Re: oops

[Ben Collins]

> 
> Sorry to have jumped the gun but I am "spring loaded" to blame libc at
> this point for any weirdies I see.
> 

which helps nothing, to say the least about making an already
overworked libc maintainer stop what he's doing and take time to
investigate half-investigated bug reports...

-- 
 ---===-=-==-=---==-=--
/  Ben Collins  --  ...on that fantastic voyage...  --  Debian GNU/Linux   \
`  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  '
 `---=--===-=-=-=-===-==---=--=---'

Re: latest libc does NOT fix everything

[Jeronimo Pellegrini]

On Sat, Sep 30, 2000 at 09:28:58PM -0700, George Bonser wrote:
> 
> Exim is still broken. As a matter of fact, I had it working until I loaded
> the latest libc tonite:
> 
> I have a bazillion of these in my exim paniclog:
> 
> 2000-09-30 21:25:09 queue run: process 31436 crashed with signal 15 while
> delivering 13dnoh-xq-00

Indeed. Exim was broken by the upgrade a few minutes ago... I believe that the
post-install script didn't restart all the necessary services. Restarting
*all* services seemed to work...

(Not only exim; I tried installing smail and it was also broken)

J.

-- 
Jeronimo Pellegrini
Institute of Computing - Unicamp - Brazil
http://www.ic.unicamp.br/~jeronimo
mailto:[EMAIL PROTECTED]mailto:[EMAIL PROTECTED]

IGNORE latest libc does NOT fix everything

[George Bonser]

My bad ... jumped the gun.

oops

[George Bonser]

I guess I should have looked at it for more than three seconds :-)

Those errors were from a kill sent to exim when it was stopped and
restarted. I have a ton of connections to a site that is down right now
and these got logged when exim shut down.

Sorry to have jumped the gun but I am "spring loaded" to blame libc at
this point for any weirdies I see.

latest libc does NOT fix everything

[George Bonser]

Exim is still broken. As a matter of fact, I had it working until I loaded
the latest libc tonite:

I have a bazillion of these in my exim paniclog:

2000-09-30 21:25:09 queue run: process 31436 crashed with signal 15 while
delivering 13dnoh-xq-00

Re: libdb.so.3 missing

[Pollywog]

On Sat, 30 Sep 2000 19:23:36 -0700
Michael Smith <[EMAIL PROTECTED]> wrote:

> Ihad the same problem a couple of days ago--the hard party was that the 
> library was
> missing, so I couldn't upgrade to a working version.  It tried to 
> preconfigure, and
> then perl tanked with the same error message that you got.  What I did was 
> ftp to
> another machine, grab /usr/lib/libdb.so.3, and move it to the right place on 
> the
> broken box.  Then I immediately did an upgrade and the problem fixed itself 
> with the
> upgrade (the package in question got upgraded, overwriting the bad version.

I upgraded libc6 libc6-dev locales and the libdb2 packages and all is well.
The buggy libc6 is no more.


--
Andrew

proposed 'bouncing email' procmail recipe

[will trillich]

hmm. i'm wondering if there's a simple procmail recipe
that'll head these bouncing emails off at the pass--

:0 H
* ^From:.*(mailer|daemon|webmaster)
* ! ^To:[EMAIL PROTECTED]
{
:0 B
* ^X-Loop: [EMAIL PROTECTED]
/dev/null
}

anybody else care to add some tweaks? (what'd i miss?)

what's the official pattern to search for to know whether or not
it  came directly from the debian list server?

-- 
mailto:[EMAIL PROTECTED]
http://www.dontUthink.com/

Re: libdb.so.3 missing

[Michael Smith]

Ihad the same problem a couple of days ago--the hard party was that the library 
was
missing, so I couldn't upgrade to a working version.  It tried to preconfigure, 
and
then perl tanked with the same error message that you got.  What I did was ftp 
to
another machine, grab /usr/lib/libdb.so.3, and move it to the right place on the
broken box.  Then I immediately did an upgrade and the problem fixed itself 
with the
upgrade (the package in question got upgraded, overwriting the bad version.

I think we need a bumper sticker that says "Use Unstable Debian--live by the 
seat of
your pants."

"P.J.Walsh" wrote:

> dpkg is choking on some upgrades, showing libdb.so.3 missing... dpkg
> -S doesn't help.  To what does it belong?
>
> --
> Patrick Walsh
> Edmonton AB CA
>
> --
> Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null

--
Michael J. Smith [EMAIL PROTECTED]
2250 Patterson #25 Eugene, OR 97405
(541)346-7562

bouncing mails, third instance -- fixable?

[will trillich]

i'm getting bounces from ngi.de now --

Return-Path: <[EMAIL PROTECTED]>
Received: (qmail 15792 invoked by alias); 30 Sep 2000 04:53:27 -
Delivered-To: [EMAIL PROTECTED]
Received: (qmail 15763 invoked from network); 30 Sep 2000 04:53:26 -
Received: from sam.julianhaight.com (207.12.88.58)
  by 195.243.0.245 with SMTP; 30 Sep 2000 04:53:26 -
Received: (from [EMAIL PROTECTED])
by sam.julianhaight.com (8.10.0/8.10.0) id e8U4rPt20340
for [EMAIL PROTECTED]; Sat, 30 Sep 2000 00:53:25 -0400
Received: from kp8.kpsnet.net (kp8.kpsnet.net [216.121.191.51])
by sam.julianhaight.com (8.10.0/8.10.0) with ESMTP id e8U4rNe20295
for <[EMAIL PROTECTED]>; Sat, 30 Sep 2000 00:53:23 -0400
Received: (from [EMAIL PROTECTED])
by kp8.kpsnet.net (8.9.3/8.9.3) id GAA21349
for [EMAIL PROTECTED]; Sat, 30 Sep 2000 06:53:34 +0200
Received: from murphy.debian.org (murphy.debian.org [216.234.231.6])
by kp8.kpsnet.net (8.9.3/8.9.3) with SMTP id GAA21336
for <[EMAIL PROTECTED]>; Sat, 30 Sep 2000 06:53:33 +0200
Resent-Date: Sat, 30 Sep 2000 06:53:33 +0200
Received: (qmail 19959 invoked by uid 38); 30 Sep 2000 04:53:11 -
X-Envelope-Sender: [EMAIL PROTECTED]

looks like the intended party is
[EMAIL PROTECTED]
(which forwards? to
[EMAIL PROTECTED]
which forwards? to
[EMAIL PROTECTED]
which bounces...)

if we can remove that one email address from the list,
[EMAIL PROTECTED]
would this bounce be solved?

-- 
mailto:[EMAIL PROTECTED]
http://www.dontUthink.com/

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[montefin]

Gregory,

It was all my fault. I pulled the plug out of the wall when the lamp was
still on.

Regrets,

montefin

"Gregory T. Norris" wrote:
> 
> On Sat, Sep 30, 2000 at 09:33:55AM -0400, Ben Collins wrote:
> > [1] Debian Chaos Events are scheduled on a haphazard basis. We make no
> > gurantees about show times.
> 
> Hey, don't drag me into this!  I didn't do it!!! :-)
> 
> Sorry, couldn't resist.
> 
> --
> Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Gregory T. Norris]

On Sat, Sep 30, 2000 at 09:33:55AM -0400, Ben Collins wrote:
> [1] Debian Chaos Events are scheduled on a haphazard basis. We make no
> gurantees about show times.

Hey, don't drag me into this!  I didn't do it!!! :-)


Sorry, couldn't resist.

xfmail seems broken after Woody upgrade today

[Pollywog]

I upgraded libc6 and her friends a little while ago, and now xfmail starts but 
will not retrieve mail.  Sylpheed is retrieving mail, so the problem does not 
seem to be my pop server.  I get the same result with either qpopper or 
solid-pop3d.

Any other xfmail users having problems after upgrading libc6 today?
I did NOT upgrade xforms.

tnx

--
Andrew

Eterm, /dev/null, and segfaults

[rich]

Howdy all,

I just installed Slink onto an HP pentium 166 from CD. I'm using
Windowmaker... When I tried to use Eterm as root, it worked... when I
tried to use it as a normal user, it said something about /dev/null and
segfaulted...

/dev/null is set up like so... crw-rw-rw-

xterm, however, worked for both root and normal users.


I then used apt to upgrade to Potato over ethernet Eterm still works
for root, but now when I try to run Eterm OR xterm as normal user,
absolutely nothing happens. No window, no disk-noises, nothing. However,
if I switch to twm, I can get a cheesy-looking xterm. Strange?

One last clue when I log out of Windowmaker, before switching to
back
to wdm, a screen of strange SVGA-looking chaotic graphics briefly
flashes - it does NOT do this for root.

Any clues?

Thanks in advance,

Rich

Re: machines (ii)

[George Bonser]

> 
> Do you have a link or know of a good book that describes how to do this?  I'd 
> love to give this a try at my house.
> 
> Jesse

Well, I would first look at the  CD-Writing-HOWTO which has some basic
stuff on creating CDROMs and some information on making bootable
CDs.  Then I would direct you to the ../Documentation/initrd.txt file in
your linux source tree so you can learn a little about initrd
booting. Then, while you are there, have a look at the ramdisk.txt file.  

These should keep you busy for a while :-)

Re: machines (ii)

[Jesse Goerz]

On Sat, 30 Sep 2000, George Bonser wrote:
> On Sat, 30 Sep 2000, Mark Simos wrote:
> > I am looking to put together a Debian based firewall and a mail server
> >   -how bad of an idea is it to host them on the same machine?
> >(please explain how dumb it is, if so)
> >
> > How much power would I need (CPU/RAM/HD) to make it (or each of them)
> > work?
> >
> > I just browse at home and download the occasional files and would like
> > to learn how to configure various mail and firewall packages. not too
> > much strain.
> >
> > Thanks!
> >
> > Mark
>
> Mark, I would not put the mail server on the firewall. This is because of
> disk requirements. It is pretty easy for a lot of users on a machine to
> fill up disk fairly quickly. About the only time I would put a mailer on a
> firewall is for use as a transparent proxy to handle outbound mail or as a
> relay machine for inbound mail to route it to your real mailhost.
>
> You do not need much in the way of CPU power. One method is to get a BUNCH
> of RAM, create a bootable CDROM that boots initrd and runs from a
> ramdisk. Set up syslog to log to a remote system that has a hard disk,
> remove ALL hard disks from the system. Now if someone roots your box, you
> just reboot and everything they did evaporates with the freshly loaded
> binaries. Once you discover how they got in, you create another bootable
> CDROM and reboot the box to load the more secure stuff.
>
> I would use ECC RAM in this configuration. Since logs are being sent to a
> remote system, and since nothing is persistant over a reboot, and since
> there are no hard disks to fail, you have a fairly secure and robust
> firewall.


Do you have a link or know of a good book that describes how to do this?  I'd 
love to give this a try at my house.

Jesse

Re: libdb.so.3 missing

[Jesse Goerz]

On Sat, 30 Sep 2000, P.J.Walsh wrote:
> dpkg is choking on some upgrades, showing libdb.so.3 missing... dpkg
> -S doesn't help.  To what does it belong?


To find out which package it belongs to go to www.debian.org and go to the 
packages page.  You can look up which package it belongs to.

Re: newbie: ps

[Christian Pernegger]

On Sat, Sep 30, 2000 at 07:47:29PM -0400, jason lee wrote:
> hello all,
>   pls help a newbie -- i accidentally removed ps (and its manpages), how can 
> i get them back?  (can't seem to find them on the debian homepage, or on the 
> CDs (official debian 2.2)).

It's in the procps package

Christian

quota utils don't show grace time

[Christian Pernegger]

THe subject pretty much sums it up. When I run repquota or quota,
the grace time column is always empty.

Why ist this?

Christian

Re: libdb.so.3 missing

[Pollywog]

On 30-Sep-2000 Ben Collins wrote:
> 
> This is no longer true. Follow the threads. libc6 2.1.94-3 has zero
> issues.

Thanks, I will upgrade now.

--
Andrew

newbie: ps

[jason lee]

hello all,
 pls help a newbie -- i accidentally removed ps (and its manpages), how can 
i get them back?  (can't seem to find them on the debian homepage, or on the 
CDs (official debian 2.2)).

_
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.

PS/2 Mouse

[Jack Morgan]

I just installed a new motherboard. I'm running woody on the only HDD. When i 
plug in the PS/2 mouse the system hangs during the boot strap process. If I 
unplug the PS/2 mouse it boots fine, but hangs when I use X-windows. Is this an 
irq issue or should i reinstall?

Thanks In Advance,
Jack

Re: libdb.so.3 missing

[Ben Collins]

On Sat, Sep 30, 2000 at 09:54:56PM +, Pollywog wrote:
> 
> On 30-Sep-2000 P.J.Walsh wrote:
> > dpkg is choking on some upgrades, showing libdb.so.3 missing... dpkg
> > -S doesn't help.  To what does it belong?
> 
> It belongs to the libdb package but the problem is with libc6.
> I had to revert to the previous versions of libc6, libc6-dev, and locales.
> 
> The new libc6 packages are buggy.  Your ldconfig is probably missing after the
> libc6 upgrade.  Mine just disappeared.

This is no longer true. Follow the threads. libc6 2.1.94-3 has zero
issues.

-- 
 ---===-=-==-=---==-=--
/  Ben Collins  --  ...on that fantastic voyage...  --  Debian GNU/Linux   \
`  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  '
 `---=--===-=-=-=-===-==---=--=---'

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Ben Collins]

> 
> Ladies and Gentlemen, Welcome to another televised Major League Baseball
> game.  In an effort to match the viewership of Monday Night Football, now
> that Dennis Miller is part of it, we looked long and hard for someone,
> someone with a great personality, wonderful people skills, and above all,
> geek appeal.  We've found him in Richard M Stallman!!!  RMS, say hello.
> 
> RMS:  That's Gnu/Baseball.
> 

Lol!

-- 
 ---===-=-==-=---==-=--
/  Ben Collins  --  ...on that fantastic voyage...  --  Debian GNU/Linux   \
`  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  '
 `---=--===-=-=-=-===-==---=--=---'

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Seth Cohn]

On Sat, 30 Sep 2000, Ben Collins wrote:

> [1] Debian Chaos Events are scheduled on a haphazard basis. We make no
> guarantees about show times.

Coming soon, Watch the Amazing Horse that dives into a Package Pool!! 
See the unstable freaks who apt-get update every 10 minutes!!!
Hear the weird wildman Bruce Perens! (appears courtesy of Technocrat.net)
The scary flamewars of Debian-devel... cover your kid's eyes!!!
Witness the key juggling and mirror manipulation of trained professionals!

> [2] Debian Chaos Events are Copyright 2000 Software in the PublicInterest.
> Any reproduction or rebroadcast of events in whole or in part requires
> express written consent by SPI. 

and in a parallel reality:

Ladies and Gentlemen, Welcome to another televised Major League Baseball
game.  In an effort to match the viewership of Monday Night Football, now
that Dennis Miller is part of it, we looked long and hard for someone,
someone with a great personality, wonderful people skills, and above all,
geek appeal.  We've found him in Richard M Stallman!!!  RMS, say hello.

RMS:  That's Gnu/Baseball.

Re: Good Book for setting up T-1?

[George Bonser]

> 
> All my information dates from approximately 1997.  At the time there were
> many T1 cards with integrated CSU/DSU's in development, but I didn't
> consider any of them quite ready for prime time yet.  You might be able to
> save more money by finding one of them.

Sangoma makes a capable T1 card with integrated CSU/DSU, I have used it in
production. Debian contains the wanrouter package which contains the stuff
to make it work and the Sangoma drivers are included in the Linux kernel,
you just need to turn them on. The problem with that is when you use a
Linux box for your WAN router, the network is only as reliable as the
machine. If you should have to reboot it, the entire network is toast. I
have found that you either get a router when you need a router, or set up
a Linux box to do NOTHING but be a router, remove all hard disks and run
out of an initrd image booted from cdrom. Don't forget to log to a remote
system or your RAMDISK will fill up with log messages.

The benefit of a router is that you have no disks, they tend to run for
years and years.

You might also try a firewall from someplace like Protectix  ... they run
Linux and boot from flash so there is no disk drive. I don't think they
include a T1 option, though.

Re: Good Book for setting up T-1?

[William T Wilson]

On Sat, 30 Sep 2000 [EMAIL PROTECTED] wrote:

> I have been using deb linux for some 5 years now and am quite happy
> with it.  It has been a webserver for me for only 1 of those years and
> that is on a DSL.  As it trns out, some of the people I've done some
> contract work with wish to install a t1 line and run debian as the OS
> on all the systems.

The hard part isn't setting up the Linux systems, it's setting up the
T1.  Once you have a plan in place for making that work, then worry about
the hosts on the network.

You already set up a webserver, so you know how to configure systems on
the network.  T1 gives you some advantages- you don't have to worry about
DHCP or anything, you just set your IP address and leave it.  If you have
a whole class C network your DNS gets a lot simpler, otherwise you need
some assistance from your upstream provider.  You can still handle
everything in 'yourdomain.com' fine but reverse DNS will not work.  But
the DNS HOWTO I believe has the trick for the reverse.  Mail is easy to do
too, you just have everyone deliver mail to the mailserver.  To handle
case of '[EMAIL PROTECTED]' you can either use an A record for
'yourdomain.com' pointing to the mail server, or you can use an MX record.  
Never use a CNAME for anything having to do with mail.  Make sure your
mail server knows it has to handle mail for 'host.yourdomain.com' as well
as 'yourdomain.com'.  Of course you will have to set up POP3 or IMAP, but
these are not harder than installing any other program.

> Anyway, I have not done this before... maybe someone could point me in
> the direction of a list of hardware needed.. CSU/DSU, routers, etc...

You'll need a router and a CSU/DSU.  :} The Cisco 2500 series is the
'canonical' single T1 router.  But you can do this with Linux, too, if you
want.  Total cost is about the same, the Cisco has better routing but a
Linux system is more expandable and makes a much better firewall (get a
cheap Pentium to do the routing, the expensive V.35 serial hardware will
make up the difference in cost between the cheap PC and the more expensive
Cisco).  Bat Electronics CSU/DSU's are cheap (I paid like $400 for mine)
and easy to set up, they have a RJ-45 on one side for the T1 and a V.35
serial port on the other to go to the router.  The Bat CSU/DSU has a love
it/hate it reputation - they have essentially no features and a high
defect rate, but Bat will replace any defective ones and they are super
cheap, and they do perform all the required functions for a simple setup.

All my information dates from approximately 1997.  At the time there were
many T1 cards with integrated CSU/DSU's in development, but I didn't
consider any of them quite ready for prime time yet.  You might be able to
save more money by finding one of them.

Re: APT::Force-LoopBreak ??

[Jason Gunthorpe]

On Sat, 30 Sep 2000, Thomas J. Hamman wrote:

> Okay, so I can't dist-upgrade because of a problem that I shouldn't work
> around because it's "A GRAVE BUG" and I don't "-really- know" what I'm
> doing -- so, what am I supposed to do?

In this case you may activate the option, however, I recommend doing
'install libc6' then proceeding with the dist-upgrade, try not to crash
your computer during that time because login will stop working for a
little while..

Jason

RE: libdb.so.3 missing

[Pollywog]

On 30-Sep-2000 P.J.Walsh wrote:
> dpkg is choking on some upgrades, showing libdb.so.3 missing... dpkg
> -S doesn't help.  To what does it belong?

It belongs to the libdb package but the problem is with libc6.
I had to revert to the previous versions of libc6, libc6-dev, and locales.

The new libc6 packages are buggy.  Your ldconfig is probably missing after the
libc6 upgrade.  Mine just disappeared.

--
Andrew

libdb.so.3 missing

[P.J.Walsh]

dpkg is choking on some upgrades, showing libdb.so.3 missing... dpkg
-S doesn't help.  To what does it belong?

-- 
Patrick Walsh
Edmonton AB CA

Re: Why isn't .Xdefaults read?

[Eric G . Miller]

On Sat, Sep 30, 2000 at 08:50:56PM +0200, Preben Randhol wrote:
> Why do I have to run xrdb .Xdefaults manually when I start X ?

Rename .Xdefaults to .Xresources and you won't.

-- 
/bin/sh ~/.signature:
Command not found

APT::Force-LoopBreak ??

[Thomas J. Hamman]

After running an apt-get update in woody, I ran an apt-get -s
dist-upgrade to see which packages were going to be upgraded, and I was
given this error message:

E: This installation run will require temporarily removing the essential
package libpam-modules due to a Conflicts/Pre-Depends loop. This is
often bad, but if you really want to do it, activate the
APT::Force-LoopBreak option.
E: Internal Error, Could not early remove libpam-modules

Looking in man apt.conf, I see this:

Force-LoopBreak
   Never Enable this option unless you  -really-  know
   what  you  are doing. It permits APT to temporarily
   remove   an   essential   package   to   breaka
   Conflicts/Conflicts  or  Conflicts/Pre-Depend  loop
   between two essential packages. SUCH A LOOP  SHOULD
   NEVER  EXIST  AND  IS A GRAVE BUG. This option will
   work if the essential packages are not  tar,  gzip,
   libc,  dpkg,  bash  or anything that those packages
   depend on.

Okay, so I can't dist-upgrade because of a problem that I shouldn't work
around because it's "A GRAVE BUG" and I don't "-really- know" what I'm
doing -- so, what am I supposed to do?

-- 
Tom
"No idea is so antiquated that it was not once modern; no idea is so
modern that it will not someday be antiquated."
-Ellen Glasgow

Re: machines (ii)

[Ray Percival]

It depends on how secure you want it to be if all you need is basic NAT and 
some packet filtering it is not a *really* bad idea. If you want anything more 
than that I would not do it.

-- Original Message --
From: Mark Simos <[EMAIL PROTECTED]>
Date: Sat, 30 Sep 2000 12:13:47 -0400

>I am looking to put together a Debian based firewall and a mail server
>  -how bad of an idea is it to host them on the same machine?
>   (please explain how dumb it is, if so)
>
>How much power would I need (CPU/RAM/HD) to make it (or each of them)
>work?
>
>I just browse at home and download the occasional files and would like
>to learn how to configure various mail and firewall packages. not too
>much strain.
>
>Thanks!
>
>Mark
>
>p.s. please cc to me directly on replies
>
>
>-- 
>Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
>
>

Trying to fix exim from woody upgrade

[Bryan K. Walton]

I am running the latest version of woody (including the brand new version of 
libc6 : 2.1.94-3).

I am also running the latest version of exim (3.16-3.1).  Ever since upgrading 
to this version of exim, I can't send any mail.  When I try, this is what I 
get: 

2000-09-30 14:42:10 13fSWQ-0001GF-00 Failed to create spool file
/var/spool/exim/input//13fSWQ-0001GF-00-D: Permission denied
2000-09-30 14:42:10 13fSWQ-0001GF-00 Failed to create spool file
/var/spool/exim/input//13fSWQ-0001GF-00-D: Permission denied
2000-09-30 14:42:10 13fSWQ-0001GF-00 Failed to create spool file
/var/spool/exim/input//13fSWQ-0001GF-00-D: Permission denied

I can't downgrade to the stable version of exim because I get a message saying 
that it conflicts with the new version of libc6.  Is this simply a permissions 
issue in my mail spool at the error sugggests?  I haven't altered the 
permissions manually (don't know if apt-get did it for me when upgrading exim).

Here is the permissions on my /var/spool/exim directory:
drwxr-x---5 mail mail 4096 Sep 30 12:41 exim

and my input directory:
drwxr-x---2 mail mail 4096 Sep 30 12:42 input

Thanks,
Bryan Walton

Re: linux disk editor/fat tarbal recover [time running out :( ] (fwd)

[Richard E. Hawkins]

While I'm at it:  I expect it would make more sense to search for magic
numbers for the tarballs.  But how in the world would I do this.  Does
anyoen actually understand lde (it's man page doesn't have directions
for
what it claims to do through ncurses . . .)


-- 
Prof. Richard E. Hawkins, Esq.   Smeal 178(814) 375-4700
[EMAIL PROTECTED]
These opinions will not be those of Penn State until it pays my
retainer.

Re: machines (ii)

[Nate Amsden]

shouldnt be a problem, most machines i build are very multipurpose and
usually all have their own individual firewalls. e.g. my home network is
3 machines, with 1 of them acting as:

firewall
gateway(hooked directly to the dsl router)
NAT
NFS server
NIS server
www server
POP3 server
SMTP server
DNS
ftp server
Unreal tournament server
runs solaris on top of vmware
port forwarding for a couple ports
X server
x font server(true type)
ssh server
dhcp server
nocol server
snmpd server(for gatherning network stats)
vpnd client
proxy/caching server(squid)

i use it for irc too

its a k6-3 400 256MB with a 640k dsl connect and 36GB of hd. soon to be
upgraded to a 1mb connect. my firewall rules are quite extensive, and it
works great. thats one of the things i like best about linux is it's
ability to do tons of things at the same time.

some people are more paranoid and like to have a dedicated box for a
firewall, but for my situation a dedicated box is not needed.

you can see the webserver and mrtg/ipac stats at
http://portal.aphroland.org

or join me in unreal tournament sometime at portal.aphroland.org:12001
(private server, not advertised to the world since im limited on b/w)

if there was more services that i felt were useful id be the first to
try them :)

nate

Mark Simos wrote:
> 
> I am looking to put together a Debian based firewall and a mail server
>   -how bad of an idea is it to host them on the same machine?
>(please explain how dumb it is, if so)
> 
> How much power would I need (CPU/RAM/HD) to make it (or each of them)
> work?
> 
> I just browse at home and download the occasional files and would like
> to learn how to configure various mail and firewall packages. not too
> much strain.
> 
> Thanks!
> 
> Mark
> 
> p.s. please cc to me directly on replies
> 
> --
> Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null

-- 
:::
ICQ: 75132336
http://www.aphroland.org/
http://www.linuxpowered.net/
[EMAIL PROTECTED]

linux disk editor/fat tarbal recover [time running out :( ]

[Richard E. Hawkins]

Once more, I'm trying to recover data from a tarball on a fat partition,
but
time is running out--I have to file my tax return by October 15, and so
I need 
to recover at least a week ahead of that (or I'll have to do it all
again,
and this is a big, ugly, return with lots of schedules :(   )

Roughly, I created a primary fat partition, tarred /home onto a tarball
on it,
deleted the partition, and tried to resurrect it as an extended
partition.  

I've learned so far that this wipes out the fat, whcih is a bad thing :)

The good news is that (I think) the tarball, which was uncompressed,
should be 
the first file on the partition, so the file should be contiguous.  No
files 
have been written to the partition since then. I've come to understand
that 
if I find one piece of it, I should be able to find therest. 

I've installed the hard drive into another machine to work with. I've
used dd
to copy a primary fat partition placed in the same place as the lost
partition.
I've used dd to copy the file to "recovery.2", and split that into 100M
pieces 
[having discovered that you can kill the system by trying to load a 1.7G
file 
into beav :) ]

I am now running a script "grepall" which reads

#! /sbin/sh

for file in `ls x*` ; do
  echo $file
  strings $file | grep -e "/home"  -e"hawk" > grep.${file}
  done

This is based on the premise that the first two lines of the tarball
should be
something like

/home
/home/hawk

It's clear that I'll get a match in grep.xaa like this.  It doesn't have
all the entries I'd expect, though.  Once I hit the tarball, I should
get lots.

This locates the file to within 100MB..  I could just recursively search
at
that point, but that would just beg the question.   So what do I do once
I 
find which one has my tarball (or at least the beginning?)

I've compiled lde, the only disk editor I could find on freshmeat.  But 
I have no idea what to do with the information.  I assume that somehow I
create a new file starting at the correct location and untar it.  But
how do 
I do this?

hawk, running out of time



-- 
Prof. Richard E. Hawkins, Esq.   Smeal 178(814) 375-4700
[EMAIL PROTECTED]
These opinions will not be those of Penn State until it pays my
retainer.

Re: machines (ii)

[Mark Brown]

On Sat, Sep 30, 2000 at 12:13:47PM -0400, Mark Simos wrote:

> I am looking to put together a Debian based firewall and a mail server
>   -how bad of an idea is it to host them on the same machine?
>(please explain how dumb it is, if so)

Well, if someone cracks your firewall then they'll also get your mail
and the mail server may provide an additional way into your firewall.
OTOH, how much do you care?

> How much power would I need (CPU/RAM/HD) to make it (or each of them)
> work?

How much load do you have?  For a home system on a modem a 486 should
handle the load from both quite happily.  With a broadband connection
you might want a somewhat more powerful CPU.

-- 
Mark Brown  mailto:[EMAIL PROTECTED]   (Trying to avoid grumpiness)
http://www.tardis.ed.ac.uk/~broonie/
EUFShttp://www.eusa.ed.ac.uk/societies/filmsoc/


pgpK7JKqlcxRi.pgp
Description: PGP signature

Why isn't .Xdefaults read?

[Preben Randhol]

Why do I have to run xrdb .Xdefaults manually when I start X ?

-- 
Preben Randhol - [EMAIL PROTECTED] - http://www.pvv.org/~randhol/   ._.
Debian 2.2 |"Don't think about domination, think about freedom,  / _,\
Potato | it doesn't dominate." - Richard M. Stallman| (_./
GNU/Linux  | To learn more visit => http://www.debian.org/   \,

Re: Problem with Lucent winmodem on debian 2.2

[Phil Brutsche]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

A long time ago, in a galaxy far, far way, someone said...

> Hi,
> 
> I am trying to configure Lucent Winmodem on my HP Pavillion (6735) box
> with Debian 2.2.  I have followed the instructions from
> www.linmodems.org for installing the binary only driver provided by
> Lucent, but still have problems in loading the driver.
> 
> The following bits should tell the story.. Can someone help me out? The
> modem is working fine with Windows ME.
> 
> I am not able to understand what exactly the problem is. 

> 1. Why kernel module is not getting loaded. (Lucent's driver is
> supposed to support shared IRQ - Shouldn't it probe for the IRQ?
> Windows ME uses IRQ 3)

Detecting the IRQ isn't the problem.  The driver is built for RedHat 6.1
(who has a long history of heavily patching their kernels), and what
you're using is a *hell* a lot more recent.

I was going to suggest "insmod -f", but I see you alreay tried that :)  
Since that didn't work, I have to say that you're really out of luck.  I
hope you didn't pay much for the modem...

As to why it doesn't work: A good analogy is you're trying to load a
driver for Win95 into WinME - the kernels are wy to different for the
driver to load.

In Linux, the driver is very sensitive to the kernel it's being loaded in
to.  If someone uses a different compiler from you, or, more importantly,
has a different set of patches from you when they built the driver,
there's a very good chance that the driver won't load.

> 2. Why setserial complains about "No such device"

The driver wasn't loaded.

> >From insmod -f ltmodem
> --
> Using /lib/modules/2.2.17/misc/ltmodem.o
> Warning: kernel-module version mismatch
>  /lib/modules/2.2.17/misc/ltmodem.o was compiled for kernel version
> 2.2.12-20
>  while this kernel is version 2.2.17
> 
> /lib/modules/2.2.17/misc/ltmodem.o: init_module: Device or resource busy
> Hint: this error can be caused by incorrect module parameters, including
> invalid IO or IRQ parameters

[...]

- -- 
- --
Phil Brutsche   [EMAIL PROTECTED]

GPG fingerprint: 9BF9 D84C 37D0 4FA7 1F2D  7E5E FD94 D264 50DE 1CFC
GPG key id: 50DE1CFC
GPG public key: http://tux.creighton.edu/~pbrutsch/gpg-public-key.asc
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE51jYI/ZTSZFDeHPwRApQ7AKDUe++f4tXNJ2I8H9UYvjWZUbmtigCdHg2A
ey7Rw3+6z22vkTf63QuflxY=
=wrZJ
-END PGP SIGNATURE-

Eterm, /dev/null, and segfaults

[rich]

Howdy all,

I just installed Slink onto an HP pentium 166 from CD. I'm using
Windowmaker... When I tried to use Eterm as root, it worked... when I
tried to use it as a normal user, it said something about /dev/null and
segfaulted...

/dev/null is set up like so... crw-rw-rw-

xterm, however, worked for both root and normal users.


I then used apt to upgrade to Potato over ethernet Eterm still works
for root, but now when I try to run Eterm OR xterm as normal user,
absolutely nothing happens. No window, no disk-noises, nothing. However,
if I switch to twm, I can get a cheesy-looking xterm. Strange?

I last clue when I log out of Windowmaker, before switching to back
to wdm, a screen of strange SVGA-looking chaotic graphics briefly
flashes - it does NOT do this for root.

Any clues?

Thanks in advance,

Rich

Re: Good Book for setting up T-1?

[Robert Waldner]

On Sat, 30 Sep 2000 10:59:32 PDT, George Bonser writes:
>On Sat, 30 Sep 2000, [EMAIL PROTECTED] wrote:
>
>> i think your best off contacting your local telco company and asking them
>> what they reccomend as far as CSU/DSU, and as far as routers, depends on
>> your needs, i usually use cisco 2500 series for t1s.
>
>The 1600's are several hundreds of dollars cheaper. A 1601 is about $500
>cheaper than a 2501.

and a 1601 is perfectly enough for a single T1, you don´t need the 
 extra power of a 25xx for running a single line. Just make sure you go 
for a 1601_R_, not for an old CH-model.

just my 2 cents,
&rw
-- 
/ Robert Waldner <[EMAIL PROTECTED]> | Phone: +43 1 89933 0 Fax x533 \
\KPNQwest/AT tech staff| Diefenbachg. 35   A-1150 Wien / 

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Ben Collins]

On Sat, Sep 30, 2000 at 01:56:33PM -0400, Greg Stark wrote:
> 
> Ben Collins <[EMAIL PROTECTED]> writes:
> 
> > For those bitten by The Great Glibc Update of 2000, welcome to our annual
> > ritual. Please stay tuned during the next few months where we install a new
> > gcc. This will be followed by self inflicted pain of stricter C++ syntax,
> > macro collisions, binary incompatability and general chaos among kernel
> > builds[1]. Thank you for your patronage, and please come again[2].
> 
> It's still the case that binaries compiled on woody or RedHat 7 will crash
> randomly on potato or Redhat 6? If so in the current system that still
> requires changing the soname but of course that's a losing battle, right?

Backward compatible binaries are not guaranteed, I don't think. As for
soname, GLibc uses versioned symbols for a finer grain control of this. No
reason to change the soname if only two symbols change their ABI.

> Hmph, I wonder if ld.so/ld could be patched to have a list of sonames provided
> by a library for run-time but a single one to use for compile-time linking.

RH gets around this with a gross hack (IMHO) by making binaries depends on
library symbols as opposed to the library itself. Not something that is
very intuitive (do you have all the symbols of all the libraries ever
memorized?).

Debian gets around this by having strict dependencies on the packages.
That says nothing for straight binary compiles, but it works for package
installs. There's not much we can do about binary tarballs.

> That way glibc2.2 could provide sonames 2.0, 2.1, and 2.2 for run-time linking
> but when programs are compiled against it they would record 2.2 as the version
> they should be linked with.
> 
> That seems like it would fix the binary incompatibility problems that glibc is
> introducing in linux recently. 

Recently? :) It also doesn't help when the compiler breaks this aswell. We
do what we can without overcompilcating things.

-- 
 ---===-=-==-=---==-=--
/  Ben Collins  --  ...on that fantastic voyage...  --  Debian GNU/Linux   \
`  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  '
 `---=--===-=-=-=-===-==---=--=---'

Re: machines (ii)

[George Bonser]

On Sat, 30 Sep 2000, Mark Simos wrote:

> I am looking to put together a Debian based firewall and a mail server
>   -how bad of an idea is it to host them on the same machine?
>(please explain how dumb it is, if so)
> 
> How much power would I need (CPU/RAM/HD) to make it (or each of them)
> work?
> 
> I just browse at home and download the occasional files and would like
> to learn how to configure various mail and firewall packages. not too
> much strain.
> 
> Thanks!
> 
> Mark

Mark, I would not put the mail server on the firewall. This is because of
disk requirements. It is pretty easy for a lot of users on a machine to
fill up disk fairly quickly. About the only time I would put a mailer on a
firewall is for use as a transparent proxy to handle outbound mail or as a
relay machine for inbound mail to route it to your real mailhost.

You do not need much in the way of CPU power. One method is to get a BUNCH
of RAM, create a bootable CDROM that boots initrd and runs from a
ramdisk. Set up syslog to log to a remote system that has a hard disk,
remove ALL hard disks from the system. Now if someone roots your box, you
just reboot and everything they did evaporates with the freshly loaded
binaries. Once you discover how they got in, you create another bootable
CDROM and reboot the box to load the more secure stuff.

I would use ECC RAM in this configuration. Since logs are being sent to a
remote system, and since nothing is persistant over a reboot, and since
there are no hard disks to fail, you have a fairly secure and robust
firewall.

Re: half-installed packages

[Jesse Goerz]

On Fri, 29 Sep 2000, Thomas J. Hamman wrote:
> I was walking my gf through an install of Debian (it didn't scare her,
> and she's no techie), and everything went fine until she tried
> installing some stuff.
>
> Installing task-x-window-system basically choked, and now she is left
> with a bunch of half-installed packages.  I can't tinker with it myself
> since she is annoyingly located hundreds of miles away from me.
>
> What is the official correct way to deal with lots of half-installed
> packages?
>
> Also, while I'm asking, if you install a task, what's the quickest and
> easiest way to just get rid of everything it installs?  She installed a
> couple other tasks in tasksel before realizing it would probably be
> better to start with a minimal system and apt-get as needed..

I had this happen a couple of times because my apt sources.list file didn't 
have everything it needed to get the appropiate dependencies.  It returned an 
error and the next time I ran apt-get install it told me to run...

apt-get -f install

...with no packages listed.  I guess this forces it to complete the install 
and then you can try to grab the rest of the dependencies.  I don't know if 
this will work for you but good luck.

Jesse

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Daniel Jacobowitz]

On Sat, Sep 30, 2000 at 09:33:55AM -0400, Ben Collins wrote:
> If you had problems with the last few 2.1.94 packages (and even if you
> didn't), get these. They are currently in incoming (to be installed later
> this afternoon, then mirrors will get it some hours later). Make sure you
> get the libdb2 that goes along with it.
> 
> I don't have time to list everything it fixes, but I can tell you that it
> fixes EVERY issue that EVERYONE was having. Installing this should bring
> you back to a worry free stable setup again (remember this is unstable, so
> if your system breaks, it's probably someone elses fault, so I can't
> guarantee everything).

A good other breakage to be aware of is bug #72803 against sshd.  You
may need to use ssh -X (no x11 forwarding) if you do not have ipv6
enabled until this is fixed.

Dan

/\  /\
|   Daniel Jacobowitz|__|SCS Class of 2002   |
|   Debian GNU/Linux Developer__Carnegie Mellon University   |
| [EMAIL PROTECTED] |  |   [EMAIL PROTECTED]  |
\/  \/

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Roland Bauerschmidt]

On Sat, Sep 30, 2000 at 07:00:59PM +0200, Marcelo E. Magallon wrote:
>  I thought this was chance, but Christian Kurz said the same happened
>  to him.

It didn't happen to me. FYI, I am using postfix and not exim...

Roland

-- 
Roland Bauerschmidt <[EMAIL PROTECTED]>

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Daniel Jacobowitz]

On Sat, Sep 30, 2000 at 01:56:33PM -0400, Greg Stark wrote:
> 
> Ben Collins <[EMAIL PROTECTED]> writes:
> 
> > For those bitten by The Great Glibc Update of 2000, welcome to our annual
> > ritual. Please stay tuned during the next few months where we install a new
> > gcc. This will be followed by self inflicted pain of stricter C++ syntax,
> > macro collisions, binary incompatability and general chaos among kernel
> > builds[1]. Thank you for your patronage, and please come again[2].
> 
> It's still the case that binaries compiled on woody or RedHat 7 will crash
> randomly on potato or Redhat 6? If so in the current system that still
> requires changing the soname but of course that's a losing battle, right?
> 
> Hmph, I wonder if ld.so/ld could be patched to have a list of sonames provided
> by a library for run-time but a single one to use for compile-time linking.
> 
> That way glibc2.2 could provide sonames 2.0, 2.1, and 2.2 for run-time linking
> but when programs are compiled against it they would record 2.2 as the version
> they should be linked with.
> 
> That seems like it would fix the binary incompatibility problems that glibc is
> introducing in linux recently. 

I'd think you could just do this with symlinks - the problem is that
you'd need to change the soname away from libc6, which is a whole world
of chaos.

Dan

/\  /\
|   Daniel Jacobowitz|__|SCS Class of 2002   |
|   Debian GNU/Linux Developer__Carnegie Mellon University   |
| [EMAIL PROTECTED] |  |   [EMAIL PROTECTED]  |
\/  \/

Re: DHCP renumbering and NAT

[Leen Besselink]

On 30 Sep 2000, David Z Maze wrote:

> I can take it as a fact of life that I'll occasionally be renumbered;
> I can deal with this.  I guess my question is this: is the ipmasq
> package clueful enough to recognize when this happens, and tweak the
> firewall rules appropriately?  (Experience this morning suggests not.)
> Failing this, is there a good way to put some sort of appropriate hook 
> into dhclient to make the right thing happen?  TIA...
> 

What you can do is just not use the outside _address_ when masq. but the
interface name, you can do the same for blocking.

Hope this help,
Lennie.

Re: Good Book for setting up T-1?

[George Bonser]

On Sat, 30 Sep 2000, [EMAIL PROTECTED] wrote:

> i think your best off contacting your local telco company and asking them
> what they reccomend as far as CSU/DSU, and as far as routers, depends on
> your needs, i usually use cisco 2500 series for t1s.

The 1600's are several hundreds of dollars cheaper. A 1601 is about $500
cheaper than a 2501.

Re: Anyone want a utility to find the best mirror?

[Seth Cohn]

> Debian has a apt-cache program that works great for exactly
> that purpose.

Correction:  it's called apt-proxy.  apt-cache is entirely different. :)

BTW, while I'm at it:   You folks who wrote "sign me up" and "me
too" - _ever_ heard of private replies?  This isn't AOL, thank you very
much. 

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Greg Stark]

Ben Collins <[EMAIL PROTECTED]> writes:

> For those bitten by The Great Glibc Update of 2000, welcome to our annual
> ritual. Please stay tuned during the next few months where we install a new
> gcc. This will be followed by self inflicted pain of stricter C++ syntax,
> macro collisions, binary incompatability and general chaos among kernel
> builds[1]. Thank you for your patronage, and please come again[2].

It's still the case that binaries compiled on woody or RedHat 7 will crash
randomly on potato or Redhat 6? If so in the current system that still
requires changing the soname but of course that's a losing battle, right?

Hmph, I wonder if ld.so/ld could be patched to have a list of sonames provided
by a library for run-time but a single one to use for compile-time linking.

That way glibc2.2 could provide sonames 2.0, 2.1, and 2.2 for run-time linking
but when programs are compiled against it they would record 2.2 as the version
they should be linked with.

That seems like it would fix the binary incompatibility problems that glibc is
introducing in linux recently. 

-- 
greg

Re: Anyone want a utility to find the best mirror?

[Jesse Goerz]

On Sat, 30 Sep 2000, will trillich wrote:
> On Thu, Sep 28, 2000 at 07:37:28PM -0400, Jim Lisi wrote:
> > > Joe Emenaker wrote:
> > > Anyway, if anybody wants to give it a whirl, I'm open to some
> > > beta-testing.
> > >
> > > - Joe
> >
> > Sounds interesting, I'd be willing to beta-test it.
>
> sign me up!

Me too!

DHCP renumbering and NAT

[David Z Maze]

I'm hiding a small apartment network behind a potato box running IP
masquerading.  This all works fine for the most part; I use dhclient
and the ipmasq package, and everything just works.

This morning, I couldn't reach the outside world.  Some poking around
revealed that I couldn't even reach the machine my firewall believed
was its gateway.  I restarted dhclient, and got a completely new IP
address on a different network.  (I assume this is a mistake on the
part of my provider: my understanding is that a DHCP lease should be
valid for the entire length of the lease, and breaking things under me 
is Just Wrong.)  Even after this, though, things still didn't work;
looking at my syslogs suggested that I might need to reinitialize the
firewall rules (which blocked outgoing packets not on the local
network or from the gateway's public IP), and in fact running
/etc/init.d/ipmasq restart got things running again.

I can take it as a fact of life that I'll occasionally be renumbered;
I can deal with this.  I guess my question is this: is the ipmasq
package clueful enough to recognize when this happens, and tweak the
firewall rules appropriately?  (Experience this morning suggests not.)
Failing this, is there a good way to put some sort of appropriate hook 
into dhclient to make the right thing happen?  TIA...

-- 
David Maze [EMAIL PROTECTED]  http://www.mit.edu/~dmaze/
"Theoretical politics is interesting.  Politicking should be illegal."
-- Abra Mitchell

Re: Print filter/tool recommendations?

[John Carline]

Svante Signell wrote:

> Anyone who can recommend good print filters for ordinary text?
>
> lpr gives lousy output: pages are too long and wide, the font is too large.
> lpf or pr is not much of use either.
>
> man -k print|grep filter gives:
> lpf (1)  - general printer filter
> magicfilter (8)  - automatic configurable printer filter
> magicfilterconfig (8) - writes interactively a printcap file.
> fixpspps (1) - filter to fix PSPrint PostScript so PSUtils work
> vprint (1)   - print filter to break long lines
>
> Not much here either. Pointers, suggestions?

magicfilter didn't work? Or did you not try magicfilterconfig?

I'm currently using a deskjet (HP890C) and magicfilter is working fine. If 
you've
tried it and it still doesn't size right maybe (only a wild guess) fonts?.

John


--

Powered by the Penguin

Re: Anyone want a utility to find the best mirror?

[Seth Cohn]

On Sat, 30 Sep 2000, Joe Emenaker wrote:
> > This is a nice idea.  You are doing both http _and_ ftp, right?
> 
> Not yet. Do you have any numbers on how many people actually use http for
> dselect? Strangely, I've almost always used FTP for transferring files and
> HTTP for transferring hypertext. Call me silly

These days, http is the preference of anyone behind a firewall (more and
more people all the time).  FTP is a complex protocol, using multiple
ports (and active versus passive diferences, since only passive works
behind the firewall).  Http is allowed through most firewalls, and caches
really well.  Debian has a apt-cache program that works great for exactly
that purpose.  You are doing a one way transfer, file at a time, and http
shines for that.

> I'll take a look at netselect and see if I can get it to give me the top X
> in the list. Last time I ran it, it just kinda sat there for a minute and
> then it just blurted out one host name.

You can make it return the top X, you can also make it return stats.
Using perl, you can then look at the results and figure out which boxes to
check further.

Re: bitchx and forward delete key

[[EMAIL PROTECTED]]

i've been using bitchx for many years and never had a need for either
forward delete or ctcp cloaking. same goes for scrollz or ircii or epic4
..

nate

On Sat, 30 Sep 2000, C. Falconer wrote:

criggi >At 03:34 PM 9/29/00 -0400, you wrote:
criggi >>Is there any reason that anyone would object if I changed the forward 
delete
criggi >>key to do something sane, like delete forward, instead of toggling 
cloaking?
criggi >>
criggi >>Replies just to me, please.
criggi >
criggi >I don't have an answer to your question sorry - but why request replies 
not 
criggi >go to the list?  I happen to have started using bitchx at work, and I 
might 
criggi >be interested in someones answer... also the thread can diverge into 
criggi >another semi-related area which other people will find interesting.
criggi >
criggi >Just a thought.
criggi >
criggi >-- 
criggi >Criggie
criggi >
criggi >
criggi >-- 
criggi >Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
criggi >

:::
http://www.aphroland.org/
http://www.linuxpowered.net/
[EMAIL PROTECTED]
10:02am up 14 days, 17:58, 1 user, load average: 0.16, 0.06, 0.02

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Marcelo E. Magallon]

>> Ben Collins <[EMAIL PROTECTED]> writes:

 > If you had problems with the last few 2.1.94 packages (and even if you
 > didn't), get these. They are currently in incoming (to be installed later
 > this afternoon, then mirrors will get it some hours later). Make sure you
 > get the libdb2 that goes along with it.

 Impatient as I might be (and wanting to check if the plaympeg problem
 went away) I upgraded libc6, libc6-dev, locales and libdb2 to the
 versions in incoming.  When configuring libc6 it asked if I wanted to
 restart running services.  I said yes, it did so with exim and ssh...
 it was about to do the same with apache and the machine froze.  I
 don't have remote access to this box, so I couldn't check how bad
 this was.  Magic-SysRq KSUB did work.

 I thought this was chance, but Christian Kurz said the same happened
 to him.


 Marcelo

Re: HELP! xconfig won't give access to modules

[[EMAIL PROTECTED]]

did you enable expermintal driver support? that should make the rtl8139
driver show up ..i use htat card in a lot of systems too.

nate

On Sat, 30 Sep 2000, Cam Ellison wrote:

cam_el >I have an ethernet card that requires the rtl8139 driver.  The driver
cam_el >is present in the kernel source (2.2.17), but xconfig (not to mention
cam_el >config and menuconfig) does not make it available for selection, i.e.
cam_el >it is greyed out.  To forestall the obvious question, yes, I have a
cam_el >number of networking options set to "Y", Network device support is set
cam_el >to "Y", Loadable module support is Enabled, the Kernel module loader is
cam_el >set to "Y", Networking support and PCI support are on (it is a PCI
cam_el >card).  IF you need to know more, I have attached the entire setup.
cam_el >
cam_el >The card is installed, and ifconfig displays it correctly.
cam_el >
cam_el >Why is this happening (or not happening, depending on your
cam_el >perspective)?
cam_el >
cam_el >What do I do about this?  
cam_el >
cam_el >Thank you, anyone, for whatever help you can provide.  This is driving
cam_el >me nuts!
cam_el >
cam_el >Cam
cam_el >
cam_el >
cam_el >
cam_el >Cam Ellison, Ph.D., R.Psych.
cam_el >
cam_el >[EMAIL PROTECTED]  or [EMAIL PROTECTED]
cam_el >
cam_el >>From the lovely Sunshine Coast, where it only SEEMS to rain.
cam_el >
cam_el >
cam_el >
cam_el > 
cam_el >

:::
http://www.aphroland.org/
http://www.linuxpowered.net/
[EMAIL PROTECTED]
10:02am up 14 days, 17:58, 1 user, load average: 0.16, 0.06, 0.02

Re: Good Book for setting up T-1?

[[EMAIL PROTECTED]]

i think your best off contacting your local telco company and asking them
what they reccomend as far as CSU/DSU, and as far as routers, depends on
your needs, i usually use cisco 2500 series for t1s.

and i'd probably reccomend having the telco setup the CSU/DSU and have
your isp setup the router. make it easier and much faster.

i run an isp that used to run offa t1(now its colocated) and we did that
..worked fine for years.

nate

On Sat, 30 Sep 2000 [EMAIL PROTECTED] wrote:

jack >Hello all,
jack >I have been using deb linux for some 5 years now and am quite happy with 
it.  It has been a webserver for me for only 1 of those years and that is on a 
DSL.  As it trns out, some of the people I've done some contract work with wish 
to install a t1 line and run debian as the OS on all the systems.  
jack >
jack >Anyone have a line on some good documentation on how do go about this?  I 
know some additional hardware is needed, and here's what I want.  Two 
nameservers, one webserver, and a mail server right off the bat.  We wish to 
have access to all the machines on a 24H basis for peace of mind <- why we 
don't use a hosting service.
jack >
jack >Anyway, I have not done this before... maybe someone could point me in 
the direction of a list of hardware needed.. CSU/DSU, routers, etc...
jack >Thanks,
jack >Jack
jack >
jack >
jack >-- 
jack >Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
jack >

:::
http://www.aphroland.org/
http://www.linuxpowered.net/
[EMAIL PROTECTED]
10:02am up 14 days, 17:58, 1 user, load average: 0.16, 0.06, 0.02

Re: Problems with AC97 PCI sound card

[David Z Maze]

Francois Fayard <[EMAIL PROTECTED]> writes:
FF> I own a AC97 PCI Audio sound card built in my main board.

I've had decent luck with a similar configuration using ALSA, using
the snd-card-via686a module.

FF> I've tried to load the ac97_codec module with modconf.
FF> Here is the list of my modules loaded in my kernel
FF> 
FF> Module  Size  Used by
FF> soundcore   2628   0  (autoclean) (unused)
FF> ac97_codec  7044   0  (unused)

This pretty clearly indicates that the kernel OSS sound driver isn't
noticing your card; I don't know a good way to work around that, though.

-- 
David Maze [EMAIL PROTECTED]  http://www.mit.edu/~dmaze/
"Theoretical politics is interesting.  Politicking should be illegal."
-- Abra Mitchell

Re: Problem with Lucent winmodem on debian 2.2

[[EMAIL PROTECTED]]

good luck getting it to work, your best bet is to use the kernel that it
is asking for 2.2.12-20 which probably comes with some version of redhat..

thats one of the problems with binary kernel drivers 99% of the time they
do NOT work on other kernel versions, there is nothing you can do except
ditch the hardware that uses it, or change kernels.

nate

On Sat, 30 Sep 2000, Shaji N V wrote:

nvshaj >Hi,
nvshaj >
nvshaj >I am trying to configure Lucent Winmodem on my HP Pavillion (6735) box 
with
nvshaj >Debian 2.2. I have followed the instructions from www.linmodems.org for 
installing the binary only driver provided by Lucent, but still have problems 
in 
nvshaj >loading the driver.
nvshaj >
nvshaj >The following bits should tell the story.. Can someone help me out? The
nvshaj >modem is working fine with Windows ME.
nvshaj >
nvshaj >I am not able to understand what exactly the problem is. 
nvshaj >1. Why kernel module is not getting loaded. (Lucent's driver is 
supposed to support shared IRQ - Shouldn't it probe for the IRQ? Windows ME 
uses IRQ 3)
nvshaj >2. Why setserial complains about "No such device"
nvshaj >
nvshaj >
nvshaj >Thanks in advance,
nvshaj >Shaji
nvshaj >
nvshaj >
nvshaj >>From insmod -f ltmodem
nvshaj >--
nvshaj >Using /lib/modules/2.2.17/misc/ltmodem.o
nvshaj >Warning: kernel-module version mismatch
nvshaj > /lib/modules/2.2.17/misc/ltmodem.o was compiled for kernel version
nvshaj >2.2.12-20
nvshaj > while this kernel is version 2.2.17
nvshaj >
nvshaj >/lib/modules/2.2.17/misc/ltmodem.o: init_module: Device or resource busy
nvshaj >Hint: this error can be caused by incorrect module parameters, including
nvshaj >invalid IO or IRQ parameters
nvshaj >
nvshaj >
nvshaj >
nvshaj >>From cat /proc/pci
nvshaj >--
nvshaj >
nvshaj >Communication controller: Lucent (ex-AT&T) Microelectronics Unknown
nvshaj >device
nvshaj >(rev 0).
nvshaj >  Vendor id=11c1. Device id=44e.
nvshaj >  Medium devsel.  Fast back-to-back capable.  Master Capable.
nvshaj >Latency=64.
nvshaj >Min Gnt=252.Max Lat=14.
nvshaj >  Non-prefetchable 32 bit memory at 0xf410 [0xf410].
nvshaj >  I/O at 0x3400 [0x3401].
nvshaj >  I/O at 0x3000 [0x3001].
nvshaj >
nvshaj >>From /etc/serial.conf
nvshaj >-
nvshaj >
nvshaj ># These are two spare devices you can use to customize for
nvshaj ># some board which is not supported above
nvshaj >#
nvshaj >/dev/ttyS14 uart 16450 port 0x0260 irq 3
nvshaj >#/dev/ttyS15 uart X port  irq X
nvshaj >
nvshaj >>From setserial -agv /dev/ttyS*
nvshaj >--
nvshaj >/dev/ttyS0, Line 0, UART: 16550A, Port: 0x03f8, IRQ: 4
nvshaj > Baud_base: 115200, close_delay: 50, divisor: 0
nvshaj > closing_wait: 3000
nvshaj > Flags: spd_normal skip_test
nvshaj >
nvshaj >/dev/ttyS1, Line 1, UART: unknown, Port: 0x02f8, IRQ: 3
nvshaj > Baud_base: 115200, close_delay: 50, divisor: 0
nvshaj > closing_wait: 3000
nvshaj > Flags: spd_normal skip_test
nvshaj >
nvshaj >/dev/ttyS14: No such device
nvshaj >/dev/ttyS2, Line 2, UART: unknown, Port: 0x03e8, IRQ: 4
nvshaj > Baud_base: 115200, close_delay: 50, divisor: 0
nvshaj > closing_wait: 3000
nvshaj > Flags: spd_normal skip_test
nvshaj >
nvshaj >/dev/ttyS3, Line 3, UART: unknown, Port: 0x02e8, IRQ: 3
nvshaj > Baud_base: 115200, close_delay: 50, divisor: 0
nvshaj > closing_wait: 3000
nvshaj > Flags: spd_normal
nvshaj >
nvshaj >>From ls -l /dev/ttyS*
nvshaj >-
nvshaj >crw-rw1 root dialout4,  64 Jul  5 23:14 /dev/ttyS0
nvshaj >crw-rw1 root dialout4,  65 Jul  5 23:14 /dev/ttyS1
nvshaj >crw-rw1 root dialout62,  78 Sep 28 05:42 /dev/ttyS14
nvshaj >crw-rw1 root dialout4,  66 Jul  5 23:14 /dev/ttyS2
nvshaj >crw-rw1 root dialout4,  67 Jul  5 23:14 /dev/ttyS3
nvshaj >
nvshaj >

:::
http://www.aphroland.org/
http://www.linuxpowered.net/
[EMAIL PROTECTED]
10:02am up 14 days, 17:58, 1 user, load average: 0.16, 0.06, 0.02

Re: tr '\verb|\|000' '\verb|\|\n'?

[David Z Maze]

Johann Spies <[EMAIL PROTECTED]> writes:
JS> Can somebody explain the subject line to me please.  I have read it in
JS> a linux training document and it is not doing what the document says
JS> it should do.  What I do not understand and do not know where to find
JS> documentation on it is the '\verb|\|000' '\verb|\|\n' part.  I know tr
JS> but what is '\verb|\|000'?  And the use of |\|?  

Are you reading this out of the source of a LaTeX document?  In LaTeX, 
the notation \verb|foo| means to print the foo verbatim, in a monotype 
font, without interpreting any of the contents as "special" (and
backslashes are *very* special in LaTeX).  More likely, the formatted
output would look like

tr '\0' '\n'

which would try to transform null bytes in the input into newlines.

JS> The whole command according to the document is
JS> 
JS> cat /proc/$$/environ | tr '\verb|\|000' '\verb|\|\n' 
JS> 
JS> which is supposed to put each variable it shows on a separate
JS> line.

Right, that makes sense here.

-- 
David Maze [EMAIL PROTECTED]  http://www.mit.edu/~dmaze/
"Theoretical politics is interesting.  Politicking should be illegal."
-- Abra Mitchell

Re: Print filter/tool recommendations?

[Thomas J. Hamman]

On Sat, Sep 30, 2000 at 03:26:59PM +0200, Svante Signell wrote:
> Anyone who can recommend good print filters for ordinary text?
> 
> lpr gives lousy output: pages are too long and wide, the font is too large.
> lpf or pr is not much of use either.
> 
> man -k print|grep filter gives:
> lpf (1)  - general printer filter
> magicfilter (8)  - automatic configurable printer filter
> magicfilterconfig (8) - writes interactively a printcap file.
> fixpspps (1) - filter to fix PSPrint PostScript so PSUtils work
> vprint (1)   - print filter to break long lines

Magicfilter (configured with magicfilterconfig) works fine for me and my
HP DeskJet 672C.  Apsfilter seems okay too.

-- 
Tom
"It is impossible for a man to learn what he thinks he already knows."
-Epictetus

Customizing menus

[Christopher S. Swingley]

Hello!

I'm trying to alter what happens when I choose "Programs | XShells | 
Eterm" from my window manager's pop-up menu (sawfish) so it brings up 
a transparent Eterm with some other customizations instead of the 
themed Eterm that normally appears.  What is _The Right Way_ to do 
this?  The docs for menu suggest a ~/.menu directory with files 
underneath.  I copied the entry for Eterm into my ~/.menu directory 
and tweaked it, but nothing changed.

Any suggestions / hints?

Thanks,

Chris
-- 
Christopher S. Swingley   tel: 907-474-2689 cell: 322-1889
Programmer / Analyst  email: [EMAIL PROTECTED]
University of Alaska Fairbankswww.frontier.iarc.uaf.edu/~cswingle
Fairbanks, AK  99775

PGP2 key:   http://www.frontier.iarc.uaf.edu/~cswingle/pgpkey.asc
GNUPG key:  http://www.frontier.iarc.uaf.edu/~cswingle/gnupgkey.asc

Re: Good Book for setting up T-1?

[George Bonser]

> 
> Anyway, I have not done this before... maybe someone could point me in the 
> direction of a list of hardware needed.. CSU/DSU, routers, etc...
> Thanks,
> Jack

Well, You can get a small Cisco router, say a 1600 series with a WIC card
that has an integrated CSU/DSU. You just plug the T1 line directly into
the card. It will be farily simple to configure, as a matter of fact, the
provider may be able to give you a sample config for it or even provide
the router. A 1600 series router is about the same size as a standard DSL
or cable modem box.

informazioni

[Michele Savastano]

Potete dirmi per piacere dove posso acquistaare in 
Italia i Vs. CD.
Grazie Michele Savastano

Re: Glibc 2.1.94-3, fixes all issues with db libraries

[Decklin Foster]

Ben Collins writes:

> I don't have time to list everything it fixes, but I can tell you that it
> fixes EVERY issue that EVERYONE was having.

Indeed, it's working perfectly here. Cheers! (I will no longer have to
bug anyone about recompiling Perl...)

> For those bitten by The Great Glibc Update of 2000, welcome to our annual
> ritual.

As a visiting teacher of Zen told me last week, "pain is an essential
part of practice." It's good to know we're looking out for the
spiritual health of our users. ;-)

-- 
There is no TRUTH. There is no REALITY. There is no CONSISTENCY. There
are no ABSOLUTE STATEMENTS. I'm very probably wrong. -- BSD fortune(6)

Re: tr '\verb|\|000' '\verb|\|\n'?

[Brad]

On Sat, Sep 30, 2000 at 01:37:40PM +0200, Johann Spies wrote:
> I know tr but what is '\verb|\|000'?  And the use of |\|?  

First, there's apparently a bug: there should be a backslash before the
0s.

That said, the "\verb|\|" is apparently redundant, since it appears at
the same place in both sets. It'll translate vertical tab (\v) to
vertical tab, e to e, r to r, and so on.

Maybe it's an attempt to confuse beginners? Or possibly, it's some
obscure feature for a non-GNU version of tr?


-- 
  finger for GPG public key.


pgpxah3E4jzJd.pgp
Description: PGP signature

machines (ii)

[Mark Simos]

I am looking to put together a Debian based firewall and a mail server
  -how bad of an idea is it to host them on the same machine?
   (please explain how dumb it is, if so)

How much power would I need (CPU/RAM/HD) to make it (or each of them)
work?

I just browse at home and download the occasional files and would like
to learn how to configure various mail and firewall packages. not too
much strain.

Thanks!

Mark

p.s. please cc to me directly on replies

machines

[Mark Simos]

I am looking to put together a Debian based firewall and a mail server
  -how bad of an idea is it to host them on the same machine?
(please explain how dumb it is, if so)

How much power would i need (CPU/RAM/HD) to make it (or each of them)
work?

I just browse at home and download the occasional files and would like
to learn how to configure various mail and firewall packages. not too
much strain.

Thanks!

Mark

Re: HELP! xconfig won't give access to modules

[Brad]

On Sat, Sep 30, 2000 at 08:03:17AM +0800, Cam Ellison wrote:
> I have an ethernet card that requires the rtl8139 driver.  The driver
> is present in the kernel source (2.2.17), but xconfig (not to mention
> config and menuconfig) does not make it available for selection, i.e.
> it is greyed out.

As you may know, the way the kernel determines what modules are
available is through the Config.in files in various subdirectories.

First, by checking the Configure.help file, we find that the variable
for the rtl8139 driver is called "CONFIG_RTL8139". Searching
linux/drivers/net/Config.in, we find it specified in the following
section:
bool 'Ethernet (10 or 100Mbit)' CONFIG_NET_ETHERNET
if [ "$CONFIG_NET_ETHERNET" = "y" ]; then
 ...SNIP...
 if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then
   tristate 'RealTek 8129/8139 (not 8019/8029!) support' CONFIG_RTL8139
 fi
 ...SNIP...
fi

Apparently, the rtl8139 driver is considered experimental, so you'll
need to specify "Prompt for development and/or incomplete code/drivers"
   


-- 
  finger for GPG public key.


pgp9fQbMKImEu.pgp
Description: PGP signature

Re: stripping linux from a box

[Mike Leone]

>I'm presuming that LILO wrote to the master boot record.  So how does one go
>about fully expunging Linux?

fdisk /mbr to clen out the master boot record.



--
*-^-*-^-*-^-*-^-*-^-*-^-*-^-*-^-*-^-*-^-*-^-*-^-*-^-*
Michael Leone 

PGP Fingerprint: 0AA8 DC47 CB63 AE3F C739 6BF9 9AB4 1EF6 5AA5 BCDF
PGP Key ID:  0x5AA5BCDF
--

Problem with Lucent winmodem on debian 2.2

[Shaji N V]

Hi,I am 
trying to configure Lucent Winmodem on my HP Pavillion (6735) box withDebian 
2.2. I have followed the instructions from www.linmodems.org for installing the binary 
only driver provided by Lucent, but still have problems in 
loading the 
driver.The following bits should tell the story.. Can someone help me 
out? Themodem is working fine with Windows ME.
 
I am not able 
to understand what exactly the problem is. 
1. Why kernel 
module is not getting loaded. (Lucent's driver is supposed to support shared IRQ 
- Shouldn't it probe for the IRQ? Windows ME uses IRQ 3)
2. Why 
setserial complains about "No such device"
Thanks in advance,ShajiFrom insmod -f 
ltmodem--Using 
/lib/modules/2.2.17/misc/ltmodem.oWarning: kernel-module version 
mismatch /lib/modules/2.2.17/misc/ltmodem.o was compiled for kernel 
version2.2.12-20 while this kernel is version 
2.2.17/lib/modules/2.2.17/misc/ltmodem.o: init_module: Device or 
resource busyHint: this error can be caused by incorrect module parameters, 
includinginvalid IO or IRQ parametersFrom cat 
/proc/pci--    Communication 
controller: Lucent (ex-AT&T) Microelectronics Unknowndevice(rev 
0).  Vendor id=11c1. Device 
id=44e.  Medium devsel.  Fast back-to-back 
capable.  Master Capable.Latency=64.Min Gnt=252.Max 
Lat=14.  Non-prefetchable 32 bit memory at 
0xf410 [0xf410].  I/O at 0x3400 
[0x3401].  I/O at 0x3000 [0x3001].From 
/etc/serial.conf-# These are two spare devices 
you can use to customize for# some board which is not supported 
above#/dev/ttyS14 uart 16450 port 0x0260 irq 3#/dev/ttyS15 uart 
X port  irq XFrom setserial -agv 
/dev/ttyS*--/dev/ttyS0, Line 0, UART: 
16550A, Port: 0x03f8, IRQ: 4 Baud_base: 115200, close_delay: 50, 
divisor: 0 closing_wait: 3000 Flags: spd_normal 
skip_test/dev/ttyS1, Line 1, UART: unknown, Port: 0x02f8, IRQ: 
3 Baud_base: 115200, close_delay: 50, divisor: 0 closing_wait: 
3000 Flags: spd_normal skip_test/dev/ttyS14: No such 
device/dev/ttyS2, Line 2, UART: unknown, Port: 0x03e8, IRQ: 
4 Baud_base: 115200, close_delay: 50, divisor: 0 closing_wait: 
3000 Flags: spd_normal skip_test/dev/ttyS3, Line 3, UART: 
unknown, Port: 0x02e8, IRQ: 3 Baud_base: 115200, close_delay: 50, 
divisor: 0 closing_wait: 3000 Flags: spd_normalFrom ls 
-l /dev/ttyS*-crw-rw    1 
root dialout    4,  64 Jul  5 
23:14 /dev/ttyS0crw-rw    1 root 
dialout    4,  65 Jul  5 23:14 
/dev/ttyS1crw-rw    1 root 
dialout    62,  78 Sep 28 05:42 
/dev/ttyS14crw-rw    1 root 
dialout    4,  66 Jul  5 23:14 
/dev/ttyS2crw-rw    1 root 
dialout    4,  67 Jul  5 23:14 
/dev/ttyS3

Good Book for setting up T-1?

[jack]

Hello all,
I have been using deb linux for some 5 years now and am quite happy with it.  
It has been a webserver for me for only 1 of those years and that is on a DSL.  
As it trns out, some of the people I've done some contract work with wish to 
install a t1 line and run debian as the OS on all the systems.  

Anyone have a line on some good documentation on how do go about this?  I know 
some additional hardware is needed, and here's what I want.  Two nameservers, 
one webserver, and a mail server right off the bat.  We wish to have access to 
all the machines on a 24H basis for peace of mind <- why we don't use a hosting 
service.

Anyway, I have not done this before... maybe someone could point me in the 
direction of a list of hardware needed.. CSU/DSU, routers, etc...
Thanks,
Jack

HELP! xconfig won't give access to modules

[Cam Ellison]

I have an ethernet card that requires the rtl8139 driver.  The driver
is present in the kernel source (2.2.17), but xconfig (not to mention
config and menuconfig) does not make it available for selection, i.e.
it is greyed out.  To forestall the obvious question, yes, I have a
number of networking options set to "Y", Network device support is set
to "Y", Loadable module support is Enabled, the Kernel module loader is
set to "Y", Networking support and PCI support are on (it is a PCI
card).  IF you need to know more, I have attached the entire setup.

The card is installed, and ifconfig displays it correctly.

Why is this happening (or not happening, depending on your
perspective)?

What do I do about this?  

Thank you, anyone, for whatever help you can provide.  This is driving
me nuts!

Cam



Cam Ellison, Ph.D., R.Psych.

[EMAIL PROTECTED]  or [EMAIL PROTECTED]

>From the lovely Sunshine Coast, where it only SEEMS to rain.



 


Kernel_Config
Description: Binary data

ISDN/ipppd problems

[Wolfram Kruschel]

Hi!
I'm trying to configure my ISDN-Card with Debian 2.2 . I didn't have any
problems until I tried to dial the first time. /var/log/messages tells
me that "ipppd:no pap/chap-secrets defined for this user" although
pap-secrets exists. Everything else looks fine. May be somebody could
help me to fix this. Thanks a lot, Wolf

Re: Search for NAT or http proxy

[John Hasler]

thingels writes:
> It's, in fact, no bigger problem than:
> pppconfig

> I want to set up my "modem" for automatic dialup upon detection of
> non-local IP-requests.  That is, make it act like a dialup router with
> NAT, and there's something with the IP setup I can't get to work. Help,
> anyone?

Not without more information.  Exactly what have you tried and exactly what
happened?  Be verbose.  Post config files and logs.
-- 
John Hasler
[EMAIL PROTECTED] (John Hasler)
Dancing Horse Hill
Elmwood, WI

Glibc 2.1.94-3, fixes all issues with db libraries

[Ben Collins]

If you had problems with the last few 2.1.94 packages (and even if you
didn't), get these. They are currently in incoming (to be installed later
this afternoon, then mirrors will get it some hours later). Make sure you
get the libdb2 that goes along with it.

I don't have time to list everything it fixes, but I can tell you that it
fixes EVERY issue that EVERYONE was having. Installing this should bring
you back to a worry free stable setup again (remember this is unstable, so
if your system breaks, it's probably someone elses fault, so I can't
guarantee everything).

For those bitten by The Great Glibc Update of 2000, welcome to our annual
ritual. Please stay tuned during the next few months where we install a new
gcc. This will be followed by self inflicted pain of stricter C++ syntax,
macro collisions, binary incompatability and general chaos among kernel
builds[1]. Thank you for your patronage, and please come again[2].

Ben

[1] Debian Chaos Events are scheduled on a haphazard basis. We make no
gurantees about show times.

[2] Debian Chaos Events are Copright 2000 Software in the Public Interest.
Any reproduction or rebroadcast of events in whole or in part requires
express written consent by SPI. 

-- 
 ---===-=-==-=---==-=--
/  Ben Collins  --  ...on that fantastic voyage...  --  Debian GNU/Linux   \
`  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  --  [EMAIL PROTECTED]  '
 `---=--===-=-=-=-===-==---=--=---'

Re: tr '\verb|\|000' '\verb|\|\n'?

[Vee-Eye]

Can't offer copious elucidation, but
> Can somebody explain the subject line to me please.  I have read it in
> a linux training document and it is not doing what the document says
> it should do.  What I do not understand and do not know where to find
> documentation on it is the '\verb|\|000' '\verb|\|\n' part.  I know tr
> but what is '\verb|\|000'?  And the use of |\|?  
> 
> The whole command according to the document is
> 
> cat /proc/$$/environ | tr '\verb|\|000' '\verb|\|\n' 
  cat /proc/$$/environ | tr '\verb|\|\000' 'verb|\|\n'
 ^
works.
And why not simply   cat (...) | tr '\000' '\n'
without the \verb part. 
\verb|\| must be, obviously, the unchanged part, but it's superfluous.

I would need also some elucidation about the \verb part, didn't find anything.
 
MH
-- 
(Dr.) Michael Hummel
mailto: [EMAIL PROTECTED]
[EMAIL PROTECTED]
--
fprint = F24D EAC6 E3D7 372C 9122 D510 EB24 01CA 0B56 B518
key: http://www.seitung.net/key


pgpk51WFctNUH.pgp
Description: PGP signature

Print filter/tool recommendations?

[Svante Signell]

Anyone who can recommend good print filters for ordinary text?

lpr gives lousy output: pages are too long and wide, the font is too large.
lpf or pr is not much of use either.

man -k print|grep filter gives:
lpf (1)  - general printer filter
magicfilter (8)  - automatic configurable printer filter
magicfilterconfig (8) - writes interactively a printcap file.
fixpspps (1) - filter to fix PSPrint PostScript so PSUtils work
vprint (1)   - print filter to break long lines

Not much here either. Pointers, suggestions?
At work I use something called lpp, but I don't know if lpp is generally
available. 

Are a2ps or djscript (have a deskjet printer) usable?  Can they be
used in connection with printcap?

Problems with AC97 PCI sound card

[Francois Fayard]

Hi,

I own a AC97 PCI Audio sound card built in my main board.
I've tried to load the ac97_codec module with modconf.
Here is the list of my modules loaded in my kernel

Module  Size  Used by
ppp_deflate38988   1  (autoclean)
bsd_comp3828   0  (autoclean)
ppp20684   2  (autoclean) [ppp_deflate bsd_comp]
slhc4436   1  (autoclean) [ppp]
soundcore   2628   0  (autoclean) (unused)
nls_cp437   3904   2  (autoclean)
lockd  31112   1  (autoclean)
sunrpc 52420   1  (autoclean) [lockd]
serial 19564   1  (autoclean)
ac97_codec  7044   0  (unused)
vfat9008   1 
unix   10212  93  (autoclean)

But when I try to play a sound (as root or as user in the audio group) 
with mpg123 I have a:
Can't open /dev/dsp! 

Thanks for your help
Francois
-- 

Re: Canon BJC-1000 problem

[John Hasler]

Willy Lee writes:
> I installed magicfilter, set up the BJC-600 driver, scratched my head for
> awhile, then installed a2ps and enscript, but now when I send off my
> newly PostScriptized files to be printed, I get ... silence.

I had to edit /etc/magicfilter/bj600-filter and add this as the last line:

default fpipe /usr/bin/a2ps --silent --user-option=lp -o -
-- 
John Hasler
[EMAIL PROTECTED] (John Hasler)
Dancing Horse Hill
Elmwood, WI

Re: PPP problems with the Debian 2.2

[John Anderson]

An easier way then to manually add the user into dialup group is to
adduser option in the pppconfig program.  Start pppconfig and select
"Advanced options," then select "adduser."

Hope this helps!

===
John Kerr Anderson
Powered by Debian GNU/Linux 2.2
===

On Sat, 30 Sep 2000, Martti Hamunen wrote:

> How can I use pppconfig as user (not only as root)?
> (I don't want to use the netconnection as root.)
> Why gnome-ppp dial-up gives the message "the pppd daemon died unexpectedly"?
> I am a newuser with the Debian.
> 
> Martti Hamunen
> 
> 

Re: PPP problems with the Debian 2.2

[John Hasler]

Glyn Millington writes:
> Now there's a thing!  I did all that (and added myself to group
> dialout) and when I tried to run "pon" I got this message

> /usr/sbin/pppd: must be root to run /usr/sbin/pppd, since it is not
> setuid-root

Looks like a bug in the ppp package (and adding yourself to dialout is
irrelevant).

> Can you tell me, have I done a BAD THING?

You have made it possible for anyone at all to run pppd as root.

> And what then is the right way?

Pppd should be:

-rwsr-x--- root dip
-- 
John Hasler
[EMAIL PROTECTED]
Dancing Horse Hill
Elmwood, Wisconsin

Re: Search for NAT or http proxy

[thingels]

 Mark Brown <[EMAIL PROTECTED]>:

> On Sat, Sep 30, 2000 at 11:41:42AM +, jblanche wrote:
> 
> > Dynamic IP address) I would need  NAT module for Linux PPPC  which
> > would work with dynamic IP allocations for ppp (The NAT rule cannot be
> > static..)
> 
> The standard Linux kernel NAT can do this.  If you tell the kernel to
> masquerade all packets with source addresses on your local network and
> leave the destination address unspecified then it will do the right
> thing.
> 
>

It's, in fact, no bigger problem than:
pppconfig


I want to set up my "modem" for automatic dialup upon detection of non-local
IP-requests. 
That is, make it act like a dialup router with NAT, and there's something with
the IP setup I can't get to work. Help, anyone?
I use a normal ISP account which assigns random IP adresses in both ends.

Re: /var/spool/mail on NFS

[Mark Brown]

On Fri, Sep 29, 2000 at 10:12:18PM -0500, oneiros wrote:

> Use nfs-kernel-server, the userspace nfsd does not have the needed locking
> functionality.  Plus, it's tons faster and a lot more reliable, which is
> always a good thing(tm).

Better yet, use Maildir mailboxes.  Locking over NFS isn't reliable.

-- 
Mark Brown  mailto:[EMAIL PROTECTED]   (Trying to avoid grumpiness)
http://www.tardis.ed.ac.uk/~broonie/
EUFShttp://www.eusa.ed.ac.uk/societies/filmsoc/


pgpPTj1gNu2vb.pgp
Description: PGP signature

Re: Search for NAT or http proxy

[Mark Brown]

On Sat, Sep 30, 2000 at 11:41:42AM +, jblanche wrote:

> Dynamic IP address) I would need  NAT module for Linux PPPC  which
> would work with dynamic IP allocations for ppp (The NAT rule cannot be
> static..)

The standard Linux kernel NAT can do this.  If you tell the kernel to
masquerade all packets with source addresses on your local network and
leave the destination address unspecified then it will do the right
thing.

-- 
Mark Brown  mailto:[EMAIL PROTECTED]   (Trying to avoid grumpiness)
http://www.tardis.ed.ac.uk/~broonie/
EUFShttp://www.eusa.ed.ac.uk/societies/filmsoc/


pgpt9DRbBrWSj.pgp
Description: PGP signature

Re: PPP problems with the Debian 2.2

[Ethan Benson]

On Sat, Sep 30, 2000 at 02:02:20PM +0100, Glyn Millington wrote:
> On Sat, Sep 30, 2000 at 06:38:36AM -0500, thus spake John Hasler:
> > Martti Hamunen writes:
> > > How can I use pppconfig as user (not only as root)?
> > 
> > By using the sudo command.  man sudo.
> > 
> > > I don't want to use the netconnection as root.
> > 
> > You don't need to.  Pppconfig is just for configuring ppp, not for running
> > it.  To be able to use pon as a user to bring up ppp run pppconfig, go to
> > 'Advanced', go to 'Add User', and enter your username.  This adds you to
> > the dip group, all of whose members can run ppp.
> 
> 
> Now there's a thing!  I did all that (and added myself to group
> dialout) and when I tried to run "pon" I got this message

dip, not dialout, remove yourself from group dialout and add yourself
to group dip:

gpasswd -d yourusername dialout
adduser yourusername dip

> /usr/sbin/pppd: must be root to run /usr/sbin/pppd, since it is not 
> setuid-root

hmm

> 
> so I changed pppd to setuid-root
> 
> chmod u+s /usr/sbin/pppd
> 
> Can you tell me, have I done a BAD THING?  And what then is the
> right way?

well i think its permissions were screwed up, now EVERYONE can run
pppd, which is not good IMO, do this to fix it:

chown root.dip /usr/sbin/pppd
chmod 4754 /usr/sbin/pppd
ls -l /usr/sbin/pppd 

the last command should look like this:

-rwsr-xr--1 root dip252392 Apr  1  2000 /usr/sbin/pppd

-- 
Ethan Benson
http://www.alaska.net/~erbenson/


pgpnN6rOQ2kmV.pgp
Description: PGP signature

Re: PPP problems with the Debian 2.2

[Glyn Millington]

On Sat, Sep 30, 2000 at 06:38:36AM -0500, thus spake John Hasler:
> Martti Hamunen writes:
> > How can I use pppconfig as user (not only as root)?
> 
> By using the sudo command.  man sudo.
> 
> > I don't want to use the netconnection as root.
> 
> You don't need to.  Pppconfig is just for configuring ppp, not for running
> it.  To be able to use pon as a user to bring up ppp run pppconfig, go to
> 'Advanced', go to 'Add User', and enter your username.  This adds you to
> the dip group, all of whose members can run ppp.


Now there's a thing!  I did all that (and added myself to group
dialout) and when I tried to run "pon" I got this message

/usr/sbin/pppd: must be root to run /usr/sbin/pppd, since it is not setuid-root


so I changed pppd to setuid-root

chmod u+s /usr/sbin/pppd

Can you tell me, have I done a BAD THING?  And what then is the
right way?

TIA

Glyn M









-- 
   **
   * "The soul is greater than the hum of its parts. "  *
   * Douglas Hoftstatder*
   **

Canon BJC-1000 problem

[Willy Lee]

Hello,

I found archived posts to this list discussing this printer, which I
happen to have been cursed with.  Apparently this printer is supposed
to work with the BJC-600 driver, but not for ascii.  So I installed
magicfilter, set up the BJC-600 driver, scratched my head for awhile,
then installed a2ps and enscript, but now when I send off my newly
PostScriptized files to be printed, I get ... silence.  'lpq' gives
me:

geldar$ lpq
waiting for lp to become ready (offline ?)
Rank   Owner  Job  Files Total Size
1stwilly  2(standard input)  19661 bytes
2ndwilly  3(standard input)  15382 bytes

It just stays like that until I 'lprm' them.

Has anyone else had this problem or know of a way to get around it?
It sucks to have to boot into Win95 in order to print 8^( 

=wl

-- 
Albert ``Willy'' Lee, Emacs user, game programmer
"They call me CRAZY - just because I DARE to DREAM of a RACE of 
SUPERHUMAN MONSTERS!"

tr '\verb|\|000' '\verb|\|\n'?

[Johann Spies]

Can somebody explain the subject line to me please.  I have read it in
a linux training document and it is not doing what the document says
it should do.  What I do not understand and do not know where to find
documentation on it is the '\verb|\|000' '\verb|\|\n' part.  I know tr
but what is '\verb|\|000'?  And the use of |\|?  

The whole command according to the document is

cat /proc/$$/environ | tr '\verb|\|000' '\verb|\|\n' 

which is supposed to put each variable it shows on a separate line.

Johann.
-- 
J.H. Spies - Tel/Faks +27-21-876-2337 Sel/Cell +27-82 898 1528
 "Beloved, think it not strange concerning the fiery 
  trial which is to try you, as though some strange 
  thing happened unto you; But rejoice, inasmuch as ye 
  are partakers of Christ's sufferings; that, when his 
  glory shall be revealed, ye may be glad also with 
  exceeding joy."   I Peter 4:12,13 

Re: PPP problems with the Debian 2.2

[John Hasler]

Martti Hamunen writes:
> How can I use pppconfig as user (not only as root)?

By using the sudo command.  man sudo.

> I don't want to use the netconnection as root.

You don't need to.  Pppconfig is just for configuring ppp, not for running
it.  To be able to use pon as a user to bring up ppp run pppconfig, go to
'Advanced', go to 'Add User', and enter your username.  This adds you to
the dip group, all of whose members can run ppp.

-- 
John Hasler
[EMAIL PROTECTED]
Dancing Horse Hill
Elmwood, Wisconsin

Re: lvm

[Christoph Gaitzsch]

> "JM" == Julio Merino <[EMAIL PROTECTED]> writes:

   JM> Hi all,
   JM> I have now a currently working Potato system organized in simple
   JM> partitions as I posted in another message. Well, my question is about
   JM> how can I use LVM.

   JM> Can I convert my current installation structure in a LVM one ? Do I
   JM> have to reinstall all my system ? HOW ?
   JM> Potato install doesn't contemplates LVM.

Hi,

that´s the way I did it:

I installed lvm deb

[EMAIL PROTECTED]:~$ dpkg -l lvm
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name   VersionDescription
+++-==-==-
ii  lvm0.8i-3 The Logical Volume Manager for Linux

Then I patched my Kernel 2.2.17 with

patch-2.2.16-LVM-0.8i ( http://linux.msede.com/lvm/ )
linux-2.2.17-reiserfs-3.5.25-patch ( http://devlinux.com/projects/reiserfs/ )

and recompiled with lvm and reiserfs compiled in ( not as modules ).

The I did a vgscan (after reboot), pvcreate on spare partitions ( I
repartitioned a disk with parted to 500MB - Partitions, so I can reassign
them to different volume groups), vgcreate and lvcreate, and copied my
existing partitions to the new ones, e.g.

# vgscan
# pvcreate /dev/hdb[12345]
# vgcreate root_vg /dev/hdb[12345]
# lvcreate -L 1G -n usr_lv root_vg
# mkreiserfs /dev/root_vg/usr_lv
# mkdir /usr_new
# mount -t reiserfs /dev/root_vg/usr_lv /usr_new
# cd /usr
# find . -print | cpio -dpm ../usr_new

and also for all other partitions

Then I edited /etc/fstab to point to the new partitions, rebooted and ready.

The I pvreated the old partitions, added them to the right volume group, and
pvmoved the /dev/hdb[12345] back to /dev/hdb. Here´s my mount - output:

[EMAIL PROTECTED]:~$ mount
/dev/hda1 on / type ext2 (rw,errors=remount-ro,errors=remount-ro)
proc on /proc type proc (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
/dev/root_vg/var_lv on /var type reiserfs (rw)
/dev/root_vg/usr_lv on /usr type reiserfs (rw)
/dev/root_vg/home_lv on /home type reiserfs (rw)
/dev/root_vg/tmp_lv on /tmp type reiserfs (rw)
/dev/local_vg/local_lv on /usr/local type reiserfs (rw)
/dev/local_vg/apt-mirror_lv on /usr/local/apt-mirror type reiserfs (rw)

The / - partition can also be brought under lvm-control, but the you need to
boot in a initial ramdisk before mounting the physical partiton on your HD.

This system works without any hassles, and I moved and resized very much. I
recommend using reiserfs, it does journaling, and comes up very fast after a
power failure or cold reboot.

If you have any questions, send me a mail.

Here some konfiguration outputs:

[EMAIL PROTECTED]:/home/cg# pvscan
pvscan -- reading all physical volumes (this may take a while...)
pvscan -- ACTIVE   PV "/dev/hda10" of VG "local_vg" [500 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda11" of VG "local_vg" [500 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda12" of VG "local_vg" [500 MB / 476 MB free]
pvscan -- ACTIVE   PV "/dev/hda13" of VG "local_vg" [500 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda14" of VG "local_vg" [500 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda15" of VG "local_vg" [500 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda16" of VG "local_vg" [500 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda5"  of VG "root_vg"  [996 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda6"  of VG "root_vg"  [2.93 GB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda7"  of VG "root_vg"  [996 MB / 0 free]
pvscan -- ACTIVE   PV "/dev/hda9"  of VG "root_vg"  [500 MB / 500 MB free]
pvscan -- inactive PV "/dev/hdb1"  is in no VG  [502 MB]
pvscan -- inactive PV "/dev/hdb2"  is in no VG  [502.03 MB]
pvscan -- inactive PV "/dev/hdb3"  is in no VG  [502.03 MB]
pvscan -- inactive PV "/dev/hdb5"  is in no VG  [502 MB]
pvscan -- inactive PV "/dev/hdb6"  is in no VG  [502 MB]
pvscan -- inactive PV "/dev/hdb7"  is in no VG  [502 MB]
pvscan -- inactive PV "/dev/hdb8"  is in no VG  [502 MB]
pvscan -- inactive PV "/dev/hdb9"  is in no VG  [619.66 MB]
pvscan -- total: 19 [12.83 GB] / in use: 11 [8.79 GB] / in no VG: 8 [4.04 GB]

[EMAIL PROTECTED]:/home/cg# lvscan
lvscan -- ACTIVE   "/dev/local_vg/local_lv" [1.95 GB]
lvscan -- ACTIVE   "/dev/local_vg/apt-mirror_lv" [1 GB]
lvscan -- ACTIVE   "/dev/root_vg/usr_lv" [2.73 GB]
lvscan -- ACTIVE   "/dev/root_vg/var_lv" [152 MB]
lvscan -- ACTIVE   "/dev/root_vg/home_lv" [1 GB]
lvscan -- ACTIVE   "/dev/root_vg/tmp_lv" [1012 MB]
lvscan -- 6 logical volumes with 7.82 GB total in 2 volume groups
lvscan -- 6 active logical volumes

Greetings, Christoph

Re :strange xkb problem

[guillaume bouichet]

Hi!
I had the same pb and I found a symbolic link that was
wrong. I did :
rm /usr/X11R6/lib/X11/xkb/compiled
ln -s ../../../../../var/state/xkb compiled
and it works.

Guillame B.


___
Do You Yahoo!? -- Pour dialoguer en direct avec vos amis, 
Yahoo! Messenger : http://fr.messenger.yahoo.com

Re: stripping linux from a box

[Jeff Green]

DOS boot disk 
then 
fdisk /mbr

Agner-Nichols wrote:
> 
> My childrens' school has a machine with what looks to be a Debian
> installation
> with boot from the hard drive.  They want to strip Linux from the box and
> install the
> other OS.  Did the drop partition through fdisk.  But on reboot, instead of
> getting
> a no operating system found message, LILO kicks in and then dies with a
> kernel panic.
> I'm presuming that LILO wrote to the master boot record.  So how does one go
> about
> fully expunging Linux?
> 
> --
> Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null

stripping linux from a box

[Agner-Nichols]

My childrens' school has a machine with what looks to be a Debian
installation
with boot from the hard drive.  They want to strip Linux from the box and
install the
other OS.  Did the drop partition through fdisk.  But on reboot, instead of
getting
a no operating system found message, LILO kicks in and then dies with a
kernel panic.
I'm presuming that LILO wrote to the master boot record.  So how does one go
about
fully expunging Linux?

Search for NAT or http proxy

[jblanche]

Hi all

I am new in this mailing liste. ..

I am using  LinuxPPC on Imac box and I have 3 other hosts at home. As I
need to use all this hosts with only one PPP connection (and only one
Dynamic IP address) I would need  NAT module for Linux PPPC  which
would work with dynamic IP allocations for ppp (The NAT rule cannot be
static..)

Does any body know if this kind of NAT module existe for Linux
PPC ??? If not I will chose only a proxy http server. Does somebody as
already tests some http  proxy on Linux PPC box ?

Thanks for answer

Jerome.

Re: PPP problems with the Debian 2.2

[Steve Traylen]

 Set up pppconfig as root and then add 
 each user to the groups dip and dialout.
 so

 adduser username dialout
 adduser username dip

 This allow access to the devices.  
   Steve
On Sat, 30 Sep 2000, Martti Hamunen wrote:

> How can I use pppconfig as user (not only as root)?
> (I don't want to use the netconnection as root.)
> Why gnome-ppp dial-up gives the message "the pppd daemon died unexpectedly"?
> I am a newuser with the Debian.
 

--
Steve Traylen
[EMAIL PROTECTED]
http://nrich.maths.org/~smt32/

PPP problems with the Debian 2.2

[Martti Hamunen]

How can I use pppconfig as user (not only as 
root)?
(I don't want to use the netconnection as 
root.)
Why gnome-ppp dial-up gives the message "the pppd 
daemon died unexpectedly"?
I am a newuser with the Debian.
 
Martti Hamunen
 

Re: Anyone want a utility to find the best mirror?

[Joe Emenaker]

> On Thu, 28 Sep 2000, Joe Emenaker wrote:
>
> >   o It only tests ping times, not actual transfer rates of data
>
> Right, because ping times are a good first step.  If the box pings really
> slowly, the odds are the transfer rate will be greatly affected.
> I don't care if the lines you and the mirror are on are T1s, if a critical
> link in the middle is causing slow pings, the odds are the transfer will
> be slow.

True. But the odds aren't always right. I once saw an article in SysAdmin
where the author noted that, since pings are icmp echo requests, it's a
really low-level request and the author put forth the assertion that the TCP
layer of the pinged host isn't even involved. Ultimately, the author argued
that there are cases where the machine can be pretty much hung yet the
network card can be responding to pings.

And, more apropos to this particular discussion, I've seen several cases
where the transfer rates from quick-pinging hosts were pretty bad and vice
versa. One way you can get results like this is to compare, say, two
machines that are separated by a single ISDN connection, and two hosts that
are separated by 50 T1's (in series, of course). The ISDN link is going to
have pretty good ping times because a ping packet is so darned small and
they'll get over the ISDN link in no time. However, even though the T1's are
15 times faster, there are 50 routers that need to figure out what the hell
to do with that packet.

Now, this latency problem with 50 routers would cripple your transfer of
larger files if you didn't have a receive window. But we do have receive
windows, so there's a chance of the latency only creeping in to affect the
transfer at the very beginning. Kinda like the sun. When the sun "turned on"
at the dawn of time, it took 7 minutes for that first ray of sunshine to hit
us... but we've been receiving light from it at the speed of light ever
since.

> > These last two are a real pain... especially the last one. Many of the
sites
> > I tried first in the mirrrors list didn't have the debian tree anymore.
> > Hrmph!
>
> So the list is out of date.  That isn't netselect's fault.

I never said it was. My point was that netselect only goes about 25% of the
way that most of us would like to go. That's not to say that netselect
sucks. It's not even to say that netselect isn't among the tools for the
job. I guess my point is there needs to be something more whether or not
netselect is employed as well.

> Can I suggest you focus on that side of the problem

By cleaning up the mirrors list? Well, I kinda thought of that... where the
script would maintain a history of the sites... and it would just write off
the ones that had been inaccessible for over, say, two weeks. What would be
really neat is to have this running on the central Debian server and then
have it publish "today's list of working mirrors"... instead of it being
such a static list.

Still, that doesn't really address the problem of finding the fastest mirror
for your place on the net. What I'd ultimately like to do is have it
(optionally, of course) automatically update /var/lib/dpkg/methods/ftp/vars
to point to the best server(s) that it finds.

> > So, anyway I broke down and wrote a little perl script
>
> This is a nice idea.  You are doing both http _and_ ftp, right?

Not yet. Do you have any numbers on how many people actually use http for
dselect? Strangely, I've almost always used FTP for transferring files and
HTTP for transferring hypertext. Call me silly

> > and then it actually tries to download a
> > file from the site to see what kind of transfer rate it gets.
>
> I suspect the mirrors admins would prefer this:
>
> Use netselect to narrow down to 5 or sites from a list (or even 2-3?) and
> _then_ do a transfer rate test.  I guarantee you: the results will be the
> same, and more so, you will have skipped transfering data from dozens or
> more other sites.

There's a thought. It would save me from having to make my utility
parallelizing, lest it take a hour to run.

However, keep in mind that the file I'm grabbing from each one is the
mirrors list itself... about 20k or so so it's not like I'm killing the
server. However, the problem with small files is kinda the same problem with
ping packets... where you're not leveraging the benefit of receive windows
to test the steady-state throughput betweent the two hosts.

> > It does *not* check for *currency* of the mirror. In other words, if the
> > files in the mirror are all 6 months old, the script will not pick that
> > up yet. I do plan, however, to make it sense that and take that into
> > account.
>
> Actually, I think the answer to this should be done on a server level: the
> mirror _should_ have a timestamp of last update.  This would be all that
> is needed.  If a site stopped mirroring, the timestamp would be way off.

Is there a file in the tree that's always being regenerated?

> > Anyway, if anybody wants to give it a whirl, I'