Re: Locale-related questions

[Martin Str|mberg]

In article  Nicolas George  
wrote:
> wchar_t portably. For starters, the i4s at microsoft decided that 64k
> characters should be enough for everyone, so if your cross-platform includes
> microsoftisms, you can not use wchar_t to represent an Unicode code point.
> The i4s at sun had other interesting ideas on how to make the coding for
> wchar_t itself depend on the locale.

I understand the reference to (the mess from) MS (although at that
time everyone in America did think 64k would be enough characters for
everyone, didn't they?), but not the one to Sun.

Were they the ones that made sure that C's wchar_t is standardized as
it it? Care to enlighten me?


-- 
MartinS

Re: Xorg replaces TTY1

[Martin Str|mberg]

In article  Chris Bannister 
 wrote:
> On Thu, Nov 26, 2015 at 06:54:34PM +, Brian wrote:
> > There are still users (an example is in
> > this thread) who believe ctrl-alt-backspace no longer works in Debian.
> > It does.

> So it does! Wonder why it didn't work for me on another machine. :(

It's because those nice and user-friendly DE have it disabled by
default (IIRC). I always have to go search for where they've hidden
that option this time, when I try another one or a new version.


-- 
MartinS

Re: debianlive iso with xfce and clamav

[Martin Str|mberg]

In article  rlhar...@oplink.net wrote:
> I am attempting to recover data from a virus-damaged Windows machine.  For
> this work, I wish to run debianlive with xfce desktop and clamav.

> I found at:

> ttps://www.debian.org/CD/live/

> ISO images for Jessie with xfce desktop.

> And debianlive includes synaptic, so it it fairly easy to install clamav;
> but my task would be simpler if I could create a debianlive ISO image
> which includes both xfce and clamav.  Is there a HOWTO for this?

There are documentation for this and . Install
live-build and company, read, read, build, boot, build and boot.


-- 
MartinS

Re: LVM info - OTHER than HOWTO's

[Martin Str|mberg]

In article <qwids-8p...@gated-at.bofh.it> Joel Rees <joel.r...@gmail.com> wrote:
> On Thu, Nov 19, 2015 at 2:42 PM, Martin Str|mberg <a...@ludd.luth.se> wrote:
> > [...]
> >
> >> No information on dual boot.
> >
> > If with not Linux, it won't work.

> That's news to me.

> I've mulit-booted openBSD, Fedora in a non-VM LVM, debian, SUSE, and a
> previous version of the OSS fork of Solaris. Not all at once, but
> three or four at a time.

Ok, I don't know if the BSDs can handle LVM or not, but when did
Fedora and Suse leave the Linux kernel? (That's news to me.) Don't know
about "OSS fork of Solaris".

> And I'm pretty sure I've dual-booted MSWindows 7 and Fedora with LVM.

I don't believe you. I'm pretty sure that WINDOWS can't read LVM.

Even if it did, I wouldn't want it to muck around in it.

> Haven't tried LVM in a GPT mapped disk yet.

Obviously that'll work (although I don't remember if I tried
that). It's just partitions...


-- 
MartinS

Re: LVM info - OTHER than HOWTO's

[Martin Str|mberg]

In article  Richard Owlett  
wrote:
> When searching for more information all I'm finding are 
> essentially HOWTO's with only a couple of paragraphs on "Whats" 
> and "Whys". Essentially nothing on "Why not".

One good use is when you're encrypting / (and /home if it's own) and swap.

I use luks to encrypt the partition, then put LVM on it and then /,
/home and swap on LVM. This was you only need to give the
passwd/unlock once.


> No information on dual boot.

If with not Linux, it won't work. With Debians it will. Other Linuxes
probably will. I'm running several Debian installations in the above
encrypted LVM scenario. (You have to fight the installer for the
non-first installations though.)


-- 
MartinS

Re: how execute a script

[Martin Str|mberg]

In article  David Wright 
 wrote:
> As for script-file extensions in DOS, there was really only .BAT
> wasn't there?, so the idea of distinguishing .bash, .csh, .py, .pl,
> .sh, .zsh etc as being inherited from DOS is difficult for me to
> understand.

Perhaps it's because (MS)DOS begat WINDOWS that only knew how to run
something based on the extension?

And that is why we shudder on the sight of a (unnecessary?) extension?


-- 
MartinS

Re: Charsets v grep

[Martin Str|mberg]

In article  R. Clayton  
wrote:
> and I've been getting a lot of this lately:

>   $ grep ^Subject: cbtm 
>   Binary file cbtm matches

> whereas before (a month or so ago) I used to get actual matches on std-out.
> It's easy enough to work around like so

>   $ sed -n -e '/^Subject:/p' < cbtm 
>   Subject: Re: PTFACULTY: FTFACULTY: When saying "Nous sommes Paris" is not
>   Subject: FTFACULTY: When saying "Nous sommes Paris" is not enough
>   Subject: Lowered Reserve Prices

> but I'd like to grep working like it used to.  What is the way for me to get
> grep back?  Some other points that may be useful:

>   $ file cbtm 
>   cbtm: ISO-8859 text, with very long lines

>   $ ba env | grep -i utf
>   LANG=en_US.UTF-8
>   XTERM_LOCALE=en_US.UTF-8

Trying "grep -a ..." might work. Or "LANG=C grep ...".


-- 
MartinS

Re: Installer partitioning problem (was: System Dorked -- Help! (Interim solution!))

[Martin Str|mberg]

In article <qnkxq-5oc...@gated-at.bofh.it> Gene Heskett <ghesk...@shentel.net> 
wrote:
> On Monday 26 October 2015 02:29:04 Martin Str|mberg wrote:
> > I don't understand why you can't. I can create (and did) partitions of
> > 200 MB size. I use the text based installer and manual partitioning.
> >
> On a 4k/sector, 2 terabyte disk?  I tried from 500m to 2g, it would not 
> accept it.  Finally I said to use 5% of the disk, and that worked. Since 

Mmm... Lately, usually the installation is onto an SDD, so lately no.

But earlier I'm quite sure I've done that. That should be around/with
Debian 6 and then probably not 4k/sector disk.

But some SDD has claimed 4k/sector physical block size (but then again
they are not 2TB).

Frankly, I don't see what the 4k/sector physical block size has to do
with it.

> > Supposing your variant of installer can't, then why don't you boot a
> > live CD/stick and partition the hard drive before any booting the
> > installer and then you should be able to tell the installer to just
> > use the partitions without any repartitioning.

> The key phrase is "should be able to tell it to use what it finds", 
> applying only mount point labels.  I have not been allowed to do so by 
> any linux installer over the last 7 or 8 years.  That limitation has 

Again I don't understand why you can't. I have done that successfully
too.

> Thats also when I parted company with fedora, forever.  I was tired of 
> being one of their development lab rats. Always something broken, screw 
> it.

??? I'm talking about Debian.


-- 
MartinS

Re: installing/using grub-legacy

[Martin Str|mberg]

In article  Curt  wrote:
> https://wiki.debian.org/Grub

It seems that page is somewhat outdated/broken:

1. If I follow the "grub2 variants" link below "Up to Debian Lenny" I get
"Error

two or more packages specified (grub2 lenny)"

2. If I follow the "GRUB 2 variants" link below "Debian Squeeze and later"
I get
"Source Package: grub (0.97-64)
Links for grub

Debian Resources:
..."
which obviously is about GRUB1.

I'm using Lynx Version 2.8.6rel.5 (09 May 2007).
(This is not Debian.)


-- 
MartinS

Installer partitioning problem (was: System Dorked -- Help! (Interim solution!))

[Martin Str|mberg]

In article  Gene Heskett  
wrote:
> The worst part of that is that the partitioner will not accept a 1 
> gigabyte partition, which is a great plenty, so I was forced to use 5% 
> of the disk as a boot partition.

I don't understand why you can't. I can create (and did) partitions of
200 MB size. I use the text based installer and manual partitioning.

Supposing your variant of installer can't, then why don't you boot a
live CD/stick and partition the hard drive before any booting the
installer and then you should be able to tell the installer to just
use the partitions without any repartitioning.

HTH.


-- 
MartinS

Re: debian-user list problem

[Martin Str|mberg]

In article <qiin0-4j...@gated-at.bofh.it> to...@tuxteam.de wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1

> On Mon, Oct 12, 2015 at 11:00:17AM +0200, Martin Str|mberg wrote:
> > In article <qin8s-74s...@gated-at.bofh.it> Lisi Reisz 
> > <lisi.re...@gmail.com> wrote:
> > > D'oh!  I was forgetting that this list is not subscribers only.
> > 
> > Really?!
> > 
> > I hope this goes through. I sure have been wanting to reply to some things.
> > I read this as the news group linux.debian.user on aioe.org with tin.
> > 
> > 
> > Now I hope somebody can tell me if there's an easier way to reply to the
> > list using tin, hitting r (for reply to sender) and then replacing the
> > mail address of the post I'm replying to with d-u@l.d.o, so I don't forget
> > to do that in my excitement in replying to something.

Yay! It got through,

> Long time I didn't use tin, but if my hazy memory serves, "followup" is
> what you are looking for (instead of "reply"). So perhaps try an "f".

That doesn't work. It tried several times (three?) first before realising
that my replies didn't get through, so I gave up thinking I needed to be
subscribed.

So for me follow-up in tin on aioe.org doesn't work.


-- 
MartinS

Re: debian-user list problem

[Martin Str|mberg]

In article  Lisi Reisz  
wrote:
> D'oh!  I was forgetting that this list is not subscribers only.

Really?!

I hope this goes through. I sure have been wanting to reply to some things.
I read this as the news group linux.debian.user on aioe.org with tin.


Now I hope somebody can tell me if there's an easier way to reply to the
list using tin, hitting r (for reply to sender) and then replacing the
mail address of the post I'm replying to with d-u@l.d.o, so I don't forget
to do that in my excitement in replying to something.

Thanks!


-- 
MartinS

Re: Partitioning questions.

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
[Klippa, klapp, kluppit prelude.]
: How should I do the partition ? i've read the Partition 
: mini-Howto. Therefore I will anyway have:
: 
:   '/' - a root partition, 
:   '/swap' - a swap partition (32Mo).
: 
:   -1- To achieve what I want to do, I'll have to make a '/home' 
:   partition. Correct ?

Do you have any means to backup? No? Then that's a good idea.

:   -2- my friend an I have a 3Go and 4Go HDs resp. How many 
:   space should I give to the root partition (M$ mu$hroom will also 
:   be present: ~400Mo) ? That's the bootable partition and therefore, it 
:   will contains all the system file. True ?   

Yes and no. If you're only planning to have / and /home, then
yes. Personally I have a small root (for less corruption likelyhood
when the system crashes or the power goes (no crash yet, which wasn't
self-inflicted, for two years)), and then mount /home /usr and /var,
which are bigger. In this case, the answer is no. / is 64MB (not
anywhere near full), /home 500MB (always full - hehe - the usual
problem), /usr 500MB (a little cramped, next time I'll let it be
700MB) and /var 120MB (plenty of room, but if you expect a lot of
mail/news traffic perhaps bigger is necessary).

:   -3- Once I've my Debian CDs, how to overwrite the RH files ?

Just let the installation re-initialise your partitions (not /home of 
course).


Cocteau Twins, Blue Bell Knoll,

MartinS


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

Re: 8 GB limit on cfdisk?

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: 
: I'm having a problem partitioning my 9.6 GB harddrive on my Dell P-II
: 400.  This is an EIDE drive.  When I use cfdisk, it sees only 8 GB.
: I believe that this is due to a limit in cfdisk which sets the max
: number of sectors to 1024.  Mine should have 1227 (approx).  
: 
: When I boot Linux it identifies my hard drive and says it has 9.6 GB.
: Also, when I used Partition Magic to reformat my windoze 95 area, it
: saw all of my disk.  
: 
: I am running linux 2.0.34, cfdisk 0.8l (from util-linux-2.8), on
: Debian 1.3.1r8.  
: 
: Is this a program limitation?  Is there a workaround?  Is there
: another program for Linux that I could use?

Are you using the setting LBA in the BIOS? I think I do and I have a
8.5GB disk with heads 255 sectors 63 and cylinders 784. Those are not
near any 1024 limit.


Throwing Muses, Limbo,


MartinS


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

Re: Sound with wine

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: [EMAIL PROTECTED] wrote:
:  I'm having trouble getting sound to work under wine.
:  I've installd libwine0.0.971116_0.0.980601-1.deb
:  and wine_0.0.980601-1.deb from slink.
:  I downloaded the files ossfree38s9-linux20x.tar.gz and
:  ossfree38s-linux20x.README and follow the instructions
:  in this README. Rebuilt the kernel (2.0.34) with sound
:  support (seems to work except with wine).
:  one error I've seen when running wine is:
:  fixme:sound:OpenSound16 (void): stub
:  
:  Any ideas?

Yes. Get the source of wine and fix the fixme. You might not know it
but wine is far from finished. Perhaps you would like to help? There's
a newsgroup comp.emulators.ms-windows.wine.


Right,

MartinS


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

Re: Automatic reply

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: On Sat, 13 Jun 1998 [EMAIL PROTECTED] wrote:
:  Thanks for your email. This is an automatic reply.

Me too.


Throwing Muses, Limbo,

MartinS


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

Re: user can delete kernel images

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: Something very strange has happened to my system. I have my kernels in
: /boot (the usual setup ) with permission 644. I have never touched that
: after they are created by the kernel-package. I am doing some experiments
: concerning security. So I tried as a user with no root privileges and no
: root group privileges to delete the files /boot/vmlinuz.2.0.0 and
: /boot/vmlinuz.2.0.27 I was asked whether 644  should be overrided I said
: yes and it removed the files  Why ?? This is of no immediate risk
: since I have several more kernels which I use but this is scary. Can
: anayone explain this ??

What are the permissions of the directory /boot?


Right,

MartinS


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

Re: Help w/ Linux install

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: I have a question which I can not seem to find an answer from the online
: documentation, FAQs, etc.
: 
: I would like to install Linux on a PC which has 3, 2-GB partitions.
: Partition C contains my required corporate (ugh) Microsoft Software.
: 
: Partition D holds the Debian distribution and Partition E is empty. I
: would like to install Linux on the E partition. Is there any way I can
: specify access to that specific partition when I am installing from a
: floppy?

Yes. But after installing you probably won't have an E partition in
DOZE anymore as Microsnoft's product won't understand the ext2 file
system.


Right,

MartinS


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

Re: startx giving error message

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: This might sound like a dumb question, but how do you check where a symbolic 
link
: points to?  I understand what you're talking about, and I can see X in the
: /usr/X11R6/bin directory, but I don't know what to do from there to get it to 
point
: to my current SVGA server (I'm using a Hercules Terminator 64/3D S3 virge 
card,
: FYI).  Thanks for any help you can give.

Try ls -la file_or_directory. To change the link try ln -s target
symbolic_link_that_points_to_target.


Right,

MartinS


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

Re: How do I - Directory manipulation

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: 3 questions
: 
: How do I delete a directory with all subdirectories and files (similar to 
deltree)?

rm -fr directory.

: AFter using emacs some mysterious directories have appeared that I cannot 
remove
: #this dir# How do I get rid of these?

This is auto-save files, IIRC. What does rm #this dir# say?

: I downloaded a file using lynx and now cannot find it.  I tried 'find' but 
could not use
: woldcards.  Is there a program similar to 'whereis' that will allow me to use 
wildcards and
: search all directories and subdirectories?

How about find . -name '*file*name*' -print?


Right,

MartinS


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

Re: Compiling a kernel (fwd)

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: avail. Who is Manjo?

Manoj Srivastava. Maintainer of kernel-package (and perhaps other
things as well).


Right,

MartinS


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

Re: persist not working

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:
: 
: I use the persist option in my pppd to keep my link alive, and it
: used to work very well, but lately, sometimes (not always) the link
: goes down, and is not restablished again. Any clues why would this
: happen? 

No. Turn on debugging with the debug option in ppp.options_out or by
sending the right signal to the pppd process (the man page says which
it is, I don't remember), and see if you get any clues.


Right,

MartinS


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

Network routing

[Martin Str|mberg]

Hello.

I'm running Debian 1.something. I've some trouble with my network
setup.

First a picture on my network:

-  eth0  -   eth1  -
I  192.168.1.2  I  I  192.168.0.1  I   I  192.168.0.2  I
-I  192.168.1.1  I -
 -

Now on 192.168.1.2 I run /usr/sbin/traceroute 192.168.0.2:
traceroute to 192.168.0.2 (192.168.0.2), 30 hops max, 40 byte packets
 1  * * *
 2  * * *
 3  * * *
[The other 27 are just the same.]

On 192.168.1.2 /sbin/route -n reports:
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
192.168.1.0 0.0.0.0 255.255.255.0   U 0  0 3 eth0
127.0.0.0   0.0.0.0 255.0.0.0   U 0  0 2 lo
0.0.0.0 192.168.1.1 0.0.0.0 UG1  0   872 eth0

On 192.168.0.1/192.168.1.1 /sbin/route -n reports:
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
192.168.1.0 0.0.0.0 255.255.255.0   U 0  0 4 eth0
192.168.0.0 0.0.0.0 255.255.255.0   U 0  0 0 eth1
127.0.0.0   0.0.0.0 255.0.0.0   U 0  0 6 lo

or this after I've tried to add gateways:
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
192.168.1.0 192.168.1.2 255.255.255.0   UG0  0 2 eth0
192.168.1.0 0.0.0.0 255.255.255.0   U 0  0 4 eth0
192.168.0.0 192.168.0.2 255.255.255.0   UG0  0 2 eth1
192.168.0.0 0.0.0.0 255.255.255.0   U 0  0 0 eth1
127.0.0.0   0.0.0.0 255.0.0.0   U 0  0 6 lo


Can somebody advice me how to find out what's wrong? Should I add
gateways or not?


Thanks,

MartinS


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

Re: Network routing

[Martin Str|mberg]

According to Nils Rennebarth:
 On Mon, May 25, 1998 at 09:08:44PM +0200, Martin Str|mberg wrote:
  Hello.
  
  I'm running Debian 1.something. I've some trouble with my network
  setup.
  
  First a picture on my network:
  
  -  eth0  -   eth1  -
  I  192.168.1.2  I  I  192.168.0.1  I   I  192.168.0.2  I
  -I  192.168.1.1  I -
   -
  
  Now on 192.168.1.2 I run /usr/sbin/traceroute 192.168.0.2:
  traceroute to 192.168.0.2 (192.168.0.2), 30 hops max, 40 byte packets
   1  * * *
   2  * * *
   3  * * *
  [The other 27 are just the same.]
  
  On 192.168.1.2 /sbin/route -n reports:
  Kernel IP routing table
  Destination Gateway Genmask Flags Metric Ref Use Iface
  192.168.1.0 0.0.0.0 255.255.255.0   U 0  0 3 eth0
  127.0.0.0   0.0.0.0 255.0.0.0   U 0  0 2 lo
  0.0.0.0 192.168.1.1 0.0.0.0 UG1  0   872 eth0
  
  On 192.168.0.1/192.168.1.1 /sbin/route -n reports:
  Kernel IP routing table
  Destination Gateway Genmask Flags Metric Ref Use Iface
  192.168.1.0 0.0.0.0 255.255.255.0   U 0  0 4 eth0
  192.168.0.0 0.0.0.0 255.255.255.0   U 0  0 0 eth1
  127.0.0.0   0.0.0.0 255.0.0.0   U 0  0 6 lo
 This is strange, because up to now, everything is ok, and traceroute should
 at least proceed from 1.2 to 1.1
 
 - what does ping 192.168.1.1 give when issued on 192.168.1.2 ?

kant:~ ping 192.168.1.1
PING 192.168.1.1 (192.168.1.1): 56 data bytes
64 bytes from 192.168.1.1: icmp_seq=0 ttl=64 time=0.2 ms
64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=0.2 ms


 - Are you sure, IP-Forwarding is enabled on 192.168.1.1 ? You need to
   compile a custom kernel for this, the debian ones don't have it enabled.

It seems it's not compiled in (found out by checking the .config file;
if it was the right file I found). Can this be verified in some other
way? Like cat /proc/net/something?

 - The routes on 192.168.0.2 need to be
   ifconfig eth0 192.168.0.2 netmask 255.255.255.0 broadcast 192.168.0.255
   route add -net 192.168.0.0
   route add default gw 192.168.0.1

They are.

 - what does ping 192.168.0.1 give when issued on 192.168.0.2 ?

skrot:~ ping 192.168.0.1
PING 192.168.0.1 (192.168.0.1): 56 data bytes
64 bytes from 192.168.0.1: icmp_seq=0 ttl=64 time=3.6 ms
64 bytes from 192.168.0.1: icmp_seq=1 ttl=64 time=3.4 ms
64 bytes from 192.168.0.1: icmp_seq=2 ttl=64 time=3.4 ms

--- 192.168.0.1 ping statistics ---
3 packets transmitted, 3 packets received, 0% packet loss
round-trip min/avg/max = 3.4/3.4/3.6 ms


To conclude it seems I need to enable IP-Forwarding in the kernel.


Thanks,

MartinS


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

Re: umsdos

[Martin Str|mberg]

Jason Ish said

 I have to install Linux onto a umsdos file system on one of my computers,
 even though I know it isn't a suggested practice.

 What steps should I take to install a fresh debian system onto umsdos.  I
 have a up and running Debian system to make a new kernel and what not.

 Thanks for any info or pointers to info.
 Jason

I've asked the same myself, and got some feedback:

- Forwarded message from Giuliano Procida -

From [EMAIL PROTECTED]  Sat Aug 23 15:02:52 1997
To: Martin Str|mberg [EMAIL PROTECTED]
Subject: UMSDOS support for boot-floppies
Date: Sat, 23 Aug 1997 14:02:48 +0100
From: Giuliano Procida [EMAIL PROTECTED]
Message-Id: [EMAIL PROTECTED]

I have done this already. However, there are bugs in UMSDOS [yes, the
Linux kernel] which make it impossible to use reliably.

In particular there is one bug of unknown origin that completely
crashes Linux (it never returns from a syscall). This bug is triggered
during dpkg -i ncurses-term* . Another bug caused dpkg -i libc* to
fail [but this has been isolated].

For the moment, I have suspended further development work on UMSDOS
support. It is possible that the UMSDOS stuff may be cleaned up when
it is converted to work with new dcache scheme in the 2.1 kernels.

If you are interested in my work to date, please feel free to take a
look at ftp://hilfy.magd.cam.ac.uk/pub/ boot-floppies/*patch* and
current/PROBLEMS .

I can make a more recent patch available if you are interested.

Giuliano.

- End of forwarded message from Giuliano Procida -

Please note that this info is somewhat dated, but the ftp directory is
still there.

There might be some info in the debian-user mailing list archives as well.


HTH,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: Help me interpret error messages

[Martin Str|mberg]

In article [EMAIL PROTECTED] gerdoisa-no-spam wrote:
: Also make sure the following directories are set to have a group of man
: see: chgrp
: 
: rootvdir /var/catman -d
: drwxrwxr-x  14 root man  1024 Mar 21  1997 /var/catman
: 
: rootvdir /var/catman   
: total 12
: drwxrwxr-x  12 root man  1024 Mar 21  1997 X11
: drwxrwxr-x   2 root man  1024 Oct  4 13:49 cat1
: drwxrwxr-x   2 root man  1024 Sep 14 11:20 cat2
: drwxrwxr-x   2 root man  1024 Oct  4 13:37 cat3
: drwxrwxr-x   2 root man  1024 Jun 11 02:39 cat4
: drwxrwxr-x   2 root man  1024 Oct  1 08:44 cat5
: drwxrwxr-x   2 root man  1024 Oct  3 14:41 cat6
: drwxrwxr-x   2 root man  1024 Dec 20  1996 cat7
: drwxrwxr-x   2 root man  1024 Oct  4 13:26 cat8
: drwxrwxr-x   2 root man  1024 Aug  8  1995 cat9
: drwxrwxr-x   2 root man  1024 Sep 14 11:21 catn
: drwxrwxr-x  12 root man  1024 Mar 21  1997 local
: 
: This is so the man program can set it's group id to man when it
: is run, and thus have write permissions to those directories for when you
: are reading the manual pages.
: 
: rootvdir /usr/bin/man
: -rwxr-sr-x   1 root man 28920 Sep 22 19:05 /usr/bin/man

Hmm. Are you sure about that? Chez moi it's like this:
descartes:~ vdir /var/catman/ -d
drwxr-xr-x  13 man  root 1024 Sep 28 07:01 /var/catman/
descartes:~ vdir /var/catman/
total 260
drwxr-xr-x   6 man  root 1024 Sep 28 07:12 X11R6
drwxr-xr-x   2 man  root 1024 Sep 29 06:45 cat1
drwxr-xr-x   2 man  root 1024 Aug 24 18:18 cat2
drwxr-xr-x   2 man  root 1024 Aug 24 18:18 cat3
drwxr-xr-x   2 man  root 1024 Aug 24 18:18 cat4
drwxr-xr-x   2 man  root 1024 Sep  2 06:46 cat5
drwxr-xr-x   2 man  root 1024 Aug 24 18:18 cat6
drwxr-xr-x   2 man  root 1024 Aug 24 18:18 cat7
drwxr-xr-x   2 man  root 1024 Sep 30 23:00 cat8
drwxr-xr-x   5 man  root 1024 Sep 28 06:48 de_DE
-rw-r--r--   1 man  root   253952 Sep 28 07:01 index.bt
drwxr-xr-x   5 man  root 1024 Sep 28 06:48 it_IT
descartes:~ vdir /usr/bin/man
-rwsr-xr-x   1 man  root71204 May 21 11:40 /usr/bin/man

I'm running Debian 1.3.1.


Ho-hum,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: Boot Problems

[Martin Str|mberg]

In article [EMAIL PROTECTED] Russ Cook wrote:
[Klippa, klapp, kluppit.]

: I have built a custom kernel (2.0.30) to remove unnecessary file system
: support, and add IP Masquerading. I've followed the instructions in
: /usr/src/linux/README.  I've made a boot floppy with the new kernel, and I
: ran rdev /dev/fd0 /dev/sdb6 where sdb6 is the Linux partition on my
: second SCSI drive.  I then ran rdev /dev/fd0 to verify the floppy was
: pointing to /dev/sdb6, and it reported correctly.  When I reboot the
: machine from the floppy, a screen of preliminary configuration info is
: displayed, then I get a kernel panic with the message: VFS: can't mount
: boot device 8:16.

This is in hex. In decimal it becomes 8, 22, which is sdb6, just as
you have said you have set it to.

Perhaps you said no to some neccessary config options, which you
didn't think you'd be needing?

I think that you need e. g. CONFIG_BLK_DEV_RAM and
CONFIG_BLK_DEV_INITRD set to y. Perhaps somebody else care to
comment?

: I've been reading the messages regarding 16 partitions per SCSI drive.  I
: specifically configured (using rdev) to boot to /dev/sdb6.  I don't yet
: understand all about the major/minor stuff.  I plan to read more this
: weekend.  When I run lilo on my kernel on the hard drive, and reboot, I get
: the same message as when I boot from floppy. 

As long as you don't have more than 11 extended partitions there's
nothing to worry about (more than the usual things).

: The originally installed kernel (from the 2 CD set) boots fine.  I have
: remembered to set /vmlinuz to point to the correct kernel, and I ran lilo
: to update the System.map.

The package kernel-package is a very neat one if you want to rebuild
your kernel.


Hth,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: NFS expert, please diagnose this

[Martin Str|mberg]

Hello.


I think I see the problem with regard to setuid programs and nfs.
It's in the kernel.

It's in the file kernel-source-2.0.27/fs/nfs/proc.c.
Alas the same code is repeated in several places, so I'm not sure how to
fix it: there must be a reason for exact this coding.

Does anybody here know what to do? Or know whom I might contact to get a
second opinion?


Thanks,

MartinS


Here follows a snippet from the above mentioned file and an possible scenario
(the one that makes at fail if /var/spool/cron/atjobs is on a nfs mounted
partition) to illustrate the problem:
 Start of proc.c snippet. 
[...]
int nfs_proc_lookup(struct nfs_server *server, struct nfs_fh *dir, const char *n
ame,
struct nfs_fh *fhandle, struct nfs_fattr *fattr)
{
int *p, *p0;
int status;
int ruid = 0;

PRINTK(NFS call  lookup %s\n, name);
#ifdef NFS_PROC_DEBUG
if (!strcmp(name, xyzzy))
proc_debug = 1 - proc_debug;
#endif
if (!(p0 = nfs_rpc_alloc(server-rsize)))
return -EIO;
retry:
p = nfs_rpc_header(p0, NFSPROC_LOOKUP, ruid);
p = xdr_encode_fhandle(p, dir);
p = xdr_encode_string(p, name);
if ((status = nfs_rpc_call(server, p0, p, server-rsize))  0) {
nfs_rpc_free(p0);
return status;
}
if (!(p = nfs_rpc_verify(p0)))
status = -errno_NFSERR_IO;
else if ((status = ntohl(*p++)) == NFS_OK) {
p = xdr_decode_fhandle(p, fhandle);
p = xdr_decode_fattr(p, fattr);
PRINTK(NFS reply lookup\n);
/* status = 0; */
}
else {
if (!ruid  current-fsuid == 0  current-uid != 0) {
ruid = 1;
goto retry;
}
PRINTK(NFS reply lookup failed = %d\n, status);
status = -nfs_stat_to_errno(status);
}
nfs_rpc_free(p0);
return status;
}

[...]

static int *nfs_rpc_header(int *p, int procedure, int ruid)
{
return rpc_header(p, procedure, NFS_PROGRAM, NFS_VERSION,
(ruid ? current-uid : current-fsuid),
current-egid, current-groups);
}

[...]
 End of proc.c snippet. 

Now first ruid is initialised to 0. First time around we fail looking up the
file and status (what is supposed to be propagated to errno) is set to 2
(ENOENT), so the statements ruid = 1; and goto retry; is executed.

Second time around the lookup will be performed as current-uid, which
obviously fail (as the directory is only writeable for root), hence
status (and after that errno) is set to 13 (EACCES).

Comments?


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: NFS expert, please diagnose this

[Martin Str|mberg]

Thanks for the response, I've been feeling abandoned lately!

According to joost witteveen:
  Appearently, stat-ing a file in a directory that is only readable for
  root from a program that is setuid root, that is on a nfs mounted
  partition fails.
 
 Does adding a no_root_squash (like below) change anything?
 
 /directory   host(no_root_squash)

It's already there:

Line from exports at descartes:
/var/clients/kant   kant.dcd.se(rw,no_root_squash)

Line from output of mount on kant:
descartes.dcd.se:/var/clients/kant/var on /var type nfs 
(rw,rsize=8192,wsize=8192,addr=192.168.1.1)

 Old versions of nfsd used to have no_root_squash by default, but
 more recent ones don't (that's a lot more secure). When root_squash
 is in effect, any request from uid=0 will be mapped to uid=nobody,
 and thus root isn't allowed to access the directory.
 
  Here are some version that might be relevant (please ask if you need
  another package's version number):
  
  base,   1.1.0-14
  libc5,  5.4.33-3
 
 While you are at it, why not tell the version of nfsd (netstd)?

Oops, I thought I did with base, now I see I missread it. Here's netstd:
netstd, 2.13-1

Does my test program work for you?


Thanks,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: installing twin and other questions

[Martin Str|mberg]

In article [EMAIL PROTECTED] tmalloy wrote:
: I know someone asked about this recently.  Sorry for not paying more
: attention.  I got twin from the caldera archive.  It does not seem to have
: compiled correctly.  Can anyone tell me how they got it up and running?  I
: am trying to install it to /twin because I have a lot of room on /.  I
: know that it is hard to tell what the problem I am having is. Is it
: appropriate to attach a script file of the compile?  If it is I can do
: that.

That's ok, I think. However you might want to look at the mailing list
archives at www.debian.org. Select Support and then mailing list
archives.

:  I try to do almost everything except administration as user.  I only
: proform administration such as installing all debian stuff as root.  Is it

Good.

: safe to install/compile non-debian programs as root.  I tried to install

In the most general sense, no. You have to go through the code to see
it doesn't do things you don't want to happen.

In the case of twin, I'd think it'd be safer. They obviously don't
want to wreck you system. But there might be bugs that might, but
that's the same for all code.

For compiling debian packages I don't think you need to be root, except
for a very limited set. For libraries it ought to go well to compile
them as non-root and then install them as root.

: twin as both and neither worked well.  Why is it that certain programs do
: not execute as root even though x permission is set. (xboard for example)
: I am not advocating playing chess as root I would just like to understand
: the mechanics involved.

Because the s bit isn't set, cf.:
descartes:~ ls -la /usr/games/xboard /usr/bin/at
-rwsr-xr-x   1 root root31836 Jun 17 18:05 /usr/bin/at
-rwxr-xr-x   1 root root   283280 Jan 16  1997 /usr/games/xboard
   ^
   Here!

:  And lastly, like most all of you I love poking around my system.  I have
: no objection to having to fiddle around a little in order to make things
: work. In fact I love it.  This is one of the charms of linux and something
: will be lost when or if it becomes a fully polished product.  However
: having said that, it really annoys me when I cannot try an application
: deselect says I have installed because I cannot figure out what the name
: of the executable is. Many new linux users are not coming with strong unix
: backrounds. And may in fact know little or nothing about the program
: they are trying out or learning.  Please keep this in mind. At least let
: us know how to start the application.  Anyway dselect says visual-tcl is
: installed, but where is it?

I don't know anything about visual-tcl, but you could try man -k
visual-tcl and if that doesn't give any hints man -k tcl. And you
could look in /usr/doc to see if there's a directory named something
like tcl.


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: NFS expert, please diagnose this

[Martin Str|mberg]

According to joost witteveen:
 I don't really think this has anything to do with setuid stuff
 or anything, as ls running as root itself doesn't see anything
 eighter.

??? I don't understand. As root:
kant# ls -la /var/spool/cron/atjobs/
total 7
drwx--   2 daemon   daemon   1024 Sep 28 15:15 .
drwxr-xr-x   5 root root 1024 Feb 19  1997 ..
-rw---   1 daemon   daemon  6 Sep 24 22:12 .SEQ
[Other entries removed.]
kant# ls -la /boot/no/
total 2
drwx--   2 daemon   daemon   1024 Sep 22 00:54 .
drwxr-xr-x   4 root root 1024 Sep 22 00:52 ..
-rw---   1 daemon   daemon  0 Sep 22 00:54 a

Of course as an ordinary user I don't see anything, but then I shouldn't
and I'm not surprised as ls is not setuid. 

 Yes, this looks as a bug to me.

How do I find where it is so I can fix it? Do you think it's libc or nfsd?

 Recently, netstd_3.00 has been released (libc6, unstable), that one
 may behave differently. Unfortunately, due to a bug in it (it sefaults
 on squach_uid stuff), I cannot install it on my machine.

Ugh! I wouldn't want to install that.

  Does my test program work for you?
 
 Well, just ls doesn't work eigther (and ls filename doesn't much
 but stat eighter), so I don't think your program would make much
 difference.
 
 Does ls work for you? (I hope not!)

As I said above it does as root, but not as an ordinary user. All as
expected.


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: NFS expert, please diagnose this

[Martin Str|mberg]

According to joost witteveen:
  According to joost witteveen:
   I don't really think this has anything to do with setuid stuff
   or anything, as ls running as root itself doesn't see anything
   eighter.
  
  ??? I don't understand. As root:
  kant# ls -la /var/spool/cron/atjobs/
 
 And /var/spool/cron/atjobs is mounted over nfs?
 
 Strange, I see (on the client):
 
 rulvsa:/home/joostje/tmp# ls -al root/file
 ls: root/file: Permission denied
 
 with /home/joostje/tmp mountd via nfs. (At the moment I've got the
 no_root_squash removed again, but it was the same with it enabled).
 
 
 On the server: 
 rulcmc:/tmp# ls -al root
 total 9
 drwx--   2 root root 1024 Sep 28 12:43 .
 drwxrwxrwt  24 root root 8192 Sep 28 15:41 ..
 -rw-r--r--   1 root root0 Sep 28 12:43 file
 
   Yes, this looks as a bug to me.
  
  How do I find where it is so I can fix it? Do you think it's libc or nfsd?
 
 Well, what _I_ see surely comes from nfsd. (netstd source).
 running nfsd with
   /usr/sbin/rpc.nfsd -F -d call
 clearly showed that it refused to serve /tmp/root/file (result: 13,
 which is access denied). But you seem to see something different
 (if /var/spool/cron/atjobs is mounted over nfs, apparetnly ls
 can see it).
 
 But really, ls doens't do much else than just lstat (ok, not stat but lstat):
 
 # strace root/file
 [...]
 brk(0x8058000)  = 0x8058000
 lstat(tmp/root/file, 0x8054d04)   = -1 EACCES (Permission denied)
 write(2, ls: , 4ls: ) = 4
 write(2, tmp/root/file, 13tmp/root/file)   = 13
 [...]
 
 I don't hope changing the stat to lstat in your programme makes
 any difference (if it does, you're messing up some symlinks somewhere),
 but anyway, I cannot test, as the behaviour I see is at least consistent
 (lstat (from ls) and stat from your stat programme both returning error):
 
 # ./test/st
 stat(/boot/no/a successful.
 stat(/home/joostje/root/file unsuccessful.
 
 (yes, I compiled it now).
 
 -- 
 joost witteveen, [EMAIL PROTECTED]
 #!/usr/bin/perl -sp0777iX+d*lMLa^*lN%0]dsXx++lMlN/dsM0j]dsj
 $/=unpack('H*',$_);$_=`echo 16dio\U$kSK$/SM$n\EsN0p[lN*1
 lK[d2%Sa2/d0$^Ixp|dc`;s/\W//g;$_=pack('H*',/((..)*)$/)
 #what's this? see http://www.dcs.ex.ac.uk/~aba/rsa/
 


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: NFS expert, please diagnose this

[Martin Str|mberg]

Oops. Disregard my last message, which didn't contain anything new. I
slipped on the keys. 

According to joost witteveen:
 And /var/spool/cron/atjobs is mounted over nfs?

Yes, but it's /var that's mounted. On the client (kant, descartes is the
server):

kant:~ mount
[...]
descartes.dcd.se:/var/clients/kant/var on /var type nfs 
(rw,rsize=8192,wsize=8192,addr=192.168.1.1)
[...]

 Strange, I see (on the client):
 
 rulvsa:/home/joostje/tmp# ls -al root/file
 ls: root/file: Permission denied
 
 with /home/joostje/tmp mountd via nfs. (At the moment I've got the
 no_root_squash removed again, but it was the same with it enabled).

Are you sure that you restarted all the necessary deamons?

 On the server: 
 rulcmc:/tmp# ls -al root
 total 9
 drwx--   2 root root 1024 Sep 28 12:43 .
 drwxrwxrwt  24 root root 8192 Sep 28 15:41 ..
 -rw-r--r--   1 root root0 Sep 28 12:43 file
 
   Yes, this looks as a bug to me.
  
  How do I find where it is so I can fix it? Do you think it's libc or nfsd?

So now, I have TWO problems? Oh, dear!

 Well, what _I_ see surely comes from nfsd. (netstd source).
 running nfsd with
   /usr/sbin/rpc.nfsd -F -d call
 clearly showed that it refused to serve /tmp/root/file (result: 13,
 which is access denied). But you seem to see something different
 (if /var/spool/cron/atjobs is mounted over nfs, apparetnly ls
 can see it).

Part of nfsd log while doing ls -la /var/spool/cron/atjobs/ as root on
the client:

[...]
nfsd[3338] 09/28/97 16:13 lookup [1 97/9/28 16:13:17 kant 0.0+0]
fh:/var/clients/kant/var/spool/cron n:atjobs
nfsd[3338] 09/28/97 16:13   new_fh = /var/clients/kant/var/spool/cron/atjobs
nfsd[3338] 09/28/97 16:13 result: 0
nfsd[3338] 09/28/97 16:13 getattr [1 97/9/28 16:13:17 kant 0.0+0]
/var/clients/kant/var/spool/cron/atjobs
nfsd[3338] 09/28/97 16:13 result: 0
nfsd[3338] 09/28/97 16:13 readdir [1 97/9/28 16:13:17 kant 0.0+0]
/var/clients/kant/var/spool/cron/atjobs
nfsd[3338] 09/28/97 16:13 result: 0
nfsd[3338] 09/28/97 16:13 lookup [1 97/9/28 16:13:17 kant 0.0+0]
fh:/var/clients/kant/var/spool/cron/atjobs n:..
nfsd[3338] 09/28/97 16:13   new_fh = /var/clients/kant/var/spool/cron
nfsd[3338] 09/28/97 16:13 result: 0
nfsd[3338] 09/28/97 16:13 lookup [1 97/9/28 16:13:17 kant 0.0+0]
fh:/var/clients/kant/var/spool/cron/atjobs n:.SEQ
nfsd[3338] 09/28/97 16:13   new_fh = 
/var/clients/kant/var/spool/cron/atjobs/.SEQ
nfsd[3338] 09/28/97 16:13 result: 0
[...]

 But really, ls doens't do much else than just lstat (ok, not stat but lstat):
 
 # strace root/file
 [...]
 brk(0x8058000)  = 0x8058000
 lstat(tmp/root/file, 0x8054d04)   = -1 EACCES (Permission denied)
 write(2, ls: , 4ls: ) = 4
 write(2, tmp/root/file, 13tmp/root/file)   = 13
 [...]

Part of log from strace ls -la /var/spool/cron/atjobs/ /var/tmp/ls.log
21 as root on the client:

[...]
brk(0x8058000)  = 0x8058000
lstat(/var/spool/cron/atjobs/, {st_mode=S_IFDIR|0700, st_size=1024, ...}) = 0
stat(/var/spool/cron/atjobs/, {st_mode=S_IFDIR|0700, st_size=1024, ...}) = 0
open(/var/spool/cron/atjobs/, O_RDONLY) = 3
fcntl(3, F_SETFD, FD_CLOEXEC)   = 0
brk(0x805b000)  = 0x805b000
getdents(3, /* 8 entries */, 8192)  = 184
lstat(/var/spool/cron/atjobs/., {st_mode=S_IFDIR|0700, st_size=1024, ...}) = 0
lstat(/var/spool/cron/atjobs/.., {st_mode=S_IFDIR|0755, st_size=1024, ...}) = 0
lstat(/var/spool/cron/atjobs/.SEQ, {st_mode=S_IFREG|0600, st_size=6, ...}) = 0
[...]


 I don't hope changing the stat to lstat in your programme makes
 any difference (if it does, you're messing up some symlinks somewhere),
 but anyway, I cannot test, as the behaviour I see is at least consistent
 (lstat (from ls) and stat from your stat programme both returning error):
 
 # ./test/st
 stat(/boot/no/a successful.
 stat(/home/joostje/root/file unsuccessful.

No difference:
kant:/var/tmp ./st
stat(/boot/no/a successful.
lstat(/boot/no/a successful.
stat(/var/spool/cron/.SEQ unsuccessful.
lstat(/var/spool/cron/.SEQ unsuccessful.


I'll gladly (well, not gladly exactly, more like if it's necessary then it's
necessary) supply you with a complete dump of configuration files and such
but then I prefer we do it off away from the list.

I'll post a summary to the list what we conclude, when we know what's
happening.


Thanks a lot,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: NFS expert, please diagnose this

[Martin Str|mberg]

[Mailed to debian-user as well as this relevant (bug in my test program).]

According to joost witteveen:
 Are you sure you didn't have a typo there? didn't you mean
 /var/clients/kant/var/spool/cron/atjobs/.SEQ
 instead of 
 /var/spool/cron/.SEQ
 
 (there's at least a atjobs missing).

Ah! You're very right. Thank you!

No I meant /var/spool/cron/atjobs/.SEQ.

Sorry about that. Now I have:
kant:/var/tmp ./st
stat(/boot/no/a successful.
lstat(/boot/no/a successful.
stat(/var/spool/cron/atjobs/.SEQ successful.
lstat(/var/spool/cron/atjobs/.SEQ successful.

So that wasn't the problem. Back to the original one. There seems to be bug
in the at package.

You see I've this problem running at as an ordinary user on the client:

kant:/var/tmp at 2:2
Cannot access /var/spool/cron/atjobs: Permission denied

Looking at the at code, I found the line where it says the above, and
whipped up a simple test case that showed the same problem: my st program.

Unfortunately, it had (at least) one bug. The one you've pointed out.

So now back to the original problem:
Can you please try running at 2:2 as an ordinary user on a client
that nfs mounts it's /var directory (or at least it's /var/spool/cron/atjobs
directory)

Turning on logging for nfsd as before, I run this as an ordinary user on
the client:
kant:/var/tmp at 2:2
Cannot access /var/spool/cron/atjobs: Permission denied

In the nfs log on the server, I see:
[...]
nfsd[3388] 09/28/97 17:42   new_fh = /var/clients/kant/var/spool/cron/atjobs
nfsd[3388] 09/28/97 17:42 result: 0
nfsd[3388] 09/28/97 17:42 lookup [1 97/9/28 17:42:20 kant 0.100+100,4,24,25,29]
fh:/var/clients/kant/var/spool/cron/atjobs n:.SEQ
nfsd[3388] 09/28/97 17:42   new_fh = 
/var/clients/kant/var/spool/cron/atjobs/.SEQ
nfsd[3388] 09/28/97 17:42 result: 0
nfsd[3388] 09/28/97 17:42 getattr [1 97/9/28 17:42:20 kant 0.100+100,4,24,25,29]
/var/clients/kant/var/spool/cron/atjobs/.SEQ
nfsd[3388] 09/28/97 17:42 result: 0
nfsd[3388] 09/28/97 17:42 read [1 97/9/28 17:42:20 kant 0.100+100,4,24,25,29]
/var/clients/kant/var/spool/cron/atjobs/.SEQ: 4096 bytes at 0
nfsd[3388] 09/28/97 17:42 result: 0
nfsd[3388] 09/28/97 17:42 write [1 97/9/28 17:42:20 kant 0.100+100,4,24,25,29]
/var/clients/kant/var/spool/cron/atjobs/.SEQ: 6 bytes at 0
nfsd[3388] 09/28/97 17:42 result: 0
nfsd[3388] 09/28/97 17:42 lookup [1 97/9/28 17:42:20 kant 0.100+100,4,24,25,29]
fh:/var/clients/kant/var/spool/cron/atjobs n:a0001d00dea622
nfsd[3388] 09/28/97 17:42 result: 2
nfsd[3388] 09/28/97 17:42 lookup [1 97/9/28 17:42:20 kant 
1000.100+100,4,24,25,29]
fh:/var/clients/kant/var/spool/cron/atjobs n:a0001d00dea622
nfsd[3388] 09/28/97 17:42 result: 13
nfsd[3388] 09/28/97 17:42 lookup [1 97/9/28 17:42:27 kant 0.0+0]
fh:/var/clients/kant/var n:log

As root on the client I see this (ls -la /var/spool/cron/atjobs/):
kant# ls -la /var/spool/cron/atjobs/
total 3
drwx--   2 daemon   daemon   1024 Sep 28 17:41 .
drwxr-xr-x   5 root root 1024 Feb 19  1997 ..
-rw---   1 daemon   daemon  6 Sep 28 17:42 .SEQ

So what is that a0001d00dea622 file that somehow makes at fail? Is it the
script file that at is creating? 

Hmm. Running a script like this as root an the client, while I'm at 2:2-ing
several times doesn't reveil anything:
 script start ---
#!/usr/bin/tcsh

while(1)
ls -la /var/spool/cron/atjobs/
end
 script end ---

Neither does the equivalent script on the server.

Do you have any idea how to trace this?


 Oh, yes, I am sortof soure I started the nessecary daemons!

Of course! Just making sure, you know. It didn't make sense to me.


Thanks,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: Basic Debian hd mounting...

[Martin Str|mberg]

In article [EMAIL PROTECTED] Lawrence wrote:
: Carey Evans wrote:

[Klippa, klapp, kluppit.]

:  % ls -l /mnt/win95/autoexec.bat
:  -rw-r--r--   1 root root  269 Sep 12 10:26 
/mnt/win95/autoexec.bat
: 
: well not all people can execute/erase files but still not good enough,
: people can read all files in this partition.

I'm sorry if I'm poking my nose into somewhere it doesn't belong but 
what is there on a windodows partition other shouldn't be able to
read?

hint Now if somebody only could answer or explain my question and
problem about stat-ing nfs files while running suid. /hint


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: formatted floppy type??

[Martin Str|mberg]

In article [EMAIL PROTECTED] lucier wrote:
: Howdy Will...thanks for the reply. :-)  Seems it's just the four 1.44 b=
: ase install disks
: that don't respond to either:
: 
: mount -t msdos /dev/fd0 /mnt 
: 
: or
: 
: mount -t ext2 /dev/fd0 /mnt 
: 
: VFS gives an error message stating Can't find valid MSDOS/ext2 filesyste=
: m on dev 02:00.

They are parts of a tar file. At least during the installation there's
a very stripped tar called star. If you want to know more about this
try the boot-floppies package.

: BTW, would you happen to know what the 02:00 describes?

Yes it is indeed the major:minor device numbers, as an earlier post
said. However you should be aware of that the numbers you're asking
about is in hexadecimal, and the numbers you'll see if you do an ls
-al of /dev is in decimal. Hence 03:11 refers to hda17 and not to
hda11. 


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: dpkg other assorted questions...

[Martin Str|mberg]

In article [EMAIL PROTECTED] jd? wrote:
: i have some .deb packkages and cant seem to install them because i get the
: following error when i run dpkg -i filename.deb;
: 
: dpkg: parse error, in file/'var/lib/dpkg/updates/0024' near line 1:
: newline in field name '#padding
: 
: i have no idea what this is...any clues on how to fix it???
: 
: all that file contains is a number of '#padding each on one line like
: this ;
: 
: #padding
: #padding
: #padding
: #padding
: #padding

Do df to check how much free space you have on on your hard
disk(s). I have seen this happen when when you run out of hard disk
space, you see.

I'm not sure how to fix it the correct way, but making sure you have
space on the hard drive and editing that file and removing all
#padding entries in it fixed it for me. Or perhaps I simply removed
that file?

: Also how do i make fvwm2  my default windows manager and how do i add
: programs to my menus.or where do i find documentation on how to
: customize X...

I don't know, but perhaps /etc/X11/window-managers is what you should
edit? 

: one last thinghow do i force my machine to start up to the x windows
: login box..thanx in advanced.

I think you want to edit /etc/X11/config.


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

NFS expert, please diagnose this

[Martin Str|mberg]

I've mailed to debian-user twice earlier about this but I haven't
gotten any responses at all.

If you know NFS, please try this an tell me if it works for you or
not.

As it doesn't work for me and I think it wont work for you, I suppose
I'd file a bug report but I don't know if it's the nfsd, libc, atd or
some other package that isn't doing what it should. 

Appearently, stat-ing a file in a directory that is only readable for
root from a program that is setuid root, that is on a nfs mounted
partition fails.

Here are some version that might be relevant (please ask if you need
another package's version number):

base,   1.1.0-14
libc5,  5.4.33-3
at, 3.1.7-3


Of course, if you need more information, I'll gladly supply it.


Thanks,

MartinS


Here follows the major part of my last posting on this subject:

In the appended program, make sure that FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR
points to a LOCAL file that is in a directory which is only readable
for root.
Likewise, make sure that FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR points to a
file that is in a partition that is NFS MOUNTED that is in a directory
which is only readable for root (/var/spool/cron/atjobs/.SEQ is
perfect if you have it NFS mounted).

Compile the program (gcc -o st st.c), chown daemon.daemon st and
chmod u+s st. Then run it as an ordinaty user.

At me it is like this (/boot is a local directory and /var is nfs
mounted):
kant# ls -al /boot/no
total 2
drwx--   2 daemon   daemon   1024 Sep 22 00:54 .
drwxr-xr-x   4 root root 1024 Sep 22 00:52 ..
-rw---   1 daemon   daemon  0 Sep 22 00:54 a
kant# ls -la /var/spool/cron/atjobs/
total 6
drwx--   2 daemon   daemon   1024 Sep 22 00:58 .
drwxr-xr-x   5 root root 1024 Feb 19  1997 ..
-rw---   1 daemon   daemon  6 Sep 22 00:42 .SEQ
[Some other files (jobs) removed.]

Running this program as an ordinary user (NOT root) I get:
kant:/var/tmp ./st
stat(/boot/no/a successful.
stat(/var/spool/cron/.SEQ unsuccessful.


Now I hope somebody can test this and report if they have the same
problem.

Then there's the question: what's wrong? Is it a nfs problem? Can
anybody tell me?


- Start of st.c -
#include sys/stat.h
#include unistd.h
#include stdio.h

#define FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR /boot/no/a
#define FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR /var/spool/cron/.SEQ

main(void)
{
  struct stat statbuf;

  if(stat(FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR, statbuf) == 0
 )
  {
fprintf(stderr, stat(\ FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR \ 
successful.\n);
  }
  else
  {
fprintf(stderr, stat(\ FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR \ 
unsuccessful.\n);
  }

  if(stat(FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR, statbuf) == 0
 )
  {
fprintf(stderr, stat(\ FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR \ 
successful.\n);
  }
  else
  {
fprintf(stderr, stat(\ FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR \ 
unsuccessful.\n);
  }

  return(0);

}
- End of st.c -


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: ftp.debian.org mirror

[Martin Str|mberg]

In article [EMAIL PROTECTED] Timothy Phan wrote:
: Hi,
: 
:   I've been doing mirror of off the ftp.cdrom.com for a while now.
:   Yesterday,  I've just found out that this site was not up to today.
:   So I started to mirror from the ftp.debian.org.  I found out that 
:   there were a lot of files in the bo/ directory with difference in
:   the timestamp. i.e. On my system, fileXYZ is at 3:00AM and on
:   ftp.debian.org, the same would have 6:00AM.  Most of them were about
:   exactly 3 or 4 hours difference.  The minutes, date and size were all
:   the same.  Anybody knows what is going?

They are probably in different time zones. Run mirror once with -T,
when you are switching source.


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

alien: is it used to install or convert rpms?

[Martin Str|mberg]

Hello.

I was reading about running software on clusters of machines. There's 
(at least) two different ways: PVM and MPI.

Neither of them seems to be available as a debian package, but I have
found a rpm package of MPI.

Now my questions. Is alien supposed to install rpms, or to convert
them to debs, which can be installed? If it's converting them to debs,
or not and I have to repackage them, is there interest for a MPI
package?


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Powerpc: how long until installation is possible?

[Martin Str|mberg]

I've got a friend that has been trying to install Red Hat on his
Powerpc. No end of problems there. Anyway I though I'd check out the
status of debian for Powerpcs. And perhaps convert my friend...

Alas, there are no installation instructions, but there are a lot of
debs - hurrrm, how does that compute?

Do you think it is possible to crunch together something, that makes
it possible to install, or at least trying to install?


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Problem stat-ing nfs directories (atd and nfs mounted var partition, part 2)

[Martin Str|mberg]

Hello.


As some of might have read about a week ago, I thought I had a problem with
atd and it's pid file.

After recompiling at and hacking a little I've discovered that the problem is
in at or perhaps with nfs.

You see, stat-ing a file in a directory that is only readable for root from
a program that is setuid root, that is on a nfs mounted partition fails.

If you didn't follow me in that last sentence, I understand you, so here is
an example:

In the appended program, make sure that FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR
points to a LOCAL file that is in a directory which is only readable for root.
Likewise, make sure that FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR points to a file
that is in a partition that is NFS MOUNTED that is in a directory which is
only readable for root (/var/spool/cron/atjobs/.SEQ is perfect if you have
it NFS mounted).

Compile the program (gcc -o st st.c), chown daemon.daemon st and chmod
u+s st. Then run it as an ordinaty user.

At me it is like this (/boot is a local directory and /var is nfs mounted):
kant# ls -al /boot/no
total 2
drwx--   2 daemon   daemon   1024 Sep 22 00:54 .
drwxr-xr-x   4 root root 1024 Sep 22 00:52 ..
-rw---   1 daemon   daemon  0 Sep 22 00:54 a
kant# ls -la /var/spool/cron/atjobs/
total 6
drwx--   2 daemon   daemon   1024 Sep 22 00:58 .
drwxr-xr-x   5 root root 1024 Feb 19  1997 ..
-rw---   1 daemon   daemon  6 Sep 22 00:42 .SEQ
[Some other files (jobs) removed.]

Running this program as an ordinary user (NOT root) I get:
kant:/var/tmp ./st
stat(/boot/no/a successful.
stat(/var/spool/cron/.SEQ unsuccessful.


Now I hope somebody can test this and report if they have the same problem.
Then there's the question: what's wrong? Is it a nfs problem? Can anybody
tell me?


Wrong,

MartinS

- Start of st.c -
#include sys/stat.h
#include unistd.h
#include stdio.h

#define FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR /boot/no/a
#define FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR /var/spool/cron/.SEQ

main(void)
{
  struct stat statbuf;

  if(stat(FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR, statbuf) == 0
 )
  {
fprintf(stderr, stat(\ FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR \ successfu\
l.\n);
  }
  else
  {
fprintf(stderr, stat(\ FILE_IN_LOCAL_UNREAD_AND_UNEXEC_DIR \ unsuccess\
ful.\n);
  }

  if(stat(FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR, statbuf) == 0
 )
  {
fprintf(stderr, stat(\ FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR \ successful.\
\n);
  }
  else
  {
fprintf(stderr, stat(\ FILE_IN_NFS_UNREAD_AND_UNEXEC_DIR \ unsuccessfu\
l.\n);
  }

  return(0);

}
- End of st.c -


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: debian-disent

[Martin Str|mberg]

According to Arthur Jerijian:
 Er...I immediately unsubscribed from that list because of all the negative
 energy that it conveyed.  Although I can't recommend it, you can try
 sending subscription requests to [EMAIL PROTECTED], or if that
 doesn't work, try [EMAIL PROTECTED]

Yes, I suspected as much. Why do you think I want an archive? I thought that
it might good seeing what they are up to, though. No-no-no, no spying!


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: dpkg-source error ?

[Martin Str|mberg]

In article [EMAIL PROTECTED] Marc Fleureck wrote:
: Hi,
: I ran dpkg-source (dpkg-dev 1.4.0.8) on 3 files: *.diff.gz, *.dsc and
: *.orig.tar.gz, as you described in the readme (dpkg-source -x *.dsc).
: 
: Its says:
: 
:   Can't locate POSIX.pm in @INC at /usr/bin/dpkg-source line 8.
:   BEGIN failed--compilation aborted at /usr/bin/dpkg-source line8.
: 
: My system is Debian1.3/kernel 2.0.30.
: Note: dpkg-dev was installed without any errors.

Hmm. What's the output from ls -la /usr/lib/perl5/POSIX.pm?
Chez moi it's:
-r--r--r--   1 root root17806 Apr 22 06:42 /usr/lib/perl5/POSIX.pm

And version of package perl? I have:
Package: perl
Status: install ok installed
Priority: important
...
Version: 5.003.07-10
...


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: Compiling TWIN

[Martin Str|mberg]

In article [EMAIL PROTECTED] Paul Serice wrote:
: Has anyone compiled TWIN successfully on a Debian system?
: 
: I've followed the instructions (I'm pretty sure), but I
: keep getting errors that ldconfig couldn't find libtwin
: during the make process because it is trying to link with
: the library.  However, I can't link with the library
: until I create it, and I can't create the library, according
: to the make file, until I link it.  So it appears to me
: to be a catch-22.

Well, I don't know anything about twin, but usually you do an ar and
ranlib as the last step of creating a library. You don't need the
library to create the library - thus no catch-22.

Perhaps it's a Windoze term that you link the library, which might be
equvivalent to running ar and ranlib in Unix?

Does anybody else have any enlightingly comments?


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: configure network

[Martin Str|mberg]

In article [EMAIL PROTECTED] dada wrote:
: how is the script name that Debian uses to configure network?
: 
: ...the script that ask me about how configure network when I installed
: Debian 1.3.1 and that create the file /etc/init.d/network.

I dont think there is one. To find the files needing a change you can
do find /etc/ -type f -exec grep '192.168' {} /dev/null \;, where
the '192.168' part is a prefix of the network addresses you gave when
configuring the first time. To find that prefix have a peek in
/etc/init.d/network. 


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

debian-disent

[Martin Str|mberg]

I remember a message saying that a new list debian-disent had been
formed. Is that list archived somewhere, like the other debian lists
at http://www.debian.org/Lists-Archives/?


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: IP address and ipfwadm

[Martin Str|mberg]

In article [EMAIL PROTECTED] Lawrence wrote:
: I am using dialup PPP and need to run ipfwadm everytime I connect to my
: ISP.  I know that I can put ipfwadm into /etc/rc file.  What is the
: best/easy way to get the IP address other than using ifconfig or route
: (they are not good because the output is more than one line)?
: 
: ipfwadm -A -a -P My IP address go here -D 0/0

Look in /etc/ppp/ip-up!


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: 1st install, dselect error on at package, can't go further

[Martin Str|mberg]

[Posted and mailed. Eh - mailed and mailed.]

In article [EMAIL PROTECTED] you wrote:
: Subject pretty well says it all... even tried both mini and full
: installs.

Yes but not in any way near enough to make us able to help you: you
say what goes wrong but not the error messages.

: Installing from official Debian 1.3(.1) obtained recently from
: CheapBytes.
:
: How do I find out what's wrong with the package and bypass it?
:
: Richard Nelson

Considering the amount of information you are giving us, you could try
to run the Install part of dselect again (and again and again and
again; seriously this works if you're upgrading).

What's that (.1) part? Is it Debian 1.3.1 or not?
What is the file name of the at deb package?

If it looks like this, which is version 3.1.7-3 of at in Debian 1.3.1,
I've mananaged to install the package; but I was upgrading:

at_3.1.7-3.deb


To by-pass it, just deselect the at package in dselect, but as the package
is Priority: important you might have to de-dselect a lot of other packages.
I don't know what will happen, but if you're desperate you could try setting
the at package on hold and see what breaks.


Good luck,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: fdisk vs cfdisk to partition 3.1G drive

[Martin Str|mberg]

In article [EMAIL PROTECTED] you wrote:

[Klippa, klapp, kluppit fdisk vs cfdisk question.]

: Now I have /home on a very large separate primary partition, and /usr/local
: on another fairly large partition.  However, I linked /usr/src to the
: directory (in a logical partition mounted as /usr2) /usr2/src.  Is there
: something special I can do so this is transparent to the ls and dir
: commands?  When I type dir /usr/src I get a nice listing of the link, but
: when I type dir /usr/src/ a listing is printed.  This seems lit must be an
: FAQ, and I apologize, but I haven't run across it.

I don't know about dir but this behaviour from ls -al is a feature.

[Klippa, klapp, kluppit logical partition q.]

: Alan

I hope somebody else can answer your other questions, as I don't know.


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: Linux floppy disk problem

[Martin Str|mberg]

[Posted and mailed - erh! Mailed and mailed.]

In article [EMAIL PROTECTED] you wrote:
: I'm trying to install Debian Linux 1.3.1 and everything works fine until
: I try
: to make a boot disk from the install menu. The installation program
: prompts me to insert a blank floppy in the drive and I do so and press
: enter.
: The installation program then tries to format the disk and almost
: immediately exits with the following messages:
: 
: Verifying track 0, head 0 end_request: I/O error, dev 02:00, sector 17
: end_request: I/O error, dev 02:00, sector 17

[Klippa, klapp kluppit repeatingly errors.]

: It doesn't matter which disk I use, it gives the exact same messages. I
: have tried with different floppy
: disks, reformatted them in DOS and Minix first (no errors when
: formatting there) but it doesn't help.
: 
: I have configured my system according to the instructions in install.txt
: (disabled Shadow Ram etc.)
: 
: I haven't tried with a new blank non DOS/Minix -formatted disk yet,
: would it
: make any difference?

Oh - do I regocnise this! I stumble on the _exact_ problem. I didn't
have another choice than to skip the making of the boot floppy.

There's somthing fishy about superformat, but it seemed that only I
was having any problems with it. Beside you, now.

My problem is very repeatable, as soon as try to use superformat, I
normally get this error.

I seriously doubt it will work how many floppies you'd try, but you 
never know.

An alternative is to hold you breath and only do the make linux
bootable from hard disk and skip the make a boot floppy and
reboot and hope. It worked for me - YMMV.

When (if?) you finish installing try superformat -d /dev/fd0 (or
/dev/fd1 or whatever) and see if you wont get this problem at that
time as well.

To try to pin-point the problem: is it possible that you check what
kind of floppy drive you have? I checked but now all I remember is
that it was a Teac something.


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

atd and nfs mounted var partition

[Martin Str|mberg]

Hello.

I seem to have a problem with atd and its pid file (/var/run/atd.pid).
I have a client that mounts all directories from a server.
I can't find any differences except that at's pid file seems to be empty on
the client, while for example crond's pid file does contain something.

Does anybody have a clue?
Let me know if more inforamtion is neccessary.


Package: at
Version: 3.1.7-3

On the server:
descartes:~ ls -la /var/run/{crond,atd}.pid
-rw-r--r--   1 root root4 Sep 13 18:06 /var/run/atd.pid
-rw-r--r--   1 root root4 Sep 13 02:01 /var/run/crond.pid
descartes:~ ps -agux|grep atd
daemon 789  0.0  0.5   820   320  ?  SSep 13   0:00 /usr/sbin/atd
martin   10384  0.0  0.4   908   312  p0 S 22:520:00 grep atd
descartes:~ at now + 1 min
at EOT
warning: commands will be executed using /bin/sh
job 8 at 1997-09-15 22:54
descartes# ls -la /var/spool/cron/atjobs/
total 3
drwx--   2 daemon   daemon   1024 Sep 15 22:54 .
drwxr-xr-x   5 root root 1024 Nov  2  1996 ..
-rw---   1 daemon   daemon  6 Sep 15 22:53 .SEQ

On the client:
kant:~ ls -la /var/run/{crond,atd}.pid
-rw-r--r--   1 root root0 Sep 15 22:01 /var/run/atd.pid
-rw-r--r--   1 root root4 Sep 15 22:01 /var/run/crond.pid
kant:~ ps -agux|grep atd
daemon 110  0.0  1.0   828   336  ?  S 00:02   0:00 /usr/sbin/atd
martin 199  0.0  1.0   908   312  p6 S 00:54   0:00 grep atd
kant:~ at now + 1 min
Cannot access /var/spool/cron/atjobs: Permission denied
kant# ls -la /var/spool/cron/atjobs/
total 3
drwx--   2 daemon   daemon   1024 Sep 15 22:01 .
drwxr-xr-x   5 root root 1024 Feb 19  1997 ..
-rw---   1 daemon   daemon  6 Sep 15 22:53 .SEQ


Right,

MartinS


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . 
Trouble?  e-mail to [EMAIL PROTECTED] .

Re: How to migrate a Debian system to another hard drive?

[Martin Str|mberg]

Bruce Perens wrote:

[Klippa, klapp, kluppit]

   cd /old_partition
   tar cf - . | (cd /new_partition; tar xvlpf -)
 

Watch out so you don't run into an infinite loop:
Suppose that you mount the target disk on /mnt, then 
cd /
tar cf - . | (cd /mnt; tar xvlpf -) 
would copy some things to /mnt then copy /mnt/* to /mnt/mnt/ and so on.



Recursive,

MartinS

Re: lilo installation on IDE disk 500 megabytes

[Martin Str|mberg]

Hello.

 
 I am trying to install the 7-14-96 debian release on a machine
 with over 500 megabytes on an IDE hard disk.  I want to have a DOS
 partition and a linux partition.  At present fdisk shows:
 
 Device Boot   BeginStart  End   Blocks   Id  System
  /dev/hda111  356   179392+   6  DOS 16-bit =32M
  /dev/hda2   *  357  357  966   307440   83  Linux native
  /dev/hda3  967  967  99916632   82  Linux swap
 
 fdisk also displays a warning:
 
  The number of cylinders for this disk is set to 1025.
  This is larger than 1024, and may cause problems with:
  1) software that runs at boot time (e.g., LILO)
 
 The lilo documentation says:
 
 Note that large partitions that only partially extend into the forbidden 
 zone are still in jeopardy even if they appear to work at first, because 
 the file system does not know about the restrictions and may allocate disk 
 space from the area beyond the 1024th cylinder when installing new kernels. 
 

I think that your hda2 partition is under the 1024 cylinder boundary. Roar 
those of you debianites that disagree (let us know, i. e.)! 
Is there someway to check this with a program or can you be sure that when you 
have a partition taking up 16MB (the swap) at the end of a disk of this size
that that last cylinder (1025) is only used by that partition?

 Lilo suggests either using LOADLIN or booting from the DOS partition:
 
  In order to accomplish this, the DOS partition is mounted read-write, a 
  directory (e.g. /dos/linux) is created, all files from /boot are moved to 
  that directory, /boot is replaced by a symbolic link to it, the kernels are 
  also moved to the new directory, their new location is recorded in 
  /etc/lilo.conf, and finally /sbin/lilo is run.
 
I tried creating a subdirectory c:\linux in DOS, then from linux:
 
  mkdir ./dos
  mount /dev/hda1/linux ./dos
  cp /boot ./dos
 
 At this point, I got some warning or error messages.  It looks as though
 file names got truncated, and in some cases discarded completely.
 
 From ls /boot:
 
 System.map-2.0.6 chain.b  
 any_b.b  map  
 any_d.bmbr.b
 boot.0302  os2_d.b  
 boot.b vmlinuz-2.0.6
 
 From ls ./dos:
 
 any_b.bmap   
 any_d.b  mbr.b 
 boot.030   os2_d.b   
 boot.b   system.map
 chain.b
 

Yes FAT handles only 8.3 names as I'm sure you know. There is a file system
called umsdos that adds this functionality to FAT. But to use this I suppose 
that you would need it compiled into the kernel as we are talking booting 
here. Perhaps a little too inconvenient and wasting of space.

 I would appreciate some hand-holding at this stage.  Exactly
 what commands do I type to carry out the lilo instructions (specifically,
 the symbolic link and copying the kernel)?  Or would I be better off
 trying loadlin?
 

I saw that there was another reply as well. (S)He (I don't remember the name, 
sorry) had some good ideas as well, like the name of the kernel doesn't 
matter.

 Sorry this has been such a long post.  Thanks for your patience!
 
 


I hope my reach is long enough,

MartinS

Re: OS/2 HPFS File System - Is this a Bug? + a note for the list maintainers

[Martin Str|mberg]

Hello.

First a little note if you have seen this message already:
I'm having problems getting the mail go futher than the network the machine
I'm on, so sorry if you see this for the nth time.

[EMAIL PROTECTED] said:
 
 I'm running OS/2 Warp with HPFS on several of my drives.
 
 I noticed that Linux 1.1 fdisk reveals two different file system
 identifiers for these
 HPFS partitions:
 
   /dev/sda5  id 7  OS/2 HPFS
 
   /dev/hda2  id 17 Unknown
 
 
 The Unknown partition type (id 17) was created by OS/2 Warp fdisk during
 the installation process.  It is also a primary partition.
 
 The HPFS partition type (id 7) was created by OS/2 Warp after
 installation.  Note that this is an extended partition.
 
 Has anybody else observed the two different identifers for HPFS
 filesystems?  Is this a Bug?  Linux produces some error messages when
 mounting the id 17 filesystem but it everything seems to work ok.  I
 didn't observe any error messages when mounting the id 7 filesystem
 which also works fine.
 
 OS/2 Warp doesn't complain at all.
 
 

I already sent a mail about this but it didn't reach the debian list to my 
knowledge, so here I go again. Sorry if somebody gets this twice.

Yes. I have this behaviour in my system as well, although in my case it's
a FAT partition. I have two primary ones on my first hard disk, so one is 
always hidden (for DOZ and OS/2).

When I check out the partion types with fdisk under Linux I see that the 
partition that is considered hidden has id 16 in contrast to the one that 
isn't hidden, id 6. Then if I make the hidden one unhidden (and the other 
one hidden) the ids have changed places.


How about that,

MartinS

Re: which..

[Martin Str|mberg]

[EMAIL PROTECTED] (Guy Maor) said:
 
 Argh!  What a can of worms this has opened up.  I've tested which on
 several other systems and also looked at the source of the 2 examples
 posted.  The problem is that they all have DIFFERENT behavior.
 
 What do you do given two arguments, which ls cat, for example?  Print
 them both out?  Print only the first?
 
 What do you do if nothing is found?  Print nothing?  Print something?
 exit 1?
 
 I'll just pick some reasonable behavior.  I guess there is no real
 convention.  Real shells use type.
 
 
 Guy
 
 

Well as I have metioned earlier, IMHO which is from tcsh, so the correct
behaviour is what tcsh does. However I realise that different implementations
of tcsh might have different behaviour, so this might not solve the problem.

This is what happens at work, on a SUN running Solaris:
mpw wtree 78  which ls cat
ls:  aliased to ls -F
/usr/bin/cat
mpw wtree 79  which arne
arne: Command not found.


Hopefully helpful,

MartinS

Re: which..

[Martin Str|mberg]

On the subject on which, why not use:

-- which script for bash starts --
/bin/tcsh
# or where-ever tcsh is

which $*
# or what-ever prints all input parameters in tcsh
-- which script for bash ends --
 
As you all surely understand I'm not well versed in the writing of shell 
scripts, but hopefully you get the idea. 

I would just like to add, as which is originally from tcsh (IMHO), why not 
use tcsh to run which and we'll have the same behaviour in bash as in tcsh.


Suckled on tcsh as an two-month old baby,

MartinS