Re: Boot so slow it never completes, while Windows boots fine

[Marcin Owsiany]

On Thu, Apr 23, 2020 at 04:57:03PM +0200, Marcin Owsiany wrote:
> TL;DR: My laptop starts ~20x slower than normal. Booting Debian hangs
> before the kernel starts. Windows 10 boots slow, but then works fine.

Many thanks to everyone who responded! As usual, one can count on the
Debian community!

There is a new exciting (for me at least) development:

I noticed, that in order to prevent this slowness, I just need to have
some USB device plugged into any USB port while the laptop is starting!
It does not need to be used, its mere presence helps.

Also:
- it does not matter which USB port,
- it needs to be a device (I tried a flash drive and a headset, both
  help), a mere cable not connected to anything is not enough
- an SD card inserted into the built-in reader is not enough either,
- once booting nice and fast, it still goes slow as soon as I disconnect
  the device. The opposite is not true (while booting slow, connecting
  something has no effect).

This seems to point blame at the USB controller? Perhaps it's flooding
the CPU with interrupts for whatever reason?

Some responses to the suggestions posted:


David, thanks for your suggestion to try booting from a flash drive,
this let me find the workaround :-)


David and Kent, the computer is sealed (still on warranty) so I'd rather
not take it apart just yet. It expires in a few days so I might try this
soon, although USB controller does not seem to be something
disconnectable easily (or rather, not connectable back easily :-)


Alexis, as for memtest-ing, it seems like neither memtest86 nor
memtest86+ seems to be able to start on this system, even with secure
boot off. Perhaps these are not UEFI-compatible?

Speaking of secure boot, I wonder why it keeps showing:

 Secure Boot Status: Disabled

even though it's turned on:

 Secure Boot: Enabled

I think both were enabled initially, but this changed around the time I
upgraded BIOS.


I did run the Lenovo diagnostics test tool for Windows though. It took
several hours and the only warning had something to do with the RTC. The
test was not eager to say what was wrong, it just displayed a warning
sign next to "RTC".


Alexander, I did install the latest available BIOS update. That did not
help. I also reset the settings, to no effect either. FWIW, the model
name is Lenovo ideapad FLEX 5-1570 (81CA0010US). More details are on the
Information tab of the BIOS SETUP visible in the video.


kind regards,
-- 
Marcin Owsiany   http://marcin.owsiany.pl/
GnuPG: 4096R/CDFB68E9  59F4 A7DE D37D 95C5 8939  FDDB EFDE D44B CDFB 68E9

Boot so slow it never completes, while Windows boots fine

[Marcin Owsiany]

TL;DR: My laptop starts ~20x slower than normal. Booting Debian hangs
before the kernel starts. Windows 10 boots slow, but then works fine.
Hardware problem?

More details:

I bought a Lenovo IdeaPad Flex 5 in April 2019. It came with Windows 10, I
installed Debian buster on it (dual-boot with GRUB 2). Worked flawlessly
(with secure boot!) for ~8 months.

In January 2020 after some reboot for the first time it started slow, and
hung while booting Debian. Windows boots slow too, but once it's up it
works just fine. BIOS SETUP UI is also very slow.

Since it might not be obvious what I mean, here is a recording, with
detailed timeline in video description:
https://youtu.be/HCgO9UblqtI

After a few reboots it came back to normal. Then this effect came and went
a few times, and now it's here for good. I cannot boot Debian at all. I
tried leaving it booting overnight once and it didn't show any progress.
Windows still works OK. Updating BIOS to the most recent version did not
help.

I asked on the Polish Lenovo support forum
,
the response was that "this model does not support dual boot (sic!), they
have had no similar reports in the past, and installing Linux on it might
have caused the problem". I'm shocked.

Initially I thought this might have something to do with Windows updates,
because on the first occasion it seems to have disappeared after Windows
completed its scheduled update. But now I think it was just a red herring.

In my almost 30-year experience I have not encountered a problem which does
not go away after a cold reboot, but does go away after Windows starts :-O

My only theories now are:
- a hardware problem (but why does it go away once Windows boots?)
- a botched CPU microcode update (but I suppose there are checksums which
would prevent it from happening, so not likely).

Any thoughts?

Marcin

Re: Wheezy: pressing power button initiates a suspend in parallel with system halt

[Marcin Owsiany]

2014-07-28 16:58 GMT+02:00 Marcin Owsiany :

> I'm running stable on an amd64 PC since wheezy release. Until recently, I
> was able to nicely suspend and resume the system by pressing the hardware
> "power" button on the chassis.
>
>
> After a recent update (not sure which packages are involved, as a lot of
> time has passed since the last update on that machine), I noticed that when
> I press the power button the system starts to suspend, but at the same time
> a shutdown is started - I can get a glimpse of the message:
>
> Power button pressed.
>
> as it appears on all terminals just before the screen blanks. As a result
> of this, as soon as the system completes the resume, it goes on and
> continues with the shutdown and the machine powers off.
>
> I'm using GNOME classic session.
>
> How do I debug this? Where do I begin?
>

In case someone finds it useful:

I ended up searching for "Power button pressed" on the internets and with
rgrep in /etc, found /etc/acpi/powerbtn-acpi-support.sh, and from there
/usr/share/acpi-support/policy-funcs, ran the various commands in
CheckPolicy by hand, and basically independently discovered
http://bugs.debian.org/755969 :-)
That bug has the 3-character fix.

Marcin

Wheezy: pressing power button initiates a suspend in parallel with system halt

[Marcin Owsiany]

I'm running stable on an amd64 PC since wheezy release. Until recently, I
was able to nicely suspend and resume the system by pressing the hardware
"power" button on the chassis.


After a recent update (not sure which packages are involved, as a lot of
time has passed since the last update on that machine), I noticed that when
I press the power button the system starts to suspend, but at the same time
a shutdown is started - I can get a glimpse of the message:

Power button pressed.

as it appears on all terminals just before the screen blanks. As a result
of this, as soon as the system completes the resume, it goes on and
continues with the shutdown and the machine powers off.

I'm using GNOME classic session.

How do I debug this? Where do I begin?

-- 
Marcin Owsiany   http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216

"Every program in development at MIT expands until it can read mail."
-- Unknown

Wireless USB webcam for Debian?

[Marcin Owsiany]

Hi,

I'm looking for a webcam that does not have a cable so I can move it
around the room, but can still can be connected to a PC over USB (i.e.
NOT a standalone WiFi/ethernet webcam). Does anyone know any that works
with Linux, preferrably Debian stable?

I've managed to find this one - Konig electronic CMP-WEBCAM100
http://www.konigelectronic.com/en_us/55839440
but of course the manufacturer only mentions Windows.

regards,
-- 
Marcin Owsiany  http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20111017135843.GO3796@beczulka

Re: Admit that the typical Debian machine has tons of cruft(8)

[Marcin Owsiany]

I'm not sure if this post is serious, but assuming that it is:

On Fri, Apr 24, 2009 at 09:22:12AM +0800, jida...@jidanni.org wrote:
> Admit that the typical Debian machine has tons of cruft(8)
> $ man cruft
>   cruft - Check the filesystem for cruft (missing and unexplained files)

It's more correct to admit that cruft is incorrectly reporting things as
cruft, rather than Debian being full of of it. See e.g. #522108

-- 
Marcin Owsiany  http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Dell PowerEdge 2950 server with PERC 5/i RAID controller

[Marcin Owsiany]

On Mon, Oct 29, 2007 at 12:13:37AM +0100, Andraz Sraka wrote:
> On Sun, 2007-10-28 at 10:49 +0000, Marcin Owsiany wrote:
> > afaapps_2.6-3_i386.deb from http://www.brandl.net/pe2550/ works for me
> > on the etch kernel. Also, the kernel logs any problems that appear, so
> > logcheck will do for healh monitoring.
> 
> Are you sure that this tool works with PERC 5/i?! I think it was written
> for older version PERC controllers or I am typing wrong command in? :(

Oops, I checked properly now, and you are probably right. I have a
different controller :-/

Sorry for the confusion.

-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


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

Re: Dell PowerEdge 2950 server with PERC 5/i RAID controller

[Marcin Owsiany]

On Sat, Oct 27, 2007 at 06:19:54PM +0200, Andraz Sraka wrote:
> instance) for monitoring Dell PERC 5/i RAID controller. I found that
> there are some utilities for RHEL/SuSE enterprise distribution. Has
> anyone managed to see status of controller in Debian? What tools do I
> need to use/install and are any kernel patches required? It would be
> nice to see raid status as also the faulty disks :)

afaapps_2.6-3_i386.deb from http://www.brandl.net/pe2550/ works for me
on the etch kernel. Also, the kernel logs any problems that appear, so
logcheck will do for healh monitoring.


-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


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

Compaq R1500H UPS with Debian?

[Marcin Owsiany]

Hi,

Has anyone had any success using Compaq R1500H UPS with Debian?
Information on the internet suggests that you can connect it to the
computer using a serial cable, but I cannot find any information about
management software for Linux for this UPS..

thanks,

-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


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

Block read "resulted in short read" when dumping root fs, but all else is well

[Marcin Owsiany]

Hi all,

The box is a debian sarge, standard Debian kernel. Every morning I run
dump on all filesystems, level 0 on Sundays, 1 on Mon and Thu, 2 on Tue
and Fri, and 3 on Wed and Sat. I do this on live filesystems, mounted
rw. All is well except occasionally (two times over the last four weeks
- once on Friday and once on a Wednesday two weeks later) I get the following
error when dumping the root partition:

[...]
|   DUMP: Dumping volume 1 on /srv/backup/2006-07-26_Wed_level_3/_001
|   DUMP: Volume 1 started with block 1 at: Wed Jul 26 06:25:08 2006
|   DUMP: dumping (Pass III) [directories]
| /dev/mapper/lv-pv: Attempt to read block from filesystem resulted in short 
read while reading inode #2
| 
|   DUMP: error reading command pipe: Connection reset by peer
|   DUMP: Broken pipe

and dump exits with code 3

I could not find this message in dump source, so this is probably a libext2fs
message, which all the more probable that most finds on Google are reports of
running fsck on some seriously mangled filesystems.

However my filesystem seems OK. I haven't run fsck on it yet, but the fact that
the next dump (on the subsequent day) runs just fine, suggests that this is not
an error in the filesystem.

The strange thing is that inode #2 is the root directory itself, and stat
reports:

|   File: `/'
|   Size: 1024Blocks: 2  IO Block: 4096   directory
| Device: fd01h/64769dInode: 2   Links: 21
| Access: (0755/drwxr-xr-x)  Uid: (0/root)   Gid: (0/root)
| Access: 2006-07-30 14:41:03.845422864 +0200
| Modify: 2006-07-13 09:12:30.663241896 +0200
| Change: 2006-07-13 09:12:30.663241896 +0200

Which means it wasn't modified nor changed (I never remember the difference)
neither around the time dump ran, nor after that.

Does anyone have a clue on what might be happening?

Marcin
-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


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

lm-sensors support for SE7501BR2 ?

[Marcin Owsiany]

Hi!

[ sorry for the cross-post, but both lists seem relevant ]

I have an Intel SE7501BR2 server motherboard, and using lm-sensors
2.6.3-5+ only detects successfully four chips like this: (using eeprom
driver)

  * Bus `SMBus I801 adapter at 0580' (Non-I2C SMBus adapter)
Busdriver `i2c-i801', I2C address 0x51
Chip `Serial EEPROM (PC-100 DIMM)' (confidence: 8)

But does not find any thermal sensors. Has anyone had more luck with
such hardware? The manual says the sensors are managed by Heceta chip
U5F10, for which google finds no good hits.

Marcin
-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


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

Intel SRCU42X SCSI RAID contoller

[Marcin Owsiany]

Hi!

Forgive me the cross-post, but this is rather urgent for me :-/

Does anyone know if the Debian kernel in woody-proposed-updates (2.4.22)
supports Intel SRCU42X SCSI RAID contoller?

Intel's web page says that it is supported by Suse and RH, but they make
a binary driver available for download (megaraid.o). The source is
included, so probablu it the same as in stock kernel, but could anyone
confirm this?

regards

Marcin
-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


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

Re: xinetd /etc/host.deny ALL:PARANOID

[Marcin Owsiany]

On Fri, Jan 11, 2002 at 12:11:13AM +0100, martin f krafft wrote:
> it's not really a security measure anymore, i find. feel free to
> disagree...

Disabling PARANOID mode only means that you shouldn't trust the logged
hostnames, because thay may be faked, no?

Marcin
-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216

procmail's child killed?

[Marcin Owsiany]

Hi!

Recently I got this error message:

procmail_pipe transport failed: Child process of procmail_pipe
transport (running command "/usr/bin/procmail -d
${local_part}") was terminated by signal 9 (Killed)

1. Does procmail_pipe call fork() before execing procmail?
(That would mean that procmail itself was killed, not the
process ran by procmail from .procmailrc)

2. Does anyone have a clue about what and why could kill it?
The cmd ran from .procmailrc is a few simple echos and formail
to pipe a few lines into $SENDMAIL (sending an SMS message) The
particular mail is only composed of an attachment (30KB).

Marcin
-- 
Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216

Re: Debian installation screens (cont)

[Marcin Owsiany]

On Mon, Jan 29, 2001 at 01:44:42PM -0200, Paulo Henrique Baptista de Oliveira 
wrote:

> I forgot the mail. It is something like get boot-floppies
> package and use dbootstrap. I tried to do a "make release"
> and the following error occurs:

> 
> find . -name \*~ | xargs rm -f
> make resc1440.bin resc1440-s.bin
> find: /archive/debian/Incoming/: No such file or directory
[...]

Read the documentation in the root directory of boot-floppies
tree - you need to edit the config file to contain a pointer to
your local archive mirror.

(Yeah, this is a PITA :-( )

regards

Marcin
-- 
Marcin Owsiany <[EMAIL PROTECTED]>
http://student.uci.agh.edu.pl/~porridge/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216

LILO and 2 disks

[Marcin Owsiany]

Hi *!

Does anybody know a checked way to do the following?

hda1 and hdb1 have the same files structure. hdb1 is mounted under /mnt
How to make LILO boot after removing hda completly and connecting hdb as
primary master (so it becomes hda)?
I know I can make a boot floppy, switch disks, boot from floppy and then run
"lilo", but is there a way to make without the floppy? I tried

cd /mnt
sbin/lilo -r /mnt -b /dev/hdb1

But that complained about hdb (or hdb1 - I don't remember now) not being on
the first disk and did not work - I got 

LI 01 01 01 01 01 01 01 01 01 ..

after reboot.

thanks in advance

Marcin

-- 
++ The reason we come up with new versions
|Marcin Owsiany  | is not to fix bugs. It's the stupidest
|[EMAIL PROTECTED]| reason to buy a new version
++ I ever heard.- Bill Gates

Re: package compilation madness

[Marcin Owsiany]

On Wed, Feb 09, 2000 at 11:04:27AM -0800, Chris Waters wrote:
> Drew Parsons <[EMAIL PROTECTED]> writes:
> 
> > Running on the theory that it might help speed up my Pentium II system
> > (debatable, but that's beside the point), I've tried
> > recompiling a selection of packages with pentium optimisation, downloading
> > the debian source and running `./debian/rules binary`. 
> 
> > However, after having done this, every time I `apt-get upgrade`, these
> > packages get downloaded from the debian ftp server
> 
> Ok, 1) this isn't really a laptop issue, and 2) yes, you're correct.
> Apt will assume you want the most up-to-date package, and, since the
> package on the ftp server is newer than the one you built, it gets
> preferred.  The easy solution is to put the package on hold -- but
> this has the disadvantage that you *don't* get newer versions at all.
> 
> Actually, it's not really about "newer", it's about having a
> version number that sorts later, 

not really
Once I compiled 'mirror' program myself (i had to apply a patch) and
installed the resulting package. I don't remember what version it was, so
let's assume it was version 'X'.

After doing a apt-get upgrade apt downloaded and installed the debian
'mirror' package even though it had _the same_ version 'X' (it was stable
dist) _and_ it was older (I compiled my patched version _after_ the debian
package was available).


regards
Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Kernel upgrades = security upgrades - a possible solution?

[Marcin Owsiany]

On Wed, Sep 29, 1999 at 05:24:54PM +0300, Martin Fluch wrote:
> On Wed, 29 Sep 1999, Marcin Owsiany wrote:
> 
> > I guess this kind of kernel packages would be for people quite concerned
> > about security but also quite lazy :)
> 
> I guess, this is mutual exclusive. People which are lazy will leave many
> (and I think also bigger) security holes some where else on the system, so
> that it won't matter, if you keep your kernel that much secure...

well, yes you are right.
:)
I guess i didn't really think of it before writing :(

> > Also if you administer a lot of boxes, and if they work ok with the default
> > kernel you will find it _a lot_ more convenient to automatically upgrade
> > kernel than to compile it for each box...
> 
> Ever considerd the package 'kernel-package'. This makes out of any kernel
> source debian packages, which then can be installed with dpkg, apt-get or
> what ever ... 

sure, since i had discovered it, i've never made a kernel without using it.
But still you have to make the kernel, and if you compile it, you can't
resist tweaking it to each particular system's needs, can you? :)

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Kernel upgrades = security upgrades - a possible solution?

[Marcin Owsiany]

On Wed, Sep 29, 1999 at 02:42:38AM -0700, Seth R Arnold wrote:
> On Wed, Sep 29, 1999 at 10:27:43AM +0000, Marcin Owsiany wrote:
> > On Tue, Sep 28, 1999 at 09:41:26PM -0500, Ashley Clark wrote:
> > > On Tue, 28 Sep 1999, Marcin Owsiany wrote:
> > > > the way to solve the problem would be to create a package called e.g.
> > > > "secure-kernel", which would depend on the most secure 
> > > > "kernel-image-".
> > > > Then if the security team has newer kernel with security bugfixes, they
> > > > would make a new version of "secure-kernel" which would depend on the 
> > > > fixed
> > > > kernel.
> > > 
> > > I, for one, wouldn't want my kernel upgraded automatically, no matter
> > > what the fixes involved are. Here's why: I have compiled my own
> > > kernel with my hardware selected (sound, tape drive, scsi card,
> > > network card) and Debian simply can't afford to make all possible
> > > combinations of kernel configurations to provide an easy upgrade path
> > > for users. Now, possibly there could be some kind of secure-kernel
> > > package which would do nothing more than simply inform you during
> > > upgrade that a newer kernel with such-and-such security patches is
> > > available and recommend how to upgrade, that's seems more reasonable
> > > to me at least.
> > 
> > That is the point of this idea. If you want your kernel to be upgraded
> > automatically, you install secure-kernel, if you only want to be informed,
> > you install secure-kernel-info, if you don't care at all, you instal
> > neither.
> 
> I am still very leery of automatic kernel updating... I do rather like the
> idea of secure-kernel-info, as Marcin has described it, but it needs a
> better name; secure-kernel just won't do it. kernel-update-watcher perhaps.

but of course, i know the names need improving

> However, if security is enough of an issue for you that you think a kernel
> package should be made around it, maybe you should keep an eye on bugtraq
> and freshmeat, or a cron-job to grab the LATEST-VERSION-IS file from the
> kernel.org servers -- no matter which approach is taken, it will be faster
> than waiting for a new kernel package to come along...

I guess this kind of kernel packages would be for people quite concerned
about security but also quite lazy :)
Also if you administer a lot of boxes, and if they work ok with the default
kernel you will find it _a lot_ more convenient to automatically upgrade
kernel than to compile it for each box...
Just my 0.02

Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Kernel upgrades = security upgrades - a possible solution?

[Marcin Owsiany]

On Tue, Sep 28, 1999 at 09:41:26PM -0500, Ashley Clark wrote:
> On Tue, 28 Sep 1999, Marcin Owsiany wrote:
> > the way to solve the problem would be to create a package called e.g.
> > "secure-kernel", which would depend on the most secure "kernel-image-".
> > Then if the security team has newer kernel with security bugfixes, they
> > would make a new version of "secure-kernel" which would depend on the fixed
> > kernel.
> 
> I, for one, wouldn't want my kernel upgraded automatically, no matter
> what the fixes involved are. Here's why: I have compiled my own
> kernel with my hardware selected (sound, tape drive, scsi card,
> network card) and Debian simply can't afford to make all possible
> combinations of kernel configurations to provide an easy upgrade path
> for users. Now, possibly there could be some kind of secure-kernel
> package which would do nothing more than simply inform you during
> upgrade that a newer kernel with such-and-such security patches is
> available and recommend how to upgrade, that's seems more reasonable
> to me at least.

That is the point of this idea. If you want your kernel to be upgraded
automatically, you install secure-kernel, if you only want to be informed,
you install secure-kernel-info, if you don't care at all, you instal
neither.

regards

Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Installing debian to another hard drive

[Marcin Owsiany]

On Tue, Sep 28, 1999 at 10:42:15PM +0200, Hugo van der Merwe wrote:
> Hello,
> 
> How can I install the debian "base" system to another hard drive in the
> same computer? I have a working Debian / DOS+Win95 dual boot system, and
> would like to install Debian to a laptop, whose floppy disk drive is
> disfunctional, and has no CD-ROM drive.
> 
> I have connected the laptop's hard drive to my Debian machine, using an
> adapter. There must be some way to install Debian to this hard drive,
> using my desktop machine. How can I do this? Will the normal
> install-from-dos-partition procedure work?
> 
> Thanks in advance,
> Hugo van der Merwe
> 
> ps. I would appreciate if replies could be CC:'ed to [EMAIL PROTECTED]
> I am not subscribed to the list, and sometimes my ISP's news server messes
> up...

could you please forward a solution, if you find one, to me?
i also have a laptop with a broken floppy drive. :(

regards,
Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Flat-panel monitors?

[Marcin Owsiany]

On Tue, Sep 28, 1999 at 05:57:27PM -0400, B. Szyszka wrote:
> Hi,
> 
> I was wondering what luck, if any, you guys have had with flat panel
> monitors in Debian. Has everything worked as it should?

not right from the start...
X docs are your friends :)
don't ask me where have i found the right info, because i don't remember :(

regards,
Marcin

-- 

------------
Marcin Owsiany
[EMAIL PROTECTED]

Re: Kernel upgrades = security upgrades - a possible solution?

[Marcin Owsiany]

On Tue, Sep 28, 1999 at 05:05:21PM -0500, Brian Servis wrote:
> *- On 28 Sep, Fraser Campbell wrote about "Re: Kernel upgrades = security 
> upgrades"
> > Brian Servis wrote:
> > 
> >> Notice that the version is part of the package name.  Thus a
> >> kernel-image-2.0.34 and kernel-image-2.0.36 are two totally different
> >> packages as far as Debian is concerned, except that they both provide
> >> the virtual package kernel-image and that fact is not determined until
> >> it is being installed.
> > 
> > Ok.  To my way of thinking it should be called kernel-image_2.0.34,
> > kernel-image_2.0.36-3, etc.  That way apt-get upgrade would grab updated
> > kernels for the user.  IMO kernels are a very critical part of security and
> > that they should be upgradeable as part of the normal process.
> > 
> > I realize the kernel is a very special piece of software but still see no
> > reason why it is treated differently from normal software.  Perhaps the
> 
> If kernel-images did not have the version in the package name then you
> could not have two different versions of the kernel installed at the
> same time.  There are instances where having two different kernels is
> needed.  Such as for development machines where code must be checked
> under various kernel versions.  This is especially true for the case
> when a new stable kernel branch is released and not all programs are
> compatible with the newer kernel, such as was the case with 2.0.x and
> 2.2.x kernels. 
> 
> 
> > upgrade process depends on the virtual package kernel-image which I don't
> > seem to have installed?
> > 
> 
> The Debian-policy best describes how virtual packages work:
> 
> 2.3.5. Virtual packages
> ---
> 
>  Sometimes, there are several packages doing more-or-less the same job.
>  In this case, it's useful to define a _virtual package_ who's name
>  describes the function the packages have. (The virtual packages just
>  exist logically, not physically--that's why they are called
>  _virtual_.) The packages with this particular function will then
>  _provide_ the virtual package. Thus, any other package requiring that
>  function can simply depend on the virtual package without having to
>  specify all possible packages individually.

the way to solve the problem would be to create a package called e.g.
"secure-kernel", which would depend on the most secure "kernel-image-".
Then if the security team has newer kernel with security bugfixes, they
would make a new version of "secure-kernel" which would depend on the fixed
kernel.

any objections?

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: upgrade 1.3 -> 2.1 question

[Marcin Owsiany]

On Tue, Sep 28, 1999 at 01:57:44PM -0600, Bruno Melli wrote:
> 
> Hi Folks,
> 
> I just upgraded (from CD) my 1.3 system to 2.1. 
> After installing 1.3 a couple of years ago I haven't done much
> sysadmin so I'm a little rusty when it comes to kernel
> configuration...
> 
> Everything seems to have updated OK but I'm puzzled that the machine 
> was never rebooted during the upgrade. 
> What is the mechanism to upgrade the kernel ? (I haven't
> kept up to speed with kernel version but I assume 2.1 uses a newer
> kernel than 1.3 did...)
> 
> Does it mean that I have to upgrade the kernel sources and rebuild one
> (as well as making a new boot disk ?)

no, just install newest slink kernel-image-x.x.xx deb package, configure
your lilo.conf to allow you boot both your present (in case something goes
wrong) and the upgraded kernel, rerun lilo and reboot.

that's it :)

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: libgnome32 1.0.16 where ?

[Marcin Owsiany]

On Mon, Sep 27, 1999 at 01:02:34PM +0200, J.H.M. Dassen Ray" wrote:
> On Mon, Sep 27, 1999 at 11:32:01 +0200, Ingo Reimann wrote:
> > what happened with gnome? since friday, some packages require ( for

> These library packages are currently stuck in Incoming; get them from an

I just want to ask: what is the whole Incoming thing? why does it take for
the packages to go from Incoming to usual archive? Aren't they moved
automatically? If not, then why does it need to be done manually?

Marcin

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Remote Login

[Marcin Owsiany]

On Sun, Sep 26, 1999 at 11:01:43AM -0700, j way wrote:
> Hi, again.
> I tried the suggestions provided(& tnx) but without change. My login
> looks like this at the remote terminal:
> 
> Debian GNU/Linux 2.1 Struggles ttyS1
> Struggles login: jway
> Password: (keyed in)
> (time passes.)
> Login timed out after 60 seconds.
> Any more suggestions?  Thanks. -John.

Did you try looking at your logs?

Marcin

-- 

--------
Marcin Owsiany
[EMAIL PROTECTED]

Re: Inchon International Airport Information

[Marcin Owsiany]

On Thu, Sep 23, 1999 at 09:10:03AM -0600, eric drayer wrote:
> hey stop that spam 

You may not say so!
This is because Debian approves spam on its mailing lists. This gives mooney
to the Debian project.
:)

Marcin

-- 

----
Marcin Owsiany
[EMAIL PROTECTED]

Re: Gateway hangs on accesses from LAN

[Marcin Owsiany]

On Thu, Sep 23, 1999 at 11:50:07PM +1000, Andrew J.F. Clark wrote:
> I have a problem with my IP Masq gateway (slink, 2.0.36) hanging during
> accesses via the LAN.  What happens is that whenever one of the machines
> on the LAN (NT4, Win98 or Debian Slink) accesses anything on the gateway a
> couple of times, the gateway freezes totaly and won't respond to anything.
> 
> Currently the gateway has the following hardware:
> 
> intel P200 MMX 
> 32mb RAM
> s3 pci video card
> NE2000 Isa network card
> SB Vibra 16 Isa PNP sound card
> Unknown MB (M752 is the Model, Intel i430TX chipset)
> Generic ATAPI Cdrom
> 1gb Seagate HDD
> Mitsumi 1.44mb FDD
> 
> and the following software:
> debian 2.1 base
> apache-ssl
> proftpd
> qpopper
> squid
> distributed-net client
> uptime daemon
> linuxlogo
> ssh
> mysql (server, client, docs etc)
> xringd
> pine
> mutt
> pgp5i
> pgpi
> gnupg
> python (base, net, misc)
> fetchmail
> minicom
> bind
> irquery
> linbot
> 
> I assume that the problem has to be the gateway machine as the other 3
> machines all work fine.
> 
> If anyone has any ideas as to what could be causing this, please mail me
> at [EMAIL PROTECTED]

what about the logs? try increasing verbosity level of the daemon
corresponding to the service the other machines are accessing and monitor
the logs with "tail -f "
Also you don't mention what services are causing trouble... any of them? If
any, than this is probably a hardware issue.

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: awk or sed?

[Marcin Owsiany]

On Thu, Sep 23, 1999 at 12:12:57PM +1200, Tim Thomson wrote:
> Hi,
> 
> I want to write a script, that gets the local ip address, and does a
> reverse ns lookup, then sets the hostname to the dns response.
> 
> I can use grep to get the line I want from ifconfig and nslookup. But how
> do I get the numbers and name out of the output? Was it awk or sed, or am
> I going along the wrong track?

if it is not too complicated, try "cut"

hope this helps

Marcin


-- 

--------
Marcin Owsiany
[EMAIL PROTECTED]

Re: hda: irq timeout:

[Marcin Owsiany]

On Tue, Sep 21, 1999 at 04:08:29PM +0200, J Horacio MG wrote:
> Hi,
> 
> this just happened to me while while the daily cron was running, I tried
> a connection:
> 
> h0rus:~$ pon provider
> hda: irq timeout: status=0x50 { DriveReady SeekComplete }
> h0rus:~$
> 
> and the connection request was resumed, and got back to the prompt.  I
> recall this is just got to do with the system not being able to read the
> disk for a moment but, is it normal that it's echoed on screen?
> 
> I got no messages from cron though.  Should I worry?

I don't think so, this is normal since this is not a program message, but
direct kernel message, so it should go to klog and to the console, if
console log level is high enough.

I would worry however about the cause of the message... it could be nothing,
but could also mean that your hard drive's life is coming to an end...
BTW such messages begin showing up on my console more frequently as well.

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: plip problems ahoy

[Marcin Owsiany]

> 
> > 
> > > Saturn also acts as a gateway for IP-Masqerading, and so
> > > already has support for IP forwarding. I don't want to compile support
> > 
> > I think that's the point. If I were you, I would check if earth isn't
> > masqeraded, because the symptoms are like it is.
> 
> I'm not sure what you mean by this. Earth is masqueraded when saturn's ppp 
> link
> is up. I can't see how the masquerading would make any difference though...
> saturn is the gateway for both earth and comet, and should be forwarding 
> packets
> between the eth0 and plip0 interfaces.

are you sure it does masquerading both for 192.168.1.x and 192.168.2.x and
not just for 192.168.1.x?
I'm not sure, but i think the most basic masquerading setup does
masquerading of the local network for _all_ other hosts. And you have to
remember that a masqueraded host canot be pinged nor connected to in any
way! (as the name suggests, unless you have port forwarding rules to enable
this)

> 
> > However I'm not sure
> > because you have provided no informaion about your other boxes, other
> > saturn's interfaces and the masquerading setup???!
> 
> All of saturn's interfaces are on the diagram, except for the ppp link which 
> is
> a dialup internet connection. Saturn masquerades this connection to earth and
> any other computer on 192.168.1.x.
> 
> As it happens, last night I managed to get it working. I just don't know how 
> it
> happened... I turned everything on, and suddenly I was able to telnet and ftp
> from comet to earth (I did not change any settings, and used the same scripts 
> I
> had previously written to bring up the plip link). I managed to install the
> packages I wanted, after fiddling with anonftp on earth and apt.sources on
> comet. Curiously, later on the plip link once again refused to work. Could the
> order that things initialise be influencing plip?

I repeat, i know nothing about things specific to PLIP, but the only thing
that would change the situation seems to be some modification of routing
information. (read: i still suspect the masquerading - wasn't it when the
modem connected when the plip refused to work?)
Also: if A side of a PLIP link sees the B side but not the other way round,
it's not the fault of PLIP link. It's the routing information.

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: telnet to my machine (cont)

[Marcin Owsiany]

On Tue, Sep 21, 1999 at 05:22:43PM +0200, Manuel Arenaz Silva wrote:
> Hello,
> 
> I have taken a look at the file "/var/log/syslog", which registers the
> same information every time a telnet conection is refused:
> 
> Sep 21 17:11:49 cambados in.telnetd[8102]: connect from 193.144.50.23
> Sep 21 17:11:49 cambados in.telnetd[8102]: error: cannot execute
> /usr/sbin/in.telnetd: No such file or directory
> 
> Efectively, the file "/usr/sbin/in.telnetd" doesn't exist. How can I
> solve this problem?

Well, it would certainly help us if you wrote what is wrong about this
situation, i.e. if you actually _want_ the telnet service enabled or _not_?
 - if you want telnetd running, then install package telnetd
 - if not, then comment out in /etc/inetd.conf the line with in.telnetd -
the messages will disappear

BTW: if you want telnet, i would suggest switching to SSH - it's more secure
and has a few nice things about it.

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: plip problems ahoy

[Marcin Owsiany]

On Mon, Sep 20, 1999 at 01:04:00PM +1000, Matthew Dalton wrote:
> Hi there,
> I've been trying to install Debian 2.1 on my
> laptop, which has no cdrom or network card. Debian's base system was
> easy to install (using floppies). I've been trying to set up a PLIP
> link, so I can ftp the rest of the system over.
> 
> My setup is unusual:
> 
> +---+
> | laptop: comet |
> +---+
>|  192.168.2.2 (plip1)
>|
>  laplink cable
>|
>|  192.168.2.1 (plip0)
> +--+
> | headless box: saturn |
> +--+
> |  192.168.1.1 (eth0)
> |
> ethernet
> |
> |  192.168.1.13 (eth0)
>++
>| desktop: earth |
>++
> 
> 
> comet is running Debian 2.1 (kernel 2.0.36, glibc 2.0.7)
> saturn is running RH5.2 (kernel 2.2.10, glibc 2.0.7)
> earth is running RH5.2 (kernel 2.2.6, glibc 2.0.7)
> 
> Earth is the only computer which has a cdrom drive. I wish to connect
> comet to saturn though, because PLIP requires that IP forwarding be
> enabled in the kernel (according to the PLIP mini-HOWTO... is it
> correct?).

I know almost nothing about PLIP, so I can't tell you if it's required, but
if it is similar to PPP, then it isn't if you just want to connect _two_
boxes with a cable. However, your situation requires IP forwarding simply
beacause you have to forward packets from earth to comet. It would be
necessary in any case provided earth and comet don't have a direct link.

> Saturn also acts as a gateway for IP-Masqerading, and so
> already has support for IP forwarding. I don't want to compile support

I think that's the point. If I were you, I would check if earth isn't
masqeraded, because the symptoms are like it is. However I'm not sure
because you have provided no informaion about your other boxes, other
saturn's interfaces and the masquerading setup???!

Please do so that we can say something more.

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Compilation prob

[Marcin Owsiany]

On Mon, Sep 20, 1999 at 01:53:36PM +0200, Menno Scholten wrote:
> Hi there,
> 
> I don't know if this te right place to ask this question, but I'll try
> anyway :)
> 
> I am having a little trouble compiling the new Kernel. I did everything
> according to the manual ( HOWTO ) and after "make config", "make dep",
> and "make clean" the kernel started compiling with "make zImage".
> However during the process I get an error. This is what is sais:
> 
> Make[1]: Entering directory '/usr/src/linux/arch/i386/boot'
> as86 -0 -a o bootsect.o bootsect.s
> Make[1]: as86: Command not found
> Make[1]: *** [bootsect.o] Error 127
> Make[1]: Leaving directory '/usr/src/linux/arch/i386/boot'
> Make: *** [zImage] Error 2

You need the 16 bit assembler program in order to make bootsector for Linux.
Just install the package bin86

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: startx (after upgrade) dials modem before running X

[Marcin Owsiany]

On Sun, Sep 19, 1999 at 10:35:15AM -0600, [EMAIL PROTECTED] wrote:
> Greetings:
> 
> After upgrading some X-window packages last night I cannot start X without
> waiting for the modem to dial out and connect. It seems to be doing a name
> lookup since, if I get back out of X and hang up the modem and then do a
> startx immediately, there is no dial-out since the name lookup is still
> cached. A few hours later, though, the problem is back. The packages that
> updated last night were mostly fonts, but xbase-clients updated also.
> 
> I am running slink, and my X packages come fronm:
> 
> deb http://samosa.debian.org/~branden/ xfree86-334-slink/
> 
> Has anyone else seen this?  Any ideas on how to stop it?

If it is really a DNS lookup then the easiest way would be to put some more
information to your /etc/hosts. You have to think out yourself which host
could it be looking up :)

hope this helps

Marcin


-- 

----
Marcin Owsiany
[EMAIL PROTECTED]

Re: netscape error (won't display images)

[Marcin Owsiany]

On Sat, Sep 18, 1999 at 08:04:18PM -0700, David Karlin wrote:
> Hello,
> I just debianized a friend's old '486 (16MB ram) with a fresh slink
> installation.  Everything seems to be running fine, except for netscape.
> 
> I did 'apt-get install navigator-smotif-45', and navigator comes up fine,
> displays text, background colors, etc., but does not display images
> (they show up as a grey box).  This is an old box; on-board VGA video.
> 
> When I do 'navigator' from an xterm, error messages from navigator go
> to that xterm, and when I try to load an image, it shows a bunch of
> (repeated) error messages.  Here is a snippet:
> 
> navigator-smotif_real:
> X Error of failed request:  BadMatch (invalid parameter attributes)
>   Major opcode of failed request:  72 (X_PutImage)
>   Serial number of failed request:  26570
>   Current serial number in output stream:  26584
>   Widget hierarchy of resource: unknown
> 
> Can anyone help me to interpret this?

I'd say this is because the color palette is to small for it.

just a thought, I'm probably wrong... to see if I'm right try making a small
.xsession like this:

--
#!/bin/bash

exec netscape
--

don't forget chmod +x it :)

then see if it still shows only grey boxes... if it does, then i'm almost
surely wrong in my assumption


Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Dying services due to low memory?

[Marcin Owsiany]

On Sat, Sep 18, 1999 at 11:17:13AM +0200, Ralf G. R. Bergs wrote:
> Hi there,
> 
> can anyone point me to a solution for the following problem?
> 
> I have several machines running as Internet servers, mainly FTP and HTTP. 
> They're relatively low-end machines (P100 and 486-133 with 48 resp. 64 MB 
> RAM.) Every couple of days I have to restart inetd or other stand-alone 
> services (like syslogd, klogd, snmpd, apache.)
> 
> I'm pretty sure the reason why the processes fail is that memory usage is 
> too high (it's *definitely* not due to memory problems, like failing RAM 
> modules or overclocked CPUs.) Memory usage is permanently about 99%, swap 
> usage only a few percent. But obviously processes are dying because they 
> can't allocate "real" memory?!

As far as I know the Unix processes don't care if something is physical
memory or not. They simply use virtual memory and it's kernel's job to
proviede it to them.

As for the memory amount, i don't think that running a box at its full
memory capabilities does something bad to Linux. I myself used to administer
a box which had as little as 8 MB of RAM (Pentium 90Mhz) with the following
services: httpd, ftpd, squid and sometimes even X!
It was trashing horribly, but was usable and hardly ever crashed.
So it's not the amount of physical memory you have.. i'd rather think of
making more swap (don't know how much your box has...), since some peaks in
memory usage can be lethal to Linux.
Really, i've heard of guys who have a GB of swap on their boxes and are
thinking of increasing it...

just a thought

Marcin


-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: ppp failure under debian (slakware with pppd 2.2 patch level 0 works)

[Marcin Owsiany]

On Sat, Sep 18, 1999 at 05:06:18PM +1000, [EMAIL PROTECTED] wrote:
> Hi,
>I'm typing this from my old slackware install.  I connect on this
> installation using cu -l /dev/cua1 then negotiating my way through my
> ISP's login manually, then start pppd in another window, then quitting
> from cu.  This worked fine whne this machine was a 486 and now works with
> a new motherboard & cpu.  
> 
> I tried installing potato on the machine, and most things are going fine,
> except that sound isn't working on the new setup, and ppp won't work on
> it.  
> 
> I'm fairly sure that the serial conifg is OK, as I can get out with 
> cu -l /dev/ttyS1, but when I try to start pppd, it complains about cu
> having the serial line and won't let it at it.
> 
> As far as using pon, I get the following :
> 
> 
> Aug  4 19:16:00 castle kernel: PPP: version 2.2.0 (dynamic channel 
> allocation) 
> Aug  4 19:16:00 castle kernel: PPP Dynamic channel allocation code copyright 
> 199
> 5 Caldera, Inc. 
> Aug  4 19:16:00 castle kernel: PPP line discipline registered. 
> Aug  4 19:16:00 castle kernel: registered device ppp0 
> Aug  4 19:16:00 castle pppd[620]: pppd 2.3.8 started by robert, uid 1000
> Aug  4 19:16:01 castle chat[621]: abort on (BUSY)
> Aug  4 19:16:01 castle chat[621]: abort on (NO CARRIER)
> Aug  4 19:16:01 castle chat[621]: abort on (VOICE)
> Aug  4 19:16:01 castle chat[621]: abort on (NO DIALTONE)
> Aug  4 19:16:01 castle chat[621]: abort on (NO DIAL TONE)
> Aug  4 19:16:01 castle chat[621]: abort on (NO ANSWER)
> Aug  4 19:16:01 castle chat[621]: send (ATZ^M)
> Aug  4 19:16:01 castle chat[621]: expect (OK)
> Aug  4 19:16:46 castle chat[621]: alarm
> Aug  4 19:16:46 castle chat[621]: send (AT^M)
> Aug  4 19:16:46 castle chat[621]: expect (OK)
> Aug  4 19:17:31 castle chat[621]: alarm
> Aug  4 19:17:31 castle chat[621]: Failed
> Aug  4 19:17:32 castle pppd[620]: Exit.
> Aug  4 19:18:00 castle kernel: PPP: ppp line discipline successfully 
> unregistere

first thing: should there be any chat script at all if you authenticate
using cu? (i've never used cu, so i may be wrong)
Should the chat script try to reset the modem if it has already connected??
(the ATZ and AT)

I would try to edit /etc/chatscripts/provider to contain only one line:
'' ''
if I were you.

> One thing that concerned me was that the kernel messages from dmesg don't
> mention the ppp0 device as happens on a working debian ppp machine I've seen.
> The relevant part of dmesg looks like this 
> 
> PPP: version 2.3.7 (demand dialling)
> PPP line discipline registered.
> 
> and that's it.
> 
> What is happening?

this is how it looks like on my system (slink, 2.2.12):

PPP: version 2.3.7 (demand dialling)
PPP line discipline registered.
registered device ppp0
PPP BSD Compression module registered
PPP Deflate Compression module registered
PPP: ppp line discipline successfully unregistered

and as far as i can remember, the "registered device ppp0" appears only
after chat script has finished successfully:

Sep 18 12:56:04 pecet pppd[166]: Serial connection established.
Sep 18 12:56:05 pecet pppd[166]: Using interface ppp0

hope this helps,
Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: dselect broke when updating to potato

[Marcin Owsiany]

On Fri, Sep 17, 1999 at 04:32:48PM -0400, Edward Di Geronimo Jr. wrote:
> Hi,
> 
> The other day I updated my system from slink to potato, and now dselect
> and apt are broken. First off, if I go into dselect and choose Access,
> Update, or Install, it exits out to the shell with the error:
> 
> dselect: unable to access method script
> `/usr/lib/dpkg//methods/charset/setup': Not a directory
> 
> The other dselect choices work.
> 
> "apt-get update" works, however, "apt-get install" always says:
> 
> 0 packages upgraded, 0 newly installed, 0 to remove and 11 not upgraded.
> 
> regardless of what I chose in dselect.

This is not the way to use apt.
You either apt-get install packagename
or use dselect and then apt-get dselect-upgrade

> I can't install or update anything
> with apt right now. I can still install packages with dpkg -i. I tried
> downloading a new version of the dpkg package a few minutes ago, however,
> it didn't make a difference.

don't know about the first problem you mention, though

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: How does it...

[Marcin Owsiany]

On Fri, Sep 17, 1999 at 08:47:19PM -0500, Fredrick Schmitt wrote:
> Is the setup text based or graphial? What are the extra features this
> opperating system has over my RH6 box? What system would be the
> cheeziest I could run it on?

there has been a LARGE discussion on it about a week ago.
The Subject was: Why use Debian, why not RedHat?
Check the archives out. (there's a link on the web page - just have to look
close)

just my 0.02 zł :)

Marcin

-- 

--------
Marcin Owsiany
[EMAIL PROTECTED]

Re: Apache+PHP3

[Marcin Owsiany]

On Fri, Sep 17, 1999 at 09:43:50AM -0700, Seth R Arnold wrote:
> Now I am lost. :)


I don't have the original question post handy, but i can remember the
configuration/compilation/installation commands were all OK, but there was a
lacking "make install" in apache directory, only a "make" was issued there.
This might have been a typo, but... also might not.

Marcin

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: System slows to REAL slug on boot

[Marcin Owsiany]

On Fri, Sep 17, 1999 at 08:16:49AM -0700, [EMAIL PROTECTED] wrote:
> In <[EMAIL PROTECTED]>, on 09/17/99 
>at 11:36 AM, Marcin Owsiany <[EMAIL PROTECTED]> said:
> 
> >single mode: (runs only rcS)
> > single
> 
> >emergency mode: (runs only init, which only runs sulogin, / is ro)
> > emergency
> 
> Thanks Marcin and Lex.  Last ?? of the day... to pinpoint the problem child,
> I presume I run through each script in the rc.2 dir and see when things
> freeze??

after going to single user, yes, it's a good idea
Also, remember to give an argument: start to the scripts

Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Apache+PHP3

[Marcin Owsiany]

On Thu, Sep 16, 1999 at 06:02:38PM -0700, Seth R Arnold wrote:
> On Fri, Sep 17, 1999 at 12:47:02AM +0200, Marcin Owsiany wrote:
> > On Thu, Sep 16, 1999 at 04:44:21PM +0700, Usef Saiful Ulum wrote:
> > > # make; make install
> > 
> > err... shouldn't it be make && make install instead?
> 
> The && means "execute the next one only if this one worked". If you really
> trust the source, make ; make install works fine. If you are not so sure it
> will compile, make && make install might save some headaches by preventing
> a not-fully-compiled program from being installed.
> 
> Unfortunatly, it is unlikely to help Usef directly in this case, though it
> *is* good practice. :)

but I'm not saying:
do make && make install   instead of make;make install
but:
do make && make install   instead of just make



Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: System slows to REAL slug on boot

[Marcin Owsiany]

On Thu, Sep 16, 1999 at 02:24:11PM -0700, [EMAIL PROTECTED] wrote:
> On Fri, Sep 17, 1999 at 12:55:04AM +0200, Marcin Owsiany wrote:
>  >On Wed, Sep 15, 1999 at 09:53:01PM -0700, [EMAIL PROTECTED] wrote:
>  >> 
>  >> I'm having problems with my system on rebooting to a new kernel.  I don't 
> know 
>  >> what is causing the problem, and have found nothing yet in the archives 
> about 
> [...]
>  >> Anyone else see this kind of behavior??  Any thoughts??  Would it help if 
> I
>  >> got a list of the file names in /etc/init.d to list services in place?
>  >
>  >I would say something is eating up your resources so much you can't log in.
>  >I would boot system into emergency or single mode and try switching off some
>  >services and see if it/what helps... dhcpd would be number 1.
>  >
> 
> Thanks Marcin.  I have already tried changing the dhclient (ISC) to dhcpcd
> 1.3 and still had the same slowdown.  
> 
> Is 2.2.9 known to be an ugly kernel w.r.t. networking or other??

Nothing that i would know of, but again - I'm no guru...

> When suggesting that services be switched off, do you mean to remove them
> fron init.d?  My own approach to this has been to modify the names of the
> daemons so the script fails... make sense?  

I don't know "The Right Way" to disable init.d scripts.
The proper way seems to be to run update-rc.d with appropiate arguments so
that all the links in rcx.d disappear, but than, to restore them you have to
remember the number (the sequence is important in some places)
Some people chmod -x /etc/init.d/script, but this gives nasty messages at
boot.
I personally add a "exit 0" line near the top of the /etc/init.d/script...

> Rookie question--how can I boot into single user mode... can that be done at
> the LILO prompt?

single mode: (runs only rcS)
 single

emergency mode: (runs only init, which only runs sulogin, / is ro)
 emergency


hope this helps

Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: System slows to REAL slug on boot

[Marcin Owsiany]

On Wed, Sep 15, 1999 at 09:53:01PM -0700, [EMAIL PROTECTED] wrote:
> 
> I'm having problems with my system on rebooting to a new kernel.  I don't 
> know 
> what is causing the problem, and have found nothing yet in the archives about 
> similar circumstances. The new kernel is 2.2.9, the old was 2.0.36.
> 
> Once on reboot the system halted while starting lpd.  Another time on gpm,
> which seemed to respond to a ^X.  All times, though, things ground down to a

[cut a description of system sluggishness]

> Are there incompatibilities between slink and potato which need to be
> managed?  Especially with respect to networking or other daemons, which I
> may encounter on this upgrade?  Is this sluggishness possibly due to the
> dhcp client, which is said to not work with the 2.2 kernels (though it seems
> to, here)??
> 
> Anyone else see this kind of behavior??  Any thoughts??  Would it help if I
> got a list of the file names in /etc/init.d to list services in place?

I would say something is eating up your resources so much you can't log in.
I would boot system into emergency or single mode and try switching off some
services and see if it/what helps... dhcpd would be number 1.

just a thought

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: installing debian

[Marcin Owsiany]

On Thu, Sep 16, 1999 at 09:12:41PM +0200, Tino van Uffelen wrote:
> Hello,
> 
> I want to try and install debian. As I have a rather old computer (i486, 4mb
> ram and 60 mb Hard disk), I don't think I have room for both dos/windows and
> linux so I want to remove dos completely first. This means I don't have
> anything to fall back on when installing linux doesn't work. Could you
> please advise as what to do and which debian packages I should install.
> 
> I have read the intallation guide to debian/linux. So I think I am
> comfortable with the first few steps. But what then.
> 
> I've read that on a computer as old as mine, debian version 1.3.2 will work
> a bit quicker then the latest version, so I am intending to install that
> version.

If I were you, I would install 2.1.
I think the small performance boost isn't comparable to the bugfixes made
since 1.3.

Just a thought

> Will debian/linux partition my harddrive to my needs (I don't have a clou as
> to how to do that!)

I don't think it will do it without your help :)
But don't worry, it's simple when you use the cfdisk tool!
If I were in your shoes, I would just wipe all the partitions existing
and then make a 16 MB swap partition (you have to change the type) and a
linux partition from what is left. Also activate the linux partition (mark
it as bootable)
Probably making just one linux partition and then adding a swap file instead
of a swap partition would be better on such small disk, but this requires
some hand tweaking:
after the first reboot, when the installation process wants to start
dselect, switch to another virtual terminal (Alt-F2), log in as root and
issue the following commands:

dd if=/dev/zero of=/swapfile bs=1048576 count=n
(put number of megabytes of swap in place of "n")
mkswap /swapfile
sync
chmod 600 /swapfile
swapon /swapfile

if you want the swap to be activated every boot (if you are not sure, then
you want :)) you also should do:

echo "/swapfile none swap swap" > /etc/fstab

after the "swapon" command above

now, if for any reason you want more disk space just do:
swapoff /swapfile
rm /swapfile

and repeat the above commands with another amount of megabytes (no need to
repeat the echo command)
NOTE: changing the size of swap partition is impossible without deleting whole
harddrive

> I want to use my computer as a simple wordprocessor (maybe something
> compareble to works for windows 3.0??? a bit wordprocessor, a bit database
> and a bit spreadsheet.), nothing fancy and I want to be able to surf the
> internet, send and receive e-mail, and make my own webpages.

wow! you will probably need a larger hard drive if you want to do all of
these.
Recently I have installed an unstable version of Debian with the following:
1) X (including only 1 Xserver, xterm, icewm as a window manager and Polish
fonts)
2) Netscape Navigator
3) Myslq database server and client
4) apache Web server
5) PHP3
6) wget, MidnightCommander and joe - an editor

and it all took some 130 MB

Even not counting apache and PHP3 it will be hard for you to do all the
tasks you want with only 60 MB minus swap size of disk...

but again, I may be wrong

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Apache+PHP3

[Marcin Owsiany]

On Thu, Sep 16, 1999 at 04:44:21PM +0700, Usef Saiful Ulum wrote:
> Hi there,
> 
> I have installed Apache 1.3.9 with PHP3:
> 
> # tar xvzf apache_1.3.9.tar.gz
> # tar xvzf php-3.0.12.tar.gz
> # cd apache_1.3.9
> # ./configure --prefix=
> # cd ../php-3.0.12
> # ./configure --with-mysql=
>   --with-apache=../apache_1.3.9 --enable-track-vars
> # make; make install
> # cd ../apache_1.3.9
> # ./configure --prefix=
>   --activate-module=src/modules/php3/libphp3.a
> # make

err... shouldn't it be make && make install instead?

> # cd ../php-3.0.12
> # cp php3.ini-dist /usr/local/lib/php3.ini
> 
> The problem is: I didn't get any files on   ???
> # ls -al 
> total 0
> #
> 
> How this could be, and how to solve the problem?

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: "RAMDISK: Compressed image found at block 0"

[Marcin Owsiany]

On Wed, Sep 15, 1999 at 01:45:46PM -0500, ray wrote:
> Gary Young wrote:
> > 
> > In attempting to install Debian 2.1 from resc1440
> > floppy onto an HP intel box with SCSI disks that
> > had run Windows-NT, I get the error:
> > "RAMDISK: Compressed image found at block 0"
> > and the boot stops.
> > 
> > Apparently the computer had a compressed C: drive or
> > partition when it had windows-NT as the operating
> > system.
> > 
> > (I attempted to do a "boot:floppy" but after
> > I insert the root diskette, it gets a protection fault.)
> > 
> > Is there a work around for the "compressed image"
> > problem, or am I missing something about Linux with
> > SCSI drives.
> > 
> > Gary Young
> 
> Gary,
> 
> I will respond quickly in order to give you something 
> to try while some GURU gives you the real answer later
> in the day.

Hi!
I'm not a guru, but here's what i think
:)

"compressed image" does not mean the compressed disk of WIN, it
just means that a compressed image of a root device was found in memory.
Nextly, the kernel should use it as a root partition, but apparently failed
to use it. This probably means that the image on floppy is corrupted.
Try using another floppy, as suggested.
Don't be upset! It could also mean that your floppy drive is broken :)

hope this helps
Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: apt-get dselect-upgrade

[Marcin Owsiany]

On Tue, Sep 14, 1999 at 12:15:04PM +0200, Mirek Kwasniak wrote:
> Hi,
> 
> on one of my debian (potato) machines I have:
> 
> $ apt-get dselect-upgrade
> Reading Package Lists... Done
> Building Dependency Tree... Done
> Segmentation fault (core dumped)
> 
> apt-get upgrade works ok. How to cure/diagnose it.

Hmm... try strace-ing it

just a thought

Marcin


-- 

------------
Marcin Owsiany
[EMAIL PROTECTED]

Re: ftpd and proftpd with package management

[Marcin Owsiany]

On Mon, Sep 13, 1999 at 08:22:08PM +0200, Olaf Conradi wrote:
> Hi,
> 
> I had proftpd installed (on the standerd port) and ftpd on a
> non-standard port. But the ftp daemons now have a dependency on
> ftp-server, which allows only one to be installed on a system at the same
> time.
> 
> How can I get around this so that apt will install both packages and
> keep updating then when newer versions appear?

well, this is really annoying...
If I were you, I would file a bugreport on both ftp servers, because the
conflict seems not to be very well designed.
And as for the quick and dirty fix I would:

1) get mc - it will be helpful
2) backup one of the servers (use dpkg -L packagename to list its
 contents) 
3) _remove_ it (luckily dpkg supports both --purge and --remove)
 , so that you don't loose your config files, and then restore
 the package from the backup
4) as for the upgrades, i would use mc on a new package file to copy the
files from it to the system

The other option is to:
get the source packages of the server(s) and compile your own versions of
ftp servers that don't conflict each other (or one another? I can't really
remember the rule). This however will require a certain amount of hacking
(don't ask me how to do this... i've never done that)

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: General kernel questions

[Marcin Owsiany]

Low!

On Mon, Sep 13, 1999 at 09:12:11AM -0500, David Kanter wrote:
> I wanted to customize my kernel over the weekend, but after looking at what I 
> assumed to be the current kernel configuration (i.e., the options selected 
> when viewing make xconfig), I was left with some questions.
> 
> For instance, during boot-up I get the PPP messages. But in make xconfig, PPP 
> ability is turned off---the radio button selected is "n," not "y" or "m." In 
> fact, I think all modules are turned off in make xconfig.
> 
> How can this be? Why am I able to use PPP (via pon or wvdial) when the module 
> isn't apparently there? However, during installation, I did select PPP as a 
> module. I'm quite confused.

looks like you did configure your new kernel sources for compilation, but
did you actually compile it? Did you read /usr/src/linux/README ?

just a thought,
Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Sleeping in a shell

[Marcin Owsiany]

Low!
Does anybody of you guys know a way to sleep for an amount of time less than
a second in a shell (bash/sh) script?

sleep refuses arguments like: 0.5 0,5 1/2

Maybe there is a nice perl command to do this?

I really don't feel like writing a my_sleep.c ...

TIA

Marcin


-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: weird entry in netstat

[Marcin Owsiany]

On Sun, Sep 12, 1999 at 05:40:36PM -, Pollywog wrote:
> I don't have Licq or other ICQ clone running, and I ran 'netstat' and got
> this:
> 
> udp0  0 pollywog.sunset.ne:1062 fes-d018.icq.aol.:gicqd
> ESTABLISHED 

you can use:
"fuser 1062/udp"
to get the PID of the process using this port on your host...

hope this helps

Marcin


-- 

----
Marcin Owsiany
[EMAIL PROTECTED]

Re: Zombie process using pppd

[Marcin Owsiany]

On Wed, Sep 08, 1999 at 09:47:19PM +0200, Laurent PICOULEAU wrote:
> Hi,
> 
> On Wed, 08 Sep, 1999 ? 08:30:35AM -0400, Mark Buda wrote:
> > >>>>> "Alex" == Alex V Toropov <[EMAIL PROTECTED]> writes:
> > 
> > Alex> In what situation process becomes Zombie ?
> > 
> > When a process exits, it can return an exit status code to its parent
> > process. The parent process retrieves this exit status code by
> > invoking the wait() system call. A zombie process is a process that
> > has exited, but whose parent hasn't used wait() to get its exit status
> > code yet. It isn't doing anything, it's just there in the list of
> > processes, waiting to be wait()ed for, unable to really die until that
> > happens.
> > 
> > Eventually, even if the neglectful parent process never wait()s for
> > it, it will go away, for this reason: When a parent process exits,
> > init (process id 1) inherits all its child processes. And one of
> > init's less well-known jobs is wait()ing for orphaned zombie
> > processes.
> 
> Is there a way to determine the neglecting parent of a zombi ? Such as to be
> able to kill the parent if it's not a necessary program and letting init
> inherits the zombi process.

try

ps axuf

- gives you a nice process tree

just a thought

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Modules

[Marcin Owsiany]

On Wed, Sep 08, 1999 at 11:43:24AM -0700, Brian E. Lavender wrote:
> When you recompile the kernel, the filesystem support you want to
> support as a module, you hit the M key for module when you do either
> 
> # make config
> 
> or 
> 
> # make xconfig
> 
> or 
> 
> # make menuconfig
> 
> I am assuming you know how to compile the kernel. Here are the basics
> in case you have never done it. 
> 
> # make menuconfig
> # make dep
> # make clean
> # make bzImage (long part)
> # make modules
> # make modules_install
> # make zlilo

or install kernel-package and:
1% cd 
2% make config   # or make menuconfig or make xconfig and configure
3% make-kpkg clean
4% fakeroot make-kpkg --revision=custom.1.0 kernel_image
5# dpkg -i ../kernel-image-X.XXX_1.0_.deb
6# shutdown -r now # If and only if LILO worked or you have a means of
   # booting the new kernel. YOU HAVE BEEN WARNED!!

and forget the stuff about modules... dpkg will handle this for you :)

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Startup hangs because of bad partition

[Marcin Owsiany]

On Wed, Sep 08, 1999 at 03:10:16PM -0400, B. Szyszka wrote:
> Hello everyone,
> 
> I wanted to move /var to another partition since it kept on
> filling up, but I accidentally forgot to make its partition a
> Linux one and now Linux won't start. When I boot it, it
> asks me for my root password for maintenance or press

But what exactly does it say BEFORE running this sulogin? (you can use
SHIFT+PGUP to see what was printed to the console before)

> Crtl+D to continue booting. If I put in the password, I get a
> limited set of programs including no text editor so I can't
> edit /etc/fstab 

try mounting our /usr partition - some editor should be there...


-- 

--------
Marcin Owsiany
[EMAIL PROTECTED]

Re: Help! Hang on boot, with letters LI

[Marcin Owsiany]

Low!!

On Wed, Sep 08, 1999 at 08:42:49AM -0500, Brian Boonstra wrote:
> Hi
> 
>   I can no longer boot into Debian!  I installed IE5 in my WinNT  
> partition, and after that Lilo would boot WinNT at the prompt, but hang on  
> trying to boot Linux.  Then after some fiddling it broke completely, giving  
> me the letters LI (in the same position as I previously saw a whole LILO  
> prompt), and the system hung.
>   I reckoned Lilo was the problem, and installed a different boot  
> loader (PQBoot from Powerquest).  Now I can still boot into WinNT, but if I  
> try to boot Linux, I get the same LI hang I had when Lilo was the boot  
> loader.
> 
> 
> 
>   I tried booting off the CD, activating my partitions, and  
> reinstalling LILO, but it win't install.  I then tried to create a boot  
> floppy, and I can't do that either.
> 
> 
>   I have a 7.5 GB SCSI system with 4 primary partitions.  In order  
> they are: DOS, WinNT, swap, ext2.  I was running potato with a  
> custom-compiled kernel that I had set up with dpkg-kernel.

from lilo Manual:

   LI   The first stage boot loader was able to load the second stage boot
loader, but has failed to execute it. This can either be caused by a
geometry mismatch or by moving /boot/boot.b without running the map
installer.

This is all i am able to say...
It is strange that you use LILO to load your NT...
As it is explained in a special HOWTO (NT-LOADER-HOWTO ?? Linux-and-NT-HOWTO
?? can't remember its name...) it says that you should use the NT loader to
boot NT and a special entry in a .ini file + a copy of Linux partition
bootsector put in C:\ will handle the case...
It seems that your setup differs from what is recommended and NT has garbled
some of carefully setup LILO sectors...

hope this helps,
Marcin


-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Executing programs with a keypress under X/wmaker

[Marcin Owsiany]

Low!!

On Wed, Sep 08, 1999 at 12:06:13AM -0500, Brad wrote:
> -BEGIN PGP SIGNED MESSAGE-
> 
> On Fri, 3 Sep 1999, Brad wrote:
> 
> > I'm looking for a way to execute a command under X when certain key
> > combinations are pressed. Specifically, i want to make something like
> > windowbutton-X bring up an xterm. I don't want this command to attempt to
> > execute in the console.
> > 
> > I noticed the xkbevd program in xbase-clients, but even looking at the
> > source i can't figure out how to get it to do this. Is there a program out
> > there that will do what i want, or do i have to dig into xbooks and try to
> > write my own? (The X Keyboard Extentions look promising...)
> 
> Well, since no one even responded to my query, i did go ahead and write my
> own. If anyone would like to take a look, help debug it, just email me
> privately and i'll send you the sources.
> 
> The only problem i have is that the keystroke is delivered to the focused
> app, as well as launching the command. For example, if you used ctrl-D,
> and pressed it in a bash shell in an xterm, the shell will exit at the
> same time the command is launched...

That is why I guess the hot-keys is a job of window-manager...

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: HD Space Left, dselect

[Marcin Owsiany]

Low!!

You can also use "du" (abbr. from Disk Usage)

it has some nice options so using it like:
du -hs *
in a directory will list all subdirectories with their sizes in
human-readable units

hope this helps

Marcin

On Tue, Sep 07, 1999 at 08:06:52PM -0700, Seth R Arnold wrote:
> Tom -- df will do what you want. :)
> 
> On Tue, Sep 07, 1999 at 07:16:47PM -0600, Tom Lineman wrote:
> > Hello again.
> > 
> > I installed linux on a 75 mb partition and have no others to spare,
> > so obviously space is a problem for me.  All I'd like to know is how I can 
> > see how much space there is left on my system, other than going through 
> > each 
> > directory and ls'ing and then adding up space used.
> >One more thing.  I have all the required packages from dselect 
> > installed. 
> >   If I run dselect again, should I leave the present status for those 
> > packages to "install" as opposed to "remove"?  Just how much space do all 
> > the "recommended" packages take up?
> > 
> > Any help is greatly appreciated.
> > Tom Lineman
> > 
> > P.S.  Thanks to all those who answered my question about the prompt!
> > P.P.S.  That fellow calling himself "a," with the question about VCD, was 
> > he 
> > talking about Virtual CD?
> > P.P.P.S.  I know you're getting bored, but please bear with me :)
> > I noticed a fellow who wrote his message in spanish.  I understood his 
> > question, but didn't know the answer.  If anyone wants me to translate his 
> > message so they can reply, please let me know.  Or you can use BabelFish :)
> > 
> > Bye for now.
> > 
> > __
> > Get Your Private, Free Email at http://www.hotmail.com
> > 
> > 
> > -- 
> > Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
> 
> -- 
> Seth Arnold | http://www.willamette.edu/~sarnold/
> Hate spam? See http://maps.vix.com/rbl/ for help
> Hi! I'm a .signature virus! Copy me into
> your ~/.signature to help me spread!
> 
> 
> -- 
> Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
> 
> 

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Debian, NT

[Marcin Owsiany]

Low!!

On Tue, Sep 07, 1999 at 02:36:40PM -0600, Robert Kerr wrote:
> The NT machine is on a network, and automagically mounts our userspace
> when we log in.  I'd like to be able to do this automagically too.
> So, what, in the consensus of the list, do I need to do for this to work
> nicely?

Well, i think this would be very easy to achieve provided the userspace is
mounted either from a smb-like server namely:
 Linux samba share, 
 NT,Win9x or Win3.11 share over TCP/IP
 Novell server share

The trick would be to conditionally (unless it's already mounted) run
special mount command the share in .xsession or .profile.

You probably want to take a look at "samba" package unless you will be
mounting a Novell share in which case you probably need to load module "ncpfs.o"
(or recompile the kernel - i don't know if default one comes with IPX and this
module...) and check out "ncpfs" package :)

hope this helps

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: File system questions

[Marcin Owsiany]

On Tue, Sep 07, 1999 at 08:46:48AM +0200, [EMAIL PROTECTED] wrote:
> > Hi everybody,
> > 
> > Is there a way to read/write an NTFS drive in slink?
> > (I coud not find a driver in modconf.)

You need to recompile the kernel. I'm not sure about 2.0.x, but 2.2.x has
NTFS support for sure - AFAIK only readonly support is considered stable,
but rw is also there "for the brave" :)


> > How can I use the nls_* file system types?  I guess I need 
> > nls_cp852 and/or nls_iso8859_2 for Hungarian nls support, 
> > but I found no difference after I installed them.

Luckily I'm Polish, so i need iso-8859-2 as well :)
>From what i remember from reading the docs some long time ago these modules
are only helpful (but not essential) for use with the msdos-like
filesystems (msdos,vfat) - so you can see your national chars in file NAMES.
The modules themselves do not work after installing, you'll need special
mount options for them, to be active.
I've got:

defaults,noatime,nodev,umask=000,quiet,codepage=437,iocharset=iso8859-2,noexec

in options for my win95 filesystems

read mount(8) for more/accurate info.

hope this helps

Marcin

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: ssh

[Marcin Owsiany]

On Mon, Sep 06, 1999 at 10:41:11PM -0400, B. Szyszka wrote:
> > Does anyone know where i can download the debian package ssh (secure shell)
> 
> You can get it through 'apt-get install ssh', but I'm not sure what you should
> do if you're outside of the U.S. You can probably download it manually from 
> here:
> http://www.debian.org/Packages/stable/non-us/ssh.html

no, no, no

I'm outside US (Poland) and apt-get install ssh works just fine for me.
Just need to add a non-us section mirror
I've got:
deb http://non-us.debian.org/debian-non-US stable non-US
in my sources.list

hope this helps,
Marcin


-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: xconsole transparency

[Marcin Owsiany]

> I know how to get wterm to become transparent, but what I should have said
> was how can wterm's transparency and other options be applied to xconsole.

are you sure it was really xconsole that was transparent?
Maybe it was just an eterm with "tail -f logfilename" running in it?

Marcin

-- 

--------
Marcin Owsiany
[EMAIL PROTECTED]

Re: library lost - need help, please

[Marcin Owsiany]

On Mon, Sep 06, 1999 at 07:05:14PM -0400, Peter Mickle wrote:
> hello readers,
> 
> NEW USER question/problem, need help if possible. in trying to solve one
> shared libraries problem, i unintentionally removed the following library:
> 
>   /usr/lib/gcc-lib/i486-linux/egcs-2.91.60/libstdc++.so
> 

Hi Peter!
Actually apt and dselect are frontends to the real Debian package management
tool - "dpkg" They both assusme that if the package is installed - it does
not need to be reinstalled - which is perfectly true unless the packege is
broken - like in your case. 

(BTW - there should be an option for apt to force reinstalls, shouldn't it?)

In your case the best way is to:

dpkg -i libstdc++2.9-dev_2.91.60-5.deb

which installs the package. Also, if you have Midnight Commander you can
try tapping enter on the package file and then be able to browse/install the
package contents.

hope this helps,
Marcin

-- 

--------
Marcin Owsiany
[EMAIL PROTECTED]

Re: vfat filesystem mysteriously becomes read-only

[Marcin Owsiany]

On Mon, Sep 06, 1999 at 08:06:05AM -0700, Eric Hanchrow wrote:

[snip]

> Directory sread (sector 0x8000100b) failed
> attempt to access beyond end of device
> 03:01: rw=0, want=1073743878, limit=1048288
> Directory sread (sector 0x8000100b) failed
> attempt to access beyond end of device
> 03:01: rw=0, want=1073743878, limit=1048288
> Directory sread (sector 0x8000100b) failed
> attempt to access beyond end of device
> 03:01: rw=0, want=1073743878, limit=1048288
> Directory sread (sector 0x8000100b) failed
> 

[snip]

>   /dev/hda1 /windoze vfat rw,uid=1000 0 0

seems like your windoze partition is having problems (03:01 = hda1), namely
a corrupted directory entry.

[snip]

well, i used to have such messages on my Linux box some time ago. This could 
mean
that:
a) your hda1 needs ScanDisk
b) you need a new hard drive :(

It seems strange that kernel remounts it ro on problems... but not
impossible :). I have "errors=remount-ro" in /etc/fstab options for "/" by
default... how about you? Maybe it "spreads" over newly mounted partitions?
:)

Thus Scan your hd, if it doesn't help - backup important data and buy a new
hd :(

hope this helps

-- 


Marcin Owsiany
[EMAIL PROTECTED]

Re: Block stupid/annoying sites

[Marcin Owsiany]

On Thu, Sep 02, 1999 at 10:53:47PM -0700, George Bonser wrote:
> >> What about using REJECT instead of DENY?  That way the browser should

> there is something there that they are not allowed to access. They can simply
> adjust their activity from a different location to see if they can gain access
> to the rejected service.

Isn't it the other way round?
I can remember that "DENY" means "drop packet on the floor", while "REJECT"
means to send back an ICMP packet saying: "connection refused"
And when someone wants to connect to a port, on which nothing is listenning,
he/she will get an ICMP reply "connection refused" - for example if you
point your browser at a host without httpd running, you will get "connection
refused". But if there is a rule saying to DENY packets from you, you will
have to wait for a timeout.

correct me if i'm wrong

just my 2c

Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Ooops! [was: Re: Hi all and a question]

[Marcin Owsiany]

> > Ah! the name was "perforate"
> 
> I have Debian Hamm and this program is really undocumented :(

my apologies. The perforate package does not contain the needed program.
It is in package binstats :)

and as for perforate, i've got slink, and it has some mans


Marcin

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: 1.5 years old user

[Marcin Owsiany]

On Thu, Sep 02, 1999 at 11:07:30AM +0300, Madarasz Karoly wrote:
> Is there any programs for a 1.5 years old user?
> Something like xteddy controllable by mouse,

well, i do not know what you mean by "controllable by mouse" but there is a
package "xteddy", containing a cuddly teddy-bear.
You can also try out xteddy from unofficial slink gnome packages, it
contains several other nice toys' pixmaps.

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Hi all and a question

[Marcin Owsiany]

On Wed, Sep 01, 1999 at 01:47:48AM +, John Carline wrote:
> [EMAIL PROTECTED] wrote:
> 
> > Hi all,
> >
> > I'm new at this list and I want to ask something:
> > Is there any program to check which libraries are unused on a system, so,
> > any program doesn't depends on them ?
> >
> > Many thanks.
> >
> > 
> > Juli-Manel Merino Vidal
> > [EMAIL PROTECTED] - http://jmmv.cjb.net
> > 
> 
> Hi,
> 
> That's a great question!  Wish I knew the answer ;(
> 
> Anyone?

There is such program, it's even in a Debian package.
Don't remember the name though :(

It was contained in a package whose main program was a program to "make
holes" in files, thus non-destructively gaining space on HD.

Ah! the name was "perforate"

It contains 2 scripts that gather info about progs/libs and say which are
unneeded.

Wish you luck, i'm not 100% sure

Marcin

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Why use Debian? Why not Red Hat?

[Marcin Owsiany]

On Wed, Sep 01, 1999 at 01:43:52AM -0700, George Bonser wrote:
> 
> Patrick,
> 
> I graduated from Red Hat to Debian twice. I was hard-headed and did not
> learn the first time ... gave it a second chance. I came back to Debian.
> 
> The ONLY thing Red Hat has is an easy install ... upkeep of a Red Hat
> system is a nightmare.

The same for me.
I've tried: RedHad 4.0, Debian 1.2, Slackware, SuSE, some more diffirent
RedHats 4.2 - 5.2 , and came back to Debian and i will never leave it again.
Moreover, i've Debianized every Linux system I could (about 6).

My opinion is: Debian is not perfect, but no distro is, and Debian is
closest to what I like about Linux. You can tweak everything you can in
Debian, while RedHat seems more like M$-windows to me. (you can beat me)
RedHat is not much more foolproof than Debian. If you want a Linux for
dummies, give SuSE a try - click as much as you wish :)

Marcin

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: /etc/hosts problem

[Marcin Owsiany]

On Fri, Aug 27, 1999 at 10:40:38AM -0400, Sebastian Canagaratna wrote:
> 127.0.0.1 is the localhost ( your machine ). Unless
> your machine is acting as the name server names will not be secolved.

but what about "order hosts,bind"?

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: pppd question

[Marcin Owsiany]

On Fri, Aug 27, 1999 at 02:20:12PM -, Chittaranjan Mandal wrote:
> 
> I had a working dial-up ppp link with an ISP provider.
> Recently, I am failing to setup the connection.
> I am not able to figure up what the problem is.
> The setup I am using now is just the same as before.
> I have appended a transcript of the logged messages.
> Would you like to take a look to see what could be going wrong.
> 
> Aug 12 07:27:24 merak chat[536]: Checking authorization, Please wail ...^M
> Aug 12 07:27:24 merak chat[536]: Checking authorization, Please wail ...^M
> Aug 12 07:27:24 merak chat[536]: Username: -- got it
> Aug 12 07:27:24 merak chat[536]: send (^M)
> Aug 12 07:27:24 merak chat[536]: expect (word:)
> Aug 12 07:27:24 merak chat[536]: ^M
> Aug 12 07:27:24 merak chat[536]: send (^M)

i used to have a similar problem.
The script does not wait for "password:" prompt, but sends both username and
password. This theoretically should work, but not necessarily :)

> Aug 12 07:27:24 merak pppd[535]: Serial connection established.
This is misleading. It only means that the chatscript has ended, nothing
more...

hope this helps,
Marcin
-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Another Partition spacing...

[Marcin Owsiany]

On Wed, Aug 25, 1999 at 11:39:49AM -0400, Greg Vence wrote:
> But root would contain what is not in var, usr, and home.  would 500MB
> be enough for that?
> 
[EMAIL PROTECTED]:~ $ df -h
FilesystemSize  Used  Avail  Capacity Mounted on
/dev/hda1  20M   17M   1.8M 90%   /
/dev/hda3 581M  179M   372M 33%   /var
/dev/hda5 1.6G  274M   1.3G 18%   /usr
/dev/hda6 1.2G  164M   1.0G 14%   /home
/dev/hda7 351M   13K   333M  0%   /backup

:-)
as you can see...

however /tmp is a symlink to /var/tmp, and you have to be caucious to do as
little as root, because otherwise root's homedir would grow too large.
Actually 40 MB would be very, very fine (lot's of spare place if you either
use a /tmp symlink or force as many apps as possible to use /var/tmp as
their temp)

best regards,
Marcin

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: Another Partition spacing...

[Marcin Owsiany]

On Wed, Aug 25, 1999 at 11:19:02AM -0400, Greg Vence wrote:
> What about:
>root 1.5GB
>var  2
>usr  2
>home 2.256
>swap .256 (128 MB ram)

root is MUCH too big. It should be as small as possible (unless the only
partition is the rootpart.) so that it will be hard to crash (the
probability of something giong on on a 30 MB / on a 8GB disk is less then if
the / was 1.5 GIG), and fast to fsck.

best regards,
Marcin

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

2.2.10 network "timer" problems with Interbase

[Marcin Owsiany]

Hi

a really weird problem:

I am using Interbase 5.0 SQL server and client on Debian 2.1 with 2.2.10 kernel.
However there seems to be a bug either in interbase or in kernel.

For those who are not familiar with SQL servers: an sql  server is started
by inetd after it gets a connection on port 3050 from an sql client.
The server then accepts querries for any databases and the connection ends
after the session finishes.

The problem is that the client freezes when left inactive for some time. At the
beginning i suspected DNS timeouts, but then made sure the system does not
try to use any DNS by changing 
order hosts,bind
in host.conf to
order hosts

(though it would be strange/difficult because there was no nameserver entry in 
resolv.conf)

Anyway, this did not help so i reread NET-3-HOWTO,
/usr/src/linux/Documentation/networking/routing.txt and fine-tuned routing
tables.

This did not help either!

Then I came across such command: "netstat -n -A inet -e -o"

here's what i've found:

(BTW: here's what netstat(8) has to say about timer:

Timer
   (this needs to be written)
:(


here is the state before connection

Thu Aug 19 13:03:33 CEST 1999
Active Internet connections (w/o servers)
Proto Recv-Q Send-Q Local Address   Foreign Address State   
User   Timer
udp0  0 192.168.2.144:138   0.0.0.0:*   
   off (0.00/0) 0
udp0  0 192.168.2.144:137   0.0.0.0:*   
   off (0.00/0) 0

and here after connecting:

Thu Aug 19 13:03:57 CEST 1999
Active Internet connections (w/o servers)
Proto Recv-Q Send-Q Local Address   Foreign Address State   
User   Timer
tcp0  0 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  off (0.00/0)
tcp0  0 192.168.2.144:1030  192.168.2.144:3050  ESTABLISHED 
0  off (0.00/0)
udp0  0 192.168.2.144:138   0.0.0.0:*   
   off (0.00/0) 0
udp0  0 192.168.2.144:137   0.0.0.0:*   
   off (0.00/0) 0

((from now on i'll leave only the needed info))

the output looks all the same for a minute counting from sending last byte
to the server:
Thu Aug 19 13:04:36 CEST 1999
tcp0  0 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  off (0.00/0)
tcp0  0 192.168.2.144:1030  192.168.2.144:3050  ESTABLISHED 
0  off (0.00/0)

after a minute the timer starts:
Thu Aug 19 13:04:42 CEST 1999
tcp0  40660 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  on (0.09/1)
tcp 3992  0 192.168.2.144:1030  192.168.2.144:3050  ESTABLISHED 
0  off (0.00/0)

i do not know what the timer indicates, but i figured out its format is:
(x.xx/y)
where x.xx is number of seconds and y is some strange number
on next shots you can see it works like this:
"y" is set to 1 and x.xx to some small amount (a second, i guess)
then the seconds part decrements in real time and when it reaches zero, "y"
is incremented by one and imediately "x.xx" is set again to some time, this
time it's higher value.
After the "x.xx" part reaches zero again, "y" is incremented by one and
"x.xx" to some even larger amount, and the cycle repeats.

In case my description was too complicated: here's a sample output:
connection
(0.00/0) for a minute, then
(1.00/1)
(0.80/1)
(0.70/1)
(0.60/1)
... and so on...
(0.10/1)
(0.00/2)
(2.00/2)
(1.90/2)
(1.80/2)
(1.70/2)
...and so on until
(0.10/2)
(5.00/3)
(4.90/3)
... until
(0.10/3)
(0.00/4)
(10.00/4)
(9.90/4)
(9.80/4)
...and so on, while the part before slash is two times larger each time the
part after slash is ++

also the values in Recv-Q and Send-Q of server and client have changed.

Thu Aug 19 13:04:45 CEST 1999
tcp0  40660 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  on (1.98/4)
tcp 3992  0 192.168.2.144:1030  192.168.2.144:3050  ESTABLISHED 
0  off (0.00/0)
Thu Aug 19 13:04:50 CEST 1999
tcp0  40660 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  on (4.01/5)
tcp 3992  0 192.168.2.144:1030  192.168.2.144:3050  ESTABLISHED 
0  off (0.00/0)

then i send a command from the client to the server. the values of Recv- and
Send- Qs change and client freezes (!!!) until the current x.xx value reaches
zero! So when I'm "lucky" enough this can be a few minutes!!!

Then the server responds as if everything was OK

Thu Aug 19 13:05:13 CEST 1999
tcp   64  40660 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  on (19.49/7)
tcp0  0 192.168.2.144:1030  192.168.2.144:3050  ESTABLISHED 
0  off (0.00/0)
Thu Aug 19 13:05:16 CEST 1999
tcp   64  40660 192.168.2.144:3050  192.168.2.144:1030  ESTABLISHED 
0  on (16.07/7)
tcp0  0 192.168.2

GNOME packages for slink

[Marcin Owsiany]

Hi,
Does anybody remember the URL to those unofficial GNOME packages for slink?
they were released some two months ago, i believe...

Marcin

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: compiling error: xauth not found

[Marcin Owsiany]

On Sat, Aug 21, 1999 at 06:24:25PM +0200, J Horacio MG wrote:
> $ dpkg -l | less
> 
> ii  xbase-clients   3.3.2.3a-11miscellaneous X clients
> 
> may be xauth is somewhere else?
> 

no
check if it wasn't accidentally removed (it should be in
/usr/X11R6/bin/xauth), and if it is there, check where does the "configure
script looks for it) (maybe it accepts a --with-xauth= option or similar?)

hope this helps

-- 

---------
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: How determine all users logged into linux...

[Marcin Owsiany]

On Sat, Aug 21, 1999 at 09:20:25AM -0700, André Bell wrote:
> How do I determine all users who are logged into debain while they are
> logged in?
> 
> I've finally gotten apache up and running currectly and have logged in
> locally as well as logged in via telnet under a different user name at the
> same time.
> 
> Now I'd like to "see" each user, and ideally pick and choose who to log off
> without logging off other users. 
> 
> My "Special Edition Using Linux" doesn't tell how to log off a telnet'ed
> user. It only tells how to log oneself off.
> 
> Thanks

the ways to see logged-in users:

who
w
finger

the only way to "log off" a user is to kill his/her shell
you can see what pid a shell running on a certain vterminal has by running
"ps axf"
then kill a user's pid with 

kill 

hope this helps

-- 

-
Marcin Owsiany
[EMAIL PROTECTED]
-

Re: How to resize root partition?

[Marcin Owsiany]

Hello!

On Sat, Aug 07, 1999 at 09:06:05PM +0800, a wrote:
> I have Debian 1.2 on hda3 (350M). Is it possible to resize it to 300M?

as far as i know - it's not possible..

> If this isn't possible, is it possible to cp hda3 to hda2 and run from hda2?
this is possible (actually i did something similar yesterday :-):

mke2fs /dev/hda2
mount /dev/hda2 /mnt
cd /mnt
cp -axv /* .
edit /mnt/etc/lilo.conf and change hda3-s to hda2-s
insert a floppy into drive0 and:
cp /vmlinuz /dev/fd0
rdev /dev/fd0 /dev/hda2
reboot
lilo

that's it!! :)
> 
> Please reply to [EMAIL PROTECTED] because I will leave the list in a minute.
> 
> 
> -- 
> Unsubscribe?  mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
> 
[--- snip ---]

-- 

-----
Marcin Owsiany
[EMAIL PROTECTED]
-