Re: Can't switch to virtual console anymore

[Peter Jay Salzman]

On Thu 22 Dec 05,  9:09 AM, p  said:
> Dear all,
> 
> I can't switch to a virtual console any longer.  I'm fairly sure whatever
> happened must have happened during my last "aptitude upgrade", because I
> switch to VC's very very often.  Datapoints:
> 
> 0. This is a Debian/testing machine.
> 
> 1. My XF86Config does not contain the DontVTSwitch option.
> 
> 2. On control-alt-F1, xev reports:
> 
>KeyPress event, serial 30, synthetic NO, window 0x121,
>root 0x40, subw 0x122, time 417047, (47,38), root:(1117,396),
>state 0x0, keycode 37 (keysym 0xffe3, Control_L), same_screen YES,
>XLookupString gives 0 bytes: 
>XmbLookupString gives 0 bytes: 
>XFilterEvent returns: False
> 
>KeyPress event, serial 30, synthetic NO, window 0x121,
>root 0x40, subw 0x122, time 417495, (47,38), root:(1117,396),
>state 0x4, keycode 64 (keysym 0xffe9, Alt_L), same_screen YES,
>XLookupString gives 0 bytes: 
>XmbLookupString gives 0 bytes: 
>XFilterEvent returns: False
> 
>KeyPress event, serial 30, synthetic NO, window 0x121,
>root 0x40, subw 0x122, time 417935, (47,38), root:(1117,396),
>state 0xc, keycode 67 (keysym 0xffbe, F1), same_screen YES,
>XLookupString gives 0 bytes: 
>XmbLookupString gives 0 bytes: 
>XFilterEvent returns: False
> 
> 
> 3. Once I kill X, I can switch consoles using the familiar alt-[1-9].
> 
> 4. There's no output from xinit/startx that I don't recognize as already
>being there (for example, I've seen the message:
> 
>   Could not init font path element /usr/lib/X11/fonts/Speedo,
>   removing from list!
> 
>for at least a year.)
> 
> 5. When I hit contol-alt-F1 when a vim instance has focus, I see "7P".
> 
> 
> I will go nuts without my consoles.  Any suggestions?
> 
> Thanks,
> Pete

Problem solved.  It turned out to be "one of those warnings I recognized."

For years, X complained that there was no XkbVariant named "Microsoft" in
response to the line:

   Section "InputDevice"
  ...
  Option  "XkbVariant"  "Microsoft"
  ...
   EndSection

I'm not sure when that line appeared in my config file.  Quite possibly back
when I was using Suse 6.1, way before I switched to Debian.  I guess I just
"toted" the config file around, and only changed mode, horiz and vertical
timings when configuring a new system.

The warning didn't have any adverse effect, and I always promised myself to
look into it when time permitted me to delve into learning the nitty gritty
details of keyboards under X.

There always seemed to be more pressing issues.  And besides, it's difficult
to get excited about learning details of keyboards under X.  It seems more
complicated than it ought to be.  And dry.

Short story is, that line always caused a benign warning.

Since I was (potentially) having trouble with the keyboard, I started to
scrutinize the X output, even looking at things which I didn't think had
relevence to the problem at hand.  That trained my eye on the warning.

I'm not sure what got upgraded yesterday, but whatever it was, that warning
apparently was no longer benign.  I removed the XbdVariant, and now
everything is back to normal.

Pete


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

Re: Can't switch to virtual console anymore

[Peter Jay Salzman]

On Thu 22 Dec 05,  9:43 AM, Kent West <[EMAIL PROTECTED]> said:
> Peter Jay Salzman wrote:
> 
> >Dear all,
> >
> >I can't switch to a virtual console any longer.  I'm fairly sure whatever
> >happened must have happened during my last "aptitude upgrade", because I
> >switch to VC's very very often.  Datapoints:
> >
> >0. This is a Debian/testing machine.
> >
> >1. My XF86Config does not contain the DontVTSwitch option.
> >
> I would expect that X.org has filtered into Testing by now; perhaps the 
> problem is in xorg.conf rather than in XF86Config?
 
Yeah, you're right.  I hadn't noticed the existence of that file.
Unfortunately, the two files appear to be duplicates

   # diff xorg.conf XF86Config-4
   # 

so there's no DontVTSwitch in /etc/X11/xorg.conf, either.   xwininfo and
xdpyinfo don't seem to give any info that appears relevent to the problem.
I was hoping they'd show "DontVTSwitch" in their output.  At least then I'd
know what was causing the problem.

This is so wierd...

Any other suggestions?

Thanks!


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

Can't switch to virtual console anymore

[Peter Jay Salzman]

Dear all,

I can't switch to a virtual console any longer.  I'm fairly sure whatever
happened must have happened during my last "aptitude upgrade", because I
switch to VC's very very often.  Datapoints:

0. This is a Debian/testing machine.

1. My XF86Config does not contain the DontVTSwitch option.

2. On control-alt-F1, xev reports:

   KeyPress event, serial 30, synthetic NO, window 0x121,
   root 0x40, subw 0x122, time 417047, (47,38), root:(1117,396),
   state 0x0, keycode 37 (keysym 0xffe3, Control_L), same_screen YES,
   XLookupString gives 0 bytes: 
   XmbLookupString gives 0 bytes: 
   XFilterEvent returns: False

   KeyPress event, serial 30, synthetic NO, window 0x121,
   root 0x40, subw 0x122, time 417495, (47,38), root:(1117,396),
   state 0x4, keycode 64 (keysym 0xffe9, Alt_L), same_screen YES,
   XLookupString gives 0 bytes: 
   XmbLookupString gives 0 bytes: 
   XFilterEvent returns: False

   KeyPress event, serial 30, synthetic NO, window 0x121,
   root 0x40, subw 0x122, time 417935, (47,38), root:(1117,396),
   state 0xc, keycode 67 (keysym 0xffbe, F1), same_screen YES,
   XLookupString gives 0 bytes: 
   XmbLookupString gives 0 bytes: 
   XFilterEvent returns: False


3. Once I kill X, I can switch consoles using the familiar alt-[1-9].

4. There's no output from xinit/startx that I don't recognize as already
   being there (for example, I've seen the message:

  Could not init font path element /usr/lib/X11/fonts/Speedo,
  removing from list!

   for at least a year.)

5. When I hit contol-alt-F1 when a vim instance has focus, I see "7P".


I will go nuts without my consoles.  Any suggestions?

Thanks,
Pete


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

Grub and dual booting

[Peter Jay Salzman]

Hi all,

I'm a recent lilo convert, so I'm rather new to grub.


Summary
===

I recently had to reinstall MS Windows XP (sigh), and I'm now having trouble
getting grub to boot it.  Debian boots OK.  When booting XP, I see:

   Disk error
   Press any key to restart
   Missing operating system



The System
==

BIOS is set to boot off the PATA drive.  The drives on the system are:

   hda PATA: (all ext3)
  Various backup partitions.  Expendable.  Contains the system's MBR.

   hde SATA: MS Windows XP (all FAT32)

  hde1 (hd1,0) MS Windows OS (bootable)
  hde2 (hd1,1) D drive
  hde3 (hd1,2) E drive

   hdg SATA: GNU/Linux (all ext3)

  hdg1 (hd2,0) /boot
  hdg2 (hd2,1) /var
  hdg3 (hd2,2) swap
  hdg5 (hd2,4) /home
  hdg6 (hd2,5) /
  hdg7 (hd2,6) /usr/local



Grub Configuration
==

Grub was installed on /dev/hda.  My grub menu looks like this:

   titleDebian GNU/Linux, kernel 2.6.10
   root (hd2,0)
   kernel  /vmlinuz-2.6.10 root=/dev/hdg6 ro
   savedefault
   boot

   titleWindows NT/2000/XP
   map (hd0) (hd1)
   map (hd1) (hd0)
   rootnoverify (hd1,0)
   savedefault
   makeactive
   chainloader +1



What I tried


When I installed XP, I removed the PATA drive so XP thought it was
installing itself onto the "first drive".

After finding that XP wouldn't boot, I tried:

   1. disconnect the PATA drive
   2. run the Windows XP recovery console
   3. use "fixmbr" and "fixboot".
   4. reconnect the PATA drive

But it didn't work.  I'm a little reluctant to repeat the above procedure
without taking the PATA drive out because I don't want to lose the ability
to boot into Debian.  I'm also very unknowledgable with Microsoft's
operating systems.

I don't understand why this isn't working:

   * BIOS boots of the PATA drive.
   * Grub is installed in the MBR on the PATA drive.
   * Windows first bootloader is on the MBR of (hd1)
   * I'm not sure how Windows boots after that, but I assume this is where
 the problem is.  What exactly happens once the MS first bootloader
 is loaded?

Any ideas?  I'm on unfamiliar ground all around.  Should I try running
fixboot with the PATA drive still connected?

Thanks!
Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: Discover device drivers built into a kernel image

[Peter Jay Salzman]

On Tue 31 May 05,  1:51 PM, s. keeling <[EMAIL PROTECTED]> said:
> Incoming from Paulo M C Aragão:
> > 
> > How do I discover which device drivers are built into a given kernel
> > image, not accessing the sources ?
> 
> Check the config file that came with it?
> 
>   cd /boot
>   ls -l config*
>   uname -a

If it's enabled in the kernel, you can also do

zcat /proc/config.gz

It's been a long time since I've used a stock Debian kernel, so I don't know
if they have it enabled by default.

Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: system.map files

[Peter Jay Salzman]

On Tue 31 May 05,  1:49 PM, Jon Dowland <[EMAIL PROTECTED]> said:
> LeVA wrote:
> 
> >Hi!
> >
> >Would someone please tell me what are those /boot/System.map* files for?
> >
> >Thanks!
> >
> >Daniel
> > 
>
> The website   (via google) 
> describes it so well that any attempt to summarize would pale in 
> comparison. So I won't :)
 

Whoa!!!

Very nice complement.  Thanks, Jon.  It took me a long time to research all
that information.

You just made my morning!

Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: [xine-user] Cracking noise playing Audio CDs on kaffeine/xine

[Peter Jay Salzman]

On Mon 30 May 05,  7:42 PM, [EMAIL PROTECTED] <[EMAIL PROTECTED]> said:
> On Mon, May 30, 2005 at 06:11:54PM -0400, Peter Jay Salzman wrote:
> > Try this:
> > 
> > 1. Open kaffeine/xine and start playing something.  2. Open a mixer
> > program.  Make sure the settings are not cranked up while kaffeine is
> > playing.
> 
> Thanks for the persistence !
 
No problem!

> After I replied to your 1st e-mail, I realized that I was using OSS,
> driver sb. I then installed alsa-modules for my kernel (2.4.27.2-686)
> and driver snd-sb8 loaded Ok. Now I can still play audio CDs with kscd,
> but kaffeine/xine spit the message:
> 
> Audio output unavailable. Device is busy
> 
> I can still play internet audio streams with kaffeine/xine but not audio
> CDs.
> 
> Killed kscd, but kaffeine/xine spits the same message when I try to play
> audio CDs. I'm not quite sure which device is busy. I must be missing
> something about alsa setup. I just installed the alsa packages and
> loaded the driver manually with:
 
Perfect application for strace.  Use:

strace -o LOG.txt kaffeine

Wait until you see kaffeine issue "device is busy" and then kill kaffeine.

Load up the file LOG.txt with an editor.  Do a search for "device is busy".
There should be an open() statement somewhere before the error message is
printed.  Probably not too far away, in fact.

That should tell you not only what device file kaffeine attempted to open,
but also why it couldn't open it.

If you've never used strace, this is a good way to learn.  This is
practically what strace was invented for.  ;)

Pete


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

Re: [xine-user] Cracking noise playing Audio CDs on kaffeine/xine

[Peter Jay Salzman]

On Mon 30 May 05,  6:19 PM, Paulo M C Aragão <[EMAIL PROTECTED]> said:
> Hi Peter,
> 
> On Mon, May 30, 2005 at 04:56:13PM -0400, Peter Jay Salzman wrote:
> 
> > james was suggesting to kill arts and see if your sound is any better.  this
> > effectively kills the middleman -- your audio program will interact directly
> > with the sound hardware instead of needing to interact with another user
> > land program which in turn will interact with the audio hardware.  so:
> 
> First of all, thank you so much for taking the time for such a detailed
> and patient explanation. It really set the ground clear for me.
 
my pleasure!

> I killed arts (actually disabled Sound System at KDE's Control Center,
> but made sure the process wasn't really running) and tried to play an
> audio CD again:
> 
> With kaffeine/xine  the same cracking/skipping sound while playing
> With kscd played perfectly
> 
> So, excluded sound daemon, hardware and mixer, would be there be
> anything else to look into ?
 
Not really, but we can try to grasp at straws.


Try this:

1. Open kaffeine/xine and start playing something.
2. Open a mixer program.  Make sure the settings are not cranked up
   while kaffeine is playing.


Does kaffeine have some kind of audio output plugin?  Perhaps an audio
output thing (like using libsdl or native alsa for audio, kinda like how
mplayer does it).


Does kaffeine still crackle if you "renice -20" it?


Sorry, I didn't realize that an app actually worked OK.  That makes it much
harder.  These suggestions are all grasping at straws, so please don't
laugh.  :)

Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: [xine-user] Cracking noise playing Audio CDs on kaffeine/xine

[Peter Jay Salzman]

On Mon 30 May 05,  4:40 PM, Paulo M C Aragão <[EMAIL PROTECTED]> said:
> Hi James,
> 
> On Mon, May 30, 2005 at 06:13:29PM +0100, James Stembridge wrote:
> 
> > > I hear a constant cracking/skipping noise when I play Audio CDs with
> > > kaffeine/xine. I'm using both straight out-of-the-box, installed from
> > > Debian binary packages.
> > > 
> > > Some info about my setup:
> > > 
> > > OS:  Debian Sarge with KDE 3.3.2
> > > Sound driver:  sb
> > > Audio system:  aRts (libarts1 1.3.2-3)
> 
> > Does it still happen when you stop using arts?
> 
> Actually I didn't try it. Do you think alsa or OSS might solve the
> problem ?

alsa and oss are kernel land sound drivers that drive your audio hardware.
you either use one or the other.  oss is the old (deprecated) driver and
alsa is the "new" (it's been around for awhile) driver.  you already have
alsa (more likely) or oss (less likely) enabled.

arts is a user land sound daemon (that i think is associated with kde) that
sits in between the sound drivers (alsa and oss) and your audio programs
(like xmms, sox, audacity, etc).

people like sound daemons because normally, only one program can have access
to hardware resources at a time.  so if you're playing an ogg file, any
other program that wants to play a sound won't be able to use audio.  a
sound daemon "negotiates" use between two programs that want to play a sound
at the same time.

people don't like sound daemons because they add a level of complexity.
unlike windows, where the sound daemon is part of the kernel, there are a
few sound daemons (arts, esd, maybe some others) and so there's no there's
no standard on how to interact with a sound daemon.  a program like xmms
needs to know how to "talk" to each individual sound daemon.  the situation
is not unlike the DOS era where a game had to support individual sound
cards.  linux sound apps have to know how to interact with the various sound
daemons.  some apps know how to talk to all the major sound daemons (xmms)
and some don't (mp3blaster).  there are ways around this, (like using
esddsp) but it doesn't always work (quake3 doesn't like esddsp).

james was suggesting to kill arts and see if your sound is any better.  this
effectively kills the middleman -- your audio program will interact directly
with the sound hardware instead of needing to interact with another user
land program which in turn will interact with the audio hardware.  so:

1. kill arts (i don't know how to do this because i know nothing about
   arts.  presumably killall -9 artd will work).

2. play a sound.  see if it helps.

another thing you may want to try is to make sure your mixer settings are
about 50% - 75% of their maximum setting using a mixer program like xmix,
cam, or nmixer.  adjust your speakers appropriately.  see if that helps.

also, if you have them, try a different set of speakers.  if your speakers
are really old, the cones might be brittle or even cracked which can cause
crackling.

also, wiggle your wires.  loose audio connections can also cause crackling.

pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Dual booting with XP: trouble with grub

[Peter Jay Salzman]

Hi all,

I'm a recent lilo convert, so I'm rather new to grub.


Summary
===

I recently had to reinstall MS Windows XP (sigh), and I'm now having trouble
getting grub to boot it.  Debian boots OK.  When booting XP, I see:

   Disk error
   Press any key to restart
   Missing operating system



The System
==

BIOS is set to boot off the PATA drive.  The drives on the system are:

   hda PATA: (all ext3)
  Various backup partitions.  Expendable.  Contains the system's MBR.

   hde SATA: MS Windows XP (all FAT32)

  hde1 (hd1,0) MS Windows OS (bootable)
  hde2 (hd1,1) D drive
  hde3 (hd1,2) E drive

   hdg SATA: GNU/Linux (all ext3)

  hdg1 (hd2,0) /boot
  hdg2 (hd2,1) /var
  hdg3 (hd2,2) swap
  hdg5 (hd2,4) /home
  hdg6 (hd2,5) /
  hdg7 (hd2,6) /usr/local



Grub Configuration
==

Grub was installed on /dev/hda.  My grub menu looks like this:

   titleDebian GNU/Linux, kernel 2.6.10
   root (hd2,0)
   kernel  /vmlinuz-2.6.10 root=/dev/hdg6 ro
   savedefault
   boot

   titleWindows NT/2000/XP
   map (hd0) (hd1)
   map (hd1) (hd0)
   rootnoverify (hd1,0)
   savedefault
   makeactive
   chainloader +1



What I tried


When I installed XP, I removed the PATA drive so XP thought it was
installing itself onto the "first drive".

After finding that XP wouldn't boot, I tried:

   1. disconnect the PATA drive
   2. run the Windows XP recovery console
   3. use "fixmbr" and "fixboot".
   4. reconnect the PATA drive

But it didn't work.  I'm a little reluctant to repeat the above procedure
without taking the PATA drive out because I don't want to lose the ability
to boot into Debian.  I'm also very unknowledgable with Microsoft's
operating systems.

I don't understand why this isn't working:

   * BIOS boots of the PATA drive.
   * Grub is installed in the MBR on the PATA drive.
   * Windows first bootloader is on the MBR of (hd1)
   * I'm not sure how Windows boots after that, but I assume this is where
 the problem is.  What exactly happens once the MS first bootloader
 is loaded?

Any ideas?  I'm on unfamiliar ground all around.  Should I try running
fixboot with the PATA drive still connected?

Thanks!
Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D

- End forwarded message -

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: Is 64MB enough?

[Peter Jay Salzman]

On Sat 21 May 05,  8:37 AM, Johan Kullstam <[EMAIL PROTECTED]> said:
> "John Moore" <[EMAIL PROTECTED]> writes:
> 
> > Thanks but the old clunker's motherboard is not expandable to 256M
> > :-(
> 
> Star/Open-Office is not going to be pleasant.  TeX, on the other hand,
> will run like a treat.
 
TeX always runs like a treat.  But I have to disagree with OO.  It's always
slow.  But I don't think it'll be particularly slower at 64MB as long as the
system isn't loaded down (but perhaps that's what you meant).

> > >
> > >in that case you'd need more memory. 256MB is cheap these days.
> 
> As you know, It's not as cheap if you need 72-pin EDO SIMMs  memory.

Not all memory is cheap, as I unfortunately recently learned.  Ironically,
old DIMMs --- I'm talking about the slow single density PC100 sticks you'd
put in a Pentium II computer --- are more expensive than the DDR memory
you'd put into a modern Athlon system.

The reason, I've read, is that it takes more silicon wafer surface material
to generate single density DIMMs.

Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: /etc/modules comment is wrong

[Peter Jay Salzman]

On Sat 21 May 05,  8:30 AM, Carl Fink <[EMAIL PROTECTED]> said:
> On Sat, May 21, 2005 at 01:53:28PM +0200, David Jardine wrote:
> > On Sat, May 21, 2005 at 03:25:21AM +0200, Jacobo221 wrote:
> > > In Debian Sarge, in /etc/modules out-of-the-box, the comments say that
> > > nything after '#' is ingored. But this is wrong. Only lines which BEGIN
> > > with '#' are ignored. This should be changed.
> > 
> > Hmm.  'Comments begin with a "#", and everything on the line after 
> > them are ignored', says mine.
> 
> Yes, but in some languages the "start the comment here" string (for instance
> "//") can come at any point in the line, say after an executed statement. 
> Jacobo is saying that the "#" only works a the beginning of a line.

is that really true?   you mean I can't have a line in /etc/modules that
looks like:

   lego# device driver for the lego mindstorms USB tower

that's too bad if it's true.

pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

emergency: font problems causing gv to crash

[Peter Jay Salzman]

Hi all,

My PhD dissertation is due in a few weeks and I'm having some kind of
problem with fonts that's preventing me from viewing my thesis.

This is a new computer (my old one died at the *worst* possible moment.
Luckily, I'm *really* good about backing everything up), so it's a fresh
Debian/testing installation and I'm trying to iron out bugs as quickly as
possible so I can get back to writing.

I normally view my thesis using gv, but gv segfaults:

   $ gv dissertation.ps
   Warning: Missing charsets in String to FontSet conversion
   Warning: Unable to load any usable fontset
   Segmentation fault (core dumped)

xdvi also fails to run:

   $ xdvi dissertation.dvi
   X Error of failed request:  BadValue (integer parameter out of range
for operation)
 Major opcode of failed request:  45 (X_OpenFont)
 Value in failed request:  0xe4
 Serial number of failed request:  27
 Current serial number in output stream:  28


xfontsel also seems to have a problem:

   FreeType: couldn't open face
   /var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType/ds=y:VeraMono.ttf: 1


On the console behind X, this error message is printed:

   FreeType: couldn't open face
   /var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType/ds=y:VeraMono.ttf: 1

But this doesn't look like fatal error to me.  I really need to get writing
ASAP.  Can someone please help?

Thanks,
Peter

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: mouse clicks and virtual console acting wierd

[Peter Jay Salzman]

On Fri 20 May 05, 10:48 AM, Peter Jay Salzman <[EMAIL PROTECTED]> said:
> I've never seen this before...
> 
> On a new install, when I left click the mouse on the console, I see:
> 
> 
> 
> middle clicking gives:
> 
> E EEE
> 
> right clicking gives:
> 
> 
> 
> 
> The mouse otherwise works good on the console.  It works perfectly in X.
> 
> I've never seen this before.  Why does mouse clicking produce these
> characters (cut and paste in the console doesn't work at all).
> 
> Thanks,
> Pete

I built my own kernel instead of using the Debian stock kernel and the
problem went away.

Nevermind! :)
Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

mouse clicks and virtual console acting wierd

[Peter Jay Salzman]

I've never seen this before...

On a new install, when I left click the mouse on the console, I see:



middle clicking gives:

E EEE

right clicking gives:




The mouse otherwise works good on the console.  It works perfectly in X.

I've never seen this before.  Why does mouse clicking produce these
characters (cut and paste in the console doesn't work at all).

Thanks,
Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

remote printing isn't working - lpd not listening

[Peter Jay Salzman]

Hi all,

Just installed Debian on my new workstation, "satan".  The workstation has
an HP LaserJet 6MP that I'd like to share with other Linux computers on my
home network.  I use lpd, not lprng.

I'm pretty sure /etc/printcap is set up correctly on satan and the remote
hosts.  I'm pretty sure that /etc/hosts.lpd is set up correctly on satan.

However, on other computers in the house:

   lucifer# lpq
   connection to 192.168.0.2 is down

They can't seem to connect to lpd on satan.  Supposedly, lpd uses port 515
to accept incoming print jobs.  However, there doesn't appear to be anything
listening to that port:

   [EMAIL PROTECTED] telnet localhost 515
   Trying 127.0.0.1...
   telnet: Unable to connect to remote host: Connection refused

   lucifer# telnet satan 515
   Trying 192.168.0.2...
   telnet: Unable to connect to remote host: Connection refused

But then again, on Debian, lpd is shipped running the -s option which is
described as preventing lpd from listening to tcp ports, so I'm not exactly
sure how lpd accepts incoming remote jobs...

1. How exactly does lpd accept incoming remote print jobs if its configured
   with -s by default?

2. Should I be able to telnet to port 515 on a host that is correctly
   configured to accept incoming remote print jobs?

3. If lpd is supposed to bind to port 515, can anyone think of a reason why
   it's not?

Any help appreciated!

Thanks!
Pete

-- 
Every theory is killed sooner or later, but if the theory has good in it,
that good is embodied and continued in the next theory. -- Albert Einstein

GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D


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

Re: Need lilo help - won't boot Win2k

[Peter Jay Salzman]

Hi Greg,

I read the Grub documentation to figure out what your map commands do.
It was very well written!

My system now boots all operating systems, and thanks to you, I learned
a lot about Grub in the process.  Huzzah!  Thank you so much!!!:-)

Pete



On Tue 10 Aug 04, 10:54 AM, Greg Folkert <[EMAIL PROTECTED]> said:
> On Tue, 2004-08-10 at 10:20, Peter Jay Salzman wrote:
> > Hi all,
> > 
> > I have a triple boot system with three drives:
> > 
> >* /dev/hda - 80  GB PATA drive with Debian (root is /dev/hda5)
> >* /dev/hde - 250 GB SATA drive with Win2k (root is /dev/hde1)
> >* /dev/hdg - 200 GB SATA drive with Debian (root is /dev/hdg6)
> > 
> > # compact
> > boot=/dev/hda
> > lba32
> > install=/boot/boot.b
> > map=/boot/map
> > prompt
> > delay=3
> > timeout=3
> > vga=normal
> > 
> > image=/boot/vmlinuz-2.6.7
> > root=/dev/hdg6
> > label=bleedinglinux
> > read-only
> > 
> > # Debian installed kernel
> > image=/boot/vmlinuz-2.4.26-1-386
> > root=/dev/hdg6
> > label=newlinux
> > initrd=/boot/initrd.img-2.4.26-1-386
> > read-only
> > 
> > image=/boot/vmlinuz-2.6.7
> > root=/dev/hda5
> > label=pata_bleeding
> > read-only
> > 
> > # Debian installed kernel
> > image=/boot/vmlinuz-2.4.26-1-386
> > root=/dev/hda5
> > label=pata_new
> > initrd=/boot/initrd.img-2.4.26-1-386
> > read-only
> > 
> > other=/dev/hde1
> > label=win2k
> 
> I use grub for the exact reasons you point out. You *CAN* recover grub
> from a bad setup, without resorting to a boot disk.
> 
> here is a good example of what I have found that works for me, with grub
> (curses be that LILO, a)(wait it isn't talk like a pirate day...
> oops). These say WinXP, but are literally the same setup for W2K.
> 
> Either one of these works for me:
> 
> title   WinXP hd0<->hd1+hd1
> map (hd0) (hd1)
> map (hd1) (hd0)
> rootnoverify (hd1,0)
> makeactive
> chainloader +1
> 
> 
> title   WinXP 0x80<->0x81+hd1
> map (0x81) (0x80)
> map (0x80) (0x81)
> rootnoverify (hd1,0)
> makeactive
> chainloader +1
> 
> Of course, now that we are getting into grub proper now, I suggest that
> you do the "Right Thing"(tm).
> 
> I "apt-get install grub". Then I "grub-install /dev/hda". I then make
> sure no /boot/grub/menu.lst exists, then run update-grub.
> 
> Then add the WinXP thingers at the end of the menu.lst.
> 
> Read through the menu.lst, and be amazed.
> 
> 
> 
> -- 
> greg, [EMAIL PROTECTED]
> 
> The technology that is
> Stronger, better, faster: Linux



-- 
Make everything as simple as possible, but no simpler. -- Albert Einstein
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D


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

Re: Need lilo help - won't boot Win2k

[Peter Jay Salzman]

On Tue 10 Aug 04,  8:22 PM, P V Mathew <[EMAIL PROTECTED]> said:
> Peter Jay Salzman wrote:
> 
> >Hi all,
> >
> >I have a triple boot system with three drives:
> >
> >  * /dev/hda - 80  GB PATA drive with Debian (root is /dev/hda5)
> >  * /dev/hde - 250 GB SATA drive with Win2k (root is /dev/hde1)
> >  * /dev/hdg - 200 GB SATA drive with Debian (root is /dev/hdg6)
> >
> > 
> >
> I am not sure of this, but can windows boot from something other than than
> the first partition of the first hard drive?
> 
> p v mathew
 
I've heard this too, but I don't really know what it means (or whether
it's even true).

For instance, on a system with SCSI and PATA, how do you define first
drive?

Or in my case, a system with PATA and SATA, how do you define the first
drive?  Linux assigns drive letters like a, e, and g for convenience,
but the PATA and SATA drives are on different buses, so in what sense is
one "first"?

But I've always heard this.  I would like to know whether it's true once
and for all, and what "first drive" means in that context.  It would
REALLY be nice if MS Windows printed something like:

   "I'm sorry, but I can't boot because you didn't make the drive that
   I live on the primary master"

or something!   :-)

Pete

-- 
Make everything as simple as possible, but no simpler. -- Albert Einstein
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D


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

Re: Need lilo help - won't boot Win2k

[Peter Jay Salzman]

On Tue 10 Aug 04, 10:41 PM, John Summerfield <[EMAIL PROTECTED]> said:
> Peter Jay Salzman wrote:
> 
> >Hi all,
> >
> >I have a triple boot system with three drives:
> >
> >  * /dev/hda - 80  GB PATA drive with Debian (root is /dev/hda5)
> >  * /dev/hde - 250 GB SATA drive with Win2k (root is /dev/hde1)
> >  * /dev/hdg - 200 GB SATA drive with Debian (root is /dev/hdg6)
> >
> >lilo can boot my two Debian drives.  When I try to boot into Win2k, a
> >couple of extended ASCII characters (symbols like "yen" and "cents") get
> >printed, but nothing else happens.  Win2k doesn't boot.
> >
> >Using fdisk, I've made sure that the 3 partitions on hde exist, hold the
> >files they're supposed to hold, and hde1 (which holds the actual MS OS)
> >is marked bootable.
> >
> >The motherboard is an Abit NF7-S, and I'm using the sata_sil module, and
> >other than this, have had no problems with it.  SATA on Linux has been
> >great.  My BIOS is set to boot off of HDD-0.  There's an option to boot
> >off the SATA drives though.
> >
> >As a last note, I tried Grub, and it didn't boot Win2k either.  When I
> >chose Win2k from the Grub menu, I was dumped into the Grub bash-like
> >console and didn't go any further (I'm reasonably experienced with lilo,
> >but a Grub newbie).
> >
> >My lilo.conf follows, everything looks fine with it.
> >
> >Any help would be greatly appreciated!
> > 
> >
> 
> What was the drive configuration when you _installed_ w2k?

When Win2k was installed on hde, the system looked like:

   hde (SATA): empty
   hdg (SATA): empty

In other words, the PATA drive wasn't in the sytem.  I added that after
the Win2k install.

BTW, when I change the BIOS boot order from:

   boot off HDD-0

to

   boot off SATA

the sytem boots right into Win2k, and the boot looks healthy.

Thanks!
Pete

-- 
Make everything as simple as possible, but no simpler. -- Albert Einstein
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D


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

Need lilo help - won't boot Win2k

[Peter Jay Salzman]

Hi all,

I have a triple boot system with three drives:

   * /dev/hda - 80  GB PATA drive with Debian (root is /dev/hda5)
   * /dev/hde - 250 GB SATA drive with Win2k (root is /dev/hde1)
   * /dev/hdg - 200 GB SATA drive with Debian (root is /dev/hdg6)

lilo can boot my two Debian drives.  When I try to boot into Win2k, a
couple of extended ASCII characters (symbols like "yen" and "cents") get
printed, but nothing else happens.  Win2k doesn't boot.

Using fdisk, I've made sure that the 3 partitions on hde exist, hold the
files they're supposed to hold, and hde1 (which holds the actual MS OS)
is marked bootable.

The motherboard is an Abit NF7-S, and I'm using the sata_sil module, and
other than this, have had no problems with it.  SATA on Linux has been
great.  My BIOS is set to boot off of HDD-0.  There's an option to boot
off the SATA drives though.

As a last note, I tried Grub, and it didn't boot Win2k either.  When I
chose Win2k from the Grub menu, I was dumped into the Grub bash-like
console and didn't go any further (I'm reasonably experienced with lilo,
but a Grub newbie).

My lilo.conf follows, everything looks fine with it.

Any help would be greatly appreciated!

Thanks!
Pete


# compact
boot=/dev/hda
lba32
install=/boot/boot.b
map=/boot/map
prompt
delay=3
timeout=3
vga=normal

image=/boot/vmlinuz-2.6.7
root=/dev/hdg6
label=bleedinglinux
read-only

# Debian installed kernel
image=/boot/vmlinuz-2.4.26-1-386
root=/dev/hdg6
label=newlinux
initrd=/boot/initrd.img-2.4.26-1-386
read-only

image=/boot/vmlinuz-2.6.7
root=/dev/hda5
label=pata_bleeding
read-only

# Debian installed kernel
image=/boot/vmlinuz-2.4.26-1-386
root=/dev/hda5
label=pata_new
initrd=/boot/initrd.img-2.4.26-1-386
read-only

other=/dev/hde1
label=win2k


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

Re: Help: My 'e' and 'c' keys no longer work!

[Peter Jay Salzman]

On Sun 18 Jul 04, 12:56 AM, René Seindal <[EMAIL PROTECTED]> said:
> Peter Jay Salzman wrote:
> >Dear all,
> >
> >My "c" and "e" keys stopped working in xterms (rxvt, etc) on X.   They
> >work correctly on the console, so it's not a hw issue.  The only things
> >of note that happened recently are:
> 
> Well, you're not alone.  I've had the very same problem, but it went 
> away, and I didn't manage to figure out what went wrong.  I experimented 
>  with the keyboard layout preferences in gnome, and at some point it 
> went away.
> 
> You can fix it from an xterm with setxkblayout XX where the x'es should 
> be substituted with the layout you want.  In my case it is 'dk'.  Too 
> bad there's an 'e' in the command, but do an ls and paste an 'e' :-)
> 
> FWIW, it is not just you.

THANK YOU!!!

By the way, I didn't have setxkblayout on my system, but using dpkg -S
to look for commands of similar name, I found setxkbmap.  This fixed my
problem:

   $ setxkbmap us

BTW, I update my machine (Debian/testing) a couple of times a day, so I
can pretty much pinpoint this problem starting this morning.

Can you tell me which branch of Debian you're using and when the problem
started for you?  Looking at my /usr/share/doc directory, it appears
that both xlibs and xbase-clients DID get updated today.  So I'm fairly
sure the problem was the result of some kind of misconfiguration that
was pulled onto the machine via package update.

Again, thank you SO much.  You just took a HUGE load off my mind!

Pete

-- 
In theory, theory and practise are the same.  In practise, they aren't.
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D


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

Help: My 'e' and 'c' keys no longer work!

[Peter Jay Salzman]

Dear all,

My "c" and "e" keys stopped working in xterms (rxvt, etc) on X.   They
work correctly on the console, so it's not a hw issue.  The only things
of note that happened recently are:

1. My system froze this morning.  Required a reboot and fsck.  No
   "bad" error messages, just the usual "zero dtime" stuff.

   I looked for xkb related files in all my lost+found directories, but
   didn't find anything.

2. I apt-get upgraded today.  Can't remember what got upgraded last, but
   I don't think it was X related.

I tried restarting X, then tried shutting the machine, wiggling the PS/2
connector and restarting.  No dice.

However, xfree86 IS receiving events for these keys.  The output of xev
for "e" is shown below (including the output for "r" to compare it with
it with a working key):

   KeyRelease event, serial 25, synthetic NO, window 0xc1,
   root 0x58, subw 0xc2, time 368034, (45,50), root:(49,408),
   state 0x0, keycode 26 (keysym 0x0, NoSymbol), same_screen YES,
   XLookupString gives 0 bytes:  ""

   KeyPress event, serial 25, synthetic NO, window 0xc1,
   root 0x58, subw 0xc2, time 368342, (45,50), root:(49,408),
   state 0x0, keycode 27 (keysym 0x72, r), same_screen YES,
   XLookupString gives 1 bytes:  "r"

so SOMETHING is happening.  I killed X and started it again as a
different user.  The "e" and "c" keys still didn't work.  So whatever
happened is global for all users.

I also tried reinstalling all packages that have a file named "xkb":

   apt-get install --reinstall xlibs xlibs-static-pic xbase-clients
   libx11-6 xlibs-static-dev

but still no dice.  The only keyboard type warning I get when running
startx is:

   The XKEYBOARD keymap compiler (xkbcomp) reports:
   > Error:No Symbols named "microsoft" in the include file
   > "pc/us"
   >   Exiting
   >   Abandoning symbols file "default"
   Errors from xkbcomp are not fatal to the X server

The keyboard stanza of my XF86Config file is:

   Section "InputDevice"
  Identifier  "Generic Keyboard"
  Driver  "keyboard"
  Option  "CoreKeyboard"
  Option  "XkbRules"   "xfree86"
  Option  "XkbModel"   "pc104"
  Option  "XkbLayout"  "us"
  Option  "XkbVariant" "microsoft"
   EndSection

But I don't think this has anything to do with the "e" and "c" keys not
working in X.  Lastly, I use the enlightenment wm.  I tried running twm
instead, but 'e' and 'c' still don't work, so it's not a wm issue.


I'm completely and utterly out of ideas.  The X keyboard is an area of
Linux I know almost nothing about.

Can some kind soul please please please help me figure this out?

Thanks!
Pete

-- 
In theory, theory and practise are the same.  In practise, they aren't.
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D


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

[OT] (possibly): libglide3 / libGL errors

[Peter Jay Salzman]

Dear all,

Not sure if this is a problem with Debian packages or my own programming
blunders.  Please excuse my off-topic post in case this turns out to be
my errors.


I've been playing with GLUT and OpenGL.  I get the following run-time
error message:

libGL error: can't find Glide library, dlopen(libglide3-v5.so) and
   dlopen(libglide3.so) both failed.
libGL error: dlerror() message: /usr/lib/libglide3.so: undefined symbol:
   _trisetup_Default_win_nocull_valid

these are two different issues, and both are perplexing.


The first error is odd because libglide3.so exists (it's not a dangling
symlink) in /usr/lib.   Man ld.so says /usr/lib is automatically
searched by dlopen() without needing an entry in /etc/ld.so.conf.  So
how the heck isn't dlopen() finding it?


I'm not sure what to do about the second error.  It sounds like there
might be a missing dependency, but I've checked.  Can this be a problem
with the libglide3-dev package?


The hardware is a Voodoo 5 on Sarge.  If I try to link my program
statically, all hell breaks loose with errors that look like

   undefined reference to `XInternAtom'

This is the makefile showing how I'm building the application:

   CC  = colorgcc
   CFLAGS  = -g -W -Wall -std=c99
   
   SRCS= $(wildcard *.c)
   OBJS= $(patsubst %.c, %.o, $(SRCS))
   PROGS   = $(patsubst %.c, %, $(SRCS))
   LDLIBS  = -lglut
   
   all: $(PROGS)
   
   clean:
$(RM) $(PROGS) $(OBJS)

I've scratched my head on this for at least an hour.  Any help greatly
appreciated!

Thanks!
Pete

-- 
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D


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

Re: Configuring a new system that shipped w/ Debian?

[Peter Jay Salzman]

begin Grant Edwards <[EMAIL PROTECTED]> 
> On Thu, Apr 25, 2002 at 09:18:30AM -0700, Peter Jay Salzman wrote:
> 
> > > > ahh.  it looks like fileutils owns /bin/ls!:)
> > > 
> > > If I were worried about configuring the contents of
> > > /bin/ls, I'd be set.  :)
> > > 
> > > # dpkg -S /etc/network/interfaces
> > > dpkg: /etc/network/interfaces not found.
> > > 
> > > # dpkg -S /etc/resolv.conf   
> > > dpkg: /etc/resolv.conf not found.
> > > 
> > > # dpkg -S /etc/hosts
> > > dpkg: /etc/hosts not found.
> > > 
> > > # dpkg -S /etc/hostname
> > > dpkg: /etc/hostname not found.
> > > 
> > > # dpkg -S /etc/passwd
> > > dpkg: /etc/passwd not found.
> > 
> > that seems to indicate that these files aren't owned by any package.
> > something like /etc/hostname is no more owned by a package than, say,
> > /etc is.
> 
> I assumed that they were put there by the "configure" stage of
> some package, and that I need to run dpkg-reconfigure on the
> right package.

as i understand it (and this much i'm fairly sure of) each package file
has a list of directories it needs to install to.  in the installation
process of that package, if the directory doesn't exist, the directory
is made.

> Perhaps not -- maybe there is no built-in way
> to reconfigure a Debian system.
 
whoa there, fella.   what exactly do you mean "reconfigure"?   every
time you change a hostname or a password, you're, in a sense,
reconfiguring the system.  in fact, there's "dpkg-reconfigure" as you
pointed out, which reconfigures packages.

do you really mean "no way to generate the base filesystem"?   what
exactly happened?   what are you trying to do?

short of rm -rf'ing /var/lib/dpkg (and not having a backup), i kind of
doubt there's any problem debian can't handle.

pete


> > i'm not sure *how* they get on the system, but my guess is that
> > it's simply part of the install process, as opposed to the
> > "package installation" process which occurs after the base
> > system is installed.
> 
> Could be.


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

Re: Configuring a new system that shipped w/ Debian?

[Peter Jay Salzman]

begin Grant Edwards <[EMAIL PROTECTED]> 
> On Wed, Apr 24, 2002 at 03:32:32PM -0700, Peter Jay Salzman wrote:
> > begin Grant Edwards <[EMAIL PROTECTED]> 
> > > On Wed, Apr 24, 2002 at 05:08:54PM -0500, Grant Edwards wrote:
> > > 
> > > > Doing a dpkg --configure  won't work because the
> > > > packages have already been configured.
> > > 
> > > I just stumbled across dpkg-reconfigure (a mention of
> > > dpkg-reconfigure on the dpkg man page might be a good idea).
> > > Now, if only there was some way to figure out what package
> > > owned a particular file...
> >  
> > i wonder which package owns "/bin/ls"?   let me see...
> > 
> > % dpkg -S /bin/ls
> > fileutils: /bin/ls
> > 
> > ahh.  it looks like fileutils owns /bin/ls!:)
> 
> If I were worried about configuring the contents of
> /bin/ls, I'd be set.  :)
> 
> # dpkg -S /etc/network/interfaces
> dpkg: /etc/network/interfaces not found.
> 
> # dpkg -S /etc/resolv.conf   
> dpkg: /etc/resolv.conf not found.
> 
> # dpkg -S /etc/hosts
> dpkg: /etc/hosts not found.
> 
> # dpkg -S /etc/hostname
> dpkg: /etc/hostname not found.
> 
> # dpkg -S /etc/passwd
> dpkg: /etc/passwd not found.

that seems to indicate that these files aren't owned by any package.
something like /etc/hostname is no more owned by a package than, say,
/etc is.

i'm not sure *how* they get on the system, but my guess is that it's
simply part of the install process, as opposed to the "package
installation" process which occurs after the base system is installed.

pure conjecture, but it sounds right!   :)

pete


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

Re: Downgrade ...

[Peter Jay Salzman]

i seem to recall someone saying on this list that it's not possible to
downgrade, but that doesn't sound right to me.

change unstable to testing in /etc/apt/sources.list (or sid to woody),
do a "dselect update".   then do an "apt-get dist-upgrade".

apt will print out what it thinks you want it to do.

if you don't like what you see, you can always back off by changing
testing back to unstable in sources.list and redo "dselect update".


note: "dselect update" implies "apt-get update", but not the other way
around.

pete


begin STOJICEVIC Edi EXPSIA <[EMAIL PROTECTED]> 
> Hi,
> 
> I would like to "downgrade" my Sid to Woody but I dont know
> how to do it exactly...
> I know that we can "force" apt to upgrade with one version
> (stable/testing/unstable)
> but I dont remember what file I need to configure ?
> 
> Also I didnt do an apt-get update since 2 weeks ago ... What will happen
> if I change my source.list from Sid to Woody ? Will it be okay ? 
> 
> E.
> 
> This message and any attachments (the "message") are confidential and
> intended solely for the addressees.
> Any unauthorised use or dissemination is prohibited. 
> E-mails are susceptible to alteration.   
> Neither SOCIETE GENERALE nor any of its subsidiaries or affiliates shall be 
> liable for the message if altered, changed or falsified. 
 
these types of sigs always crack me up...   like, who do i contact for
permission to forward this email to my cousin?:)


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

Re: Spamassassin tests help please

[Peter Jay Salzman]

dude -- you are SO ready for the open relay database.

you sound pretty harried.  that's where i was a few months ago.  not to
sound overly dramatic, but www.ordb.org changed my life.

also, i've been compiling a list of networks that send spam from asian
countries like china and korea.  when i get 3 pieces of spam from the
same network, and my letters of complaint go unanswered, i block the
entire network using tcpwrappers.   my /etc/hosts.deny contains a vast
number of chinese and korean networks.   you can manage exim connections
with tcpwrappers by simply running exim as:

smtp stream   tcp   nowait   mail /usr/sbin/tcpd  /usr/sbin/exim -bs

in inetd.conf.   my 3 pronged approach to spam is:

1. using ordb.org
2. running exim from tcpwrappers and dumping IP's into /ect/hosts.deny
3. spamcop

how effective is this?  i was getting *upwards* of 40 pieces of spam per
day.

today i got simply 4 pieces of spam, and this is what i would call a
"heavy spam day".

pete


begin Patrick Kirk <[EMAIL PROTECTED]> 
> Hi all,
> 
> I have given up on using my .forward as a spam filter because I've now
> gone up to over 40 spam pieces a day and its a pain to keep adding
> conditions on each .forward on each account.
> 
> Just to make clear, my particular desire to stop stuff from Korean and
> Taiwan is that I speak neither Korean nor Chinese.  
> 
> 
> I wonder if anyone can help with these tests:
> 
> 1. I am on numerous Korean spam lists.  So I want to exclude all email
> with Korean charsets.  How do I set $h_Content-Type: contains
> "ks_c_5601-1987" to score 20?
> 
> 2. I get a lot of stuff from Taiwan.  Is it poossible to simply
> blacklist all mail relayed from ISPs with .tw tld?
> 
> 3. How can I blacklist specific names?  For example, esavingszone send
> me two messages every day and I want them automatically blocked.  But
> they use differing domain nemaes so I want to block
> [EMAIL PROTECTED] [EMAIL PROTECTED] and every other
> [EMAIL PROTECTED]
> 
> 4. The ISP that uses hanmail.net and daum.net is the single worst
> offender.  Can I block all mail relayed theough these domains?
> 
> 
> Thanks in advance,
> 
> Patrick
> 
> 
> -- 
> To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
> 


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

Re: printing problems

[Peter Jay Salzman]

begin dman <[EMAIL PROTECTED]> 
> On Wed, Apr 24, 2002 at 03:10:47PM -0700, David Smead wrote:
> | 
> | I'm apparently missing a driver.
> | 
> | knuth:~# lpq
> ...
> |  Status: cannot open '/dev/lp0' - 'No such device'
> 
> | -
> | 
> | knuth:~# ls -l /dev/lp0
> | crw-rw1 root lp 6,   0 Jun 13  2001 /dev/lp0
> 
> This is meaningless.  (well, all it means is you have an inode named
> "lp0".  You've got lots of inodes in /dev that you don't have hardware
> for)

meaningless?  the major and minor numbers are correct.  the kernel
certainly knows what major number 6 is.  the parport driver (if present)
knows what minor number 0 is.

doesn't matter what you call the *file*.  it's the major and minor
numbers which matter.


> If you use devfs this becomes meaningful (the file won't exist
> unless the device and driver do).
 
maybe.  maybe not.

depends on how devfsd is configured, doesn't it?

p


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

Re: Configuring a new system that shipped w/ Debian?

[Peter Jay Salzman]

begin Grant Edwards <[EMAIL PROTECTED]> 
> On Wed, Apr 24, 2002 at 05:08:54PM -0500, Grant Edwards wrote:
> 
> > Doing a dpkg --configure  won't work because the
> > packages have already been configured.
> 
> I just stumbled across dpkg-reconfigure (a mention of
> dpkg-reconfigure on the dpkg man page might be a good idea).
> Now, if only there was some way to figure out what package
> owned a particular file...
 
i wonder which package owns "/bin/ls"?   let me see...

% dpkg -S /bin/ls
fileutils: /bin/ls

ahh.  it looks like fileutils owns /bin/ls!:)

i have a debian package tutorial that i'm writing.  it's incomplete, but
is still very useful:

http://www.dirac.org/linux/debian

pete


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

Re: jigdo question

[Peter Jay Salzman]

ok, jigdo just worked for me.

i think the problem was that i had downloaded the .jigdo files the
previous night, then used them the next morning.  the files get
generated nightly, and woody gets updated every day.

i'm guessing that if the .jigdo file doesn't match what's on the woody
mirror, jigdo freaks out.  it starting interpreting md5sums as file
names(!).

but jeez -- the download was only a few hours for me, as opposed to over
a day for an ISO download.  anyone who downloads the woody ISO rather
than using jigdo is, well, nuts.  this is a *vast* improvement over
cd-image.  i think it still needs a little development, but we have a
good thing going on.  when jigdo becomes stable, it'll be an awesome,
awesome tool.

woo hoo!

pete


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

Re: jigdo question

[Peter Jay Salzman]

for what it's worth, i've written up my experience with jigdo:

http://www.dirac.org/linux/debian/jigdo/

note that i haven't been successful -- at least, i don't think so.  it
creates an ISO image which i can burn, but during the download process,
i got some messages like:

  --11:11:00--
  ftp://ftp.fsn.hu/pub/debian-superseded/_uPSKNHm3q3gFry50rhCkw
  => `_uPSKNHm3q3gFry50rhCkw'
  Resolving ftp.fsn.hu... done.
  Connecting to ftp.fsn.hu[195.228.253.129]:21... connected.
  Logging in as anonymous ... Logged in!
  ==> SYST ... done.==> PWD ... done.
  ==> TYPE I ... done.  ==> CWD /pub/debian-superseded ...
  done.
  ==> PASV ... done.==> RETR _uPSKNHm3q3gFry50rhCkw ... 
  No such file `_uPSKNHm3q3gFry50rhCkw'.
  
  (repeat a bunch of times)
  
  -
  Aaargh - not all files could be downloaded. This should not
  happen! Depending on the problem, it may help to restart the
  download.
  Also, you could try changing to another Debian or Non-US server,
  in case the one you used was not updated recently.
  
  However, if all the files downloaded without errors and you
  still get this message, it means that the files changed on the
  server. For the moment, in the case of Debian CD images the only
  solution when that happens is to revert to rsync.
  
  Press Return to restart.


it looks like jigdo is interpreting md5sums as filenames.  the
woody-i386-2.raw.tmp file is 659141240 bytes large, so something must
have downloaded correctly.  the /tmp directory got deleted, and i take
this to mean the download ended in some normal, sane sense.

wahhh.  i really want this to work!   :-(

pete


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

Re: jigdo question

[Peter Jay Salzman]

begin curtis <[EMAIL PROTECTED]> 
> Just as this issue was being brought up I was using jigdo for the first 
> time. Now, I'm just wondering what is the next step. Everything seems to 
> have finished perfectly and I find that I have the following files:
> woody-i386-8.raw.template
> woody-i386-8.raw.jigdo
> woody-i386-8.raw
> jigdo-file-cache.db
> 
> Am i to understand that the woody-i386-8.raw is the image file and all 
> that I need to put on a CD in order to make a bootable woody installation?
 
i believe so.  i was able to mount this image on the loopback device:

[EMAIL PROTECTED] su
Password: 
satan# mount -o loop woody-i386-2.raw.tmp  /mnt/other/
satan# ls /mnt/other/
README.html  README.mirrors.txt  README.txt  boot/dists/
pics/
README.mirrors.html  README.non-US   TRANS.TBL   debian@  md5sum.txt
pool/

i'm pretty sure it's an ISO image.

pete


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

Re: jigdo question

[Peter Jay Salzman]

begin Haroldo Stenger <[EMAIL PROTECTED]> 
> Peter Jay Salzman wrote:
> > begin Haroldo Stenger <[EMAIL PROTECTED]>
> > >
> > > Don't rely on this file size's change. It doesn't have any relation with
> > > download progress, neither with completeness-to-the-moment of the
> > > image itself.  The image will be complete when jigdo tells you that.
> > 
> > it would've been nice for the docs to mention this!
> 
> Yes, I think we could contibute some changes to the docs.
 
yes, once i get some of this figured out better, i plan on sending some
diffs.  to be honest, when it started to "work" i was really surprised.
i was just dicking around.  i can't even remember what i did.  lol!!

> > i suspect people aren't use jigdo yet.  i'm the first person to try it
> > from our user group, and we have some pretty hard core debian fans in
> > the group.
> > 
> 
> Yes, this must be the case. It happens that many people are long time Debian
> users, and don't need to make fresh full CD images of woody, they just use
> apt-get and turn their systems into full blown woody/sid ones. Just the ones 
> of
> us who need to install in many computers and haven't bought images somewhere
> *need* jigdo.
 
we (the linux users' group of davis) need the woody images.  we hold
monthly installfests, but don't have net connections where the IF's take
place.  often we get very new laptops with some wierd video chipset
which is totally unsupported by X 3.*.   either we install
redhat/mandrake/suse or we need woody.  roughly 1/4 of us use debian,
including me.   we really dislike working on any system other than debian.
the other distros don't make sense to us for various reasons.  even
small stupid things -- like searching 15 minutes for an rpm until
realizing that redhat rpm's can contain uppercase letters.  every so
often, i waste 10 minutes of my life before remembering redhat has
"XFree86.rpm" rather than "xfree86.rpm".  bleah.

but for these laptops, potato is simply unsuitable.

> Let's continue in touch
> 
> Uruguay (uy) is a country with coasts to the Atlantic Ocean and the R?o de la
> Plata, east Argentina, and south Brazil. We have a very active Linux community
> here, and some of us are Debian users. Thanks for asking!
 
cool.  i love the fact that people can be friends beyond political and
cultural boundaries.  if you forget about the spam and commercialism,
the internet is a wonderful place.   :)yes, let's keep in touch.
you sound like a very cool guy!

oops -- just noticed that the woody download just finished.  some files
didn't download.  off to investigate...   :)

pete


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

Re: jigdo question

[Peter Jay Salzman]

hi haroldo,

thank you for answering my question!

begin Haroldo Stenger <[EMAIL PROTECTED]> 
> 
> Don't rely on this file size's change. It doesn't have any relation with
> download progress, neither with completeness-to-the-moment of the
> image itself.  The image will be complete when jigdo tells you that.
 
it would've been nice for the docs to mention this!
 
> I don't understand why the Debian community seems to ignore all
> jigdo-related questions. Jigdo is being advertised in the website, but
> when one asks about a doubt on the mailing-lists, or in the #debian
> irc channel, these questions don't receive any attention at all.

i noticed this -- before posting to debian-user, i looked through the
archives of debian-user and debian-cd.  it was /very/ difficult to find
any question which got a reply.   to be frank, i wasn't expecting an
answer.  i guess i got lucky.   thank you!   :)

> I wonder what's happening here. Is it that people don't rely on Jigdo
> yet to download, ignoring the website's note on the convenience of
> using Jigdo over full CD image downloading? Is it that there isn't
> enough Jigdo knowledge yet? Is it just my own biased perception?

i suspect people aren't use jigdo yet.  i'm the first person to try it
from our user group, and we have some pretty hard core debian fans in
the group.

the connection to hungry blows big chunks.  it took almost 2 days to
download just binary-1.  after reading the debian website, i got sold on
jigdo.  but i have the oddest feeling that very few people are actually
using it.

> Not trying to be rude, just to sort of shake the issue in order to
> obtain some feedback!
> 
> Best regards, Haroldo.

thanks for the words of confidence!   one last question -- if we can't
judge the file size of the .tmp file, is there any way of gauging how
far along the download i'm at?  jigdo's been doing its thing for nearly
12 hours.  i have no concept of whether i'm at the 9% completed mark or
the 99% completed mark.

do you know of any way to tell?   at least with binary images you have
some idea of an estimated time to completion.


thanks!
pete

ps- btw, where is uy?  i don't recall seeing it before.


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

jigdo question

[Peter Jay Salzman]

hi all,

i'm using jigdo to download woody for the first time.

there's a directory ./tmp which is holding files which get flushed every
so often.  right now it holds:

[EMAIL PROTECTED] ls tmp
ccmalloc_0.3.8-1_i386.deb psutils_1.17-15_i386.deb
libmailtools-perl_1.42-2_all.deb  svgalibg1_1.4.3-7_i386.deb

wating a bit, now it holds:

[EMAIL PROTECTED] ls tmp/
ch-boot-floppy-techinfo.ca.html  x-ttcidfont-conf_11_all.deb

so the files are getting flushed somewhere.  presumably, jigdo is making
a cd image somewhere.   there's a file called woody-i386-1.raw.tmp in
the same directory as tmp, but it's size hasn't changed in awhile, so
i'm not sure this is it.

does anyone know what the default location is for jigdo-lite to place
the woody cd images?

pete


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

[OT] what ports do quake3 use?

[Peter Jay Salzman]

hi there,

i want to run a quake3 server from behind a firewall.

what ports do i need to open up and forward in order to run a quake 3
server?

pete

question about building a deb package

[Peter Jay Salzman]

dear all,

in the debian new maintainer's guide:

chapter 3:

  Note that if your program uses GNU automake and/or autoconf, meaning
  the source includes Makefile.am and/or Makefile.in files,
  respectively, you will need to modify those files.

chapter 3, section 3.1:

  Basically, you need to make the program install in debian/tmp, but
  behave correctly when placed in the root directory, ie when installed
  from the .deb package.  With programs using GNU autoconf, this will be
  quite easy, because dh_make will set up commands for doing that
  automatically.


these two statements seem contradictory.  to get the program to install
in debian/tmp, can dh_make automatically do this for you when the
upstream package uses autoconf or not?

the first paragraph definitely says no--you need to modify the
Makefile.am and Makefile.in files yourself.  the second paragraph seems
to suggest that dh_make does this for you.

which paragraph is correct?

also, section 5.4 (on man pages) isn't very clear.  the author simply
says he wrote a manpage, but doesn't say what he had to do to get the
manpage in the package.   did he simply replace debian/manpage.1.ex by
what he wrote?  or did he have to run dh_installman?

one of the problems i've run into is that the package i'm trying to
debianize comes with a manpage that gets installed by the upstream make
process.  lintian didn't seem to recognize this and reported that the
binary has no manpage in my deb package.   what is the procedure when
the upstream source comes with its own man page that gets installed
during the make process?

thanks,
pete


-- 
The mathematics [of physics] has become ever more abstract, rather than more
complicated.  The mind of God appears to be abstract but not complicated.
He also appears to like group theory.  --  Tony Zee's `Fearful Symmetry'

PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D

Re: this debconf stuff is crazy!

[Peter Jay Salzman]

begin Shaya Potter <[EMAIL PROTECTED]> 
> On Sun, 2002-01-06 at 19:36, Peter Jay Salzman wrote:
> > begin Justin R. Miller <[EMAIL PROTECTED]> 
> > > Thus spake Peter Jay Salzman ([EMAIL PROTECTED]):
> > > 
> > > > ok, my XF86Config-4 is all wrong now.  i need to edit this file.  if
> > > > we're not supposed to edit between the "BEGIN DEBCONF" and "END
> > > > DEBCONF", how the hell am i supposed to get a working XFree86 again?
> > > > 
> > > > is there a tool that allows us to change this file?
> > > 
> > > Try 'dpkg-reconfigure xserver-xfree86'. 
> > 
> > just to be very clear on the issue, this isn't acceptable to me.
> > 
> > i have very special mode timings and options that i use that aren't
> > available if i reconfigure the package using dpkg.  this stuff needs to
> > be put in by hand.
> > 
> > i like the chattr idea.  just when you thought debian does the Right
> > Thing, they start fscking up by automating somthing which really doesn't
> > require automating...
> > 
> > (debian is still the best.  just slightly less better).
> 
> what were you doing during the upgrade of the package?
> 
> It clearly asked me if I wanted debconf to configure it.
> 
> It even keeps a backup of the original, if you were doing it by hand and
> accidently said yes.
> 
> sheesh.
 
no shit.

but that's really not the point.  the point is there should've been some
kind of message saying something to the effect of:

   note: if you let debconf take over your config file, you won't be
able to modify the config file yourself.  you give up all rights
to tweak it yourself by hand

i wouldn't believe that *debian* would do this.  it's just bad medicine.
yast, yes.  linuxconf, yes.  debian?  no.

"recovering"  was no big deal.  like you pointed out, it saved a copy of
the old file.  but that's like congratulating someone for taking a shit
in toilet.  you'd *expect* it to go into the toilet.  if a copy weren't
saved, then that would be an excellent reason to switch to another
distribution.  immediately.   so i don't think giving it a "pat on the
back" for backing up the original is appropriate here.

i'm not saying it sucks completely.  the whole ordeal was resolved in
under a minute after i realized what i got myself into.  i'm just saying
i was expecting better.  perhaps a better solution would've been
something like what we do with modules.conf.  let the distribution take
it over, but give the user the opportunity to modify it at will.  i LIKE
what debian does with modules.conf.  it's one of the most intelligent
solutions iv'e seen to automation vs control.

peter

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: this debconf stuff is crazy!

[Peter Jay Salzman]

begin Justin R. Miller <[EMAIL PROTECTED]> 
> Thus spake Peter Jay Salzman ([EMAIL PROTECTED]):
> 
> > ok, my XF86Config-4 is all wrong now.  i need to edit this file.  if
> > we're not supposed to edit between the "BEGIN DEBCONF" and "END
> > DEBCONF", how the hell am i supposed to get a working XFree86 again?
> > 
> > is there a tool that allows us to change this file?
> 
> Try 'dpkg-reconfigure xserver-xfree86'. 

just to be very clear on the issue, this isn't acceptable to me.

i have very special mode timings and options that i use that aren't
available if i reconfigure the package using dpkg.  this stuff needs to
be put in by hand.

i like the chattr idea.  just when you thought debian does the Right
Thing, they start fscking up by automating somthing which really doesn't
require automating...

(debian is still the best.  just slightly less better).

pete



-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

this debconf stuff is crazy!

[Peter Jay Salzman]

after upgrading woody today, my XF86Config-4 file looks like:

   ### BEGIN DEBCONF SECTION
   #
   # If you want your changes to this file preserved by dexconf, only make
   # changes before the "### BEGIN DEBCONF SECTION" line above, and/or after
   # the "### END DEBCONF SECTION" line below.
   
   (all of the meat and potatoes of XF86Config-4.conf is snipped)
   
   ### END DEBCONF SECTION


ok, my XF86Config-4 is all wrong now.  i need to edit this file.  if
we're not supposed to edit between the "BEGIN DEBCONF" and "END
DEBCONF", how the hell am i supposed to get a working XFree86 again?

is there a tool that allows us to change this file?

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: XMMS bug

[Peter Jay Salzman]

jesus.  speak of the devil.  i just spent about an hour on this today, in
fact.  you guys are mind readers.

i'm using OSS drivers (not alsa).  xmms kept segfaulting.  using the process
of elimination, i determined it was was the alsa plugin.

perhaps we should file a bug report..

pete

begin Markus Lechner <[EMAIL PROTECTED]> 
> Am Donnerstag, 20. Dezember 2001 20:46 schrieb Steven Kurylo:
> > Running debian woody 2.4.12
> >
> > I use XMMS all the time and have never had a problem with it.  Today I
> > decided to play around with plug-ins so I did a
> > "apt-get install `apt-cache search xmms plugin | cut -f1 --delimiter=\
> >
> > |sed s#smpeg-xmms##`" to get all the plugin packages.  I did the sed
> >
> > part since smpeg-xmms wouldn't install because libsdl1.2 wasn't availible.
> >
> > Once all the packages are installed and I try to run xmms I get a
> > segmentation fault.  If I do remove with the command above to remove all
> > plug ins, then xmms runs fine.  I don't have time right now to go
> > through each plug in to see which crashes xmms, but I was wondering if
> > anyone else has thing problem.
> >
> > Steven
> 
> I had such a problem. Started XMMS and got a segfault. Ran DDD and watched 
> what was happening.
> 
> In my case i am using OSS and no ALSA. But XMMS had an ALSA library installed 
> under /usr/lib/xmms/Output/ and tried to load it and so it crashed.
> 
> I manually removed this library and from this moment on everything ran fine :)
> 
> Hope this helps.

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

unresolved symbol in svgalib - a bug?

[Peter Jay Salzman]

hi there,

whenever an application uses svgalib, i get the error message:

(application): relocation error:
/usr/lib/libc5-compat/libvga.so.1: undefined symbol _xstat

i checked google and debian-user archives -- one person posted the same
message awhile back, but no one replied to his post for help.

i also checked the debian bug system.  no bugs were filed related to this
problem.

anyone have any ideas how to resolve the problem?  if not, i'm going to file
a bug report against svgalib1.

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: exim / rbl question

[Peter Jay Salzman]

begin Noah Meyerhans <[EMAIL PROTECTED]> 
> On Thu, Dec 13, 2001 at 10:41:36PM -0800, Peter Jay Salzman wrote:
> > 
> > when i installed debian and configured exim, i was under the impression that
> > exim was configured to reject any email from an IP address listed in the 
> > maps
> > rbl database.
> > 
> 
> I don't think it's configured by default to use any kind of blacklist at
> all.  Search for 'rbl' in /etc/exim.conf.

i did, and it's definitely in there.  saw it with my own two eyes.

> Then read
> /usr/share/doc/exim/spec.txt.gz to figure out how to configure rbl use.

i've resolved the problem -- ordb.org has great instructions on how to
configure (and test!) exim to use their open relay database.  very easy to
follow.

pete

Re: exim / rbl question

[Peter Jay Salzman]

begin Mario Vukelic <[EMAIL PROTECTED]> 
> On Fri, 2001-12-14 at 07:41, Peter Jay Salzman wrote:
> > hi there,
> > 
> > when i installed debian and configured exim, i was under the impression that
> > exim was configured to reject any email from an IP address listed in the 
> > maps
> > rbl database.
> 
> > rbl_domains = rbl.maps.vix.com
> > rbl_reject_recipients = true
> > rbl_warn_header = false
> > 
> > this looks good to me.  can someone guess why i'm still getting email that
> > should be rejected by the rbl database?
> 
> How do you get the mail to exim? If you use fetchmail to pull it from
> your ISP's mail server, it won't work (FAQ at www.exim.conf)
> 
> Kind regards, M.
 
oops, sorry.  i guess that's a very important detail!

email comes directly to my site, dirac.org, which is hooked up to the net via
DSL.  no fetchmail, no ISP's, no .forward.

that's what's so mysterious to me.

do you have any ideas?

pete

-- 
   ** Please don't send me html email **
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

exim / rbl question

[Peter Jay Salzman]

hi there,

when i installed debian and configured exim, i was under the impression that
exim was configured to reject any email from an IP address listed in the maps
rbl database.

this evening i was reading the map-rbl site, and came across:

Testing Your RBLSM Configuration
Russell Nelson has put together an auto-responder that you may use to 
test
your RBLSM implementation.  His instructions are:

Send mail to [EMAIL PROTECTED] from the server whose RBLSM block
you are testing.  Expect one reply from ns.crynwr.com with the SMTP
conversation.  If you get another reply from linux.crynwr.com, your 
RBLSM
filter is broken.

i followed the instructions and apparently, i'm still getting mail from
blacklisted addresses -- the 2nd email from linux.crynwr.com arrived.

next, i looked at exim.conf:

# Exim contains support for the Realtime Blocking List (RBL) that is 
being
# maintained as part of the DNS. See http://maps.vix.com/rbl/ for
# background. Uncommenting the following line will make Exim reject mail
# from any host whose IP address is blacklisted in the RBL at 
maps.vix.com.

rbl_domains = rbl.maps.vix.com
rbl_reject_recipients = true
rbl_warn_header = false

this looks good to me.  can someone guess why i'm still getting email that
should be rejected by the rbl database?

thanks!
pete

-- 
   ** Please don't send me html email **
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: printing problems

[Peter Jay Salzman]

begin: [EMAIL PROTECTED] <[EMAIL PROTECTED]> quote
> Hi,
> I have a hewlett packard deskjet  697C.  I have been able to print from off 
> the Web on this printer before.  In the past day, every time I print, the 
> paper just run through without any printing on it.  It does even try to 
> print.  Please need help or any suggestions?

if it spits out paper, your printer is recognized by the kernel AND the
spooler is working.   this narrows down the problem considerably.

exactly what application are you using to print the blank pages?

try another application.   can other applications print ok?

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: No problem with staroffice 6.0b (was: Debian woody with Openoffice. Can not install)

[Peter Jay Salzman]

my guess is that he's going to read the subject and delete your email before
reading the content.  you might want to try again with a subject that has his
name to bring it to his attention.  he's pretty frustrated right now.

although, good luck.  i've already given him the same instructions.  he
either wants someone to do it for him or barely knows what "email" means.

peter

begin: Karsten M. Self  quote
> on Sun, Dec 02, 2001 at 03:58:38PM -0800, Tran Nam Binh ([EMAIL PROTECTED]) 
> wrote:
> > Hackers have put my user id into some redistributing list of your
> > technical forum.
> >
> > I can't unsubcribe with automated system because my user id is not on
> > the main list.
> >
> > Please help.  I received tons of unwanted mails.
> >
> > Please forward this request to the list owner
> > Thanks 
> 
> Post a copy of a received message with *full headers* to
> [EMAIL PROTECTED], in the event you aren't being directly
> subscribed to the list but are being passed messages by a "friendly"
> forwarder.



-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: help: tcpwrappers aren't working!!

[Peter Jay Salzman]

begin: Michael Heldebrant <[EMAIL PROTECTED]> quote
> On Sun, 2001-12-02 at 14:41, Peter Jay Salzman wrote:
> 
> That's ... interesting, have you looked at the output of tcpdchk -v for
> possible errors in the hosts files?  It should also explain in great
> detail the access control for cvspserver.
 
this was a great suggestion.  i forgot all about tcpdchk.  yes, the problem
was "cvspserver" instead of "cvs-pserver".

i KNEW it had to be something simple.

thanks mike!

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: via82cxxx chipset question

[Peter Jay Salzman]

begin: Chanop Silpa-Anan <[EMAIL PROTECTED]> quote
> Once in debian galaxy, I heard Cheryl Homiak say
> 
> > It's possible this might get more answers on another list, but thought I'd
> > try here first.
> > I have a via82c597 rev 4 chipset and an Award bios. I have anabled the
> > via82cxxx support in my kernel, but perhaps I do not understand what
> > numbers to put in on the command line, or for which ide to put them in,
> > because no matter what I do a 33mhz bus speed is assumed; if I try putting
> > 66 even in I get something about this being impossible.
> > I have included part of my dmesg here, and have included enough so you can
> > see my setup and a couple of other messages that puzzle me.
> > Any help would be appreciated.
> 
> 33mhz PCI bus speed is correct. Don't bother changing it. I've heard
> that someone has tried and has corrupted his HD.
> 
> To check the effect try your kernel with and without via ATA support.
> check with hdparm -tT /dev/hd? to check for performance gain. When you
> have via ATA support, you should be able to get higher performance
> trough dma tranfer.
 
approximately what would hdparm -tT report when udma is enabled?

i got the same values with and without udma support.  it was explained to me
that this is because there's a difference between a burst rate and sustained
rate of data transfer.

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: help: tcpwrappers aren't working!!

[Peter Jay Salzman]

mike, thank you!

begin: Michael Heldebrant <[EMAIL PROTECTED]> quote
> On Sun, 2001-12-02 at 12:22, Peter Jay Salzman wrote:
> > begin: Peter Jay Salzman <[EMAIL PROTECTED]> quote
> 
> Have you restarted inetd after changing your hosts files?  Have you also
> killed any reamining cvspserver processes after reloading inetd?
 
i did.   and just to make ABSOLUTELY sure, i did the unthinkable.  i rebooted
after weeks of uptime.  *sigh*.   i can still access the pserver from
belial.ucdavis.edu

> Does cvspserver try and run it's own standalone daemon from any
> scripts?  netstat -atp as root should show that inetd is listening for
> the cvsperver ports, if it's not then you know it's starting from
> somewhere else.
 
cute.  i never knew about the p switch.

unfortunately, pserver is being run from inetd:

Local Address   Foreign Address State PID/Program name   
*:cvspserver*:* LISTEN 178/inetd
*:printer   *:* LISTEN 182/lpd
*:time  *:* LISTEN 178/inetd
*:finger*:* LISTEN 178/inetd
*:sunrpc*:* LISTEN 103/portmap
*:x11   *:* LISTEN 276/X
*:www   *:* LISTEN 211/apache
*:ftp   *:* LISTEN 178/inetd
*:ssh   *:* LISTEN 191/sshd
*:smtp  *:* LISTEN 178/inetd
satan.diablo.loca:32771 belial.ucdavis.edu:ssh  ESTABLISHED 300/ssh


pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: help: tcpwrappers aren't working!!

[Peter Jay Salzman]

begin: Peter Jay Salzman <[EMAIL PROTECTED]> quote
> tcpwrappers don't seem to be working for cvspserver.
> 
> on host satan (64.164.47.8), i have the following wrappers:

bad form to reply to your own message, but someone will ask.  from
inetd.conf:

cvspserver stream tcp nowait.40 root /usr/sbin/tcpd /usr/sbin/cvs-pserver 

help: tcpwrappers aren't working!!

[Peter Jay Salzman]

tcpwrappers don't seem to be working for cvspserver.

on host satan (64.164.47.8), i have the following wrappers:


  # ## hosts.allow #
  ALL: localhost
  ALL: 192.168.0.1  # mephisto
  ALL: 192.168.0.2  # satan
  ALL: 192.168.0.3  # navalle
  ALL: 192.168.0.4  # lucifer
  
  # cvspserver: 169.237.43.86 # belial.ucdavis.edu
  cvspserver: 130.88.22.5   # John Levon (mpg cvs)


  # ## hosts.deny  #
  portmap:ALL
  lockd:  ALL
  mountd: ALL
  rquotad:ALL
  statd:  ALL
  cvspserver: ALL


yet belial.ucdavis.edu can access the pserver just fine:

   belial% cvs -d ':pserver:[EMAIL PROTECTED]:/usr/local/cvs' login
   Logging in to :pserver:[EMAIL PROTECTED]:2401/usr/local/cvs
   CVS password: 
   belial% 

i ran inetd in debug mode, and it didn't give too much information.  the
"someone" in the logs is belial.ucdavis.edu, whom i commented out of
hosts.allow, and who shouldn't be allowed since i have "cvspserver: ALL" in
hosts.deny.

   satan# inetd  -d
   ADD : time proto=tcp, wait.max=0.40, user.group=root.(null) builtin=804d094
   server=internal
   ADD : time proto=udp, wait.max=0.40, user.group=root.(null) builtin=804d0a4
   server=internal
   ADD : ftp proto=tcp, wait.max=0.40, user.group=root.(null) builtin=0
   server=/usr/sbin/tcpd
   ADD : smtp proto=tcp, wait.max=0.40, user.group=mail.(null) builtin=0
   server=/usr/sbin/tcpd
   ADD : finger proto=tcp, wait.max=0.40, user.group=nobody.(null) builtin=0
   server=/usr/sbin/tcpd
   ADD : cvspserver proto=tcp, wait.max=0.40, user.group=root.(null) builtin=0
   server=/usr/sbin/tcpd
   someone wants cvspserver
   accept, ctrl 3
   2771 execl /usr/sbin/tcpd
   2771 reaped, status 0


it's very important to me to be able to restrict pserver to only one or two
collaborators.   i don't want anyone else accessing it.

does anyone have any ideas why tcpwrappers seem to be failing?

thanks,
pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: Setting up devfs

[Peter Jay Salzman]

did you enable "mount devfs at boot time" in the kernel config?

if not, did you pass the kernel the "devfs=mount" option?

pete


begin: Frederico.S.Mu?oz <[EMAIL PROTECTED]> quote
> Hi,
> 
> 
> I've recently totally reinstalled my system from scratch after losing my
> /home partition. I'm running kernel 2.4.14 with a XFS patch and xfs
> partitions (except /boot).
> 
> I've also compile devfs support in the kernel and installed devfsd.
> 
> Now, I've red the documentation on the kernel-source and the HOWTO, but
> something isn't right here. devfs apparently is working because during boot
> I see some messages that refer to the HD as
> /dev/disks/lots-of-letters-and-numbers). I also receive messages while
> connecting my USB scanner - don't
> know if that's devfs though.
> 
> The thing is, my /dev is the same as always. It doesn't have /dev/ide or any
> of the devsf directories. Its the same splattered 3232323 files as always.
> 
> If I do $devfsd /dev it complains that there is no .devfsd in the
> directory...
> 
> I have not touch fstab or anything, I simply compiled into the kernel and
> installed devfsd.
> 
> Is there some extra steps that I need to make ? Pointers to the right
> documentation and general advice are always welcome :)
> 
> thanks,
> 
> fsm
> 
> --
> Frederico S. Mu?oz
> Cap Gemini Ernst & Young : [EMAIL PROTECTED]
> IIES : [EMAIL PROTECTED]
> Debian Project: [EMAIL PROTECTED]
> 
> **
> Ever noticed something? Unix comes with compilers. Windows comes with
> Solitaire.
> **
>   -Adep
> 
> 
> --
> To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
> 

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: Radeon 7500 XFree86 support.

[Peter Jay Salzman]

begin: Ax <[EMAIL PROTECTED]> quote
> Dne so 24. listopad 2001 10:37 Karsten Heymann napsal(a):
> > > I don't have a Radeon card (yet) but I was just looking up info about
> > > it on the XFree lists and it appears that support for it is only in
> > > the XFree CVS right now. It should be included in the next official
> > > release. I think you have to use the generic SVGA support if you have
> > > a current XFree release.
> >
> > That's strange, mine is working perfectly with the XFree4 from actual
> > sid (a Radeon VE) under 2.4.9. I'm even running opengl, glxinfo prints
> > out alot and glxgears gives ~100 fps (Athlon 1000). Is that accelerated?
> 
> No ;(
> glxgears gives me more than 800 fps with Radeon GE and Athlon 1200
 
with a radeon QE and athlon 1300, glxgears broke 1000 fps.

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

vintage lilo interface

[Peter Jay Salzman]

hi there,

awhile ago, lilo changed formats -- it was just black and white text, and
now it shows a red menu that you can choose a kernel to boot.  i'd like to
get the old interface back.

is this possible?

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

exim: misleading package description?

[Peter Jay Salzman]

satan$ dpkg -p exim
(snip)
Description: Exim Mailer
 This MTA is rather easier to configure than smail or sendmail.
 It is a drop-in replacement for sendmail/mailq/rsmtp.
 Advanced features include the ability to reject connections from
 known spam sites, and an extremely efficient queue processing
 algorithm.


i'm looking at:

Advanced features include the ability to reject connections from
known spam sites

exim can use rbl to reject msgs from spam sites.  but so can every other MTA.

exim doesn't have a "spam reject" file that i can drop IP addresses into.
exim doesn't use tcpwrappers to reject SMTP sessions.

what exactly are some of these advanced features then?

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

[p: X configuration on woody]

[Peter Jay Salzman]

i'd like to obtain a list of XF86Config generating tools for woody.   can
someone list a few?

this is for 4.*, of course.  not 3.*.

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

X configuration on woody

[Peter Jay Salzman]

i'd like to have a list of XF86Config generating tools for woody.   can
someone list a few?

this is for 4.*, of course.  not 3.*.

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

exim: misleading package description?

[Peter Jay Salzman]

satan$ dpkg -p exim
(snip)
Description: Exim Mailer
 This MTA is rather easier to configure than smail or sendmail.
 It is a drop-in replacement for sendmail/mailq/rsmtp.
 Advanced features include the ability to reject connections from
 known spam sites, and an extremely efficient queue processing
 algorithm.


i'm looking at:

Advanced features include the ability to reject connections from
known spam sites

exim can use rbl to reject msgs from spam sites.  but so can every other MTA.

exim doesn't have a "spam reject" file that i can drop IP addresses into.
exim doesn't use tcpwrappers to reject SMTP sessions.

what exactly are some of these advanced features then?

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

exim: misleading package description?

[Peter Jay Salzman]

satan$ dpkg -p exim
(snip)
Description: Exim Mailer
 This MTA is rather easier to configure than smail or sendmail.
 It is a drop-in replacement for sendmail/mailq/rsmtp.
 Advanced features include the ability to reject connections from
 known spam sites, and an extremely efficient queue processing
 algorithm.


i'm looking at:

Advanced features include the ability to reject connections from
known spam sites

exim can use rbl to reject msgs from spam sites.  but so can every other MTA.

exim doesn't have a "spam reject" file that i can drop IP addresses into.
exim doesn't use tcpwrappers to reject SMTP sessions.

what exactly are some of these advanced features then?

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: cvs security - ssh vs pserver?

[Peter Jay Salzman]

begin: Joey Hess <[EMAIL PROTECTED]> quote
> Peter Jay Salzman wrote:
> > i just found out that using method 2, you can't assign a shell of 
> > /bin/false.
> > cvs won't work.   so option 2 also means "giving a shell account on my
> > machine".
> 
> Read http://kitenet.net/programs/sshcvs
> 
> > any thoughts?  is pserver really as insecure as dpkg claims in the
> > configuration of the package?
> 
> It uses plain-text passwords, which is pretty insecure, yes.
 
joey, i have no problem with plain text passwords.

just as long as they can't get _shell access_ with that password.

one more question -- i gave that URL a brief read.  it's not clear that this
method allows people to import changes to the their local copy.   does it?

please excuse my newbieness.  i /just/ learned how to use cvs a few days ago.
i'm still struggling with concept and terminology.

thanks.
pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

cvs and security

[Peter Jay Salzman]

hi all,

i'd like to set up a cvs repository for remote users.

i recently found that cvs ssh access seems to require that i give remote
users a shell account on my computer.  

but i've heard that pserver is very insecure.

i don't know what to do.   one thing is for damn sure.  i don't want people
having shell accounts on my machine.

is pserver really as insecure as i've heard?

is there a way to use ssh for cvs without giving people shell access on my
machine?   giving user cvs a shell of /bin/false seems to screw up cvs.

help!!

pete

-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

cvs security - ssh vs pserver?

[Peter Jay Salzman]

i'd like to make some code available to collaborators via cvs.  it appears
that i have a choice to make:

1. use pserver

2. use "ext" (ssh)

i just found out that using method 2, you can't assign a shell of /bin/false.
cvs won't work.   so option 2 also means "giving a shell account on my
machine".

both these options seem insecure.  i have to admit, i'm really not crazy
about giving out shell accounts.

any thoughts?  is pserver really as insecure as dpkg claims in the
configuration of the package?

pete


-- 
PGP Fingerprint: B9F1 6CF3 47C4 7CD8 D33E  70A9 A3B9 1945 67EA 951D
PGP Public Key:  finger [EMAIL PROTECTED]

Re: can exim be started with tcp wrappers?

[Peter Jay Salzman]

hi ralf,

begin: Ralf G. R. Bergs <[EMAIL PROTECTED]> quote
> On Fri, 16 Nov 2001 08:40:56 -0800, Peter Jay Salzman wrote:
> 
> It would work -- but usually Exim already includes the tcpd stuff, so that 
> Exim *automagically* uses /etc/hosts.(allow,deny). You don't need to prefix 
> the inetd.conf line with "tcpd".
 
oh, that's very cool to know.  thanks!  :)


> But note that this is only of limited use -- spammers change IP addresses and 
> SMTP relays often, so blocking them doesn't give you any benefits. You might 
> even block legitimate messages from being delivered!

i know -- but there are some very persistant spammers, and letters /
complaints don't seem to help.

for instance, 1800flowers.com is absolutely relentless.  i've done
everything, including reporting them to maps rbl.   i'm a pretty experienced
spam combatter, so i'm aware of the issues.

> Better have a look at procmail (Debian package available.) This allows you to 
> filter messages by their (header and body) contents -- probably more what you 
> want than the above.
 
i currently use procmail, but only to sort my mail, not to filter it.   i
probably should get one of these procmail spammer lists that people collect
these days.  just haven't gotten around to it.

pete
 

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

can exim be started with tcp wrappers?

[Peter Jay Salzman]

hi all,

i would like to change

smtp  stream  tcp  nowait  mail  /usr/sbin/exim exim -bs

to

smtp  stream  tcp  nowait  mail  /usr/sbin/tcpd /usr/sbin/exim exim -bs

in inetd.conf.   i'd like to do this so i can drop spammer's IP addresses in
/etc/hosts.deny so that my host won't talk to them when they send spam.

would this work?   or would it hose my mail?  i'd like to get an opinion
before trying it.

thanks,
pete

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

wine build error: specmaker?!?

[Peter Jay Salzman]

just downloaded wine tonight, and got a build error:

# apt-get -b source wine
   (snip)
   make[2]: Leaving directory `/tmp/wine-0.20011026.033955/programs/winver'
   make[1]: Leaving directory `/tmp/wine-0.20011026.033955/programs'
   dh_testdir
   dh_testroot
   # distribute the files in debian/tmp into debian/
   # according to the .files files
   dh_movefiles
   dh_movefiles: debian/tmp/usr/bin/specmaker not found
   find: usr/bin/specmaker: No such file or directory
   make: *** [binary-arch] Error 1
   Build command 'cd wine-0.20011026.033955 && dpkg-buildpackage -b -uc' failed.
   E: Child process failed

anyone have any ideas what specmaker is?  it's not on any of my debian
systems.

thanks,

pete

Re: Debugger for C programming?

[Peter Jay Salzman]

begin: Daniel Toffetti <[EMAIL PROTECTED]> quote
> > Is there a good debugger for C programming.  You know, the kind of
> > thing that lets you step through a line at a time running your
> > program and put watches on variables etc.
> >
> > Now I know that Linux rules for C programming, so what do all you
> > programmers use to debug your code?
> >
> > Thanks.
> > Mark.
> 
> I want to add another questions on this topic, does gdb supports 
> debugging a running daemon? And what if the daemon loads a dinamic 
> library? Newbie questions, of course, so please details are welcomed.
 
yes, in fact, you can debug a running process using gdb.

i wrote a gdb tutorial which talks about this very topic

www.dirac.org/linux/gdb

the tutorial is lynx friendly.

pete

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

file conflicts between packages

[Peter Jay Salzman]

hello,

i updated woody today and found that the new xlib package wanted to
install a file over the mixviews package.

   Preparing to replace xlibs 4.1.0-5 (using .../xlibs_4.1.0-6_i386.deb) ...
   Unpacking replacement xlibs ...
   dpkg: error processing /var/cache/apt/archives/xlibs_4.1.0-6_i386.deb
   (--unpack):
trying to overwrite `/usr/X11R6/lib/X11/app-defaults', which is also in
   package mixviews
   Errors were encountered while processing:
/var/cache/apt/archives/xlibs_4.1.0-6_i386.deb
   E: Sub-process /usr/bin/dpkg returned an error code (1)

what is the standard operating procedure for this?

pete

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

HELP: UDMA/100 woes and trouble.

[Peter Jay Salzman]

dear debian-user,

i'm having a HECK of a time trying to get UDMA/100 transfer speeds on my
system.  hdparm -t reports 35 MB/s.  i've played with this for at least 4
hours now, and am coming to the realization that i definitely need help.

my system:
athlon 1.3GHz
   epox 8kha mobo with VIA k266 (vt8366 + vt8233)
   hda: IBM deskstar UDMA/100

from lilo.conf:

   append="ide0=ata66"

from dmesg:

   Kernel command line: auto BOOT_IMAGE=newlinux ro root=304 ide0=ata66
   ide_setup: ide0=ata66
   ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
   VP_IDE: VIA vt8233 (rev 00) IDE UDMA100 controller on pci00:11.1
   VP_IDE: ATA-66/100 forced bit set (WARNING)!!
 ide0: BM-DMA at 0xdc00-0xdc07, BIOS settings: hda:DMA, hdb:pio
 ide1: BM-DMA at 0xdc08-0xdc0f, BIOS settings: hdc:DMA, hdd:DMA
   hda: IBM-DTLA-307045, ATA DISK drive
   hdc: QUANTUM FIREBALLP LM30, ATA DISK drive
   hdd: WDC WD205AA, ATA DISK drive
   hda: 90069840 sectors (46116 MB) w/1916KiB Cache, CHS=5606/255/63, UDMA(100)
   hdc: 58633344 sectors (30020 MB) w/1900KiB Cache, CHS=58168/16/63, UDMA(33)
   hdd: 40079088 sectors (20520 MB) w/2048KiB Cache, CHS=39761/16/63, UDMA(33)


from hdparm /dev/hda:

   multcount= 16 (on)
   I/O support  =  1 (32-bit)
   unmaskirq=  1 (on)
   using_dma=  1 (on)
   keepsettings =  0 (off)
   nowerr   =  0 (off)
   readonly =  0 (off)
   readahead=  8 (on)
   geometry = 5606/255/63, sectors = 90069840, start = 0

from kernel config:

   CONFIG_IDE=y
   CONFIG_BLK_DEV_IDE=y
   CONFIG_BLK_DEV_IDEDISK=y
   CONFIG_IDEDISK_MULTI_MODE=y
   CONFIG_BLK_DEV_IDEPCI=y
   CONFIG_IDEPCI_SHARE_IRQ=y
   CONFIG_BLK_DEV_IDEDMA_PCI=y
   CONFIG_BLK_DEV_ADMA=y
   CONFIG_BLK_DEV_VIA82CXXX=y
   CONFIG_IDEDMA_AUTO=y
   CONFIG_BLK_DEV_IDE_MODES=y

hdparm -t reports 35 MB/s.  this is nowhere near the 100MB/s that i'm
expecting to get.  apparently, the ide0=ata66 not only doesn't give me
UDMA/100, the kernel (dmesg) actually warning me that the ata-66/100 bit is
set.   should i be worried about this warning?

any suggestions or hints?  is anybody getting anywhere close to 100MB/s?

pete


-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

Re: amd / radeon users, attention.

[Peter Jay Salzman]

hi jeff (and nicole),

it was definitely working, but i was using a generic AGP port driver.   i
just noticed that 2.4.10 has support for KT266 agp, (as well as radeon frame
buffer).  i'm now getting about 800 fps on gears at 24 bpp.

at 16 bpp color depth, i'm getting over 1100 fps.   YEAH   :)

pete


begin: Jeffrey W. Baker <[EMAIL PROTECTED]> quote
> 
> 
> On Mon, 24 Sep 2001, Peter Jay Salzman wrote:
> 
> > i just got my epox 8kha board with a radeon card to work with DRI under
> > linux.  it was rough going, and i've seen, while websurfing, that alot of
> > people have had trouble with this in the past.  many people are getting "agp
> > unsupported" messages by the agpgart driver.
> >
> > if anybody needs help (i assume this is going into the archives), feel free
> > to contact me.
> >
> > i just got a frame rate of 250 fps on gears.  mamma mia!   :-)
> 
> Are you sure it's working?  I get something like 800 fps on gears on a
> Radeon DDR using a 1.4 GHz athlon, at 16 bit color depth?
> 
> Making this stuff work on the newer boards *is* tough.  AGPGart must be
> loaded as a module, or patched to understand the AMD 760.  passing
> agp_try_unsupported=1 as a kernel parameter in the boot loader does not
> work, despite the documentation which claims it does.  I also needed a
> patch from Steven Tweedie to be able to start X at all (DRI or otherwise)
> using the Radeon.
> 
> I haven't noticed any difficulties particular to getting this to work on
> Debian.  It's just the same magnitude pain as on Slackware, Red Hat, et.
> al.
> 
> -jwb
> 
> 
> -- 
> To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
> 

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

[rml@ufl.edu: Re: report: success with agp_try_unsupported=1]

[Peter Jay Salzman]

forwarded to the debian-user list for its archives...

- Forwarded message from Robert Love <[EMAIL PROTECTED]> -

Subject: Re: report: success with agp_try_unsupported=1
From: Robert Love <[EMAIL PROTECTED]>
To: Peter Jay Salzman <[EMAIL PROTECTED]>
Cc: linux-kernel@vger.kernel.org
X-Mailer: Evolution/0.14.99+cvs.2001.09.24.08.08 (Preview Release)
Date: 24 Sep 2001 17:50:38 -0400

On Mon, 2001-09-24 at 17:40, Peter Jay Salzman wrote:
> i just built a system:
> 
>   1.3GHz amd athlon
>   epox 8kha ddr motherboard
>   via kt266 (vt8366, vt8233)
>   768MB ddr ram
>   radeon QD with 64MB video buffer, tvio
> 
> i enabled agpgart, and got the message:
> 
> 
>   Linux agpgart interface v0.99 (c) Jeff Hartmann
>   agpgart: Maximum main memory to use for agp memory: 690M
>   agpgart: Unsupported Via chipset (device id: 3099), you might want to try
> agp_try_unsupported=1.
>   agpgart: no supported devices found.
>   [drm] Initialized radeon 1.1.1 20010405 on minor 0

I will write a patch for this to add VIA KT266 support (so you don't
need to do the agp_try_unsupported=1 mess, it will be supported
natively).  Although, the patch is going to be against 2.4.10.

I'll send it out in a sec, start downloading 2.4.10 :)

> 2. i recompiled the kernel but built agpgart in rather than loading it as a
>module.  i then inserted the following line into /etc/lilo.conf:
> 
>   append="agp_try_unsupported=1"
> 
>but it didn't seem to work.  someone told me that to get agp work to work
>for my system, agpgart MUST be built as a module and you MUST pass it the
>argument agp_try_unsupported=1.  in other words, you can't build it into

Yes, I believe this is a bug.  Maybe I should take a look at it...

agpgart doesn't read the command line, or not correctly, or something.

-- 
Robert M. Love
rml at ufl.edu
rml at tech9.net

amd / radeon users, attention.

[Peter Jay Salzman]

i just got my epox 8kha board with a radeon card to work with DRI under
linux.  it was rough going, and i've seen, while websurfing, that alot of
people have had trouble with this in the past.  many people are getting "agp
unsupported" messages by the agpgart driver.

if anybody needs help (i assume this is going into the archives), feel free
to contact me.

i just got a frame rate of 250 fps on gears.  mamma mia!   :-)

pete

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

Re: apt-get wierdness: when 6 is equal to 9

[Peter Jay Salzman]

begin: Hereward Cooper <[EMAIL PROTECTED]> quote
> once upon a time Peter Jay Salzman <[EMAIL PROTECTED]> said:
> 
> 
> > problem: two systems with the same sources.list don't know about the
> > same
> >   available packages.
> 
> have you tried running the update selection from dselect. I think the
> problem is that dpkg only lists the files from the package list when
> updated via dselect. Or alternativly you could use "apt-cache search
> " which would use the apt package list.
> Hereward
hereward,

thanks for the reply.  so dselect and apt contain different views of what
packages are on the system?

just tried it -- i think you're right.  imho, this goes beyond inelegance,
and borders on a bug that needs to be fixed.  it's unthinkable that there are
two databases of available packages which are out of sync everytime we choose
apt over dselect or dselect over apt.

seems like debian should strive to make apt and dselect front ends for the
same package management system.   :(

anyway, your suggestion worked.  much thanks!

pete

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

apt-get wierdness: when 6 is equal to 9

[Peter Jay Salzman]

hello all,

problem: two systems with the same sources.list don't know about the same
  available packages.

explanation: i just built a new system, and scp'd over the sources.list file
  from an older machine.  both machines have the same /etc/apt/sources.list:


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

# Security
deb http://security.debian.org/debian-non-US woody/non-US main
deb-src http://security.debian.org/debian-non-US woody/non-US main
deb http://security.debian.org/debian-non-US woody/non-US contrib
deb-src http://security.debian.org/debian-non-US woody/non-US contrib

# Wine apt-get -b source wine
#deb http://gluck.debian.org/~andreas/debian wine main
#deb-src http://gluck.debian.org/~andreas/debian wine main

yet, on the older machine:

  satan# dpkg -l "*mesa*"
  pn  giram-gnome-me(no description available)
  pn  giram-mesa(no description available)
  un  mesa-dev  (no description available)
  ii  mesademos  3.4.2-1  Example programs for Mesa (and OpenGL in gen
  rn  mesag-dev (no description available)
  pn  mesag-glide2-d(no description available)
  pn  mesag-widgets-(no description available)
  pn  mesag3(no description available)
  pn  mesag3+ggi(no description available)
  pn  mesag3+ggi-dev(no description available)
  un  mesag3-glide  (no description available)
  pn  mesag3-glide2 (no description available)
  pn  mesag3-widgets(no description available)
  ii  xlibmesa-dev   4.1.0-5  XFree86 version of Mesa 3D graphics library 
  ii  xlibmesa3  4.1.0-5  XFree86 version of Mesa 3D graphics library
  ii  xlibosmesa-dev 4.1.0-5  Mesa/XFree86 offscreen rendering library dev
  ii  xlibosmesa34.1.0-5  Mesa/XFree86 offscreen rendering library

and on the newer machine:

  # dpkg -l "*mesa*"
  un  mesa-dev  (no description available)
  ii  mesademos  3.4.2-1  Example programs for Mesa (and OpenGL in gen
  ii  mesag-dev  3.1-17   Development library for Mesa [libc6].
  pn  mesag-glide2-d(no description available)
  pn  mesag-widgets-(no description available)
  ii  mesag3 3.1-17   A 3-D graphics library which uses the OpenGL
  pn  mesag3+ggi(no description available)
  pn  mesag3+ggi-dev(no description available)
  un  mesag3-glide  (no description available)
  pn  mesag3-glide2 (no description available)
  pn  mesag3-widgets(no description available)

i've typed apt-get clean && apt-get update a bunch of times, yet the newer
machine doesn't seem to know about the xlibmesa* packages.

how can two machines with the exact same sources.list file not know about the
same packages?

pete

-- 
"You may not use the Software in connection with any site that disparages
Microsoft, MSN, MSNBC, Expedia, or their products or services ..."
-- Clause from license for FrontPage 2002

[OT] avoiding recursive readline commands in .inputrc

[Peter Jay Salzman]

hi all,

in bash, i'd like to map the uparrow key to "escape uparrow".  the trouble is
that there doesn't seem to be a way to make readline avoid the infinite
recursion that results.  specifically, i'm trying to do:

$if bash
"OA": "OA"
$endif

why?  because i like using vi style editing in bash, but i hate the fact that
i have to press escape before the uparrow starts showing command history.

pete

[OT] what is a preemp error?

[Peter Jay Salzman]

on some cd's i get a "no preemp" error from cdrecord.

what does this error mean?

what can i do about it?

thanks!
pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

[OT] ripping cd question

[Peter Jay Salzman]

when ripping cd's with cdparanoia, the default is wav format.

but i can rip with the headerless pcm format too.

question: does the wav format lose any sound quality when compared with the
headerless PCM format?

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

avoiding recursive readline commands in .inputrc

[Peter Jay Salzman]

hi all,

in bash, i'd like to map the uparrow key to "escape uparrow".  the trouble is
that there doesn't seem to be a way to make readline avoid the infinite
recursion that results.  specifically, i'm trying to do:

$if bash
"OA": "OA"
$endif

why?  because i like using vi style editing in bash, but i hate the fact that
i have to press escape before the uparrow starts showing command history.

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

how to you reconfigure an already installed package?

[Peter Jay Salzman]

i'm surprised that dpkg doesn't have a --reconfigure option.

how are we supposed to reconfigure an already installed package?  the only
way i can think of is to uninstall the package and reinstall it.

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: Numeric Keypad in X

[Peter Jay Salzman]

begin: Karsten M. Self  quote
> on Mon, Jul 30, 2001 at 05:06:48PM -0700, Brad Rhodes ([EMAIL PROTECTED]) 
> wrote:
> > How can I type numbers on the numeric keypad in X? I turn on the Num
> > Lock and I still can't type numbers.
> 
> What keysyms/keycodes are you getting?
 
see xev if you don't know how to answer this question.

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: Undeleting files

[Peter Jay Salzman]

see the undeletion-howto.   i restored a whole hard drive this way (lost
partitions and filenames, but i got every single file back!!).

totally invaluable.  do it by hand before resorting to an automated process.
it'll be a learning experience.

pete



begin: [EMAIL PROTECTED] <[EMAIL PROTECTED]> quote
> Is there any way to get
> back deleted files in Linux?
> 
> 
> -- 
> To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
> 

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

help: mutt doesn't want to send email to ip addresses

[Peter Jay Salzman]

hi all,

mutt doesn't want to deliver mail to addresses which use an ip address.

in other words, my system is dirac.org at address 64.164.47.8.  i can send
email to [EMAIL PROTECTED], but any email to [EMAIL PROTECTED] gets bounced as
"unrouteable mail domain".

i'm pretty sure that pine was able to deliver to addresses like
[EMAIL PROTECTED]  perhaps it's just a .muttrc thing.

can someone please help me?  i'd really like to be able to send email to
addresses like [EMAIL PROTECTED]

(when replying to spam, often all i have is an ip address.  i'd rather not
take the time to use nslookup to get a host/domain name from an ip address).

thanks!

pete

Re: Removing a sound module?

[Peter Jay Salzman]

begin: [EMAIL PROTECTED] <[EMAIL PROTECTED]> quote
> On Mon, Jul 30, 2001 at 07:50:15AM -0700, Peter Jay Salzman wrote:
> > if you can't remove the sound module with rmmod or modprobe, you will 
> need to
> > reboot the system.
> 
> a fuser -v /dev/dsp /dev/audio /dev/mixer may be helpful in
> identifying and killing a process which is using the sound device and
> tying up the module.
> 
> i've also found on occasion, my sound drivers will get corrupted after
> a suspend/resume on my laptop.

this isn't a driver getting corrupted.  when a driver gets corrupted, your
kernel gets corrupted.  a reboot is necessary.

you're talking about bad behavior on the _user_ side.  not the _kernel_ side,
which is where device drivers live.

> fuser didn't show anything using the
> device, but when i checked my processes, esd was sitting there pegging
> the cpu.  when i killed esd, i could remove and reinstall the sound
> driver (maestro3) and sound would work again.
> 
> i appreciated this solution instead of rebooting because it is just
> fun to tell people you are tracking an uptime on your laptop.

very true -- however, the original poster (you should've left the entire
quote in, tsk tsk!) said that he had loaded the _wrong_ module.

meaning that, somehow, the module was able to init but simply can't
communicate with the kernel.  probing hardware is a funny business.  it's
much more than conceivable that such an occurance will corrupt kernel code.

in such a case, the module fails to load not because /dev/dsp is in use, but
because of some other reason.  perhaps the MODCOUNT gets lost.  a pointer
went wild.  who knows?

in that case, a reboot is much more than recommended.  even if the system
seems fine, the potential for true misery exists.

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: GIMP 1.2 +GIFs

[Peter Jay Salzman]

dpkg is your friend.  you should definitely learn how to use it.

see dpkg -l "gimp*"


begin: Hereward Cooper <[EMAIL PROTECTED]> quote
> 
> Hi,
> I'm trying to get the GIMP to work with .gif files. It can open them, but
> doesn't recognize the format when trying to save them (even when i've
> converted the colours). I remember ages ago having to install a seperate
> gimp-gif program, or simalar. Any help would be great,
> 
> Thanks in advance,
> 
> Hereward

Re: Removing a sound module?

[Peter Jay Salzman]

begin: staf wagemakers <[EMAIL PROTECTED]> quote
> On Mon, Jul 30, 2001 at 04:07:05PM +0300, Antti Tolamo wrote:
> > How can I remove a soundmodule if it doesn't succeed from modconf?
> > 
> > I installed wrong module and now it fails as I try to remove it.
> 
> modprobe -r module_name
> 
> example:  modprobe -r sb
 
STM that this is exactly what he's talking about.

if you can't remove the sound module with rmmod or modprobe, you will need to
reboot the system.

pete


-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: Who's using a program

[Peter Jay Salzman]

you're prolly thinking of lsof.


begin: Miguel Griffa <[EMAIL PROTECTED]> quote
> ps aux for programs?
> 
> At 07:30 a.m. 30/07/01 -0600, Robert Kerr wrote:
> >Hi all,
> >I seem to recall reading (somewhere in my wanderings) of a utility that
> >would tell me who is currently accessing a file/program.  Does anyone know
> >of any such beast?
> >Thanks
> >
> >--
> >-bob

proftp question: symlinks can't be followed.

[Peter Jay Salzman]

dear all,

i made a symlink from /data/MP3 to /home/ftp.

the trouble is that even though anonymous users can see the directory, the
can't look into the directory.  for example:

> ls -l
lrwxrwxrwx   1 root root9 Jul 30 07:16 MP3 -> /data/MP3
drwxrwxrwx   2 root root 4096 Dec 16  2000 incoming/
-rw-r--r--   1 root root  166 Mar  7  2000 welcome.msg

> ls -l MP3

> cd MP3
MP3: No such file or directory

according to the proftpd documentation, this should be ok.  obviously i'm
missing something here.

help?

thanks!
pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

question about MTRR and XFree86 4.0

[Peter Jay Salzman]

according to /usr/src/linux/Documentation/mtrr.txt...

  A patch is being written for XFree86 which will make this automatic:
  in other words the X server will manipulate /proc/mtrr using the
  ioctl() interface, so users won't have to do anything.

does anybody know if this has been done yet?  it looks like my mtrr was set:

  reg00: base=0x (   0MB), size= 128MB: write-back, count=1
  reg01: base=0x0800 ( 128MB), size=  64MB: write-back, count=1
  reg02: base=0x0c00 ( 192MB), size=  32MB: write-back, count=1
  reg03: base=0xe800 (3712MB), size=  64MB: write-combining, count=2

but i haven't echo'd anything to /proc/mtrr.

has this patch already been integrated into X 4.0?

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

wine question

[Peter Jay Salzman]

dear all,

as i understand it, wine comes with its own libraries which implement the
win32 API.

i also understand that if we somehow have access to a windows CD, all legal
issues aside, we can use files off windows to make wine work even better.

how can i do this?   is there a debian specific way of doing this?

pete


-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

woody yadex is 2 years behind the times?

[Peter Jay Salzman]

dear all,

this is from dpkg:

ii  yadex  1.0.1-1.1  WAD file editor for doom-style WADs

this is from the yadex website:

2001-06-30 - Yadex 1.5.2 is out
2000-12-12 - Yadex 1.5.1 is out
2000-09-25 - The mailing lists have moved
2000-08-27 - Yadex 1.5.0 is out
2000-04-01 - Yadex 1.4.0 is out
2000-01-14 - Yadex 1.3.2 is out
2000-01-12 - Yadex 1.3.1 is out
2000-01-11 - Yadex 1.3.0 is out
1999-11-23 - Yadex 1.2.0 is out
1999-08-28 - Yadex 1.1.0 is out

is the woody version of yadex REALLY 2 years behind the upstream version?
i can't believe this.  someone please tell me what's going on!

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

wine question

[Peter Jay Salzman]

dear all,

i put the following in sources.list since i'd like to update wine on a daily
basis from cvs:

  # Wine
  deb http://gluck.debian.org/~andreas/debian wine main
  deb-src http://gluck.debian.org/~andreas/debian wine main

i'm a little confused, because i'm not seeing wine updated when i run
apt-get.   what gives?   it appears that i have version:

  ii  wine   0.20010223.034 Windows Emulator (Binary Emulator)

which looks about 4 months old.

anyone know what's going on?

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: Unix administrator

[Peter Jay Salzman]

begin: Dimitri Maziuk <[EMAIL PROTECTED]> quote
> * Chris Parker ([EMAIL PROTECTED]) spake thusly:
> > Student here from a micro$oft school of thought and sick of it.  What
> > do I need to read...study to gain the honor of a unix admin.?  Is
> > athere any good online classes or tutorials that i should check into? 
> > Also what would be a good route to take for a beginner programming? 
> > Hopefully the Debian gurus will reply.
> > THANKS TO ALL THAT REPLY!
 
this is sheer and utter nonsense.  you said "admin", not "programmer", right?

1. you DON'T start programming assembly language to become a unix admin.

2. you DON'T have to learn C to be a fledgling unix admin, although it would
   definitely make sense to read K&R down the road, AFTER you learn basics.
   a good knowledge of make is more important.

3. you DON'T need to read knuth and aho to become a unix admin.

4. you DON'T need to learn discrete math to become a unix admin.  being able
   to use a calculator is good enough.

5. you CERTAINLY DON'T need to read stevens to become a unix admin.  a good
   book like the NAG is certainly enough to begin with.   later on, you can
   flesh out your knowledge with more detailed books.  stevens is overkill.

6. you DON'T need to read ESR's cathedral to become a unix admin.  that the
   HECK does this have to do with admining?

7. you DON'T need to start soldering shit.

this is completely insane.  he must have thought you meant
"programmer".  here are the few things i agree with:


> Read Tannenbaum, Silbershatz (sp?) and Sobell's Practical Guide. Try shell, 
> awk, sed scripts. Read the Camel book and learn Perl.
 
now this is good advice.  you MUST MUST MUST know perl and shell.

> Set up Sendmail, Bind and Inn on your quadruple-boot home peecee

only if you want to become a professional.  you never said if you want to use
linux for home use or not.  no need to know bind and inn if so.  knowing a
bit about sendmail is crucial, but you certainly don't need to, say, read the
o'reilly horror called "sendmail".

> (read everything you can find on linuxdoc.org).

good advice, but unrealistic.  read what you need to know.  don't read what
you don't need to know.  you certainly won't need to learn how to program
device drivers, for instance.  a basic understanding of what they are, how
they work and how they're configured is enough.

> Proceed to Garfinkel's Practical Guide and Zwicky's Firewalls.

there's tons of good references out there.

> Get an entry-level helpdesk job at an ISP and work there 
> until hospitalized. Alternatively, work there for a few months, walk to the 
> nearest asylum and surrender yourself to friendly nursies.

mandatory if you want to become a professional.  not a good move if you want
to administer your home system.

> Or you can skip all of those steps and go get your head examined now.
 
h

one last comment from me.  once you install linux on your system, almost
everything you need to know will be on your hard drive somewhere.  advice
which is better than mine (and MUCH better than the guy who posted this) is
to learn where to find this info on your hard drive.  once you do that,
you've taken the most important step to becoming a linux admin.

peter

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

custom spam file (ala rbl)

[Peter Jay Salzman]

is there a file that i can drop IP numbers in to keep exim from accepting
email from those sites?

kind of like using rbl, except i'd have my own custom reject file.

i can simulate such a file usign ipchains, but i'd like to know if exim has
an IP reject file.

thanks!
pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

help making debian ISO's

[Peter Jay Salzman]

i'd like to burn a copy of the debian woody distribution, so i'm following
the funky instructions about making pseudo-images and using rsync to convert
the images to official images.

the problem is that the .list file that i'm supposed to get from
http://www.uk.debian.org/debian-cd/cd-images/ is only for 2.2r3.  there
doesn't appear to be a .list for woody.

does this mean that there is no official woody cd image or does it mean i
need to look elsewhere for the woody .list file?

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: voodoo trouble - sanity is at stake here.

[Peter Jay Salzman]

begin: Alson van der Meulen <[EMAIL PROTECTED]> quote
> On Sun, Jun 03, 2001 at 09:40:21AM -0700, Peter Jay Salzman wrote:
> > hey all,
> > 
> > my girlfriend asked me to fix something on her computer, and i'm at a 
> > loss...
> > 
> > she has a voodoo3 / woody / X 4.0
> > 
> > hardware acceleration is obviously working for root (gears looks great, and
> > independent of window size; quake3 looks great).   it's obviously NOT 
> > working
> > for non-root users (gears is slw.  quake3 is painful).
> > 
> > 
> > when X boots (even as non-root), everything looks nice and healthy.  i get 
> > the
> > "DRI enabled" message.   yet, it's obvious that something isn't going right
> > for non-root users.
> > 
> > i'm at a loss here.  X looks good.  voodoo is good under root.  you don't
> > need a kernel module for non-root acces of the X4 voodoo boards.  i don't
> > think anything needs to be setuid.
> have something like this in your xf86config:
> Section "DRI"
> Mode 0666
> EndSection

perfect, thank you.

> look a the docs on dri.sf.net for more info

definitely a good site to spend some time at.  thanks for the link!

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

voodoo trouble - sanity is at stake here.

[Peter Jay Salzman]

hey all,

my girlfriend asked me to fix something on her computer, and i'm at a loss...

she has a voodoo3 / woody / X 4.0

hardware acceleration is obviously working for root (gears looks great, and
independent of window size; quake3 looks great).   it's obviously NOT working
for non-root users (gears is slw.  quake3 is painful).


when X boots (even as non-root), everything looks nice and healthy.  i get the
"DRI enabled" message.   yet, it's obvious that something isn't going right
for non-root users.

i'm at a loss here.  X looks good.  voodoo is good under root.  you don't
need a kernel module for non-root acces of the X4 voodoo boards.  i don't
think anything needs to be setuid.

what else is there to check?!?

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

compiling gtk under woody

[Peter Jay Salzman]

hello all,

today i would like to teach myself how to use gtk.

i'm running woody.  which packages should i apt-get install?

pete

-- 
"The following addresses had permanent fatal errors..."  [EMAIL PROTECTED]
   -- Mailer Daemon  www.dirac.org/p

Re: offtopic (completely): linux games

[Peter Jay Salzman]

On Sun 15 Apr 01,  1:52 PM, Nate Amsden said: 
> Peter Jay Salzman wrote:
> 
> > sorry, but people like me who refuse to run a dual-boot computer (refuse to
> > fork over $$$ to the evil empire) rely on loki software for recreation.  :)
> 
> then you should pay full price and support loki.

well, i DO buy each and every game as they come out, so you're nagging up
the wrong tree!  :)

> i bought quake 3 about a week
> after it came out. i played the hell out of the demo, a few days after i
> bought quake 3, i got the demo for unreal tournament, and never used quake

blah blah blah blah blah blah blah blah blah blah...  (snip)

you're rubber, i'm glue whatever you said is true for me too.   :-)

my point is a sale to ebworld is better than no sale at all.  and shame on
you, nate, if you say otherwise.

pete

offtopic (completely): linux games

[Peter Jay Salzman]

dear all,

i know i'll get flamed for this, but i also happen to know that some of you
will greatly appreciate this information:

if you ever wanted to get quake III but didn't want to spend the (admittedly
hefty) price of $50 or $60, now is your chance to purchase it:

   quake 3 arena:  9.99
   descent 3:  9.99
   myth 2: 9.99
   heretic 2:  9.99
   heavy gear 2:   9.99
   soldier of fortune: 9.99
   railroad tycoon 2:  9.99
   eric's solitaire:   9.99

http://www.ebgames.com/search on "linux"

these are INCREDIBLE prices.  if loki sales get good enough, maybe some day
we'll get diablo or (native) halflife for linux.   :-)

sorry, but people like me who refuse to run a dual-boot computer (refuse to
fork over $$$ to the evil empire) rely on loki software for recreation.  :)

pete


pgpXbqtzSHefN.pgp
Description: PGP signature

WTF -- klogd has disappeared from woody?!?

[Peter Jay Salzman]

dear all,

i just updated woody and klogd has disappeared into thin air.

it's not even in "dpkg -L sysklogd"

# dpkg -L "sysklogd" | grep klogd
/usr/share/doc/sysklogd
/usr/share/doc/sysklogd/changelog.gz
/usr/share/doc/sysklogd/copyright
/usr/share/doc/sysklogd/changelog.Debian.gz
/usr/share/doc/sysklogd/readme.txt.gz
/usr/share/man/man8/sysklogd.8.gz
/etc/init.d/sysklogd
/etc/cron.daily/sysklogd
/etc/cron.weekly/sysklogd

i'm using:

# dpkg -p sysklogd
Package: sysklogd
Priority: required
Version: 1.4.1-1

WTF is going on here?


pgpzOYsWCb9IQ.pgp
Description: PGP signature