Re: teTeX (was Re: dvips top margin)

[Nils Naumann]

Paul Seelig <[EMAIL PROTECTED]> writes:

> I'd prefer it to be included with Debian in any case

I second that.  I'm using tetex at my debian system without problems.

They are only some quirks with dependency conflicts if I install new
packages. But they will went away if the tetex package is debianized.

Nils

PS color separations

[ciccio]

Hi,

Is there a package that allows a device dependent and table-based conversion
of RGB to CMYK? (to prepare RGB-graphics for output on a typesetter)

Thanks,
Ciccio C. Simon
[EMAIL PROTECTED]

Work around: X11 Configuration of Elsa Winner 2000PRO/X-2

[Volker M. Goebbels]

Hi.

I managed to work around the problem of higher resolutions on a Elsa Winner
2000PRO/X in conjunction with a Philips Autoscan-Monitor.
I dropped in the 3.1.2 G Beta release XF86_S3 server and everything works 
well in every desired mode from 640x480/60Hz up to 1600x1200/80Hz. The 
server only complains:
"Couldn't load XKB keymap - falling back to pre-XKB keymap"
But I think I can live with this ill-humoured piece of software :-)

Thanks to all the nice people who helped me.

Volker M. Goebbels
System Administrator, http://www.mc.rwth-aachen.de/~vmg/
Institute of Macromolecular Chemistry, http://www.mc.rwth-aachen.de
Rheno-Westfalian Technical University, Aachen, Germany
Member of the HTML Writers Guild, http://www.hwg.org
---
Nine megs for the secretaries fair,| One disk to rule them all,
Seven megs for the hackers scarce, | One disk to bind them,
Five megs for the grads in smoky lairs,| One disk to hold the files
Three megs for system source;  | And in the darkness grind 'em.
---

Re: samba problems

[Kevin Traas]

On Sat, 14 Sep 1996, Behan Webster wrote:

> My /var/log/nmb* log files are a mixture of the following lines:
> 
> more than one master browser!
> connect error: Connection refused
> 
> The first two lines about seem to be errors, but I haven't been able
> to find any documentation of what they mean.

I don't know much about Samba on Linux, but I do know quite a bit about 
Microsoft networks.  The master browser line above indicates quite a 
problem.  

Have you recently added an NT Server to your network segment?  I've no 
idea how Samba handles elections (who gets to be master/backup/etc), but 
that may be the source of this problem and the high CPU usage.  

Do you notice unusually high broadcasts on your network?  i.e. slow 
response outside of the Linux box?

Sorry I can't help more.

Kevin Traas
ValleyNet Hardware/Software Committee Chair &   Abby:  (604) 859-9741
Head System Administrator   Chwk:  (604) 823-4763
Central/Upper Fraser Valley, BC, Canada Miss:  (604) 859-9741
Toll Free Pager:  (604) 918-2054 
Alternate E-Mail: [EMAIL PROTECTED]

Re: Q. about dosemu package

[Lazaro . Salem]

Yves Arrouye <[EMAIL PROTECTED]> wrote: 
>Subject: Q. about dosemu package
>Author:  debian-user@lists.debian.org at cclink
>Date:20.09.96 08:25
[snip]
>2. I only have Windows 95 on my machine. Is it possible to boot its DOS 
>from Dosemu? Or should I make a hard-disk file with a 6.x DOS? 
I have no experience w/Windows95 but another alternative is to have 
a hard disk image,  which is afile on your linux filesystem pretending to 
be a DOS fs.

>3. Does any of you have succeeded in using Windows 3.1 under dosemu? If 
>so, I'd like to be helped (in private mail, please).

In the accompanying documentation dosemu warns you about the danger of 
trying that. In short, Windows is not usable under dosemu.
 
 lazaro <[EMAIL PROTECTED]>
 

Re: Debian Linux Installation problem

[Lazaro . Salem]

Unless you're inserting the wrong disk, it looks like a kernel bug. 
Send it to Linus.
lazaro <[EMAIL PROTECTED]>

__ Reply Separator _
Subject: Debian Linux Installation problem
Author:  debian-user@lists.debian.org at cclink
Date:19.09.96 22:14


I have run into a strange problem while trying to install Debian Linux 
1.1. I downloaded the boot, root and base image files (3.5"), rawrote 
the image files to floppies and booted off the boot floppy. At the boot 
prompt, I ed w/ the boot floppy still inside the floppy drive. I 
inserted the root floppy disk after being prompted to do so. After 
, my box just hung up. Here is what I can copy from the screen:
 
Ramdisk driver initialized: 16 ramdisks of 4096K size
ide: 430FX (Triton) on PCI bus 0 function 57  [for some reason, VX is not]
ide0: BM-DMA at 0x9000-0x9007 [being correctly identified] 
ide0 timing: (0xa207) sample_CLKs=3, recovery_CLKs=2
master: fastDMA=off PreFetch=on IORDY=on fastPIO=on 
slave:  fastDMA=off PreFetch=off IORDY=off fastPIO=off 
ide1: BM-DMA at 0x9008-0x900f
ide1 timing: (0x8000) sample_CLKs=5, recovery_CLKs=4 
master: fastDMA=off PreFetch=off IORDY=off fastPIO=off 
slave:  fastDMA=off PreFetch=off IORDY=off fastPIO=off
hda: WDC AC21200H, 1222MB w/ 128kB Cache, LBA, CHS=621/64/63, DMA 
ide0 at 0x1f0-0x1f7, 0x3f6 on irq14
Floppy drive(s): fd0 is 1.44M
Started kswapd v1.4.2.2
FDC 0 is an 8272A
NCR53c406a: no available ports found
PPA: unable to initialise controller at 0x378, error 2 
scsi: 0 hosts
scsi: detected total
lance.c: PCI bios is present, checking for devices... 
Partition chec: 
hda: hda1
VFS: Insert root floppy disk to be loaded into ramdisk and press Enter 
RAMDISK: Compressed image found at block 0
stack segment: 
CPU: 0
EIP: 0010: []
EFLAGS: 00010246
eax:  ebx: 005f ecx: 00ef ecx: f000ef6f 
esi: 0009 edi: 00fe8020 ebp: f000ef6f esp: 00258738 
ds: .
Process swapper (pid: 1, process nr: 1, stackpage=00258000) 
Stack: ..
   ..
   ..
Call Trace: [<0016aa05>]  
Code: 0f b6 5d 00 fb 10 0f 87 6a ff ff ff 0f b6 4d 01 d3 6c 24
 
[the machine just hung up at this point]
 
Configuration: AMD K5-90 on Triton VX board, Western Digital AC21200 
1282MB IDE (one 200MB DOS partition), S3 Trio64 based video card, 16MB 
RAM
 
I would really appreciate if someone could kindly help me out of this. Thanks.
 
A. R. ABID
[EMAIL PROTECTED] 
 
 

Re: imagemagick depends on libtiff3

[Philippe Troin]

On Wed, 18 Sep 1996 15:00:08 PDT Randy Gobbel ([EMAIL PROTECTED])
 wrote:

> This brings up something that been annoying me: I've run into this and similar
> problems two or three times now, where a package is placed onto the "stable"
> area before the packages it depends on.  This is a problem for obvious
> reasons--the package isn't really available as a "stable" package until all
> the other packages it requires are also considered stable.  There's no point
> in putting stuff you can't actually install into the directory.  I'd like to
> request that in the future, packages not be placed on "stable" until all the
> pieces are ready to go out the door.  Surely this can't be that difficult to
> manage.

Err, imagemagick has never been placed in the 'stable' tree. It used to be (and 
will be until the next version) in the non-free directory which has no 
distinction between stable and unstable.
Maybe we should address this...

Phil.

Apache virtual domains.

[Fundamental]

Okay, how do people add virtual domains using apache, i tried the following
- i added this line to the S20apache runtime file;

ifconfig eth0:1 servername up

but this didnt work.

thanks

Pachi,

- mIcHaEl


  ///\  The Australian Internet Company
  c-00  ISP par Excellence
  \  >  http://www.electric-rain.net/   (mine)
  |\_-  http://www.aic.net.au/  (not mine)
  \ /
   . 

   "On the Plains of Hesitation bleach the bones of countless millions 
who,
at the dawn of victory, sat down to wait and waiting died."  
-G.W Cecil/Adlai 
Stevenson.

ncurses problems

[Randy Gobbel]

Has anyone else out there had trouble with ncurses?  I've now run across at
least four separate failures, all related to using ncurses instead of
termcap:

1) When building NEURON (big hairy biophysical simulation system), the build
fails because some symbols are the same in NEURON and ncurses.  Despite the
fact that they have different types (in C++), the linker links to the wrong
symbols, causing a segmentation fault.  Switching to termcap makes the problem
disappear.

2) In GENESIS 2.0 (another biophysical sim), the build goes ok, but the
program bombs when you try to execute regular shell commands from the command
line.  Again, switching to termcap fixes things.

3) diald-top, a monitoring program for diald, bombs with a segmentation fault,
and it looks suspiciously like it might be down in ncurses code when this
happens.  I haven't yet tried switching to termcap to see if that helps.

4) If you do a custom build of XFree86 (for the beta version), xterm bombs if
you use ncurses instead of termcap.

I have heard of a couple of other Debian users having similar ncurses
problems, but I've never heard of anyone having trouble with ncurses in other
distributions.

Any info?

-Randy
-- 
http://cogsci.ucsd.edu/~gobbel/

NOTICE: I DO NOT ACCEPT UNSOLICITED COMMERCIAL EMAIL MESSAGES OF ANY KIND.  I
CONSIDER SUCH MESSAGES PERSONAL HARRASSMENT AND A GROSS INVASION OF MY
PRIVACY.  By sending unsolicited commercial advertising/solicitations (or
otherwise on or as part of a mailing list) to me via e-mail you will be
indicating your consent to paying John R. (Randy) Gobbel $1,000.00 U.S.D./hour
for a minimum of 1 hour for my time spent dealing with it. Payment due in 30
days upon receipt of an invoice (e-mail or regular mail) from me or my
authorized representative.

dosemu 0.60.3-1 won't run

[Paul Christenson \[N3EOP\]]

I'm trying to run dosemu, version 0.60.3-1.  However, whenever I go to
run it, I get:

=
moe:~> dosmoe:~> dos
dos: can't load library '/usr/lib/libX11.so.6'
Unknown error
dos: can't load library '/lib/libX11.so.6'
Unknown error
dos: can't find library 'libX11.so.6'
moe:~>
=

However, I do have /usr/X11R6/lib/libX11.so.6 on the system.  Symlinking
does not help.

Help???

-- 
 +---+ .
 | Technical Support Engineer, Cyclades Corporation  |
 | 800/88-CYCLADES (882-9252) or (510)770-9727, x258 |
 | Maker of High Performance Multiport Serial Cards  |
 +---+

Re: Swap partition and fdisk

[Carl Johnson]

In reply to Lars Wirzenius's message:
> A R Abid:
> > I have got 16MB of RAM on my machine. Could someone tell me if it would
> > still be necessary for me to create a Linux swap partition. 
> 
> The terse form of the formula is:
> 
>   swap needed = total memory need - physical memory size
>   
> (Forget everything about "twice physical size". That is an evil
> prank that people play on the uneducated rich who have bouth
> 512 MB of RAM.)

Thank you for finally getting that right!  Actually the 'twice
physical memory' formula is about right for some Unix systems, HP-UX
in particular.  HP-UX pre-allocates swap space for physical memory,
and then pre-allocates swap space for every job when it starts or
malloc memory.  Your formula is definitely the right one for Linux,
though.

> First you need to estimate your total memory need. This depends
> very heavily on what you do and what programs you run at the
> same time. I need about 30-40 MB to run a mailer, Mosaic,
> xpat2, up to a dozen or so xterms and editors, a HTTP server,
> a news server, a mail server, compilers, makes, the X server,
> window manager, a clock, xload, desktop pager, window list,
> and a few other niceties.

Gcc also needs several megabytes for compiling, and I recently
discovered the Java compiler needs 8-10 megabytes to compile a
program.  On the other hand, if you are only using virtual consoles
with simple programs, then 8 MB total *might* be enough.  If the
original person is going to be running X11, then he probably wants
to add at least 8-16 MB of swap.

> For a somewhat more detailed explanation, read the memory
> management chapter in the System Administrators' Guide. The
> current version is 0.3, but 0.4 is imminent (I need to see
> how it looks on paper, but if there aren't any big problems 
> with the that, I will release it in a couple of days).

Thanks,  glad to hear that.

Re: Linux binary? (Tkined)

[Juergen Schoenwaelder]

[EMAIL PROTECTED] (David Engel) said:

David>  Aside to Juergen.  What do you think of Dr. Ousterhout's proposed
David>  improvement to package loading?  I will probably add it, or something
David>  similar, to the Debian version of Tcl 7.5 before packaging Scotty
David>  2.1.2.

Well, you know that I am not 100 % happy with John's solution. Anyway,
scotty will use whatever mechanism will appear in Tcl 7.6 to pickup
packages automatically. (I still hope that John modifies its scheme a
bit.)

I don't think that the patches proposed on the comp.lang.tcl list
create nasty side effects - you need a complicated setup with some
naming conflicts to make this happen. So it should be save to include
one or both of these patches in the Debian version if you are prepared
to change this scheme again when Tcl7.6 comes out.

Juergen

Re: How can I set the TERM environment variable?

[Susan G. Kleinmann]

Hi Richard --

You said:
> After installing Debian base packages I installed elvisnox.  But, I see that
> after booting the TERM environment variable is set to cons80x25. How can I 
> best set it to another value (eg linux) during boot for all users?  Where is 
> it set anyway?

In looking over some old mail, I found that having TERM set to cons80x25 is
a consequence of booting an old kernel.  You can check what kernel you're using
by executing 
   cat /proc/version
Apparently, this problem is particularly likely if one is using LILO,
and hasn't set up LILO's definition of the root directory just right.

Newer kernels won't do this.

HTH,
Susan Kleinmann

Need advise on repartitioning

[Pedro I. Sanchez]

Hello,

I have two hard disks, /dev/hda and /dev/hdb. The latter has only two
primary partitions assigned to swap and /home. I want to create a third
partition in /dev/hdb and asign it to /var, which currently lives in
/dev/hda in the same partition of / (no separate partition for /var). 

I am thinking on the following:

1. "tar cvf var.tar /var/*" to have a backup of the current contents of
   /var.
2. "tar cvf home.tar /home/*" to have a backup of home.
3. "umount /home" and use cfdisk to create the new partition.
4. Modify /etc/fstab to account for the new mapping of file systems.
5. Mount /home and restore the files from home.tar
6. Reboot Linux and once the new /var is mounted restore the files from
   var.tar

I'm not sure about the last point since during the booting process some
info in /var might be needed. Also, if it works, what happpens to the
files in the original /var (in /dev/hda)? I'd appreciate if someone can
enlighten for me the right way to do this.

Thank you,

Pedro Ivan
--

P.S. I overlooked giving /var its own partition when I first installed
debian and it lives under the root partition. However /var/spool/mail is
getting to big (to many users!) and that's why I need the extra space
available in /dev/hdb.

Re: Linux binary? (Tkined)

[David Engel]

Mark Purcell writes:
> > My Debian versions of 2.0.2/1.3.4 are available on ftp.debian.org.
> > I have a huge backlog of things I am trying to work off and hope to
> > get to version 2.1.1 in the next week or two.
> 
> Please do put together the debian 2.1.1 package of scotty/tkined!  I

I'm still planning on packaging it, but I can't make any promises as
to when.  It hopefully won't be too long.

Aside to Juergen.  What do you think of Dr. Ousterhout's proposed
improvement to package loading?  I will probably add it, or something
similar, to the Debian version of Tcl 7.5 before packaging Scotty
2.1.2.

> Thanks for Tkined and Debian.

You're welcome, at least for Debian.  Tkined is Juerhen's work.

David
-- 
David EngelOptical Data Systems, Inc.
[EMAIL PROTECTED]  1101 E. Arapaho Road
(214) 234-6400 Richardson, TX  75081

Re: Linux binary? (Tkined)

[Mark Purcell]

> David Engel ([EMAIL PROTECTED])
> Sun, 11 Aug 1996 13:09:19 -0500 (CDT) > 
> 
> My Debian versions of 2.0.2/1.3.4 are available on ftp.debian.org.
> I have a huge backlog of things I am trying to work off and hope to
> get to version 2.1.1 in the next week or two.


Please do put together the debian 2.1.1 package of scotty/tkined!  I
know that without Debian, compiling and cordinating Tk, Tcl, Scotty and
Tkined would be a major undertaking.  Debian allowed me to select Tkined
and it told me what else was needed and then downloaded the relvant
components.

Tkined certainly makes for a nifty network mangement system and I will
now need to look into getting it running on the HP at work.  Or I
suppose I could bring up a Debian box at work, which would certainly be
a lot simpler!

Thanks for Tkined and Debian.

Mark

Re: 2 questions.....

[Gilbert Ramirez Jr.]

> Second is a question with Ghostscript, I use gs to print postscript files on
> my deskjet.  Now I use a function in my .bashrc to invoke it.  The only
> problem is that when it is done printing I get the GS> prompt and have to
> type "quit" to finish printing and eject the paper.  Is there a way or
> option to use that will finish printing and eject the paper automatically?

Instead of using gs directly, I'd use lpr, especially since lpr is the
standard Unix print command. Here's my entry in /etc/printcap

-(snip)
# /etc/printcap: printer capability database. See printcap(5).
# You can use the filter entries df, tf, cf, gf etc. for
# your own filters. See /etc/filter.ps, /etc/filter.pcl and
# the printcap(5) manual page for further details.
#:of=/etc/filter.pcl:

lp|My LJIIP pretending to be a Postscript printer:\
:lp=/dev/lp1:\
:if=/etc/filter.ps:\
:sh:
-(snip)


Then my /etc/filter.ps program is:


-(snip)
#!/bin/sh
# This is a simple filter for printing postscript files, (see /etc/printcap).

# HP Deskjet 500 printer device
#/usr/bin/gs -q -dSAFER -sDEVICE=djet500 -dNOPAUSE -sOutputFile=- -

# HP LJIIP
/usr/bin/gs -q -dSAFER -sDEVICE=ljet2p -dNOPAUSE -sOutputFile=- -
-(snip)

So, to print a postscript file, I just use lpr:

lpr filename

and lpr sends it to ghostscript, and re-routes gs's output to the printer.

> Thanks for any help
> 
> Andy
> 

--gilbert
__
Gilbert Ramirez Jr. [EMAIL PROTECTED]
University of Texas http://merece.uthscsa.edu/gram
Health Science Center at San AntonioUniversity Health System

Re: 1.1 upgrade hassles

[Paul Christenson]

Fundamental wrote:
> 
> hey everyone, ive just installed debian 1.1, everything went fine
> until i
> rebooted and i get the following errors...
> 
> SIOCSFADDR: bad value
> SIOCSNETMASk: bad value
> SIOCSIFBRDADDRS: Network unreachable
> SIOCSADDRT: Network unreachable

I used to get errors like that when I'd boot my machine without the
network card installed.  Make sure that the kernel is recognizing your
network card.  (Check for eth0 in the dmesg.)
 
> Ive tried the following as broadcast address to no avail...
 
> 224.255.255.255
This looks like the broadcast address for:

> 224.0.0.0 (this one works on our solaris machines).
This is a network address.  (The netmask would be 255.0.0.0)

> 203.17.176.225
Where did this come from???

The broadcast address is, by convention, the highest allowable address
on your network.  It can be found by ORing your network address by the
inverse of the netmask.  (Short answer: where the netmask says 0, use
255.)

-- 
++
| Technical Support Engineer, Cyclades Corporation   |
| 800/88-CYCLADES (882-9252) or (510)770-9727, x258  |
| Maker of High Performance Multiport Serial Cards   |
| Unsolicited mail ads subject to a $25 handling fee |
++

Does IP aliasing work?

[John D. Amidon]

My hardware is a Gateway 2000, 3c509 ethernet card, and Debian kernel is
2.0.6 built to include `ip_aliasing'.  I see `/proc/net/alias*' files so I
take it the kernel was built correctly.

vtcs-cvs# ifconfig

eth0  Link encap:10Mbps Ethernet  HWaddr 00:A0:24:C4:11:97
  inet addr:134.0.64.126  Bcast:134.0.255.255  Mask:255.255.0.0
  EtherTalk Phase 2 addr:65280/60
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:17403 errors:0 dropped:0 overruns:0
  TX packets:11764 errors:0 dropped:0 overruns:0
  Interrupt:10 Base address:0x300 


We want to add an alias for eth0 to be 136.0.0.1.  The error is:

vtcs-cvs# ifconfig eth0:0 136.0.0.1
SIOCSIFADDR: Invalid argument

The NET-2-HOWTO makes it sound sooo easy as does the mini-howto at
http://sunsite.unc.edu/mdw/HOWTO/mini/IP-Alias.

What are we doing wrong?

-- 
---
John Amidon, jhk & Associates, 3500 Parkway Ln, Ste 600, Norcross, GA 30092
E-mail: [EMAIL PROTECTED]|   Tel: +1 770 447 6831|   Fax: +1 770 449 
7268
---

Re: 2 questions.....

[Paul Christenson]

Andy Tarkinson wrote:

> Second is a question with Ghostscript, I use gs to print postscript
> files on my deskjet.  Now I use a function in my .bashrc to invoke it.
> The only problem is that when it is done printing I get the GS> prompt
> and have to type "quit" to finish printing and eject the paper.  Is
> there a way or option to use that will finish printing and eject the
> paper automatically?

Check out apsfilter.  It does lots of neat things; you can dump
Postscript as well as ASCII (and a number of other formats) using lpr,
and it prints properly automatically.

Really easy to set up.

-- 
++
| Technical Support Engineer, Cyclades Corporation   |
| 800/88-CYCLADES (882-9252) or (510)770-9727, x258  |
| Maker of High Performance Multiport Serial Cards   |
| Unsolicited mail ads subject to a $25 handling fee |
++

Re: 2 questions.....

[Syrus Nemat-Nasser]

On Wed, 18 Sep 1996, Andy Tarkinson wrote:

> First is a problem with dircolors.  I had it working fine (I think) with
> "stable" then I upgraded to "unstable" where the color-ls package becomes
> obsolete.  In profile I have a few aliases setting up ls, dir, etc using
> color.  Also I have "eval 'dircolors -b'".  I have the file /etc/DIR_COLORS
> with the colors from Slackware, and also the same as ~/.dircolors.  Now I
> get a few colors like directories and permissions but no colors specific to
> extentions like .tgz, .jpg, .tar etc...  I've tried everything and can't get
> it to work properly.

I think you have to specify the path to your config file.  Here's some info I
keep around regarding this question:

With the integration of color-ls directly into the fileutils
package, a few things have changed.  dircolors no longer sets
up aliases or shell scripts to colorize ls, dir, and vdir.

Here is an excerpt from a .bashrc which sets up aliases after
running dircolors:

# set up color-ls
eval `dircolors /home/syrus/.dir_colors`
alias d='ls -F --color=auto'
alias v='ls -l --color=auto'
alias vdir='ls -l --color=auto'
alias dir='ls -F --color=auto'  
   

Note that color=tty has been changed to color=auto.  See the
documentation for other change information.

Luck, Syrus.


--
Syrus Nemat-Nasser <[EMAIL PROTECTED]>UCSD Physics Dept.

Re: strange problems with $MANPATH and xdm

[Martin Alonso Soto Jacome]

[EMAIL PROTECTED] said:
> So my Question is now: How can I make xdm pass the environment to the 
> window manager and all progs started? is there an equivalent of /etc/
> profile for the window manager? 

Sure.  Your .xsession file is run whenever you start (actually during all the 
execution of) your session.  You must be careful when modifying it, because it 
may prevent you from using Xdm at all (!!).  Alternatively, you can create a 
file /etc/environment and put variable definitions on it.  As far as I know, 
it is not documented, but the /etc/X11/xinit/xinitrc script uses it, so it 
should work.

Regards,

M. S.

Martin A. Soto J.   Profesor
Departamento de Ingenieria de Sistemas y Computacion
Universidad de los Andes  [EMAIL PROTECTED]

Re: where is kernel-package ?

[Susan G. Kleinmann]

Hi Chris --

You said:
> 
> I can't find kernel-package
> anywhere on ftp.debian.org... I found kernel-headers, and kernel-source,
> but I already have a tar.gz of the source. 
It's in rex/binary/misc.  It't not clear why.


> Do I even need the kernel-package? What benifits will it give me?
kernel-package allows you to make a kernel-image_whatever.deb file.

Fetch a new kernel source.  Make sure it's clean.  If not, run
make clean.

Then you can (I do) run make config, or make menuconfig, or make
xonfig, whatever you like.

Then if you run 
make-kpkg -revision local.1 kernel_image

in your top level source directory, you'll have a new .deb file
in the directory just above that, which will allow you to install
the kernel (and modules) in such a way that they can be managed
by the Debian package management system.

Personally, I use the kernel-package this way all the time for 
making custom boot floppies, since the bootdisk.sh script requires
a kernel-image_whatever.deb file for its input.

HTH,
Susan Kleinmann

Re: Swap partition and fdisk

[Susan G. Kleinmann]

Hi --
You said:
> Currently, I have a DOS partition that takes up 25% of the
> total disk space and the rest is empty. My concern was if I would lose
> any data on the DOS partition if I create two Linux partitions using
> Linux fdisk on the empty disk space. 

I suppose the installation notes are conservative because
cfdisk is not too forgiving if you incorrectly set your disk partition
table and then write the new table to disk.  But note that you have
to do two things wrong in order to mess yourself up with cfdisk.
First you have to set up a bad partition table, then you have to
write that to disk.

It sounds like you have one entry in your disk partition table now.
It probably looks something like this:
/dev/hda1   Boot  PrimaryDOS 16-bit >=32Mb   200.52

Just move your cursor down one line (so that it is NOT highlighting the
line where DOS is), then select the option [New] by pressing your
right arrow key until [New] is highlighted.  Hit CR.  Then specify how much
space you want for your new partition.  When you are asked whether you
want to put that partition at the beginning (meaning, at the beginning
of the existing unpartitioned space), just say yes.  The revised
partition table will then be presented to you.  It should show DOS
in the same place it was.  Do the same thing for 1 more partition.
Then change the type of that partition to a Linux swap by using the right
arrow key again to highlight the word [Type], then answering the prompt
by giving the number 82, which is the value for a linux swap partition.
Again, you'll see the revised partition table.  Still, nothing has
happened to your disk.  Only if and when you use the right arrow key
to highlight the [Write] button, and then hit the [ENTER] key, will
you have written your new partition table to disk.

One thing you might do before starting (and this is a good idea anyway)
is to keep a record of your existing disk partition.
You can do this with this command:

cfdisk /dev/hda -P t > part_hda.tbl

or, if you want to keep a record of the partition table using the
sector format, use:

cfdisk /dev/hda -P s > part_hda.sec

(I'm assuming the disk you're working on is the master on your primary
IDE controller.  If not, adjust hda to the appropriate value.)

HTH,
Susan Kleinmann

Re: where's man ??

[Susan G. Kleinmann]

Hi Boris --

You said:
> On Wed, 18 Sep 1996, Dale Scheetz wrote:
> 
> dwarf>If you only installed the base system then the only editor 
>   availabe is AE
> dwarf>(Andrews Editor). This was done because AE is small enough to fit 
on the 
> dwarf>base disks.
> I am sorry, but isn't vi designed for that purpose?
> I mean why should ppl learn other editors commands when there is a good old
> vi :-) ? , at least that what i felt when i installed debian on my system.

The operative word here is *learn*.  'ae' essentially takes no learning.
A person who never saw it before could use it.  'ae' is useful enough
to get your system going well enough to install whatever editor you like.

BTW, you can always edit the bootdisk.sh script that comes in the
boot-floppies package written by Bruce, and build custom boot disks 
for yourself that do contain vi.

Regards,
Susan Kleinmann

Re: List Probagation Time ???

[Jim Pick]

> Is it me or does it take sometime for messags to probagate to the
> subscribers.  I have submitted Two messages and received neither.
> What am I doing wrong? Is this characteristic of this list? Can I do 
> something to help?
> 
> Thanks...Jeff
> 

For the few posts that I made, it seemed that it took 3-4 hours for the
message to come back to me via the mailing list.  It took several days
for the message to come back via news in the linux.debian.user newsgroup.

By the time this hits the list, I'll be 400 km away in Vancouver.

Cheers,

 - Jim

Re: cfs Debian package available!

[Sven Rudolph]

In article <[EMAIL PROTECTED]> [EMAIL PROTECTED] (Bernd Eckenfels) writes:

> > > 1. I need a non US and a non Canada Web site, otherwise it would become
> > > illegal to export Debian without the permission of either government and
> > > AT&T.
> 
> I vote for a non-us (or euro or whatever) Section for things like: xntp,
> SSLeay, ssh, cfs... There is already SSLeay and ssh on the Uni-Mainz
> (GERMANY), perhaps we can build a euro-master for distributiing that part of
> debian?

This should be done soon, IMHO.

We could use the uni-mainz server. Dominik, what do you think about
this ?

Otherwise I could place it on inf.tu-dresden, of course ...

Sven
-- 
Sven Rudolph <[EMAIL PROTECTED]> ; WWW : http://www.sax.de/~sr1/

Re: How can I set the TERM environment variable?

[Susan G. Kleinmann]

Hi Richard --
You said:
> After installing Debian base packages I installed elvisnox.  But, 
> I see that after booting the TERM environment variable is set 
> to cons80x25. How can I best set it to another value (eg linux) 
> during boot for all users?  Where is it set anyway?

It looks to me as if the responsible script is part of 
/etc/rc.boot/0kbd.

In the version I have, there's a line which source's /etc/kbd/config.
If that file exists, then the value for TERM is read from it.
My /etc/kbd/config says "TERM=".  That is, there is a newline 
just afte the '='.  Alternatively, you can edit the stanza in 
/etc/rc.boot/kbd which is invoked if /etc/kbd/config does not 
exist, and set
TERM=linux
in there.

Good luck,
Susan Kleinmann

Re: Shouldn't go app-defaults in /etc/X11?

[Christian Schwarz]

Hi!

Ok, you (Michael) made me change my standpoint. It's propably better to
consider app-defaults as part of the code and then it's ok if it stays
under /usr.

BTW, I just did a `grep app-defaults *.conffiles' in /var/lib/dpkg/info
and discovered, that there are a few packages specifying the app-defaults
file as conffiles (also in my xblast package--I will fix it). I'm not sure
if we should report this as bug.

You mentioned the file debian.readme and it was pointed out, that the
passage about app-defaults is too short. How about an improvement (too
avoid threads like this ;-) :

Please note that this distribution expects you to leave app-defaults
files unchanged. The app-defaults files are considered as part of the
code. That's why the directory is located in /usr/X11R6/lib/X11 instead
of /etc/X11. If you want to customise your X applications globally, put
your customisations in /etc/X11/Xresources.


Cheers,

Chris

--  _,, Christian Schwarz
   / o \__   [EMAIL PROTECTED], [EMAIL PROTECTED],
   !   ___;   [EMAIL PROTECTED], [EMAIL PROTECTED]
   \  /
  \\\__/  !PGP-fp: 8F 61 EB 6D CF 23 CA D7  34 05 14 5C C8 DC 22 BA
   \  / http://www.informatik.tu-muenchen.de/~schwarz/
-.-.,---,-,-..---,-,-.,.-.-
  "DIE ENTE BLEIBT DRAUSSEN!"

/dev/ttyS1 root only.....

[Ed Down]

I'm trying to configure my modem and, as recommended in the Serial-HOWTO,
I've been using kermit to test it out. Problem is that /dev/ttyS1, which
is the serial port my modem is on, seems to only allow root access. Anyone
recommend a fix? Also any tips on mgetty setup would be appreciated.

Ed

Re: How can I set the TERM environment variable?

[Jeff Myers]

try... /etc/profile 

At 08:50 PM 9/18/96 -0500, you wrote:
>After installing Debian base packages I installed elvisnox.  But, I see that
>after booting the TERM environment variable is set to cons80x25. How can I 
>best set it to another value (eg linux) during boot for all users?  Where is 
>it set anyway?
>
>

Re: 2 questions.....

[Brian C. White]

> Second is a question with Ghostscript, I use gs to print postscript files on
> my deskjet.  Now I use a function in my .bashrc to invoke it.  The only
> problem is that when it is done printing I get the GS> prompt and have to
> type "quit" to finish printing and eject the paper.  Is there a way or
> option to use that will finish printing and eject the paper automatically?

$ gs --help
[...]
 -c quit (as the last switch)
   exit after last file
[...]
 
  Brian
 ( [EMAIL PROTECTED] )
 
---
In theory, theory and practice are the same.  In practice, they're not.

Re: How can I set the TERM environment variable?

[Boris Beletsky]

-BEGIN PGP SIGNED MESSAGE-

On Wed, 18 Sep 1996, Richard Heestand wrote:

rlheest>After installing Debian base packages I installed elvisnox.  But, I see 
that
rlheest>after booting the TERM environment variable is set to cons80x25. How 
can I 
rlheest>best set it to another value (eg linux) during boot for all users?  
Where is 
rlheest>it set anyway?
rlheest>
check  `/etc/rc.boot/0kbd`
Good Luck
___
Boris Beletsky <[EMAIL PROTECTED]>
For pgp public key, e-mail me 
with subject "get pgp-key."
___
In Linux veritas


-BEGIN PGP SIGNATURE-
Version: 2.6.3ia+
Charset: latin1
Comment: Boris Beletsky <[EMAIL PROTECTED]>

iQCVAwUBMkGowgz8DjY6pgpxAQF5dwP9EDkG4dTTKqi1/chcsYbssNbYV9/brWv5
MOd8SWnfqBGQqfYpEXy2uGDUqLjTeChlA3d2YA0X6ubMFANORygWdaWcWA4TIKi1
NjPARtaaooDpXNCMZw0MUTKOe6nuQIOwbrQ88UTzWbWHcyfQLv9G8Af23s7OxwQ0
OT7Yqu/XvgE=
=Dgrw
-END PGP SIGNATURE-

Getting XDM to blank the screen

[Greg Smith]

For some reason, XDM never blanks the screen.  I would like to use
xautolock to either run xlock as a screen saver or maybe even apm -S 
to suspend the machine.

How would I do this?  can I put it in one of XDM's configuration scripts?


Thanks

Greg

PS.  please respond to [EMAIL PROTECTED]   Thanks again!

Re: DNS, Sendmail, or Smail

[Mike Wood]

On Wed, 18 Sep 1996 [EMAIL PROTECTED] wrote:

> I know I'm very new to Linux, and there is quite a bit I haven't 
> caught on to yet.  But, I have a problem that just isn't getting any 
> better.  I have Debian 1.1 loaded, running kernel 2.0.6.  The machine 
> I'm working in is not directly connected to the Internet, but does 
> have a tcp/ip connection to a mail server that is connected.  I loaded 

Is the machine your connected to runnning a DNS server or is it using
annother server out on the net?

> 3. Is there a way to set-up named to work as a local DNS only ?

You could run a small caching only DNS server on the machine that
IS connected to the net.  It's a very easy thing to do (as long as you
have root).  Then you could tell your Debian box that it's DNS server is
the mail server that's connected to the net.  That or just put the
hostnames you need translated into IP addresses into the /etc/hosts file.
If you need help on setting up a caching only DNS server you can check out
the howto index at http://www.ssc.com/linux  and if all else fails send me
some email and I'll give you a hand.

l8trz,
mike...

Re: 4 man page languages now

[Christian Hudon]

On Wed, 18 Sep 1996, Bruce Perens wrote:

> Susan:
> > the manpages (in English, German (man-NNN-de), and Spanish (man-NNN-es)).
> 
> I think some Italian language man pages have also recently been uploaded.
> I communicated with the maintainer last week. They are probably in
> "unstable" or the Incoming directory.

Does anyone know if there are French manpages coming up somewhere in the
pipeline?

   Christian

RE: How can I set the TERM environment variable?

[Rik Ling]

--
From:   Richard Heestand[SMTP:[EMAIL PROTECTED]
Sent:   Wednesday, September 18, 1996 4:50 PM
To: debian-user@lists.debian.org
Subject:How can I set the TERM environment variable?

After installing Debian base packages I installed elvisnox.  But, I see that
after booting the TERM environment variable is set to cons80x25. How can I 
best set it to another value (eg linux) during boot for all users?  Where is 
it set anyway?



Isn't it specified in /etc/inittab?  Should be one of the arguments to the 
gettys that run on the console ttys.

Just my $0.02.  Hope it helps.

Rik Ling
Network Administrator
Peterborough Internet Pipeline
http://www.pipcom.com/~rling (under construction, as always)

ftp.debian.org

[Brad Roberts]

I maintain a mirror of ftp.debian.org for internal use (we've got several
linux machines), however, the last several times I've started to run
mirror its reported:

compare directories (src 8356, dest 32957)

I'm not too keen on letting 3/4 of the mirror get toasted, so I've
killed the mirror before it could start its delete phase.  After
investigating a little, it seems that most of the missing files where
in the WebPages hierarchy that is empty now.  Is this the new policy or an
accident or...?  I guess I don't mind the loss of the bug tracking system,
etc.. but the list archives were nice to have.

Thanks,
Brad

NEW: Cryptographic Filesystem (CFS) has a Home/WWW Page

[Patrick J. Edwards]

-BEGIN PGP SIGNED MESSAGE-

You can view the web page I have created (please forgive all the missing
links... but those are the packages I plan of making available for Debian
in case you are wondering):
http://www.cs.usask.ca/undergrads/pje120/linux/cfs/


OR download from its new Debian/Linux FTP site (Many thanks to Pawel T.
Jochym):
ftp://hpnz8.ifj.edu.pl/pub/linux/cfs/

Package: cfs
Version: 1.1.2-2
Architecture: i386
Maintainer: Patrick J. Edwards <[EMAIL PROTECTED]>
Description:
 Cryptographic Filesystem uses NFS to allow for a complete file system that
 is encrypted on the fly. It uses DES, and Triple-DES.
Files:
 b1993b5ead27b8ac52de79354c6ca09e cfs-1.1.2-2_i386.deb 65238
 14484681596401f6c307a749c596b7e4 cfs-1.1.2-2.diff.gz 3382
 0fc0f826d7f719695d3e971e58ad76b6 cfs-1.1.2-2.tar.gz 78976
 841f8fe084f3d3ec14b7d05e9440d8bf cfs-1.1.2.orig.tar.gz 18
Changes:
 * chnaged the debian.rules to a more, complex and useful version
 * added an extra rm



- ---
"The Governent is not above the law. They cannot withhold information."
(Scully, X-Files)
- ---
Patrick J. Edwards <[EMAIL PROTECTED]>
http://www.cs.usask.ca/undergrads/pje120/
finger [EMAIL PROTECTED] for my PGP Key
Key fingerprint =  9F 45 7D 6E C0 A4 B4 0D  48 C7 14 CA 23 B0 B4 F8

-BEGIN PGP SIGNATURE-
Version: 2.6.2i

iQEVAwUBMkF3TvVAqxHbZskNAQEVDQf+J7LnGEUkhwaIjIoonkinuYJ964Ur5qSb
T5Kcv+4PDXlsIZTD2DTFPnw+cDbEQcG4ADfXFUDCmvHoEKSYwVURvpboYUKXdle7
/f8LsRBxFhKs4iN16lK9vxnsj/PT1hw1k9KN7uZxcXTJIgm/o/saqvFGHtjA5PKF
SQDINaG0ktL9ebx6glnC3eo3A/aF22Azm2RC6essXbkFZWDXMYpNJyBL0xhENgBY
Do7Vk/EE6uSP6O5L7OS1pmTdsSoceOpHQ09zhms0KTmI3bpj//vKNCQ2n/d5vxza
LYsX1xv6ZZ5FdWbPQW1FK6+37YPnLrozyEa/6lmM5EuWd1+PQz29HQ==
=hs78
-END PGP SIGNATURE-

Re: samba problems

[Behan Webster]

Martin Budsj| wrote:
>
> >My /var/log/nmb* log files are a mixture of the following lines:
> >
> >more than one master browser!
> >connect error: Connection refused
> >Domain=[workgroup] OS=[Unix] Server=[Samba 1.9.16alpha10]
> >Domain=[workgroup] OS=[Unix] Server=[Samba 1.9.15p4]
> >
> 
> Looks like you are running both 1.9.15p4 AND 1.9.16alpha10. How many
> nmbd processes is there? Only one i hope...

One of four machines in "workgroup" was sunning Samba 1.9.15p4 while the
rest were running Samba 1.9.16alpha10.  I recently upgraded that last
machine to run Samba 1.9.16alpha10 as well, but it didn't help.  Smbd
processes are still chewing up tons of CPU on each samba host.

There is only one nmbd process running on each host.

Behan

-- 
Behan Webster
[EMAIL PROTECTED]
(613) 224-7547  or page me at (613) 2~4-7547

Re: libforms

[Bill Woodall]

>
>Anybody have any idea where I can find the libforms library.
>
Sure, if it the same as the one for XForms.  From the README;

"Documentation on XForms is available from bloch.phys.uwm.edu /pub/xforms
via anonymous ftp."

But the copyright might be restrictive;

All files distributed in this package are
Copyright (c) 1995 by T.C. Zhao and Mark Overmars
All rights reserved. 

Permission to use, copy, and distribute this software in its entirety
for non-commercial purposes and without fee, is hereby granted, provided
that the above copyright notice and this permission notice appear
in all copies and their documentation. 

This software is provided "as is" without expressed or implied
warranty of any kind.

Re: where's man ??

[Paul Christenson]

Boris Beletsky wrote:

> On Wed, 18 Sep 1996, Dale Scheetz wrote:

> dwarf>If you only installed the base system then the only editor
> dwarf>availabe is AE (Andrews Editor). This was done because AE is
> dwarf>small enough to fit on the base disks.

> I am sorry, but isn't vi designed for that purpose?

Maybe it was, but vi, as all good, small programs, suffers from
"creeping featurism", and the result:

-rwxr-xr-x   1 root root23624 Sep  4 18:27 /bin/ae
-rwxr-xr-x   1 root root67969 Mar  9  1996 /bin/ed
-rwxr-xr-x   1 root root   179360 Jul 25 12:23 /usr/bin/pico
-rwxr-xr-x   2 root root   292428 Aug 10 20:18 /usr/bin/nvi
-rwxr-xr-x   2 root root  1906900 Aug 28 17:24 /usr/bin/emacs

For comparison, Slackware has:
-rwxr-xr-x   1 root bin 97644 May 22 15:45 /usr/bin/elvis
-rwxr-xr-x   1 root bin200944 Aug  7  1995 /usr/bin/vim

Enough said?

-- 
+---+
| Technical Support Engineer, Cyclades Corporation  |
| 800/88-CYCLADES (882-9252) or (510)770-9727, x258 |
| Maker of High Performance Multiport Serial Cards  |
+---+

Re: where is compress ?

[Yves Arrouye]

 I'm trying to compile a X-Window program which provides a new set of fonts.
   To generate its fonts, it uses the following command (approximately ;) ):

   bdftopcf machin.bdf | compress > machin.pcf.Z

   But I don't have compress on my Linux box.
   I tried moving to gzip without success because mkfontdir doesn't handle gzip
   format :-( Therefore I need compress!
   Is there any Debian package I can install that includes it ?

Because BSD compress might conflict with Unisys licensing of LZW, it is
not included in the Debian distribution. There is a compress-package
available, though, that, after you have got sources for compress (for
example the compress.tar.12 file that can be found on some GNU archives),
will help you build a Debian compress package, as in:

% cd /usr/src/compress/compress-4.0
% make-cpkg
[...]
% dpkg -i ../bsd-compress_4.0-1_i386.deb

Yves.

Q. about dosemu package

[Yves Arrouye]

Hello,

I'd like to know the following about dosemu:

1. Is something doing a dosemu 0.60.4 package? There is a 0.60.4 version
on the Linux archives, dated 1995, but we only have the 0.60.3 on Debian.
Is the package not maintained anymore? And is dosemu development stopped?

2. I only have Windows 95 on my machine. Is it possible to boot its DOS
from Dosemu? Or should I make a hard-disk file with a 6.x DOS? In this
case, is it okay to read/write my FAT filesystem through emufs?

3. Does any of you have succeeded in using Windows 3.1 under dosemu? If
so, I'd like to be helped (in private mail, please).

Thanks in advance,
Yves.

-- 
Yves Arrouye  Email: [EMAIL PROTECTED]
7, avenue Leon BolleeWeb: http://www.fdn.fr/~yarrouye/
75013 Paris  Work: +33 45 95 64 59
France   Home: +33 53 61 09 55

Re: What where huh?

[Brian C. White]

> Susan> I'd agree that a search engine front end to the FAQ would be helpful.
> 
> Also, the debian-user archives would be a whole lot more useful if we had
> both more up-to-date archives and a search engine.

There is a search engine for the Debian web site:

http://insite.verisim.com/search/debian/simple

I hope to include the list archives in there eventually.  The reason they
are left out at the moment is because each message has a standard header
and that header is all that gets picked up by the summary.  Eventually I'll
strip that from the top so the summary is actually the first few lines
of the mail message.
 
  Brian
 ( [EMAIL PROTECTED] )
 
---
In theory, theory and practice are the same.  In practice, they're not.

SCSI autoprobing causing panic

[David Stirrup]

Hello,

Yes I am a newbie and I have what is probably a silly problem with the generic 
"all drivers loaded" kernel boot disk for Debian 1.1

I have an Adaptec 2742AT adapter (dual EISA SCSI, aic7770) and the boot disk 
finds this without trouble and registers the attached devices OK.

Auto probing then continues until the NCR 53c406a driver reports back that no 
ports were found.  The AIC 7770 driver reports at the same time a "spurious 
interrupt" and kernel panics.

I have installed the Slakware 2.0.0 and RedHat Rembrant without this problem 
but because of the "everything including kitchen sink + basement" approach 
taken by the generic Debian boot disk I am unable to start the Debian 1.1 
install process.

To try and get around this, from a slakware install I created a new (Debian 
src) kernel and then "rdev -r"  the new kernel and copy to the generic boot 
disk.  Did not work so I then tried to create a new boot disk to work with the 
rest of the base installation set.  

No luck.  That kernel boots OK, mounts all my existing volumes and gives me a 
logon prompt.  I cannot seem to be able to generate a boot disk that then 
automaticaly loads the Root disk image into the RAM disk.  Probably a LILO 
thing or "rdev" thing.  Read the "bootdisk, FAQ's but can't seem to get it 
together. 

My conclusion from playing around with the Debian Kernel source is that it has 
to live in a Debian system to compile and get everything working properly.  
Cross implementing from Slakware has so far yielded poor results for me (a non 
programmer techo type).

All of this because I can't get the generic disk to work.  

So what do I do now ?

Question;
Is there a way to turn off autoprobing and tell the generic boot disk what SCSI 
driver to load ? 

Suggestion to Bruce Perens;
Would it not be such a bad idea to include a few "reduced / less complex" 
versions of the kernel boot disks  ie  Adaptec, NCR, Buslogic + EATA and Misc,  
SCSI + Net,  SCSI + oddball CDROMS etc etc like is done with other 
distributions ?  Once a system is base installed then we can flesh it out with 
packages.

Thankyou for your consideration and time,

Dave (newbie) Stirrup
Sydney,  Australia.

Re: Shouldn't go app-defaults in /etc/X11?

[Michael Alan Dorman]

Christian Schwarz <[EMAIL PROTECTED]> writes:
> My suggestion of tagging the files as conffiles was thought as a solution
> to your problem, if the files change.

That's the problem, though---it's not a solution at all.  Consider if
the application introduces a new resource, without which the
application will fail?

Isn't it less work for the sysadmin if he/she can know that the
app-defaults file is _guaranteed_ correct (and managed in the postinst
to be customized for given obvious system defaults (like the nntp host
for knews)), and that if the app's not working he or she may need to
expiriment with the _few_ customised options in /etc/X11/Xresources?

What if someone maintains a package they don't use (I do for several
these days), and thus they are never aware that the program won't work
correctly (or lose data, or whatever), if you don't have a certain app
default?  I can't mention it in the postinst if I don't know about it.

> So I think the point is the following: Do we consider the
> app-defaults files as part of the program (code), or as
> configuration files. In the former case, they should stay in /usr,
> in the latter they should be moved to /etc/X11 and be tagged as
> conffiles.

They're part of the code.  They're _defaults_.  Anything else is
customization, and should go on elsewhere.

Mike.

Re: How to create a "Packages" file?

[Christian Schwarz]

This is a simple script that creates the Packages files: Suppose you want
to create the file for distrib buzz and binary-i386:

touch /tmp/noverride
cd buzz
dpkg-scanpackages binary-i386 /tmp/noverride buzz/ > Packages.n
mv Packages.n binary-i386/Packages
gzip -f -q binary-i386/Packages
cd ..

I have a script that goes over the whole Debian hierarchie on my private
FTP server whenever I make a change.

Note that the `touch /tmp/noverride' creates an empty file (if it doesn't
already exist) which seams to be necessary, but I don't know why.

I also discovered another problem: When there a several versions of a
package in the same directory, dpkg processes them in the order they
appear in the directory (not sorted, as you see them with the default ls
options). So if the newer versions comes first, the older version will be
included in the Packages file.

And it looks like dselect has the same problem with installed from a
CDROM or local dir. These `access methods' seem to process not the list of
selected packages (as ftp access method does) but process each file they
find in the directory. When there are several versions of a file that's
selected, dselect upgrades/downgrades the package as they appear! That's
why I always use FTP, even if the FTP server is the local host!

Does someone have any hints?


Cheers

Chris

--  _,, Christian Schwarz
   / o \__   [EMAIL PROTECTED], [EMAIL PROTECTED],
   !   ___;   [EMAIL PROTECTED], [EMAIL PROTECTED]
   \  /
  \\\__/  !PGP-fp: 8F 61 EB 6D CF 23 CA D7  34 05 14 5C C8 DC 22 BA
   \  / http://www.informatik.tu-muenchen.de/~schwarz/
-.-.,---,-,-..---,-,-.,.-.-
  "DIE ENTE BLEIBT DRAUSSEN!"

Re: Shouldn't go app-defaults in /etc/X11?

[Lars Wirzenius]

Christian Schwarz:
> My suggestion of tagging the files as conffiles was thought as a solution
> to your problem, if the files change.

The problem scenario:

Version 1.0 installs an app-defaults file. It defines
resource ``*Background: black''.

Admin installs version 1.0. She changes the app-defaults
file to ``*Background: white''.

Version 1.1 is released. The app-defaults file is the
same as in 1.0. Admin installs 1.1. What should be done
with the app-defaults file?

If the app-defaults file is marked as a config file, then dpkg
asks whether to keep the current file, or to replace it with the
one from 1.1. Admin keeps current file, because it does what she
wants. So far, so good.

Second scenario:

Version 1.2 is released. The app-defaults file adds a
new resource, ``*enableDWIM: true''. Application will
work if this resource is missing (and defaults to false),
but not nearly as well.

System still has the admin-modified file. Admin installs
1.2. What should be done with the app-defaults file?

Again, dpkg will allow the same choice as before. Admin chooses
to keep the current file. To make the application work well, she
has to compare the current file and the one in 1.2 (dpkg saves
it in the same directory, so this is easy), and make some more
edits. It's a small problem, but not too bad, since the changes
between 1.1 and 1.2 were small. It'd be nicer if the admin wouldn't
have to do any manual edits, but we can live with this.

Third scenario:

Version 1.3 is released. It adds a new resource,
``*importantThing: foo''. If this resource is not set,
the application will crash. This is documented only
in the middle of a five thousand line manual page. The
package maintainer did not know this, because everything
happened to work for him. The app-defaults file has
been rearranged to make it easier to understand.

Admin installs 2.0. What should be done with the
app-defaults file?

This time the differences between the currently installed and
the new app-defaults file are huge. The output of `diff' is not
easy to understand, and the admin makes a mistake and assumes
that nothing important has changed, just the arrangement of the
file. She does not add the new resource. Application crashes.
Not so good.

The Debian solution is to add a new file, /etc/X11/Xresources,
which is used in addition to the app-defaults files and each
user's own ~/.Xresources file. This way, the app-defaults file
does not have to be a config file. The package maintainer can
make any changes as is necessary. The admin makes any local
configuration in /etc/X11/Xresources. Everyone is happy.

Now that I've been made aware of this, I think it's a good
solution.

-- 
Please read  before mailing me.
Please don't Cc: me when replying to my message on a mailing list.




pgpDDlJBVttb3.pgp
Description: PGP signature

Re: Swap partition and fdisk

[Lars Wirzenius]

Boris Beletsky:
> YES! u always need a swap - no metter how much ram u have.
> I would say, create a 32swap part. - that would be the best.

If you never use more than 16 MB of memory, and you have 512 MB
of physical memory, you most definitely do not need swap.

swap needed = total memory need - physical memory size

It's really very simple...

-- 
Please read  before mailing me.
Please don't Cc: me when replying to my message on a mailing list.




pgpuWALwuimhH.pgp
Description: PGP signature

teTeX (was Re: dvips top margin)

[Paul Seelig]

On Wed, 18 Sep 1996, Billy Chow wrote:

> teTeX is a high quality LaTeX distribution which takes care of these
> settings with a nice menu-driven utility.  It would be easier to
> debianise teTeX than to fix the current Debian LaTeX system (other
> 
I installed Debian for the first time two weeks ago and didn't even bother
to install the Debian LaTeX system because i already knew teTeX from
another platform and installed this instantly in the /usr/local/*
hierarchie. To my mind teTeX is the most up to date and most easy to
install and maintain LaTeX system because of it's well thought out design.
I'd prefer it to be included with Debian in any case

 Regards, Paul *8^)
-- 
   Paul Seelig [EMAIL PROTECTED]
   African Music Archive - Institute for Ethnology and Africa Studies
   Johannes Gutenberg-University   -  Forum 6  -  55099 Mainz/Germany
   Our AMA Homepage  in  the WWW at  http://www.uni-mainz.de/~bender/

Re: where's man ??

[Lars Wirzenius]

> > This was done because AE is small enough to fit on the base disks.
> I am sorry, but isn't vi designed for that purpose?

$ ls -l /bin/ae /usr/bin/vim
-rwxr-xr-x   1 root root23825 May  9 06:05 /bin/ae
-rwxr-xr-x   1 root root   269380 Jan 14  1996 /usr/bin/vim
$ 

I don't think so...

-- 
Please read  before mailing me.
Please don't Cc: me when replying to my message on a mailing list.




pgpifWC8GuD1w.pgp
Description: PGP signature

Re: DNS, Sendmail, or Smail

[Martin Schulze]

[EMAIL PROTECTED] writes:

> I know I'm very new to Linux, and there is quite a bit I haven't 
> caught on to yet.  But, I have a problem that just isn't getting any 
> better.  I have Debian 1.1 loaded, running kernel 2.0.6.  The machine 
> I'm working in is not directly connected to the Internet, but does 
> have a tcp/ip connection to a mail server that is connected.  I loaded 
> Smail, and configured it with no DNS server, and all mail goes to the 
> smarthost.  The objective is to run Majordomo on the Debian box and 
> have all mail sent to the mail server for delivery, but Smail on the 
> Debian box reports an error that domain name lookup failed, so my questions 
> are.

> 1. Does Smail require a direct Internet connection to work ?

No, it works fine with both uucp and dialup slip/ppp

> 2. Can Majordomo run on a machine that is not directly connected to 
> the Internet ?

Surely.

> 3. Is there a way to set-up named to work as a local DNS only ?

Local means without connecting to root servers? Sure, just leave out
the '. cache' entry

Joey

-- 
  / Martin Schulze  *  [EMAIL PROTECTED]  *  26129 Oldenburg /
 / Beware of bugs in the above code; I have only /
/  proved it correct, not tried it.  -- Donald E. Knuth /

RE: X11 Configuration of Elsa Winner 2000PRO/X-2

[Volker M. Goebbels]

On Wed, 18 Sep 1996, Rik Ling wrote:

> I have neither an Elsa card or a Phillips monitor, but I remember when
> trying to get my Daytek 17" to work with my ATI Mach64 under Windows
> 3.11, I had the exact same problem: worked fine at 640x480 but the 
> higher resolutions just came up with a blank screen.  No clicking, no 
> buzzing, no smoke, just.nothing.  With the help of the tech folks at 
> Daytek I tracked it down to the sync polarity of the monitor when 
> running at higher modes.  Apparently the Daytek monitor reverses the 
> polarity when in those modes.  I went into the ATI card setup utility 
> and changed the polarity there and, voila!, it worked just fine.  This 
> was over 18 months ago, and I haven't had a problem since.
> Could your monitor have the same problem?  If forced to, I could probably
> find the monitor manual and give a more detailed explanation.
> 
> Just my $0.02.  Hope it helps.

Thanks for you fast reply. I think this is the problem because the 
monitor gives a sound when going to screen saver mode and it does the 
same when switching to a higher mode in X11. He repeatedly beeps on, I 
think because the sync says "power down" and there is a video signal so 
it tries to display.
Thank you also for your offer, so if you have some time to look for the 
sync description this would help. The Philips manual lists the sync 
modes, but they may be false ...

With best regards,

Volker M. Goebbels
System Administrator, http://www.mc.rwth-aachen.de/users/vmg/
Institute of Macromolecular Chemistry, http://www.mc.rwth-aachen.de
Rheno-Westfalian Technical University, Aachen, Germany
Member of the HTML Writers Guild, http://www.hwg.org
---
  "Why use Windows, when I can leave through the door?"
---

apache-1.1.1

[Alex Romosan]

i tried using apache_1.1.1-2 and 3 from the Incoming directory and we
started having problems viewing gzipped postscript files with both
netscape and mosaic. i don't know if this is a bug in the apache
server itself or just in the debian distribution. also
/usr/lib/apache/modules/mod_proxy.so doesn't want to load in
apache-1.1.1-3. unfortunately i don't really have a lot of time to
look into this (trying to graduate, i guess), so i just went back to
1.0.5. if anybody would care to investigate this further i would
appreciate it. thanks.

--alex--

-- 
| I believe the moment is at hand when, by a paranoiac and active |
|  advance of the mind, it will be possible (simultaneously with  |
|  automatism and other passive states) to systematize confusion  |
|  and thus to help to discredit completely the world of reality. |