nov 29, 2004 at 05:04:20PM -0500, David Mandelberg wrote:
> Mike M wrote:
> >>Also, there should be a firewall on the box before it's
> >>ever connected to the net.
> > 
> > 
> > Chicken/egg dilemma here?  Any suggestions?
> This isn't a chicken/egg dilemma (almost nothing is), all you need to do
> is install it from a cd or something, if you can't do that then just
> install from the net, disconnect, set it up, and reconnect. Or if you
> have the resources (unlikely), it's easier to set up all computers on a
> dmz (if you don't know what that is, ignore this sentence) and then
> transfer them to wherever they're needed.
> 

You might google, or ask, over on debian-firewall... I seem to recall
someone mentioning a package or setup they used that allowed one to
restore the old ruleset after a defined time period---just for these
cases where you shoot yourself in the foot and lock yourself out.

Unfortunately, I can't seem to find it now. :<


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Floppy mount & unmount clarification please

[CW Harris]

On Sun, Nov 28, 2004 at 04:36:45PM -0600, Jim Hall wrote:
> I'm looking for clarification of what I see the floppy mount/unmount 
> process doing on Sarge.
> 
> 1) A native Linux disc will mount/unmount without asking for a fs type.
> 
> 2) A DOS/Windoze disc asks for a fs type.
> 
> 3) I believe the default Sarge install for floppy support is a module.
> 
> 4) Where do I find out how to "tell" the module to recognize more file 
> types? I need the system to just deal with whatever known fs is put in 
> the drive. Without damaging anthing.

The (auto) type will try filesystems noted in /proc/filesystems,
so by default it will try the ones the kernel knows about (including
already loaded modules, but it won't load a module to try).

You can change this behaviour (including the order it tries) by putting
the filesystems you want (auto) to check for in the file
/etc/filesystems.  Mine for example is:

$cat /etc/filesystems
vfat
msdos
ext3
ext2


$man mount(8) for the explanation of this.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: MBR problem

[CW Harris]

On Tue, Nov 23, 2004 at 12:43:32PM -0800, Punit Ahluwalia wrote:
> I installed grub to /dev/hdc2 which is also /root. This is not a dual boot. 
> It is plain
> and simple "Woody".
> 
> How do I install grub to the MBR?

If I understand your setup correctly, you want to install grub to hda.

This is probably where LILO *was* installed, hence it is still trying to
boot LILO.

HTH

> 
>  
> --- CW Harris <[EMAIL PROTECTED]> wrote:
> 
> > On Tue, Nov 23, 2004 at 10:00:38AM -0800, Punit Ahluwalia wrote:
> > > Thanks for the response. But I did dpkg -P lilo. yet, nothing chnages.
> > > 
> > This sounds like your BIOS is trying to boot the old LILO boot sector
> > rather than your grub boot sector.  Are you installing grub to the MBR
> > or to a partition?  If you are installing it to a partition, make sure
> > that your MBR is trying to boot that partition.
> > 
> > Is this a dual-boot windows thing?
> > 
> > > 
> > > --- Kent West <[EMAIL PROTECTED]> wrote:
> > > 
> > > > Punit Ahluwalia wrote:
> > > > 
> > [...]
> > > > >with either of them. After the BIOS transfers the control, the
> > > > >screen displays a message
> > > > >"searching for the boot record : ok"  and then the letter "L" and
> > > > >then a series of number
> > > > >40 (40 40 40 ..). I don't know what is going on. I can boot from
> > > > >the floppy using kernel-2.2.20.
> > > > >  
> > > > >
> > > > lilo is still partially installed. You might want to "apt-get remove 
> > > > lilo" followed by "grub-install", but as I've only done it a time or 
> > > > two, I'm not sure of the exact steps, so you might want to wait to hear 
> > > > from others first.
> > > > 
> > [...]

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: MBR problem

[CW Harris]

On Tue, Nov 23, 2004 at 10:00:38AM -0800, Punit Ahluwalia wrote:
> Thanks for the response. But I did dpkg -P lilo. yet, nothing chnages.
> 
This sounds like your BIOS is trying to boot the old LILO boot sector
rather than your grub boot sector.  Are you installing grub to the MBR
or to a partition?  If you are installing it to a partition, make sure
that your MBR is trying to boot that partition.

Is this a dual-boot windows thing?

> 
> --- Kent West <[EMAIL PROTECTED]> wrote:
> 
> > Punit Ahluwalia wrote:
> > 
[...]
> > >with either of them. After the BIOS transfers the control, the
> > >screen displays a message
> > >"searching for the boot record : ok"  and then the letter "L" and
> > >then a series of number
> > >40 (40 40 40 ..). I don't know what is going on. I can boot from
> > >the floppy using kernel-2.2.20.
> > >  
> > >
> > lilo is still partially installed. You might want to "apt-get remove 
> > lilo" followed by "grub-install", but as I've only done it a time or 
> > two, I'm not sure of the exact steps, so you might want to wait to hear 
> > from others first.
> > 
[...]

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: NFS permissions question

[CW Harris]

On Tue, Nov 16, 2004 at 02:01:02PM -0500, Christian Convey wrote:
> My understanding of NFS permissions is that for any file appearing on an 
> NFS share, the username/uid and groupname/gid mappings should (ideally) 
> be identical on both the NFS client and the NFS server.
> 
> So consider my home situation: I'm running two computers, each with 
> local security files.
> 
> I have four users: "alvin", "benny", "charles", and "david".
> I have several groups: "users" and "chefs" and "busboys".
> 
> I want to define an NFS share that "alvin" and "benny" can use. My 
> *expectation* at the time I'm setting this up is that any files 
> appearing on those shares will have a group-owner of "chefs".
> 
> So I go through, and ensure that "alvin" and "benny" each have the same 
> uid on both computers. I go through and ensure that "chefs" has the same 
> gid on both computers.
> 
> Is there a good way for me to ensure that alvin doesn't create, on the 
> shares, a file owned by the busboys group?
> 
> (The reason I don't want this to happen is that I've taken no steps to 
> ensure that both computers have the same groupid for the "busboys" 
> group. I don't want the resulting permissions confusion to ensue.)
> 

You may want to look at this recent thread on debian-user, in particular
this reference:
http://lists.debian.org/debian-user/2004/10/msg03439.html

Note: I'm no NFS expert, so maybe this is not what you're looking for.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: root crash recovery

[CW Harris]

On Wed, Nov 10, 2004 at 03:15:28PM +0300, Al Nikolov wrote:
> Hi all
> 
> My root filesystem has crashed because of hardware failure. (Although, it
> was ext3, and i wasn't supposed that may happens... Where's my fault?)
> 
> But there are still alive usr and, more important, var. I plan to install
> the minimal system onto new clean root partition, mount old usr and var,
> and reinstall all packages existed before the crash.
> 
> I suppose something like
> 
> apt-cache pkgnames|xargs -n1 apt-get --reinstall install

This lists all known packages.  I think what you are wanting is
(see dpkg(8)):

   To make a local copy of the package selection states:
dpkg --get-selections >myselections

   You might transfer this file to another computer, and install it
   there with:
dpkg --set-selections 
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Going from stable to testing

[CW Harris]

On Thu, Nov 04, 2004 at 10:25:57AM +0100, Jacob Friis wrote:
> >>Can I upgrade my system to be all testing just by adjusting
> >>/etc/apt/sources.list?
> >>
> >yes, just put testing instead of stable and then run apt-get update
> >apt-get dist-upgrade
> 
> I noticed this:
> http://www.debian.org/doc/manuals/reference/ch-woody.en.html";>
> Note that tracking the testing distribution of Debian can have the side 
> effect of delaying the installation of packages containing security 
> fixes, since such packages are uploaded to unstable and only later 
> migrate to testing.
> 
> 
> Is there anything to do about that?

Perhaps what you *really* want is not testing but sarge?  If you just
want to upgrade to the (soon to be stable) sarge, then put in "sarge"
not "testing".  That way when sarge becomes the new stable you will be
tracking the stable distribution, and you can then change it to stable
again.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Using more than one driver for a laser printer?

[CW Harris]

On Wed, Nov 03, 2004 at 12:29:49PM -0800, Rich Rudnick wrote:
> On Wed, 2004-11-03 at 19:39 +, Adam Funk wrote:
[...]
> > 
> > Thanks for that info.  As I said in another post in this thread,
> > modifying /etc/printcap alone didn't seem to have any effect---does my
> > lpr command (from the packages I listed above) use it at all or not?
> > 
> No, CUPS doesn't use printcap, or at least it doesn't with the version
> in unstable.  
> 
> cupsys-bsd provides the lpr used with CUPS, which is a near replacement
> for for the bsd lpr. IIRC it is missing a few options and has a few of
> it's own, but I've never had any incompatibilities with programs
> expecting bsd's lpr. In other words, just use localhost:631/admin to
> define your printers, ignore /etc/printcap and everything should work as
> expected.

Rather than ignore it, I think it is better to link /etc/printcap to the
CUPS printcap file (/etc/printcap.cups by default IIRC), then applications
that try to use /etc/printcap will be using the correct one.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: List of packages

[CW Harris]

On Wed, Oct 27, 2004 at 02:36:03PM -0500, Jim Hall wrote:
> on Sarge, is there a way to list every installed package? I don't think 
> I need things like libs, just the package names. I need to compare two 
> systems.

dpkg --get-selections

Note this does not give installed versions.

> 
> Jim
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: CUPS Printer Sharing Problem

[CW Harris]

On Mon, Oct 25, 2004 at 05:43:50PM -0400, Thomas H. George wrote:

Haven't really done much with CUPS via ipp so I could be wrong here
but:

>hp HP DeskJet 930C - Foomatic/hpjis (recommended)
>Description: DeskJet 940C
>Location: Study
>Printer State: Idle, Accepting Jobx
>"Unable to get printer status (Client-Error-Forbidden)!"
>Device URI: http://Phoenix:631/ipp
   ^^
Shouldn't this be http://Phoenix:631/printers/ ?

Although the Client-Error-Forbidden message does seem to indicate a
permissions problem.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: optimal printer configuration

[CW Harris]

On Mon, Oct 25, 2004 at 05:08:31PM +0200, Carl Fûrstenberg wrote:
> I have a HP deskjet 940c , have some problem to conigure it perfectly,
> now even lpr doen't work, only lp. Have anyone a good setup to a

Are you using CUPS?  If so, did you install the CUPS compatible lpr in
cupsys-bsd?

> printerenvironment in linux? Using debian sid, Xfce4, connected to the
> printer via usb.
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: gtop

[CW Harris]

On Thu, Oct 21, 2004 at 02:28:00PM -0500, Brandt Dusthimer wrote:
> There is a gtop package in stable, but no package in testing or
> unstable.  Is there a reason why?  Is it in another package?
$ apt-cache show gnome-system-monitor
Package: gnome-system-monitor
Priority: optional
Section: gnome
[...]
Version: 2.6.0-5
Replaces: gtop, procman
^^^
[...]
MD5sum: e573d64746d1510d1e34850721cbaa42
Description: Process viewer and system resource monitor for GNOME 2
 This package allows you to graphically view and manipulate the running
 processes on your system.  It also provides an overview of available
 resources such as CPU and memory.
 .
 The gnome-system-monitor replaces both the "procman" and "gtop"
 packages from the GNOME 1 desktop; this package is for use under the
 GNOME 2 desktop environment.


FWIW there are several ways to search for files (google this group for
many answers). You can also search for files on the packages.debian.org
web site.  For me:

$apt-cache search gtop
libgtop2 - Libraries for gtop system monitoring library (part of Gnome 2)
drawmap - draws customized maps, using raw USGS data files
gnome-system-monitor - Process viewer and system resource monitor for GNOME 2
libgtop-daemon - gtop daemon for monitoring remote machines (part of Gnome)
[...]

gnome-system-monitor looks promising ;)

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: removing wierd files

[CW Harris]

On Wed, Oct 20, 2004 at 02:42:21PM -0400, Matt Price wrote:
> CW Harris wrote:
> >On Wed, Oct 20, 2004 at 01:33:30PM -0400, Adam Garside wrote:
> >
> >>On Wed, Oct 20, 2004 at 12:51:32PM -0400, Matt Price wrote:
> >>[Snip]
> >>
> >>>rm \-*
> >>>or
> >>>rm '-0.pnf'
> >>>
> >>>is there a way to reference files whose names begin with -?  
> >>
> >>try 'rm -- -0.pnf'
> >
> >
> >Or "rm ./-0.pnf" works too.
> >
> >
> doh!  and it's even in the manual!
> 
> thank you for the info.  anyone know what the '--' option sigifies?  I 
> notice it's not really documented in the man page.
> m

"info rm" or "man getopt".  GNU tools using the getopt function make use
of "--" to delimit options from parameters. (Note that in this case info
does not give the same page as man---at least on my system).


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: removing wierd files

[CW Harris]

On Wed, Oct 20, 2004 at 01:33:30PM -0400, Adam Garside wrote:
> On Wed, Oct 20, 2004 at 12:51:32PM -0400, Matt Price wrote:
> [Snip]
> > rm \-*
> > or
> > rm '-0.pnf'
> > 
> > is there a way to reference files whose names begin with -?  
> try 'rm -- -0.pnf'

Or "rm ./-0.pnf" works too.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: printer won't stop

[CW Harris]

On Fri, Oct 15, 2004 at 09:00:55PM +0100, Richard Lyons wrote:
> On Friday 15 October 2004 16:41, CW Harris wrote:
> > On Fri, Oct 15, 2004 at 01:26:09AM +0100, Richard Lyons wrote:
> [...]
> > >   I have one box where cups won't let me change anything and I 
> > > cannot see what I have configured wrong.  
> > 
> > Are you a member of the SystemGroup (default lpadmin) defined in 
> cupsd.conf?
> 
> # grep lpadmin /etc/group
> lpadmin:x:102:richard
> 
> so, yes.  But it wont let me change anything, either as me or as root.

This sounds more like your cupsd.conf  is wrong? But
you said earlier it compares with another machine that works?

Not an expert on cups, so I probably don't have much more to contribute
here.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: printer won't stop

[CW Harris]

On Fri, Oct 15, 2004 at 01:26:09AM +0100, Richard Lyons wrote:
> On Thursday 14 October 2004 20:10, Jacob S wrote:
> [...]
> > If I'm not mistaken, you can also delete jobs in the queue using cups'
> > web interface (if not a command line tool, as well). 
> > 
> > In your web browser, go to http://localhost:631, then click the
> > "Manage Printers" link, click on the printer with the jobs to be
> > canceled and then delete the appropriate jobs from the queue. (Looks
> > like you can click the "Manage Jobs" link as well, instead of "Manage
> > Printers".) 
> 
> Yes, that is correct.  Use Manage jobs for jobs and Manage printers for 
> printers - you can modify, configure, or even delete and install 
> printers from the web interface.  This is how I usually work.

There is also /usr/bin/cancel (cupsys-client) and /usr/bin/lprm
(cupsys-bsd) if you just want to remove jobs.

> 
> But, a small warning, you do need to have the access permissions set up 
> right.  I have one box where cups won't let me change anything and I 
> cannot see what I have configured wrong.  

Are you a member of the SystemGroup (default lpadmin) defined in cupsd.conf?

> 
> A great advantage of the web interface is you don't need X running to 
> use it -- just access it via lynx (or other text-i/f browser).
> 
> -- 
> richard

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Limiting CUPS "Completed Jobs" Listing

[CW Harris]

On Wed, Sep 29, 2004 at 05:31:28PM +0700, I_MY_MLs_j wrote:
> All -
[...]
> On the testing system, I have ~90 entries in the "Completed Jobs" page of the web 
> interface, dating from 
> June to today (September).  It takes about 5 pages to list them all.  I don't care 
> about any job that was 
> completed months ago so would like to limit the number of jobs listed in the 
> "Completed Jobs" page to 
> about a page or so (~ 10-20 jobs), or perhaps list only those jobs completed in the 
> 15-30 days.
> 
> I looked in the docs and it seemed that "MaxPrinterHistory" may do the job, so I 
> changed 
> /etc/cups/cupsd.conf, MaxPrinterHistory to 50.  After a reboot, I saw no change to 
> the number of jobs 
> listed.
> 
> I don't see a way to limit the number of jobs listed, or a means to purge the job 
> listing.  Can someone 
> help?

I think this is what you want:

# MaxJobs: maximum number of jobs to keep in memory (active and completed.)
# Default is 500; the value 0 is used for no limit.
#
#MaxJobs 500
MaxJobs 50

> 
> -TIA

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: opera and java

[CW Harris]

On Wed, Sep 29, 2004 at 05:02:45PM -0600, CW Harris wrote:
> On Wed, Sep 29, 2004 at 11:03:26AM -0400, Christopher Judd wrote:
[...]
> Also the opera web page has some troubleshooting info for java:
> 
> http://www.opera.com/support/search/supsearch.dml?name=java&maxhits=15&platform=linux

Also, as mentioned on this web site you must have java-common installed (I
don't think the opera .debs depend on java-common, I guess so it's not
required if you don't want java.)


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: opera and java

[CW Harris]

On Wed, Sep 29, 2004 at 11:03:26AM -0400, Christopher Judd wrote:
> Hi, all,
> 
>  I was wondering how to use opera with java.  When starting opera
> with the -debugjava option, I get the following message:
> 
> opera: [java] There seems to be a preloaded version of Xt.
>There is a workaround for this problem in the opera
>startup script.  If that workaround fails, opera will
>most likely crash every time it tries to use Java.
>The workaround seems to have failed.  Java will be disabled.
>Technical explanation:
>There is a problem with the order of loading Xt and
>Java.  If Xt is loaded before libawt (part of Java),
>Java will crash when it tries to access the screen.
>The workaround is based on using LD_PRELOAD to load
>libawt.so first.
> 
> opera: [java] Disabling java due to potential problems. If you know
>what you are doing, you can set the environment variable
>OPERA_FORCE_JAVA_ENABLED to '1' to override this.
>The actual problems should be described above.
> 
> Is there any way to make opera work properly with java in Debian
> (sarge)?

It works for me.  Here using opera-static 7.53 and blackdown j2re 1.4
Have used opera 6.x and earlier j2re packages.

It would seem that the opera startup script is not finding your libawt.
I think in an earlier version (6.something) I edited the script to
include my java path because it just wasn't finding it. You can look at
/usr/bin/opera (just a shell script) to see what is going on.

Alternately, you can set the Javapath in /etc/opera6rc. Something like:
; Put any default settings here that are overridable by users

[User Prefs]
Javapath=/usr/lib/j2se/1.3/jre/lib/i386/
   This should be where libawt.so is located.

Also the opera web page has some troubleshooting info for java:

http://www.opera.com/support/search/supsearch.dml?name=java&maxhits=15&platform=linux

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Can only print one document

[CW Harris]

On Tue, Sep 28, 2004 at 11:14:56AM +1000, Brendan Simon wrote:
> I'm running Debian testing (kernel-image-2.4.27-1-k7) with CUPS (cupsys 
> and cupsys-bsd 1.1.20final) and I have an HP LaserJet 3330.
> 
> I have had it working with earlier kernels and cups but now it only 
> prints one document and does not print subsequent documents.  If I use 

Is this with *any* file, or just with *some* files?  I describe a
similar issue below, but not for any file.

I am currently running a 2.6.4 kernel on the box with the HP 3330
(connected over usb), but I don't recall any differences with kernel
2.4.xx, and I am now running cupsys 1.1.20final+rc1-3 but I *think* the
issue was the same with 1.1.20final.

> "lpr file.ps" it prints fine and the queues are empty.  If I run "lpr 
> file.ps" again the printer says "Printing Document" but nothing happens.

Sounds similar to a problem I had/have.  Mine would eventually print a
timeout error from the PS engine in the printer (don't recall the exact
message now--but it was suggestive of a stop in the data stream to the
printer).  I think I would also end up with a "gs" process hanging.

It seemed to be a problem with gs processing the file, since if I grab
the postscript file from the print queue and view it with gv it
indicates errors processing it.

Personally, I have found that the PPD file that came with the printer
was the worst, seemed really slow to process a page in the printer and
was prone to these processing errors.

The one I currently use and seems pretty decent is the 3200m PPD from
the Foomatic/gimpprint (foomatic-db package I think).

(Although I think I still occasionally see one get stuck---usually
 someone printing some pages from a windows box, but that's probably not
 relevant)

> 
> Any ideas?  I've racked my brain and searched the internet but nothing 
> has worked so far.  I'm at my wits end.

If you find out more, I'd be interested to know.  For sometime now, mine
has been working well enough that I haven't looked at it any more.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [solved] Re: mounting iPod with USB

[CW Harris]

On Tue, Sep 28, 2004 at 12:16:52PM -0400, Nori Heikkinen wrote:
> on Mon, 27 Sep 2004 11:43:56PM -0700, Brian Nelson insinuated:
> > On Mon, Sep 27, 2004 at 12:34:25PM -0400, Nori Heikkinen wrote:
[...]
> > > 
> > > all appears to work now.  my files are transferring over
> > > slow-as-molassses USB 1 ...
> > 
> > modprobe ehci_hcd?
> 
> i already have uhci_hcd ... i though they were mutually exclusive?

uhci/ohci (usb 1.0) mutually exclusive. ehci is usb 2.0
>From kernel config help:

config USB_EHCI_HCD
tristate "EHCI HCD (USB 2.0) support"
depends on USB
---help---
  The Enhanced Host Controller Interface (EHCI) is standard for USB 2.0
  "high speed" (480 Mbit/sec, 60 Mbyte/sec) host controller hardware.
  If your USB host controller supports USB 2.0, you will likely want to
  configure this Host Controller Driver.  At this writing, the primary
  implementation of EHCI is a chip from NEC, widely available in add-on
  PCI cards, but implementations are in the works from other vendors
  including Intel and Philips.  Motherboard support is appearing.

  EHCI controllers are packaged with "companion" host controllers (OHCI
  or UHCI) to handle USB 1.1 devices connected to root hub ports.  Ports
  will connect to EHCI if it the device is high speed, otherwise they
  connect to a companion controller.  If you configure EHCI, you should
  probably configure the OHCI (for NEC and some other vendors) USB Host
  Controller Driver or UHCI (for Via motherboards) Host Controller
  Driver too.

  You may want to read .

> 
> 



-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: WEB-Browser with links in evolution ??

[CW Harris]

On Wed, Sep 08, 2004 at 10:05:29PM +0200, Richard Palfalvi wrote:
> Hi :-)
> 
> I am using EVOLUTION as my standard-email-programm on Debian-Sarge. It
> is working fine :-)
> 
> I just would like if somebody could tell me where I can change the
> standard-web-browser evolution uses evertime when I click on a link in
> an email - it starts EPIPHANY which does not work very properly in this
> version and I want to use FIREFOX instead. 
> 
> I couldnt find any point to change this in the preferences-tool of
> evolution nor a "conf-file" in the evolution-directory where to change
> this.
> 
> Any hints upon this?

I think you want to change the x-www-browser setting
(man update-alternatives).

I'm sure others here more knowledgeable than I will correct me if I'm
wrong.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Time

[CW Harris]

On Wed, Sep 08, 2004 at 03:30:49PM -0600, CW Harris wrote:
> 
> I believe pool.ntp.org has been set aside for this use.
> 
I should add "if you don't have a more local/appropriate source such as
a timeserver at your ISP."  See http:/www.pool.ntp.org/#use for more info.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Time

[CW Harris]

On Wed, Sep 08, 2004 at 03:35:13PM -0400, Michael Marsh wrote:
> On Wed, 8 Sep 2004 21:12:34 +0200, Paul Akkermans
> <[EMAIL PROTECTED]> wrote:
> > I have just a simple problem. During installation of my Debian system I did not 
> > set the correct time on my Debian system. Does anybody know how I can do this 
> > without having to install the entire Debian system again? 
> 
> If you selected the wrong timezone, try tzselect .
> If you set the system clock incorrectly, you can fix it by sync'ing to
> a time server:
> 
> /usr/sbin/rdate -s your.favorite.ntp.server
> /sbin/hwclock --systohc
> 
> Search Google for "public NTP servers" to get a list of potential
> replacements for
> "your.favorite.ntp.server".  Only use Stratum-2 servers, though.

I believe pool.ntp.org has been set aside for this use.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Dual Boot Problem

[CW Harris]

On Wed, Sep 08, 2004 at 02:41:34PM -0400, Patrick Moroney wrote:
> CW Harris wrote:
> >On Wed, Sep 08, 2004 at 01:28:14PM -0400, Patrick Moroney wrote:
> >>CW Harris wrote:
[...]

> >>>Why not just boot from a rescue disk (I think most/all install disks can
> >>>be used as rescue disks also--check the boot options) and fix the lilo
> >>>install so you can boot windows also?  This at least fixes your boot
> >>>options until you figure out the debian install problem.
> >>>
> >>>
> >>> 
> >>>
> >>--Will do with regard to the rescue option  - I'll look at the boot 
> >>options.  But pretty sure I tried this . . .


You might also look over the errata for the installer (although I didn't
see anything quickly related to your problem---but it may be as simple
as booting the netinstall CD and booting the CD kernel as:

linux root=/dev/hda2(IIRC this was your root - change as needed)
(or linux26 if using the 2.6 kernel)

(I have not played around a lot with the netinstall CD, but I know there
 are places where you can get to a shell and mount your partition to see
 what your lilo.conf looks like.)

Or use Knoppix or other CD runnable disks as rescue disk for a more
complete environment to troubleshoot.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Dual Boot Problem

[CW Harris]

On Wed, Sep 08, 2004 at 01:28:14PM -0400, Patrick Moroney wrote:
> CW Harris wrote:
> >On Wed, Sep 08, 2004 at 10:18:45AM -0400, Patrick Moroney wrote:
> >>I'm trying to install Debian Sarge stable as a second operating 
> >>   
> >Sarge is not yet "stable"---still "testing"
> >
> 
> --Understood - but I've successfully installed other machines.

Not what I meant.  At the moment, stable = Woody, testing = Sarge, so I
still don't know what you mean by "Sarge stable."  Do you mean you are
using the new Sarge netinstall CD to install "stable?"

> >>system on a 200 gig drive and not having a lot of success.  The 
> >>installation goes fine; I install Lilo in the MBR, the system 
> >>goes for a reboot and then I get the following errors upon reboot:
> >>
> >>
> >>request_module[block-major-3]: Root fs not mounted
> >>VFS: Cannot open root device "302" or 03:02
> >>Please append a correct "root=" boot option
> >>Kernel panic: VFS Unable to mount root fs on 03:02
> >>   
> >>
> >
> >What kernel?
> >Self compiled or from installation?
> >What are your lilo/grub lines to the kernel?
> >
> > 
> >
> --Whatever kernel with Sarge stable.

What exactly are you using to install Sarge? CD from ??? Sarge
netinstall CD (and what date)?

> --The kernel is from the installation - no customization
> --As far as lilo lines, I never get to boot into linux to see the lilo 
> lines.  I really don't want to use grub.

Why not?  Although grub is strange to use at first (if you are already
used to linux device names and lilo), it does have the advantage that
you can dynamically change what it is doing at boot time.  This can be
convenient to fix some boot problems.

Anyway, that is not really the topic, but it might be something nice to
play around with later.

[...]
> >>Should I make the system so that it will only boot into debian 
> >>with a floppy?  Any suggestions would be greatly appreciated.

It is sounding like a boot floppy might at least help you boot into
debian to finish the installation, but I wouldn't make that a permanent
situation.  The booting problems can be fixed.

> >
> >Why not just boot from a rescue disk (I think most/all install disks can
> >be used as rescue disks also--check the boot options) and fix the lilo
> >install so you can boot windows also?  This at least fixes your boot
> >options until you figure out the debian install problem.
> > 
> >
> 
> --Will do with regard to the rescue option  - I'll look at the boot 
> options.  But pretty sure I tried this . . .
> 
> >Alternately, I think XP allows you to create a boot menu that will boot
> >another OS in its own partition (never tried this so can't vouch for how
> >well it works).
> >
> > 
> >
> --I basically would trust anything windoze - I'd rather Lilo control the 
> booting.
> 
:> I second that. Although I have set up dual boot where windows has
control of the MBR and installed lilo to my / partition (marking it as
the bootable partition) to avoid having windows trash my lilo
installation when it crashed/fixed things or upgraded (I never bothered
to figure out what circumstances lead windows to re-write the mbr
).  Rescue CD/boot floppy works just as well.

P.S. No need to CC me, I read the list.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Dual Boot Problem

[CW Harris]

On Wed, Sep 08, 2004 at 10:18:45AM -0400, Patrick Moroney wrote:
> I'm trying to install Debian Sarge stable as a second operating 

Sarge is not yet "stable"---still "testing"

> system on a 200 gig drive and not having a lot of success.  The 
> installation goes fine; I install Lilo in the MBR, the system 
> goes for a reboot and then I get the following errors upon reboot:
> 
> 
> request_module[block-major-3]: Root fs not mounted
> VFS: Cannot open root device "302" or 03:02
> Please append a correct "root=" boot option
> Kernel panic: VFS Unable to mount root fs on 03:02

What kernel?
Self compiled or from installation?
What are your lilo/grub lines to the kernel?

> 
> 
> Please note, the reboot is unsuccessful, I never get to the step 
> where it asks to set up mail or create a root password.  I just 
> get the above message.  I can't scroll up to see any of the other 
> messages.
> 
> To recover and make the system so that I can at least boot back 
> into windows is using a dos boot disk, do a "fdisk /mbr" - clear 
> out the MBR, and then using debian's fdisk, blow away debian 
  
Ok, I understand fixing the MBR to boot XP again, but why rm debian?
I don't understand this problem.  It is sufficient to just restore the
MBR and mark windows as bootable.

> partions, and mark the windows NTFS as bootable.  This has worked 
> many times being as I've tried Lilo, Grub, and GAG as boot 
> loaders - all successful.
> 
> The machine is x86 box with ASUS motherboard, AMD 1800 CPU, gig 
> of RAM, ATI video card, and Audigy sound card.  Windows occupies 
> the first 35 gig of the 200 gig board.
> 
> Should I make the system so that it will only boot into debian 
> with a floppy?  Any suggestions would be greatly appreciated.

Why not just boot from a rescue disk (I think most/all install disks can
be used as rescue disks also--check the boot options) and fix the lilo
install so you can boot windows also?  This at least fixes your boot
options until you figure out the debian install problem.

Alternately, I think XP allows you to create a boot menu that will boot
another OS in its own partition (never tried this so can't vouch for how
well it works).

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Samba upgrade to 3.06 is causing problems

[CW Harris]

On Mon, Sep 06, 2004 at 09:22:27AM +0200, Remon Vos wrote:
> I already tried different security settings... none of them worked for me.
> My log.[sn]mbd files do not show any errors. And testparm runs ok too.

You may need to increase the "log level", I think you get 0 if
nothing is specified.  Maybe try 2, or 3.

> 
> But... When I mount the shared directory on my server on another linux box
> (which has not been upgraded yet, and thus runs on an older (working) samba
> version), I can browse directories (like on my Windows box). But when I try
> to open a file I get the following error:
> 
> smb_get_length: Invalid NBT packet
> 
> Does anyone know what this is?

Sorry, the only quick references to this I saw were related to encrypted
password problems.  But if you can browse the directory you should be
past the auth stage.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Samba upgrade to 3.06 is causing problems

[CW Harris]

On Fri, Sep 03, 2004 at 02:58:34PM +0200, Remon Vos wrote:
> Hi,
> 
> I upgraded my linux box using dselect.
> Also samba 3.06 was installed in this process.
> Since then i'm unable to reach files from my windows machine on my linux
> server.
> I'm still able to open directories, but not any of the files. My windows
> machine locks up until I shut down samba on the linux server.

You mean you can browse the shares but not the files, or can you
actually browse directories in the shares also?  Sounds like a
permissions issue, but it seems like a strange change.

What authentication, security=user, security=share?

> 
> Any ideas?

Do you have any errors in your samba/log.[sn]mb files?
Does "testparm" show any problems with your smb.conf file?

FWIW, I just upgraded and did not experience this (the only prob I
experienced was the not well enough documented add "smb ports = 139" if
you are using netbios aliases to appear as multiple hosts).

I see on the samba site that a version 3.0.7 is coming out soon to fix
some noticeable problems with 3.0.6, but they didn't exactly sound like
your problem.  Maybe you can search the samba lists (I only did a quick
search) for more help?

> 
> Regards, Remon Vos

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Once a month I get this gibberish.

[CW Harris]

On Wed, Sep 01, 2004 at 12:13:54PM -0400, Hendrik Boom wrote:
> Once a month, the cron daemon sends me all this gibberish.
> Is there some point to it?
> Should I be watchin gout for it?
> What is it, anyway?
> 
> On Wed, Sep 01, 2004 at 06:52:55AM -0400, Cron Daemon wrote:
> > /etc/cron.monthly/scrollkeeper:
> > scrollkeeper-update: /usr/local/share/omf: No such file or directory
> > Registering /usr/share/omf/gnome-core/gnome-feedback-C.omf
> > Registering /usr/share/omf/gnome-core/fdl-C.omf
> > Registering /usr/share/omf/gnome-core/gnome-feedback-no.omf
> > Registering /usr/share/omf/gnome-core/gpl-C.omf
> > Registering /usr/share/omf/gnome-core/lgpl-C.omf
> > Registering /usr/share/omf/gnome-core/panel-C.omf
> > Registering /usr/share/omf/gnome-core/panel-de.omf
> > Registering /usr/share/omf/gnome-core/panel-es.omf
> 
> and many many more similar lines.
> 
> -- hendrik

In /etc/cron.monthly/scrollkeeper you can change
scrollkeeper-rebuilddb
to
scrollkeeper-rebuilddb -q

The "-q" option will suppress all but the serious errors.
(man scrollkeeper-rebuilddb)

Note that this option is default on newer scrollkeeper installs.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Getting better screen refresh rate in X?

[CW Harris]

On Wed, Aug 25, 2004 at 09:48:33PM +0100, Thomas Adam wrote:
> On Thu, Aug 26, 2004 at 06:42:33AM +1000, Paul Gear wrote:
> > As you can see, the definitions are the same in terms of specs.  I'm
> > guessing there is a database of mode definitions somewhere and i need to
> > copy the relevant definition from my RHL9 database, but i can't for the
> > life of me find where it is.  Any clues?  Or am i barking up the wrong tree?
> 
> Possibly not. Have a look at using 'xidtune' to fine-tune your settings a

Typo... Thomas meant 'xvidtune' I'm sure.

One other thought I had, is there any difference in the xserver 
you are using?  Same one?  Same version?


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: building or obtaining debian non-free CDs ? - margins

[CW Harris]

On Tue, Aug 24, 2004 at 08:47:57PM -0700, Alvin Oga wrote:
> 
[...]

> - if "/var/cache/apt/archives" fits on one or 2 CD's ... i'm thinking
>   that's all the packages that is needed to clone that box ??
>   - is the archives directory the same structure as the
>   main *.deb repository ( i say/think: nope, it's not )

Kind of jumping in the middle here, but if you are looking to get a
repository structure for the packages you use...

I *think* if you use apt-proxy you end up with a structure like the
normal repository:

$ find /var/cache/apt-proxy -type d -maxdepth 5 \
| sed -e 's[/var/cache/apt-proxy/[['

debian
debian/dists
debian/dists/stable
debian/dists/stable/contrib
debian/dists/stable/contrib/binary-i386
debian/dists/stable/main
debian/dists/stable/main/binary-i386
debian/dists/stable/non-free
debian/dists/stable/non-free/binary-i386
debian/dists/sarge
debian/dists/sarge/contrib
debian/dists/sarge/contrib/binary-i386
debian/dists/sarge/main
debian/dists/sarge/main/installer-i386
debian/dists/sarge/main/binary-i386
debian/dists/sarge/main/installer-powerpc
debian/dists/sarge/non-free
debian/dists/sarge/non-free/binary-i386
debian/.apt-proxy-filelists
debian/pool
debian/pool/main
debian/pool/main/a
debian/pool/main/a/aa3d
debian/pool/main/a/anacron
debian/pool/main/a/apt-proxy
[...]

> 
>   - make an itty-bitty script to go thru apt/archives
>   and install um in order ... :-)
> 
> - fun stuff to "get around" the debian way :-0
> 
> c ya
> alvin
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Kernel compilation question

[CW Harris]

On Tue, Aug 24, 2004 at 05:45:02PM +0200, Magnus Therning wrote:
> On Tue, Aug 24, 2004 at 12:23:57PM +0100, Thomas Adam wrote:
> >On Tue, Aug 24, 2004 at 09:21:59AM +0200, Magnus Therning wrote:
> >> Howcome? (Is it related to '.config support' being compiled into the
> >> standard Debian kernels?)
> >> How do prevent it from happening? (I unmounted /boot, but that that's
> >> less than elegant.)
> >
> >You copy whatever config you want to /path/to/kernel/src/.config, and
> >'make oldconfig'.
> 
> Well, that wasn't really what I was looking for. I don't have a
> '.config' I want to use, I want the defaults! This is what I do:

make defconfig

this will generate a .config file with all default answers

"make help" for other options

HTH


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Where is libXp.so.6? [was: no subject]

[CW Harris]

On Fri, Aug 13, 2004 at 11:32:53AM +1000, Keith Davidson wrote:
> Hi,
> 
>  
> 
> Any ideas where I can find libXp.so.6 for debian.  I have looked in
> www.packages.debian.org  , but can't
> find it.

The web site lets you seach for filenames in packages too.  You didn't
say which version Debian you are using.  I have the above file in
libxp6 from Sarge, packages shows it in xlibs for Woody, not sure about
sid.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Packages needed for Cups Foomatic printer

[CW Harris]

On Wed, Aug 11, 2004 at 07:51:33PM -0400, Tong wrote:
> Hi, 
> 
> I'm trying to install my HP PSC 1210 usb printer. I was able to do it in
> previous Debian installation, but this time, I am having difficulties
> doing it...

Have you checked out linuxprinting.org? (See for example,
http://www.linuxprinting.org/show_printer.cgi?recnum=HP-PSC_1210
for comments on how others have used the printer.)  I only skimmed the
comments, since I'm not sure what kernel, etc. you are using.

> 
> When I used the web-based cups printer setup, 
> 
> I can't find the driver "Foomatic + hpijs (recommended)" when choosing a
> driver (previously I can). 
> 
> When I used foomatic-gui, there is no option for me to choose an USB
> printer (previously I can).

I have only used a different model HP using the hpoj package to support
scanning also (in which case you use the HP ptal interface to the printer
rather than plain usb), so I'm not sure about other differences without
hpoj support.

Sorry not much help.
Good luck.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Keep jobs on queue

[CW Harris]

On Fri, Aug 06, 2004 at 09:30:26PM -0200, gilbertonunes wrote:
> Hi for all...
> 
>Some body knows how can I hold all jobs in a print queue, using cups?
>Thanks.

You mean like "Stop printer" from the web administration page?

You can also use the disable/enable commands.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: version mess, help needed

[CW Harris]

Not an expert in apt management, so...

On Wed, Aug 04, 2004 at 11:56:05AM -0300, Roberto Winter wrote:
> Hi again,
> doesn't the output from "apt-cache policy" guarantee that (see my
> first message)?

I think it will *still* not downgrade a package that is already
installed.  Its primary purpose is for assisting in upgrades.
Downgrades are not really supported.

That said, I think if you ask for a specific version it will downgrade
packages as necessary (try with "-s" simulate switch first).

man apt-get for info on specifying particular versions (can search for
"downgrade" to get in the right part of the document).

I also understand that aptitude is much nicer for really trying to
control your installation, and is much smarter that apt-get in resolving
problems, but I'm still learning it, so I'm not sure if it would be
easier for this or not. (or synaptic for gtk based, since you may have a
broken curses system :)

[...]
> PS: Again, I am still not on the list, so I ask you to CC any answers to me
> too, thanks.

done.  You can also browse the list at http://lists.debian.org/debian-user

HTH


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Synchronizing between Windows and Linux: character encoding problem

[CW Harris]

On Thu, Jul 29, 2004 at 04:39:27PM +0200, Debian Users wrote:
> Hello there!
> 
> I am trying to synchronize files between a laptop running either Windows
> XP or Debian and a Debian file server. I have tried rsync and unison on
> the Linux install of the laptop and rsync (under Cygwin) on the Windows
> install, but both were unable to transfer files with special characters
> (here: German umlauts) correctly.
> 
> The situation is as follows: the file server also runs a samba server.
> Files with umlauts that are created under Windows can be transferred to
> the file server without problems. When the share on the server is viewed
> with the Windows Explorer, these files have all the umlauts intact.
> 
>   Example (I hope you can read these at all!):
> 
>   LOCAL FILE (Windows)REMOTE FILE (Debian, viewed from Windows)
>   test-äöü-from-win.txt   test-äöü-from-win.txt
> 
> When viewed locally on the file server, the umlauts are missing:
> 
>   file-server:~$ ls test-*
>   test-???-from-win.txt

Not a samba expert but it sounds like some codepage mismatching. You
might look through this reference if you haven't already:
http://us3.samba.org/samba/docs/man/Samba-HOWTO-Collection/unicode.html

Also, if you aren't using samba 3.0 it might benefit you to upgrade. I
understand it has much better support for internationalisation.  It is
available at www.backports.org if you need it for Debian stable.

You might also google for "unicode", "codepage", "international" on
samba.org for more help with earlier samba versions.

As I understand it, samba 2.x required synchronization between the
server and the clients on which codepages (character sets) were in use.

HTH


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: I/O Errors

[CW Harris]

On Fri, Jul 30, 2004 at 09:04:07AM +1200, Steven Jones wrote:
> see if the manufacturer has a disk diagnostic program available, but I would suggest 
> the disk is stuffed.
> 
I could be wrong, but isn't dev 0b:00  /dev/scd0 ?  So it's having
trouble reading a scsi cdrom?


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: DNS

[CW Harris]

On Tue, Jul 27, 2004 at 08:05:28AM -0600, Marvin Gerardo Aguero Salazar wrote:
> Thanks to all the folks who replied to my original posting. I really do
> appreciate it. What I was looking for was the /etc/resolv.conf approach.
> 
> Unfortunately, I just realized that's not what I needed. :-)
> 
> I thought the windows boxes of my co-worker's were translating names to
> ip addresses for their DNS set up.
> 
> I just realized it was not their DNS but rather the WINS server the one
> that is translating from names to IP addresses.

Perhaps the winbind package will be useful? (No experience with it
though.)

apt-cache show winbind

Package: winbind
Priority: optional
[...]
Description: service to resolve user and group information from Windows NT servers
 This package provides the winbindd daemon, which provides a
 service for the Name Service Switch capability that is present
 in most modern C libraries (like the GNU C Library - glibc.)
 .
 The service provided by winbindd is called `winbind' and
 can be used to resolve user and group information from a
 Windows NT server. The service can also provide authentication
 services via an associated PAM module.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Cups can't start with error 98

[CW Harris]

I believe I'm starting to know more about CUPS than I really want to.
If the following solves your problem, we both need to RTFM better ;-)

On Tue, Jul 27, 2004 at 12:09:56PM +1000, Tim Connors wrote:
[...]
> Nothing interesting, but I did notice one more line of output that I
> didn't before:
> 
> I [27/Jul/2004:12:04:46 +1000] Listening to 0:631
> I [27/Jul/2004:12:04:46 +1000] Listening to c0a80102:631

I believe this is a configuration error in your cupsd.conf
I can replicate this error with the following:
Port 631
Listen 192.168.1.2:631

I think the Port directive binds to *any* address on 631.  The second
directive tries to bind 192.168.1.2 on port 631, but that is already
taken with the *any* address on 631.  So I *think* you can either let it
bind to *any* address and not have a listen directive.  Or remove the
port directive and list the address/ports you want it to listen on (you
may need to allow it to listen on 127.0.0.1)

> I [27/Jul/2004:12:04:46 +1000] Sending browsing info to c0a801ff:631
> I [27/Jul/2004:12:04:46 +1000] Loaded configuration file "/etc/cups/cupsd.conf"
> I [27/Jul/2004:12:04:46 +1000] Configured for up to 100 clients.
> I [27/Jul/2004:12:04:46 +1000] Allowing up to 100 client connections per host.
> I [27/Jul/2004:12:04:46 +1000] Full reload is required.
> I [27/Jul/2004:12:04:47 +1000] LoadPPDs: Read "/etc/cups/ppds.dat", 2402 PPDs...
> I [27/Jul/2004:12:04:49 +1000] LoadPPDs: Wrote "/etc/cups/ppds.dat", 2402 PPDs...
> I [27/Jul/2004:12:04:49 +1000] Full reload complete.
> E [27/Jul/2004:12:04:49 +1000] StartListening: Unable to bind socket for address 
> c0a80102:631 - Addre
> ss already in use.

For what it's worth, a successful start in my logs looks the same up to
the "StartListening: Unable ..."

> 
> The start line says listening. Has it opened the port for listen yet,
> or does it defer the actual open until the E line, where it fails? Or
> is it indeed trying to open the port twice?

The latter I believe.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Cups can't start with error 98

[CW Harris]

I don't see any answers so I'll try (not really good with CUPS though):

On Sat, Jul 24, 2004 at 05:52:29PM +1000, Tim Connors wrote:
> I haven't used cups in a while, and tried to today. It fails at
> startup, with error code 98:
> E [24/Jul/2004:17:39:06 +1000] StartListening: Unable to bind socket for address 
> c0a80102:631 - Address already in use.
> 
> c0a80102==192.168.1.2==the box I am trying to start cups from
> 
> That was fair enough the first time, because somehow lpr was installed
> at the same time, so I stop and removed it. The error stays, and I
> have done a netstat -au, and found nothing is bound to port 631. lsof

What about tcp (netstat -at)?

> agrees with me. There is no lp* or cups* processes running.
> 
> I googled, and found that someone found a problem with nfs taking the
> port randomly, so I stopped the nfs processes.

A quick look and I don't see a similar bug (1 unreproducible from a year
ago is all).  Any more info if you set LogLevel to "debug" in
/etc/cups/cupsd.conf?

> 
> I waited 5 minutes, because I rememeber something about the kernel
> sometimes keeping ports open for a little while despite the exit of
> the owner process.
> 
> I partially upgraded sid - so now cups* and libcupsys are up to date.

Sorry, I don't have any current experience with sid.

> 
> I really can't work out why I keep getting this message.
> 
> And ideas?
> 
> -- 
> TimC -- http://astronomy.swin.edu.au/staff/tconnors/
> My cats are forbidden from walking on my computer keyboard on the desk
> when I'm asdfjjhhkl;ljfd.;oier' puyykmm4hbdm9lo9j USING IT. 
  
Just like a cat. :->

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: DNS

[CW Harris]

On Mon, Jul 26, 2004 at 03:54:28PM -0600, Marvin Gerardo Aguero Salazar wrote:
> Hi guys,
> 
> How can I specify the DNS server in debian/linux?
> 
> Just to clarify, I don't want to set up a DNS Server (which is what the
> DNS howto talks about). I just want to translate names to IPs.

You probably just want to add a nameserver line (up to 3) to your
/etc/resolv.conf e.g:
nameserver   ip.of.your.server

man resolv.conf for more info on this file.

You might also read this:
http://www.debian.org/doc/manuals/reference/ch-gateway.en.html#s-dns-resolver

for more info on the possible setup.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: modules not found after kernel recompile

[CW Harris]

On Mon, Jul 26, 2004 at 09:10:27PM +0200, [EMAIL PROTECTED] wrote:
> 
> On Mon, 26 Jul 2004 10:42:37 -0600, "CW Harris"
> <[EMAIL PROTECTED]> said:
[...]
> > 
> > Is this a typo? or are you using kernel 2.6.7 with an append-to-version
> > of 1?  I would recommend you use "-" (hyphen) before numeric
> > append-to-version so the kernel version does not look wrong (i.e. there
> > is no kernel 2.6.71).
> 
> That would be correct. I put EXTRAVERSION=1 in the top-level Makefile.
> So that 2.6.7 becomes 2.6.71. But, as you observed, it makes no 
> difference in this matter.

Ah yes, I realized afterwards I was assuming you used the make-kpkg
command from the kernel-package debian package (nice way to create a
.deb of the kernel with all modules).

> 
> > What are the first errors you get and the text above them?
> 
> I don't get any errors, aside from the ones mentioned about 
> 'FATAL: module  not found'. Not anymore, since I 
> renamed /etc/modules to /etc/modules.old and that one's empty.

So it sounds like you had an old /etc/modules with modules corresponding
to a different kernel and that was causing the error messages.

> If I probe for nonexisting modules, the old FATAL msg. appears.
> Nothing strange about that, though.
> 
> UPDATE:
> Tried running 'modprobe ', where 'module name' is one 
> of the module names in /lib/modules/2.6.71. Seems to work, but is this 
> a permanent solution? A reboot will of course tell me if it's not...

No, modules loaded manually will not be loaded upon reboot. You can use
/etc/modules to have them loaded at boot time.

AFAIK /etc/modules is still a reasonable place to add modules you want
that are not automatically loaded, although a lot of new capability is
in the works in the 2.6 kernel and the udev project, which may make
other methods preferred.

Also, there may be cases where you simply need to add an appropriate
"alias" line in /etc/modprobe.d/aliases to get a module to automatically
load when needed.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: modules not found after kernel recompile

[CW Harris]

On Mon, Jul 26, 2004 at 12:15:29PM +0200, Inge Thorin Eidsæther wrote:
> 
> Hi guys, and thanks for all your help so far!
> 
> Observations:
> 
> - I misspelled moudule-init-tools as modutils-init-tools.
>   Sorry! Did all file searches by correct name, though.
> 
> - module-init-tools was installed BEFORE kernel recompile
> 
> - A non-initrd kernel image was made
> 
> - Documentation/post-halloween-2.6.txt
>   does not exist in my kernel source tree
>   (downloaded directly from www.kernel.org)
> 
> - I upgraded module-init-tools to last available version
>   (now: 3.1-pre5) after kernel recompile (was 3.1-pre2)
> 
> - I also did an 'apt-get update' and 'apt-get upgrade'
>   after recompile (could have been smart to do before?)
> 
> - There is no modprobe.conf anywhere on my system,
>   just the manpage for it and a demo script called
>   generate-modprobe.conf.gz in doc/module-init-tools/ex..

Reading the info with the 3.1 module-init-tools this seems to be okay.
Apparently modprobe.conf is *no longer* require (was with 3.0).

> 
> - Modules are found in correct location,that is:
>   /lib/modules/2.6.71, which corresponds to the output
>   of 'uname -r'

Is this a typo? or are you using kernel 2.6.7 with an append-to-version
of 1?  I would recommend you use "-" (hyphen) before numeric
append-to-version so the kernel version does not look wrong (i.e. there
is no kernel 2.6.71).

That said, I don't think this is your problem (since you say it matches
`uname -r`).

What are the first errors you get and the text above them?

> 
> - The file modules.dep in that directory looks OK.
> 
> _ I have removed modutils as suggested, since I am
>   now running a 2.6 kernel anyway.
> 
> - There's a modules.conf.old in /etc, but no modules.conf.
>   AFAIK I didn't rename it...so a script must have.
>   So moving entries from modules.conf (nonexisting)
>   into modprobe.conf (also nonexisting) or into the
>   modprobe.d directory (how? by copying them to the alias file,
>   which is already there, and contains - from what I can see -
>   about the same things?), confuses me a tad.
> 
>  - Getting a bit weary of this... :-\
> 
> Ideas, anyone? Do I have to go back to W*nd*ws?

I hope not :-)

> Does things really have to be this convoluted?

Maybe it helps to remember with W*nd*ws you wouldn't even have the
choice to compile your own kernel, or try the latest and greatest from
the developers.

> 
> Thanks for input!
> 
> best regards,
> 
> Inge Thorin Eidsæther
> blackwings NOSPAM at NOSPAM inbox dot com

Sorry I'm not more help -- I'm not up-to-date with the latest versions
of things, and unfortunately I can't bring my play-around computer down
right now to see what has changed.

Hopefully others will have better input.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: broken exim4-config

[CW Harris]

On Thu, Jul 22, 2004 at 04:08:41AM -0700, J. Nopr wrote:
> Hi all.  I am trying to set up exim4 to send mail out through my dsl
> provider's smarthost. 
> 
[...]
[text wrapped - *please* wrap your lines.]

> Now When I insatll exim4*, I immediatly get an error, starting at
> exim4-config.
> 
> [EMAIL PROTECTED] sudo dpkg --configure exim4-config  
>  [/etc/exim4]
> Setting up exim4-config (4.32-2) ...
> Error: Unsplit config selected and /etc/exim4/exim4.conf.template
>  missing ... exiting
> dpkg: error processing exim4-config (--configure):
>  subprocess post-installation script returned error exit status 1
> Errors were encountered while processing:
>  exim4-config
> 
> Can somebody send me the exim4.conf.template file?  Also, any good links

You already have it. It is part of the exim4-config package.  It sounds
like your installation is confused.  Perhaps you need to re-install
exim4-config.  Maybe try:
apt-get --reinstall install exim4-config

> for setting up spamassasin w/ exim4 as a smarthost client?
> 
> Thanks,
> -J. Nopr

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: modules not found after kernel recompile

[CW Harris]

On Wed, Jul 21, 2004 at 08:41:31PM +0200, Wim De Smet wrote:
> On Wed, 21 Jul 2004 12:17:40 -0600, CW Harris <[EMAIL PROTECTED]> wrote:
> > On Wed, Jul 21, 2004 at 07:55:27PM +0200, Wim De Smet wrote:
> > > Hi,
> > >
> > [...]
> > >
> > > You normally don't need a modprobe.conf, everything should be in
> > > /etc/modprobe.d. modprobe.conf is just an empty file on my system. I'm
> > 
> > Is this true?  Mine (a mostly Sarge with module-init-tools 3.0-pre2-1) has:
> > 
> > # This line loads the part of the modprobe configuration managed with
> > # update-modules(8) and built from the contents of /etc/modprobe.d/.
> > include /lib/modules/modprobe.conf
> > ^^
> > Which seems very important to me (not a GURU here).
> > 
> 
> module-init-tools in sarge is 3.1-pre5. Maybe it has something to do
> with the older version, or maybe you need this if you use udev or
   ^
Yes that was it.  I see in the changelog that between 3.0 and 3.1
modprobe.conf is no longer used since /etc/modprobe.d/* files are parsed
directly without the need for update-modules to process them.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: modules not found after kernel recompile

[CW Harris]

On Wed, Jul 21, 2004 at 07:55:27PM +0200, Wim De Smet wrote:
> Hi,
> 
[...]
> 
> You normally don't need a modprobe.conf, everything should be in
> /etc/modprobe.d. modprobe.conf is just an empty file on my system. I'm

Is this true?  Mine (a mostly Sarge with module-init-tools 3.0-pre2-1) has:

# This line loads the part of the modprobe configuration managed with
# update-modules(8) and built from the contents of /etc/modprobe.d/.
include /lib/modules/modprobe.conf
^^
Which seems very important to me (not a GURU here).


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Problems with mounting RAID5 array at boot

[CW Harris]

On Mon, Jul 19, 2004 at 12:50:10PM -0700, Matt Perry wrote:
> On Mon, 19 Jul 2004, CW Harris wrote:
> 
> > > This seems like a bug in the raid setup.  Maybe it should be set at a 
> > > lower number than checkfs by default so it starts sooner.
> > 
> > Which version?
> > Here I have:
> > 
> > $dpkg -l raidtools2
> > ii  raidtools2 1.00.3-5   Utilities to support 'new-style' RAID disks
> 
> 1.00.3-7 for me.  I'm running testing not stable.

Me too, just not quite updated:)  I think this is getting over my head
for understanding it, but it seems to me you are correct this is a bug.

According to the 3-7 changelog this closed bug 240384, but looking over
the bug, I don't think I understand how.  There was a new maintainer who
took over between 3-5 and 3-7.  I guess you could ask the maintainer,
unless someone else on the list has an answer I missed?

I haven't upgraded that package yet, but it would seem a lot of people
using raid would be affected by the checkfs occuring first.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: here is a LPI question,which is the answer?

[CW Harris]

On Sun, Jul 18, 2004 at 04:26:28PM +0800,  wrote:
> What is a reasonable command to install a set of Debian packages stored in directory 
> temp?
> 
> A. dpkg -l temp
> 
> B. dpkg -iGRE temp
> 
> C. dpkg -l temp/*.deb
> 
> D. dpkg -iGRE temp/*.deb
> 
> 
> 
> the answer is B
> 
> but  why not  BD ?
> 
> why is D wrong?

man dpkg (hint: read the part about --install)


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Problems with mounting RAID5 array at boot

[CW Harris]

On Sat, Jul 17, 2004 at 11:34:26PM -0700, Matt Perry wrote:
> On Sat, 17 Jul 2004, Matt Perry wrote:
> 
> > Any ideas on how to fix this?
> 
> I found a fix.  Turns out that the raid2 script needed to be run before 
> the checkfs script at startup, but they both were set to S30 in /etc/rcS.  
> I renamed S30raid2 to S29raid2 so it would start before S30checkfs and 
> everything is working fine.
> 
> This seems like a bug in the raid setup.  Maybe it should be set at a 
> lower number than checkfs by default so it starts sooner.

Which version?
Here I have:

$dpkg -l raidtools2
ii  raidtools2 1.00.3-5   Utilities to support 'new-style' RAID disks

$ls /etc/rcS.d/*raid
/etc/rcS.d/S25raid2


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: /proc/bus/usb directory is empty

[CW Harris]

On Thu, Jul 15, 2004 at 02:44:41PM -0400, Andrew Perrin wrote:
> Is the following in /etc/fstab?
> 
> none/proc/bus/usb   usbdevfs defaults   0   0

I thought it was "usbfs"... Hmmm looking in a 2.4.23
Doc../usb/proc_usb_info.txt document:

**NOTE**: If /proc/bus/usb appears empty, and a host controller
  driver has been linked, then you need to mount the
  filesystem.  Issue the command (as root):

  mount -t usbfs none /proc/bus/usb
  ...

**NOTE**: The filesystem has been renamed from "usbdevfs" to
  "usbfs", to reduce confusion with "devfs".  You may
  still see references to the older "usbdevfs" name.

I guess both work but it is migrating to usbfs. 

 
> 
> On Thu, 15 Jul 2004, Tom Brown wrote:
> 
> > Hi,
> >
> > I have two machines that are identical in hardware. One is running RH 8.0
> > (kernel 2.4.18) and the other is running Debian Woody (kernel 2.4.26). The
> > machine running RH 8.0 shows all the USB ports in /proc/bus/usb. However, on
> > the machine running Debian, the /proc/bus/usb directory is empty. I used the
> > same .config file to build the kernels on each machine. Why is the Debian
> > machine not recognizing the USB ports?
snip

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: any last words before I install kernel 2.6?

[CW Harris]

On Thu, Jul 15, 2004 at 03:02:05PM +0200, John van Spaandonk wrote:
> On Wednesday 14 July 2004 11:39, John Summerfield wrote:
> > John van Spaandonk wrote:
> > >On Tuesday 13 July 2004 01:29, Dan Jacobson wrote:
> > >>Any last words before I
> > >># apt-get install kernel-image-2.6-k7
> > >>on my home (sid) PC?
> > >>
> > >>Will things break that used to work in 2.4?
> > >
> > >I still stick with 2.4 for the following reason.
> > >
> > >I use two ethernet cards:
> > >eth0 is connected to the cable modem (it has to be this
> > >particular card because of the MAC address)

You can also use ifconfig to lie about the MAC address - I did this on 
one machine when I changed NICs and didn't want to hassle with my ISP
regarding the change.

E.g. add a
pre-up ifconfig eth0 hw ether xx:xx:xx:xx:xx:xx
stanza to your network/interfaces file.  (Of course this doesn't solve
the problem about them behaving differently in 2.4 vs. 2.6).

> > >eth1 is my home network
> > >2.6 reverses the names eth0 and eth1
> > >Apparently the PCI scanning order
> > >changed - right, why keep something the same if you can change it? :-)
snip

ifrename is available in Sarge & Sid (note I have not used it--but it
seems suited for this purpose).  I suppose udev might do this also, but
I haven't tried it.

$apt-cache show ifrename

Package: ifrename
Priority: extra
Section: net
Installed-Size: 60
Maintainer: Guus Sliepen <[EMAIL PROTECTED]>
Architecture: i386
Source: wireless-tools
Version: 26+27pre22-1
Depends: libc6 (>= 2.3.2.ds1-4), libiw27 (>= 26+27pre10)
Filename: pool/main/w/wireless-tools/ifrename_26+27pre22-1_i386.deb
Size: 37400
MD5sum: 74b13cffda8ff9a00e779cd94db6c490
Description: Rename network interfaces based on various static criteria
 Ifrename allow the user to decide what name a network interface will have.
 Ifrename can use a variety of selectors to specify how interface names match
 the network interfaces on the system, the most common selector is the
 interface MAC address.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Mirroring testing questions

[CW Harris]

On Sun, Jul 11, 2004 at 10:26:00AM +0100, Piers Kittel wrote:
> Hello all,
> 
> Am seriously considering mirroring testing for myself (faster updates, 
> installs etc due to having several computers) but due to only having a 
> 512k download ADSL, can't share it out, and want to check if the 
> bandwidth is good enough.  How much update (as in size) on average 
> occurs each day / each week?  Which would be best, mirror every night, 
> or mirror every week night?  Also reading the mirror help file, it 
> mentions each distro can be between 5 to 8GB each - how much space 
> should I reserve for testing?

Do you really need *all* of testing?  If you are only looking at having
up-to-date packages for a LAN, I would look at apt-proxy.  It allows you
to direct apt requests through 1 local machine, thus all the LAN
benefits from its cache of packages.
 
> 
> Thanks very much for your help in advance
> 
> Cheers - Piers

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: mounting floppy with fs auto doest recognize vfat

[CW Harris]

On Sat, Jul 10, 2004 at 05:54:30PM -0700, Karsten M. Self wrote:
> on Sat, Jun 26, 2004 at 12:30:53PM +, Stephen Patterson ([EMAIL PROTECTED]) 
> wrote:
> > On Sat, 26 Jun 2004 13:10:04 +0200, Micha Feigin wrote:
> > > When I try to mount a floppy with file system marked as auto in fstab
> > > it does recognize vfat file system (disks from windows), I have to
> > > specify vfat explicitly. Any way around that?
> > 
> > You can create /etc/filesystems, a text file with 1 filesystem per
> > line. This indicates which order to attempt different filesystems for
> > devices with fs-type auto.
> 
> ...and if you don't, you get autodetection in the order specified in
> /proc/filesystems.  In which msdos (8.3) precedes vfat (long filename
> support).

I think the order depends on your setup. msdos might always precede vfat
if both compiled in kernel, but if they are modules the order depends on
the order they are loaded (and if you only load vfat, msdos won't be in
/proc/filesystems).

> 
> Peace.
Indeed.



-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Is this a bug in mutt or a malformed email?

[CW Harris]

On Tue, Jun 29, 2004 at 10:43:06PM +0100, Pigeon wrote:
> A poster to a pigeon rescue group I'm on posts in
> multipart/alternative format, HTML and plain text.
> 
> Normally, when reading his emails, mutt (woody, 1.3.28-2) happily
> displays the plain text part and ignores the HTML. 
> 
> However, when he quotes one of my PGP-inline-signed posts, and doesn't
> delete the "-BEGIN PGP SIGNED MESSAGE-" line, mutt acts as if
> the plain text part doesn't exist and gives me the "text/html is
> unsupported (use 'v'..)" message.
> 
> If I edit the received email and add a space at the beginning of
> "-BEGIN PGP", or change one character of that line, mutt is
> then happy and displays the plain text part without problems.
> 
> I am not au fait with the fine details of email formatting, so I'm not
> sure if this is a bug in mutt's handling of a correctly-formed


Nor am I, however I did save your attachment to a file and could not
duplicate the behaviour you see, but this is with Mutt 1.5.5.1+cvs20040105i
on a mostly Sarge system. (Also, just tested on a home box with mutt
1.5.3i - no problem.)

Perhaps an upgrade of mutt would fix this?

I notice there is a backport of mutt 1.5.6 on backports.org.

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: iptables start on boot

[CW Harris]

On Tue, Jun 29, 2004 at 10:14:52AM -0400, Ralph Crongeyer wrote:
> John Summerfield wrote:
> 
> >[EMAIL PROTECTED] wrote:
> >
> >>I recently installed debian testing (sarge) on a clients machine and am
> >>trying to
> >>get the firewall to load on reboot.  AFAIK there was a
> >>/etc/init.d/iptables script in
> >>previous releases of debian but it doesn't seem to be there anymore.
> >>
> >>Is this correspond to others experiences?  Has this script been replaced
> >>with a
> >>different mechanism for starting iptables at boot time?
> >> 
> >>
> >The script has been superceded: I've not discovered by what: I'm not 
> >interested. The author clearly wasn't happy with it.

> Also, could someone give me a copy of the old script 
> "/etc/init.d/iptables". I need a way to save my rules, as we all do.

I haven't used this latest version, but the file list on
packages.debian.org for iptables 1.2.9-10 shows:

usr/share/doc/iptables/examples/oldinitdscript.gz

Maybe this is what you are looking for?


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: detaching a process from an ssh session ??

[CW Harris]

On Fri, Jun 25, 2004 at 08:32:19AM +1000, Zenaan Harkness wrote:
> On Fri, 2004-06-25 at 08:06, Damian Morris wrote:
> > to do it manually, you need to use one of the special ssh escape
> > codes. from my ssh man page:
> > 
> >Escape Characters
> >  ~.  Disconnect.
> > The one you want is "~." but make sure you enter it as the input on an
> > empty line.
> 
> I discovered it even needs to be the first characters typed on a line,
> not just an empty line.
> 
> The thing with this is that it terminates the backgrounded process.

(In bash at least) you can use "disown -h " to remove a job
from your shell's job table without killing it. (man bash)

In combination with a backgrounded ssh session, maybe this will do?

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: xmllint & libxml2? Stable.

[CW Harris]

On Thu, Jun 10, 2004 at 10:34:58AM -0600, s. keeling wrote:
> I've been looking for xmllint which is supposed to be in libxml2.  I
> have libxml2 installed but no xmllint.  Do I need a backport or am I
> just not finding it?

Seems to be in libxml2-utils not in stable[1]. libxml2-utils 2.6.9-2 in
testing seems to include it.  Don't know about any backports.

[1] (Search on packages.debian.org didn't come up with anything in stable.)

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Using the cp command.

[CW Harris]

On Thu, Jun 10, 2004 at 10:21:18AM -0400, alex wrote:
> I'm trying to copy the contents of one partition to another with:
> cp -afv  (partition a)/*  (partition b)/
> 
> Normally, the command works fine except when (partition b) already 
> contains a large number of directories and files that are duplicates of 
> those in (partition a) . 
> 
> The command works but each duplicate requires permission  to be 
> overwritten so it's a tedious process, clicking on 'y' for each 
> duplicate being copied.

What version of cp are you using?  I cannot duplicate this behavior
here.  Are you sure you don't have an "-i" set?  Are you using /bin/cp
or an alias?

cp (coreutils) 5.0.91  [here]

you might check out "--reply=yes" option (man cp)


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: wvdial connects but no internet and KPPP dies

[CW Harris]

Okay, been a while since I used ppp, but I'll give it a try...

On Wed, Jun 09, 2004 at 07:28:39AM -0700, Jason Kretzer wrote:
> Here is the output from the commandline when I run
> KPPP from a shell as root.
> 
> ~# kppp
> Opener: received SetSecret
> Opener: received SetSecret
> Opener: received OpenLock
> 
> Opener: received OpenDevice
> Opener: received ExecPPPDaemon
> In parent: pppd pid 19457
> Couldn't find interface ppp0: No such device
> Kernel supports ppp alright.
> pppd: The remote system is required to authenticate itself
  

You are requiring your ISP to authenticate itself to you (probably not
what you want).  Read the docs for PPP, last I knew in a file something
like /etc/ppp/peers/provider-or-whatever-name-you-gave-it you set
options such as username.  In this file you need to set "noauth",
otherwise you are asking the ISP to authenticate themselves to you.

Again, read through the documentation regarding authentication to get
better info.

HTH


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: running fsck out of a script; drive never shows being checked

[CW Harris]

On Tue, Jun 08, 2004 at 08:51:51PM -0400, Silvan wrote:
> I'm wondering if I'm missing the point of something somehow.
> 
> I have a script that fscks hdb, then mounts it and makes a backup of hda.  I 
> run this as a nightly cron job, and it mails me a report every morning.
> 
> The associated bit of the script is:
> 
> echo running fsck on /dev/hdb partitions...
> echo start time: `date`
> 
> if ! (fsck -n /dev/hdb2); then
> echo "ERROR!  Problem encountered running fsck on /dev/hdb2!"
> echo "Abort, abort, abort!!!"
> exit 1
> fi
> 
> echo end time: `date`
> 
> The most recent report (and the 41 before it) claim the drive has never been 
> checked.
 

It sounds like a "feature".  From bug#249116:

Sender: Theodore Ts'o <[EMAIL PROTECTED]>


So even if you fsck manually, in both cases, e2fsck will know when the
fsck was last done.

There's only one exception to this.  If you use e2fsck -n (or if you
use fsck -n, which passes the -n option to e2fsck), this explicitly
requests of e2fsck that **no** changes be made the the filesystem.
This includes trying to fix any filesystem corruptions, but it also
includes changing the superblock to the reflect the fact that e2fsck
was run in the first place.

- Ted

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Help! Printing is mostly non-working!

[CW Harris]

On Tue, Jun 08, 2004 at 02:53:50PM -0400, Marc Shapiro wrote:
> I hope that someone can help me with this problem.  About two weeks ago, 
> printing stopped working in Firefox, Thunderbird, Opera, OpenOffice, and 
> almost every other program that I use.
> 
> Kate (the editor) still prints, but a simple editor that I wrote, 
> myself, does not.  I can also print a file from the command line using 
> 'lp filename', or 'lpr filename'.  Commands such as 'ls | lp' do NOT 
> work, however.  This caused me to think that there might be a problem 
> with bash and piping from one command to another, but 'ls | more' works 
> just fine.
> 
> It is really annoying to have to print to a file, and then print the 
> file from the command line.
> 
> BTW: I am running Sarge with the debian stock 2.4.26-1-k7 kernel and 
> CUPS.  I am using cupsys-bsd, so that lp and lpr still work as the 
> programs were originally configured and this had been working for 
> several months untill this new problem occured.

What version of cups, cupsys-client are you using?

I don't know if this is of any use, but http://bugs.debian.org/163663
sounds similar (problem with printing from stdin), although it was
reportedly working with a later version of cupsys-client (that I think
you would be using in Sarge).

> 
> Since I can print with 'lpr filename' I know that the problem is not 
> with the printer.  Since CUPS (and cupsys-bsd) seems to work fine (I can 
> print a test print from the CUPS admin page through either Opera, or 
> Firefox even though the browsers' print functions don't work, 
> themselves) the problem does not appear to be there, either.  Also, I 
> don't beleive that there have been any recent updates to the CUPS packages.
> 
> I am at a loss as to what else to check.  If anyone has any ideas, or 
> has had this problem themselves, I would appreciate whatever help you 
> can provide.

Does this problem exist with other users (e.g. root)?
Maybe try with a different shell?
Check that /usr/bin/{lp,lpr} are not changed from the cupsys-client,bsd
packages?

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ext3 and charsets

[CW Harris]

On Tue, Jun 08, 2004 at 04:51:33PM +0200, J. Preiss wrote:
> Am Sonntag, 6. Juni 2004 21:08 schrieb Adam Aube:
> > J. Preiss wrote:
> > > I just "updated" from Suse 9.1 to debian testing, therefore I'm wondering
> > > how to change the mount charset of ext3 devices. I tried to use the suse
> > > feature "charset=utf8" in fstab, but this seems not to be recognized.
> >
> > According to the man page for mount, that option does not exist.
> 
> I didnt find it either :-) The problem is, that on my suse 9.1 installation it 
> worked great with this option. Unfortunately I decided to change to debian, 
> so I cannot compare...
> 
> >
> > > The problem is, that I read about 200 cds with kaudiocreator and sorted
> > > them with juke, among them are about 20 russian cds with cyrillic titles.
> >
> > Are these audio or data CDs? If audio, then ext3 won't be an issue. FWIW,
> > the only filesystem I've seen have problems with UTF-8 is jfs, which has a
> > mount option to enable UTF-8.
> 
> The main problem now is that I already read a few audio cds with cyrillic 
> names (and umlauts) to my ext3 partition, but the file names are not readable 
> anymore. Its clear that I cant play them with juke...
> 
Just a thought...

Is it the iocharset option for the iso9660 filesystem that you need
(when you read the CD's)?

Just to clarify...you have the filenames on an ext3 and the charset has
already been garbled?  Or is it a problem with your locale setting that
won't properly display the characters?

(Still new to all these locale settings, so I apologize if these are
 stupid questions.)

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: /usr/src/linux/include/linux/version.h for 2.6.5 kernel?

[CW Harris]

On Mon, Jun 07, 2004 at 08:47:59PM +0200, LeVA wrote:
> 2004. j?nius 7. 20:13 d?tummal Ishwar Rattan ezt ?rta:
> > I am running kernel-2.6.5 Debian testing system. I am trying to
> > compile Mplayer-1.0pre4. The complilation stops with a complaint
> > about linux/version.h include not being there. I looked in dir
> > /usr/src/linux-2.6.5/include/linux/ and the file does not exit. Where
> > can I get this file (or it is now obsolete)??
> >
> > -ishwar
> compile the kernel and there will the version file
> 
> hth,
> 
> Daniel
> 

There was also a thread a short while back that sounds similar. See:

http://lists.debian.org/debian-user/2004/04/msg09460.html

Maybe you are running into this problem?


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: handling runlevels independently with update-rc.d

[CW Harris]

On Wed, Jun 02, 2004 at 01:16:08AM +0200, Sebastian Kügler wrote:
> Hi,
> 
> I am in the course of writing a graphical runlevel editor, and I happen to
> have the following questions:
 
> 
> 2) If I remove all symlinks in all runlevels (basically "update-rc.d -f
> proftpd remove"), create a startup link ("update-rc.d proftpd start 50
> 4 ."), creating a stop link fails with the following output:
> 
> # update-rc.d proftpd stop 50 4 .
>  System startup links for /etc/init.d/proftpd already exist.

Create the start/stop links at the same time. update-rc.d does not allow
changes if any links exists to allow the local admin to override
automatic package changes (see the manpage). For example, the default is
equivalent to:

  update-rc.d foobar start 20 2 3 4 5 . stop 20 0 1 6 .

HTH

> 
> This seems a bit odd to me, I checked twice, the startup link command did
> not create any Knnproftpd links. So why's that?
> 
> The default option of update-rc.d only lets me create links for runlevels
> 3,4 and 5, so this one's out since I want to use the respective runlevels
> independently from each other.
> 
> 
> cheers,
> sebas

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: make-kpkg --revision policy and glibc?

[CW Harris]

On Wed, May 26, 2004 at 10:43:26AM +0200, Michal R. Hoffmann wrote:
> Hello,
Hi. I don't see any answers from people more knowledgeable about
"policy", so I'll give it a try...

> 
> after compiling 2.4.26 kernel I'm in trouble upgrading debian/unstable. 
> glibc complains about kernel subversion > 255. Well, my kernel is: 
> 2.4.261.01.mrh -> due to policy I've read in man make-kpkg, revision 
> number should not contain any hyphen, special characters (only 
> alphanumerics + .   is allowed).
> 
> Well, so I did follow the manual (THEY wrote there, that default is 
> 10.00.Custom):
> 
> make-kpkg --revision 1.01.mrh --append_to_version 1.01.mrh kernel_image
> 
> It has worked good until now, as with 2.4.26 it gives 261 which is > 255, 
> so I cannot upgrade libc6/glibc.

It is not the --revision that is causing your trouble, it is the
--append_to_version.  If you look at the Makefile, you can see that
$EXTRAVERSION is appended to $SUBLEVEL without any separation character,
thus you are getting (2.4.26)(1.01.mrh) -> 2.4.261.01.mrh

> 
> What is the _proper_ solution? Yes, I know I can compile with --revision 
> mrh.1.01, probably it will work, but is there another, "common, advised 
> way"?

Note in the make-kpkg man page --append_to_version allows lowercase
alphanumerics and the 3 characters "- + ."  Also if you read
/usr/share/doc/kernel-package/README.gz there is one example:

   make-kpkg clean   
   make-kpkg -rootcmd fakeroot --append-to-version -custom.${VER} \
--revision custom.${VER} kernel_image modules-image

Note the use of the hyphen character at the start of the
--append-to-version argument. (Note in the man page both formats
--append-to-version and --append_to_version are equivalent.)

So I would suggest that you add a delimiting character to the beginning
of your kernel EXTRAVERSION variable, maybe a hyphen as in the example
above, or at least a "."

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Display Catch-22

[CW Harris]

On Mon, May 24, 2004 at 04:09:51PM -0400, Thomas H. George wrote:
> On Mon, May 24, 2004 at 10:57:52AM -0600, Bob Proulx wrote:
> > Thomas H. George wrote:
> > > My Belkin UPS probably saved my computer from the power surge but its 
> > > internal programming is probably fried.  I reinstalled the software from 
> > > the Linux tarball changing all the ownerships to root:tom and all the 
> > > permissions to 770.
> > 
> > Please be specific.  What software did you reinstall?  Did you
> > reinstall Debian?
> 
> Sorry, I took it for granted that I was reinstalling the Belkin Bulldog
> software for linux which came with the UPS.  It is the internal
> programming in the Belkin UPS that is probably fried.  The computer and
> the Debian kernel and software were unscathed.
> > 
> > > This allows me to display the monitor but I cannot 
> > > change the setup.  When I originally installed the UPS I did all this as 
> > > root and was able to change the setup.  This is no longer possible as 
> > > security changes do not allow me to open a display as root.
> > 
> > What software are you talking about here?  Belkin UPS monitoring
> > software?  From Debian or from elsewhere?
> 
> The Belkin Bulldog software.  Is there a good Debian alternative?

Maybe.  Some Belkin UPS appear to be supported by the "nut" package,
which has both linux and windows versions.  I have not used this
but you might look into it.  If you are using a Belkin universal UPS,
you might need to look into the latest version, nut 2.x which has not
yet made it into the Debian archives, but the source is available from
their homepage.  Note that the stable tree appears to have a quite old
nut package (I don't know if it supports Belkin UPS), the one in
Sarge/testing supports some Belkin UPS, the 2.x version supports more.

I don't know if there are other packages that support Belkin.  If you
have to buy another UPS, you may want to check out which ones are well
supported by some UPS packages first (it seems a lot of APC ones are
supported).

> > 
> > > The only thing to change is the com port which I was certain I had right 
> > > in the first place.  To be sure I uninstalled the programs and 
> > > reinstalled them changing from /dev/ttyS0 to /dev/ttyS1.  Neither works 
> > > though /dev/ttyS0 retrieved some information but in either case the 
> > > installed software generates periodic messages that the UPS device is 
> > > unreachable.  This is what makes me believe the UPS internal software is 
> > > fried.
> > > 
> > > Before replacing it I would like to be sure that the display problem is 
> > > resolved. I have searched the X11, xdm and icewm setup files trying to 
> > > remove the restriction preventing opening a display as root.  My plan 
> > > would be to take the computer offline (The connection is wireless so 
> > > ifdown wlan0 does this) during the time the root prohibition is overwridden.
> > 
> > I usually use 'ssh -l root -X localhost' to log into the machine as
> > root and tunnel the display back.  But other suggest other options.
> > 
> I haven't studied ssh and will do so.  I did try entering the command
> exactly as given above and got "connection refused".  I'll work on this.

You probably don't have a sshd running?

If I understand correctly, you are talking about a local X display, but
you are (correctly) prevented from logging into X as root?

You can log in as your normal user, open an xterm, "su" to root and
then run the Belkin software (as root).  It sounds like this is what
you are wanting.  You can also look into "sudo" for allowing users to
execute some commands with root privileges.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Samba and network printing

[CW Harris]

On Fri, May 21, 2004 at 02:28:01AM -0700, Alvin Oga wrote:
> 
> hi ya john
> 
> On Thu, 20 May 2004, John L Fjellstad wrote:
> 
> > CW Harris <[EMAIL PROTECTED]> writes:
> > 
> > >> Now, the smb user is my guest user. I'm not sure why it tries to log in
> > >
> > > You may be having your account mapped to you guest user.
> > > IIRC the things that are required are:
> > >   1. User is in the printer admin group
> 
> doesn't matter ... lpd or other printer daemons take care of it for the
> users including root

The OP is having problems adding printer drivers to the [print$] share for
auto downloading by windows clients.

I believe all the printing is working fine.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Samba and network printing

[CW Harris]

On Thu, May 20, 2004 at 02:24:38PM +0200, John L Fjellstad wrote:
> CW Harris <[EMAIL PROTECTED]> writes:
> 
> >> Now, the smb user is my guest user. I'm not sure why it tries to log in
> >
> > You may be having your account mapped to you guest user.
> > IIRC the things that are required are:
> > 1. User is in the printer admin group
> > 2. User has a valid smb password/account
> > 3. User must be able to write to the *nix directory where
> >samba stores the printer driver info.
> 
> But I don't understand what my account has anything to do with it, since
> I'm logging in as root, and root does have rights to the printer driver
> directory. 

By "User" I meant the user you were connecting as (root in your case),
sorry for being unclear.

Okay, re-reading the thread and some Samba docs... It seems that root
may be special and may *not* have to be in "printer admin", but then
again other docs seem to show using a "printer admin = root" global, so
I'm not certain.

Or, I found this one ref. to an auth. problem with cupsaddsmb and a
Samba PDC which may be related:

http://us3.samba.org/samba/docs/man/howto/CUPS-printing.html#id2565677

Perhaps you need to include the "domain" portion?  They indicate using 

rpcclient -U "DOMAIN\root%passwd"

if I read it correctly.

HTH

P.S.  I am really getting curious about this. Windows printer drivers
(even under Samba) seem too much like black magic.  Please let me know
when you find a solution.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Samba: assign domain group policy through Samba tools?

[CW Harris]

On Mon, May 17, 2004 at 03:38:37AM -0700, Karsten M. Self wrote:
> I'm using Samba as a PDC on a domain with ten WinXP Pro clients, on
> Debian testing/unstable.
> 
> Basic shares work great.  
> 
> Getting the domain stuff set up was a bit trickier, but the OS News
> article[1] and (once I realized the difference between 2.x and 3.x) docs
> under /usr/share/doc/samba-doc/htmldocs/ were invaluable.  Tricky bit
> was creating and mapping groups/users via 'groupadd' and 'net groupmap'.
> 
> Printing through CUPS + Samba was a nightmare, but I was under the
> delusion it worked when I left work Friday night.  Erm.  Saturday
> morning.  Post-sunrise.  Tricky bit was adding printer support via
> 'cupsaddsmb', and deciphering error output (stderr and logs).
> 
> 
> I'm stuck on creating a group profile at the domain level, though.

Okay. I haven't done this so just some info you might have missed, or
might help you.

From: http://us3.samba.org/samba/docs/man/guide/happy.html#ch6-massive
 At this time, Samba-3 requires that on a PDC all UNIX (Posix) group
 accounts that are mapped (linked) to Windows Domain Group accounts must
 be in the LDAP database.

This does not actually say it, but I think I read somewhere that Samba
as a PDC requires LDAP to support the Active Directory functions.?

Also, this might be some help:
http://us3.samba.org/samba/docs/man/howto/PolicyMgmt.html#id2577673

Apparently, part of the GPO is stored directly on the Active Directory.
See also the section: Administration of Windows 200x/XP Policies" for
some steps on editting the GPO's using the MMC snap-in. (Who at MS
thinks of these names?)

Anyway, HTH.  I was all set when we got a small number of XP boxen at my
work to play around with the PDC thing, only to realize how much MS
changed the structure with 2000/XP.  I tired out trying to figure it out
for such a small number of users.  I figured by the time I got it
working, MS would release Windows eXtra-eXtra-Pain and it wouldn't work
again.

> 
> The goal is to have a single point at which I can make
> additions/deletions to Desktop, Start Menu, "Favorites" (bookmarks),
> Startup, etc.  As well as making some registry edits (allowed/disallowed
> apps).
> 
> 
> I've copied the profile itself, through one of the XP clients, to a
> directory under my [profiles] share on the Samba server.

My quick read seems to indicate it needs to be in the [netlogon] share?

> 
> What I don't see is a way to make the association between this profile
> and the group ("members") which I'd like to have use this.

Again, seems to be in the GPO that you define as in the reference above,
but then I haven't done this so maybe I'm just background noise in the
list.

 

Good luck.


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Samba and network printing

[CW Harris]

On Tue, May 18, 2004 at 09:37:01PM +0200, John L Fjellstad wrote:
> "Karsten M. Self" <[EMAIL PROTECTED]> writes:
> 
> > cupsaddsmb -U root -a 
> 
> cupsaddsmb didn't get added until later, I think. At least, the cups
> from Woody doesn't have it. I can't install the updated cups from
> backports.org because I need the Epson drivers from the Gimp.Print
> package, which hasn't been ported yet for the newer Cups (installing
> newer cups uninstall that package because of dependency issues).
> 
> Also, it seems cupsaddsmb just go through the same steps that you can do
> manually. I did test run it, and it also fails on the last step that the

I believe that is correct, although it does automate things nicely.

> manual step failed for me.
> 
> I'm not sure why I didn't think of it before, but  did check the logs
> after you mentioned it.  I got an error message, saying
> [2004/05/17 13:05:59, 0] smbd/password.c:authorise_login(536)
>   authorise_login: rejected invalid user smb
> [2004/05/17 13:05:59, 0] printing/nt_printing.c:get_correct_cversion(1119)
>   get_correct_cversion: Unable to connect
> 
> Now, the smb user is my guest user. I'm not sure why it tries to log in

You may be having your account mapped to you guest user.
IIRC the things that are required are:
1. User is in the printer admin group
2. User has a valid smb password/account
3. User must be able to write to the *nix directory where
   samba stores the printer driver info.

HTH

> as smb, since (a) I'm logging in as root, and (b) smb doesn't have the
> rights to install the drivers anyways...
> 
> -- 
> John L. Fjellstad
> web: http://www.fjellstad.org/  Quis custodiet ipsos custodes
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: mounting windows shares

[CW Harris]

On Fri, May 14, 2004 at 02:23:41PM -0400, Harland Christofferson wrote:
> At Friday, 14 May 2004, CW Harris <[EMAIL PROTECTED]> wrote:
> 
> >On Fri, May 14, 2004 at 12:05:36PM -0400, Harland Christofferson wrote:
> >> At Friday, 14 May 2004, Kent West <[EMAIL PROTECTED]> wrote:
> >> 
> >> >Harland Christofferson wrote:
> >

> 
> windblowsserver is the server name
> backup is the windows share name
> system products is a directory
> 
> 
> i thought i could mount to a directory as well. am i mistaken?
> 
I believe so.  (Not a windows or samba expert).
AFAIK you can only mount what windows considers a "share".  I believe
this is the same limitation you have under windows with the "net use"
command.

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: mounting windows shares

[CW Harris]

On Fri, May 14, 2004 at 12:05:36PM -0400, Harland Christofferson wrote:
> At Friday, 14 May 2004, Kent West <[EMAIL PROTECTED]> wrote:
> 
> >Harland Christofferson wrote:

> >>
> >>mount chokes on the windblows directory "system products" .
> >>
> >>i assume this is b/c of the space in the directory name. what should 
> >>i do so i can mount this windblows share?
> >>
> >
> >I just did a quick test, sharing out \\winserver\My Music, and was 
> able 
> >to mount it with:
> >
> >smbmount "//winserver/My Music" mntpoint
> >
> >(notice the quotes). I tried the mount command like you're using above 
> >and it failed, but then I don't know the exact syntax and am too 
> lazy to 
> >look it up, so YMMV.
> >
> >-- 
> >Kent
> >
> >
> 
> i am not having the same luck as you are. i can mount to //windblowsserver/backup 
> but i cannot mount to "//windblowsserver/backup/system products" 
> i tried smbmount and mount -t smbfs. any other suggestions?
> 

Are you sure that is a windows share?  I mean, is the windows share
"//windslowsserver/backup" and "system products" is just a directory
under that share?  Or are they *both* valid windows shares?

What does "smbclient -L windblowsserver" list as the valid shares?

HTH

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Lilo and friends

[CW Harris]

On Fri, May 07, 2004 at 03:08:46PM -0400, Adam Aube wrote:
> CW Harris wrote:
> 
> > On Thu, May 06, 2004 at 10:41:27AM +0200, Levente KOVACS wrote:
> >> On Wed, 05 May 2004 11:21:50 -0500
> >> hugo vanwoerkom <[EMAIL PROTECTED]> wrote:
> >> 
> >> > 
> >> > Get rid of XP and your problems are over ;-)
> >> > Lilo only writes the bootrecord when YOU run "lilo"...
> >> 
> >> Unfortunatelly I can not, becouse I have to program the Rabbit MCU. They
> >> only have software for win. And it's in my workplace
> > 
> > I haven't had to try to integrate Lilo with XP, but with earlier windows
> > I found it *easier* to let windows do its own thing and keep Lilo
> > installed on the linux partition.
> 
> I don't know about Lilo, but I did setup a XP Pro/Debian dual-boot box using
> Sarge and Grub, and it is installed in the MBR and boots either operating
> system without problems. I don't know why Lilo would be any different.
> 
No, I guess I wasn't clear.  Lilo should do that fine.  I just meant
that with earlier windows (and their habit of playing around with the
MBR if you upgraded, etc) I found it easier to let windows have the MBR
and put Lilo elsewhere.  Then I would just mark the Lilo partition as
the bootable one.

I merely found it easier to fit Lilo/Linux into the windows way rather
than trying to keep windows from messing up with the Lilo installation.

IIRC it was only an issue when upgrading, or installing.

> Adam
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Lilo and friends

[CW Harris]

On Thu, May 06, 2004 at 10:41:27AM +0200, Levente KOVACS wrote:
> On Wed, 05 May 2004 11:21:50 -0500
> hugo vanwoerkom <[EMAIL PROTECTED]> wrote:
> 
> > 
> > Get rid of XP and your problems are over ;-)
> > Lilo only writes the bootrecord when YOU run "lilo"...
> 
> Unfortunatelly I can not, becouse I have to program the Rabbit MCU. They
> only have software for win. And it's in my workplace

I haven't had to try to integrate Lilo with XP, but with earlier windows
I found it *easier* to let windows do its own thing and keep Lilo
installed on the linux partition.  IIRC XP comes with its own boot
manager - so perhaps you could install Lilo on the Linux partition and
add a boot entry to that partition from XP - then if XP plays around
with the MBR it doesn't mess up Lilo.

HTH

P.S.  Does anyone know if XP is an acronym for eXtra Pains (or maybe
eXtra Panes) ?

> 
> Anyways. Thank you.
> 
> -- 
> Leva


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Hosting virtual mail domains on Debian.

[CW Harris]

On Fri, Apr 23, 2004 at 01:34:56PM -0600, David Anselmi wrote:
> David Anselmi wrote:
> 
> >I'd really like to be able to host virtual mail domains on a Sarge server.
> 
> [...]
> 
> >Any advice?  Is there anything I can do to fix 100646 or get a 2.2 
> >cyrus package submitted?  I'm happy to help but I don't know how long 
> >it will take me or how useful my efforts will be.
> 
> 
> Ok, no takers on that question.  Anyone have any ideas where else I 
> should ask?  Just go straight to the developers' lists?

I assume you have tried google?

I would try the mailer specific lists.  I haven't done this, but I would
think any of the major MTA's would be able to do this.

Pick an MTA and check out their lists?

> 
> Thanks,
> Dave
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: printing

[CW Harris]

On Fri, Apr 23, 2004 at 02:21:43PM +0200, Miguel Mazzorana wrote:
> I have installed  Debian Woody  several times but I can not use my 
> printer  HP Deskjet 720 C in any way.
> I have try a lot of things!
> I have read a lot of logs!
> 
> but every time when  I type:
>  ls -all  |  /dev/lp0

How about "ls -al | lpr"?
Have you installed any specific print packages?

Maybe you need to try reading some of the Printing HOWTOs to become
more familiar with the printing setups?

Try http://www.tldp.org/HOWTO/HOWTO-INDEX/categories.html
or  http://www.tldp.org/HOWTO/HOWTO-INDEX/os.html#OSPRINTING

If you are as new to Linux as you sound, you may want to find a
Linux user group in your area.

> 
> give me:
> 
> -bash :  /dev/lp0 Permission Denied
> 
> /dev   directory have:
> 
> drwxr-xr-x9 root root24576 abr 23 12:37 dev
> 
> /dev/lp0  have:
> 
> crw-rw1 root lp 6,   0 mar 14  2002 /dev/lp0
> 


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: grub/dual-boot problem

[CW Harris]

On Fri, Apr 23, 2004 at 05:23:53AM -0700, Richard Weil wrote:
> Just a small update, since I'm sure someone will
> someday search and find this thread ...
> 
> I got it working. The grub configuration in menu.1st
> was fine; windows had overwritten the MBR. I don't
> know if it's possible, but I *may* not have installed
> grub to the MBR the first time (I don't recall running
> install-grub /dev/hda) -- by marking the partition as
> bootable, it sort of worked.

I haven't had much experience w/ XP, but I'm wondering if you
had it booting on /boot partition (marked bootable).  On rebooting
the WinXP MBR booted the partition with the bootable flag.
Then upon subsequently booting into WinXP it removed the bootable
flag on your /boot partition, thus you never got to grub
again?

I'm mostly wondering because WinXP does have a multi-boot
manager now, and this might be how it regains control if
you haven't added your alternate OS into its menu?


-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: what bash dotfiles are read during interactive ssh?

[CW Harris]

On Mon, Apr 19, 2004 at 07:27:06AM +0800, Dan Jacobson wrote:
> What file can one put "unalias ls", or anything in fact, in for it to
> get read on the remote system upon interactive ssh?  .bashrc,
> .bash_profile are apparently not read.

On mine it executes .bash_profile (normal login shell).

According to ssh man page you should get a normal login if you don't
specify a command to execute.

Is your login shell bash?  Is there some system setup to cause bash to
look in a different login file? (e.g. .profile?)

If bash is invoked as 'sh' I understand that it looks instead in
.profile

HTH

> 
> At most I can do now is
> $ ssh porky.simonds.net
> porky$ unalias ls
> 
> I don't want the dotfile read on batch jobs preferably.
> 

-- 
Chris Harris <[EMAIL PROTECTED]>
---
GNU/Linux --- The best things in life are free.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]