non-install report

2004-11-30 Thread Hank Barta
I've been trying to install some flavor of Debian AMD-64 to a system I
upgraded last night with little success.
Hardware
Abit AV8 with VIA K8T800 Pro/ VT8237 chipset
 SATA/RAID (not in use yet)
 VT8237 IDE
 Audio (AC-97?)
3400+ processor
512MB/1GM RAM
(recycled hardware)
Some year+ old Seagate ATA drive - 130GB
Ancient #9 video card (S3968)
Ancient Tulip Ethernet card
IDE CDROM burner

Trying any of the ISOs that looked like they made sense (anything save
the netboot) the system was horribly unstable. It ranged from
rebooting instantly when hitting return from the boot prompt to
locking up while loading modules from the CD or partitioning and
formatting the hard drive. It did not recognize the on board LAN so I
put the Tulip card in. It recognized that but DHCP did not work and
manual configuration resulted in lots of error messages to the
console.

(Incidentally, it was not clear which ISO I should be using. I
typically do a network install since I'm on cable Internet.  The
HOW-TO points to two sites for boot images and there are a variety of
ISOs with no description save the name to explain what they are.
Puzzling to me, the directory
http://debian-amd64.alioth.debian.org/debian-installer/current/cdrom/
contains no CDROM nor any indication if I am supposed to build one
from the files found there.)

The machine was so horribly unstable - crapping out ad a different
place nearly each try - that I thought that there was some problem
with the hardware. I tried booting a recent Knoppix CD and it seemed
to work a *lot* better, though it was not without problem. It
recognized the on board LAN and sound with no problem, but configured
my serial mouse systems mouse as a PS/2 mouse. (There is probably a
boot option to fix that, but I didn't bother. A text console was fine
with me.) It did also report once that KDE could not start due to
insufficient RAM (with 1GB available. ;) and once when running a
command at the shell prompt, I got a bus error.

I ran the Knoppix memory test since last night and it reported no errors.

I'm in the process of installing Sarge i386 and have just finished
rebooting. So far it is running flawlessly.

So, I'm wondering how to go about installing one of the 64 bit Debian
flavors. do I identify my hardware and build a kernel with only
support for it? I presume that some driver that doesn't belong is
leading to the instability. then it's DFS for the rest, right?

Suggestions and comments welcomed!

back to my ia32 install :( 

thanks,
hank

-- 
Beautiful Sunny Winfield




Re: Bug#283595: pure64 2GB HD install

2004-11-30 Thread Christian Perrier

 I had problems installing kernel 2.6 on Seagate ST380013AS SATA drive.
 I put old 2GB Seagate ATA HD in the box, retaining SATA drive but 
 setting BIOS to boot off old ATA drive.


Have you tried modprobe sata_sil in console 2 when you're notified
the no hard disk found ?

I'm not a specialist of these hardware stuff but this seems to be the
driver to use with your SATA controller, according to discover1-data.

You can do this undestructively so you can test again even on the
installed machine.





Re: Tyan Tiger in a 2U

2004-11-30 Thread Soenke von Stamm
Hi,

there are 2U cases where you can use low profile cards in stead of a riser. Is 
the AGP slot really needed? What do you want to use that machine for? You 
could of cause also use a PCI gfx card if the machine is intended to run as a 
server...

A decent hardware forum is here: http://www.acesharware.com/forum/


Yours,

 Sönke




Re: I need explanation on the design of debian-amd64.

2004-11-30 Thread Goswin von Brederlow
[EMAIL PROTECTED] (Bob Proulx) writes:

 Goswin von Brederlow wrote:
 [EMAIL PROTECTED] (Bob Proulx) writes:
  But if 'dchroot' is configured with the ia32-linux chroot then you can
  just say dchroot apt-get install foobar to install a 32bit package.
 
 dchroot will only work if you have a chroot.

 Yes.  In particular it won't work if you are using ia32-libs.  But
 personally I think ia32-libs is the wrong direction to go.  I believe
 the chroot method to be the better alternative.

 And if you do why bother with /emul?

 Having /emul allows me to transparently run binaries outside of the
 chroot.  All of the world is not openoffice.org-bin or mozilla-firefox
 packages.  The chroot works great for those 32-bit applications.  But
 having /emul allows me to run /net/ia32fileserver/cadroot/bin binaries
 on my 64-bit system completely transparently.  Having the chroot
 allows me to install and upgrade 32-bit binaries easily.  Having a
 64-bit base system allows me to install and upgrade 64-bit binaries
 trivially.  The best of all worlds.

 In summary, I am using the chroot to manage /emul.

 Bob

The problem is that many programs have data files in /usr/share or
/usr/lib/package or even config in /etc/. All those files will be
inside the chroot instead of outside when you run the program outside.

So the number of programs you can run outside is somewhat limited.

I see the point of having a chroot to install lib packages and then
install binaries (with --force-arch for example) outside the charoot
or something. But in general it doesn't always work. Not something for
the faint of heart.

MfG
Goswin




Re: Bug#283595: pure64 2GB HD install

2004-11-30 Thread Goswin von Brederlow
Christian Perrier [EMAIL PROTECTED] writes:

 I had problems installing kernel 2.6 on Seagate ST380013AS SATA drive.
 I put old 2GB Seagate ATA HD in the box, retaining SATA drive but 
 setting BIOS to boot off old ATA drive.


 Have you tried modprobe sata_sil in console 2 when you're notified
 the no hard disk found ?

 I'm not a specialist of these hardware stuff but this seems to be the
 driver to use with your SATA controller, according to discover1-data.

 You can do this undestructively so you can test again even on the
 installed machine.

At first I would thing that the discover1-data udeb on the install CD
is outdated but you (Norval) wrote the disk still doesn't show up
after several dist-upgrades.

You don't need to boot the install CD to test the sata_sil module, in
fact it is probably better to try it with the newer kernel you are now
running.

If manual loading of the module shows the disk (you need the scsi disk
module too) then check if you have the newest discover + data
installed and check why it doesn't load the module. Once discover is
fixed then trying a new install is worth it.


One more thing. There is no need to reinstall your system again (apart
from testing if it works). You can go into single user mode and copy
(tar | tar or rsync or whatever you prefer) the old system to the new
disk, change the bootloader config and mkinitrd config, run mkinitrd
and (for lilo) reinstall the bootloader by chrooting to the new disk.

MfG
Goswin




Re: non-install report

2004-11-30 Thread Goswin von Brederlow
Hank Barta [EMAIL PROTECTED] writes:

 I've been trying to install some flavor of Debian AMD-64 to a system I
 upgraded last night with little success.
 Hardware
 Abit AV8 with VIA K8T800 Pro/ VT8237 chipset
  SATA/RAID (not in use yet)
  VT8237 IDE
  Audio (AC-97?)
 3400+ processor
 512MB/1GM RAM
 (recycled hardware)
 Some year+ old Seagate ATA drive - 130GB
 Ancient #9 video card (S3968)
 Ancient Tulip Ethernet card
 IDE CDROM burner

 Trying any of the ISOs that looked like they made sense (anything save
 the netboot) the system was horribly unstable. It ranged from
 rebooting instantly when hitting return from the boot prompt to
 locking up while loading modules from the CD or partitioning and
 formatting the hard drive. It did not recognize the on board LAN so I
 put the Tulip card in. It recognized that but DHCP did not work and
 manual configuration resulted in lots of error messages to the
 console.

Can you check your hardware please (yes I saw below). From the
hardware data the board sounds just like an Asus K8V which many of us
have running perfectly so crashes sound strange.

 (Incidentally, it was not clear which ISO I should be using. I
 typically do a network install since I'm on cable Internet.  The
 HOW-TO points to two sites for boot images and there are a variety of
 ISOs with no description save the name to explain what they are.
 Puzzling to me, the directory
 http://debian-amd64.alioth.debian.org/debian-installer/current/cdrom/
 contains no CDROM nor any indication if I am supposed to build one
 from the files found there.)

The names and meaning of the images is the same across all
architectures and described in Debian-Installer somewhere. If you have
a spare minute to dig into it, a patch to the FAQ or a small webpage for
the directory would be welcome.

In Short: /debian-installer/current contains the daily build D-I
images, which does not include a cdrom iso. The cdrom directory
contains the files needed to build a cdrom with debian-cd.

I also wrote a script (far simpler than debian-cd) that uses the same
files to build a netinst cd and I do that iregulary from time to
time. The script is in the tools directory (make-cd) if you want a
newer image.

 The machine was so horribly unstable - crapping out ad a different
 place nearly each try - that I thought that there was some problem
 with the hardware. I tried booting a recent Knoppix CD and it seemed
 to work a *lot* better, though it was not without problem. It
 recognized the on board LAN and sound with no problem, but configured
 my serial mouse systems mouse as a PS/2 mouse. (There is probably a
 boot option to fix that, but I didn't bother. A text console was fine
 with me.) It did also report once that KDE could not start due to
 insufficient RAM (with 1GB available. ;) and once when running a
 command at the shell prompt, I got a bus error.

It does sound like hardware.

 I ran the Knoppix memory test since last night and it reported no errors.

I have a dual PIII-666 system that never runs longer than 5 minutes,
except when running memtest (which runs flawless without trouble).

If something like the pci dma is broken and sometimes writes to the
wrong memory or something the system won't run long but memtest can't
find such errors. It can also just need a bios upgrade or something.

 I'm in the process of installing Sarge i386 and have just finished
 rebooting. So far it is running flawlessly.

 So, I'm wondering how to go about installing one of the 64 bit Debian
 flavors. do I identify my hardware and build a kernel with only
 support for it? I presume that some driver that doesn't belong is
 leading to the instability. then it's DFS for the rest, right?

 Suggestions and comments welcomed!

Try installing the 64bit kernel-image-2.6.9-amd64-k8 (from i386) on
your system and boot it. Does it still work then (guessing not). But
if it does you can create a 64bit chroot and see if any of the
binaries trigger the instability and so on.

 back to my ia32 install :( 

 thanks,
 hank

MfG
Goswin




CardBus problem

2004-11-30 Thread Lorenzo Milesi
Hi

I'm having problems with CardBus on my ASUS A2526KUH.
The pc has
:02:01.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ac)
:02:01.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ac)

but I can't have it work properly.
Yenta_socket module loads correctly, 
#cardctl status
Socket 1:
  3.3V CardBus card
  function 0: [ready]

seems ok, but lspci then doesn't show the connected card!
I tried different cards, also an old 3com ethernet, nothing happens.
I'm using custom 2.6.9 kernel, but I tried 2.6.10rc2 custom and 2.6.8
debian with the same result.

The CardBus works, I tried it on Windows. 
But I know it works also because I tried Knoppix3.6 with 2.4 kernel
and finally I managed to see the connected card in lspci (knoppix with
2.6 didn't work!!).
It mustn't be a Debian-kernel's problem because Knoppix uses vanilla kernel. 

I'd like to know of any of you have the same Cardbus (or pc) and if it works.
I've seen, googling around, that Ricoh bus is very commonnly used on
laptops, so I can't figure out why I am the only one experiencing this
problem.

any help welcome
thanks
maxxer 
debian-amd64 (installed pretty fine...)




Re: non-install report

2004-11-30 Thread Kenan Esau
Am Montag, den 29.11.2004, 22:13 -0600 schrieb Hank Barta:
 I've been trying to install some flavor of Debian AMD-64 to a system I
 upgraded last night with little success.
 Hardware
 Abit AV8 with VIA K8T800 Pro/ VT8237 chipset
  SATA/RAID (not in use yet)
  VT8237 IDE
  Audio (AC-97?)
 3400+ processor
 512MB/1GM RAM
 (recycled hardware)
 Some year+ old Seagate ATA drive - 130GB
 Ancient #9 video card (S3968)
 Ancient Tulip Ethernet card
 IDE CDROM burner
 
 Trying any of the ISOs that looked like they made sense (anything save
 the netboot) the system was horribly unstable. It ranged from
 rebooting instantly when hitting return from the boot prompt to
 locking up while loading modules from the CD or partitioning and
 formatting the hard drive. 

Hmm -- to me this doesn't sound like an debian-amd64-related problem.
Have you tried another memory module. Or try clocking your memory
slower. 

I've expierenced similar errors on an ia32-machine. Clocking down the
memory solved the problem.

[...]





Re: I need explanation on the design of debian-amd64.

2004-11-30 Thread Bob Proulx
Goswin von Brederlow wrote:
 The problem is that many programs have data files in /usr/share or
 /usr/lib/package or even config in /etc/. All those files will be
 inside the chroot ...

Yes.

 ... instead of outside when you run the program outside.

Any program that needs to be *installed* will need to run from the
chroot.  Agreed.  But any program that can simply be run can be run
either place.  Many (most?) random programs can be run outside the
chroot.  It really depends upon what it is you are trying to do.

 So the number of programs you can run outside is somewhat limited.

I think this is just a difference in world view between your
environment and mine.  In mine only those two binary packages I
mentioned really need to run in the chroot.  A short list.  Just
openoffice and firefox.  In the chroot.  Done.

However outside the chroot I have hundreds of CAD/EDA programs that
are legacy 32-bit applications.  They run from an NFS fileserver.
They don't actually ever get installed on the local machine[1].  But
the binaries are run just the same.  That does not even mention the
random programs that users would use from their $HOME directory or
copy to /usr/local/bin.  The list is open ended.  The choice would be
a 32-bit OS to be able to run those programs.  Except that amd64 can
run both 32-bit and 64-bit applications transparently.  That means I
can also benefit from the large memory model available in 64-bit for
the smaller number of 64-bit CAD/EDA programs that need it.  A good
result.

Another alternative I am working through the details for right now is
the exact reverse.  A 32-bit base system with a 64-bit /emul layer.
Then the small number of programs that need the larger memory space
run transparently.  But in general the system is simpler to administer
for the casual user in my lab.

 I see the point of having a chroot to install lib packages and then
 install binaries (with --force-arch for example) outside the charoot
 or something. But in general it doesn't always work. Not something for
 the faint of heart.

I never use --force-arch and would advise against doing that.  (shudder)

I agree that having a multi-root system is more complicated than
having a single-root system.  I agree that some people will be
confused by it beyond being able to administer one well.  So I also
would not recommend it for everyone.  But for most system
administrators it is a useful tool in the toolbox.  One that I use to
good effect daily.

Bob

[1] CAD vendors usually install their software in a shared
filesystem location.  Their installation processes are usually
terrible shell scripts written without any real thought.  Someone at
the vendor wrote the script to just slams files out into a directory
somewhere and they expect you to be able to run them on your machine.
We have 30+ different applications all with completely different
processes to manage them.  No thought is given to system dependencies
such as required libraries or programs.  Well, actually they just say
all of the world is a VAX, er I mean all of the world is MS-Windows,
er I mean all of the world is RHEL3.0.  You get the idea.  People like
myself who administer these environments just deal with it.


signature.asc
Description: Digital signature


Re: non-install report

2004-11-30 Thread Hank Barta
Thanks all for the help.


On Tue, 30 Nov 2004 11:14:37 +0100, Goswin von Brederlow
[EMAIL PROTECTED] wrote:

 Can you check your hardware please (yes I saw below). From the
 hardware data the board sounds just like an Asus K8V which many of us
 have running perfectly so crashes sound strange.

How to check? memtest is the only thing I see offhand. I know that a
'no error' from that is not conclusive, but if it did detect errors,
that would be conclusive.

Debian i386 seems to do fine except for not noticing the built in
ethernet. It works fine with the Tulip card.

 In Short: /debian-installer/current contains the daily build D-I
 images, which does not include a cdrom iso. The cdrom directory
 contains the files needed to build a cdrom with debian-cd.

 I also wrote a script (far simpler than debian-cd) that uses the same
 files to build a netinst cd and I do that iregulary from time to
 time. The script is in the tools directory (make-cd) if you want a
 newer image.

I'm still clueless here. I can't find the tools directory or make-cd.
I'm sure this is documented somewhere, but I cannot find it. The link
on the HOW-TO points to the section on the installer describing how to
partition a hard drive.

 
 It does sound like hardware.

I can't rule that out at this point, but I do get different results
with different S/W.

 
 Try installing the 64bit kernel-image-2.6.9-amd64-k8 (from i386) on
 your system and boot it. Does it still work then (guessing not). But
 if it does you can create a 64bit chroot and see if any of the
 binaries trigger the instability and so on.

I could not find that kernel image. I did install
kernel-image-2.6.8-9-amd64-k8 and it does not work with the Tulip
card. dhclient reports checksum errors trying to establish the
connection. I think that via-rhine is the correct module for the on
board ETH and that does not load (Gentoo liveCD can load it, but
ifconfig etho reports no such device, but I don't know Gentoo and I
may be leaving something else out.)

thanks,
hank

-- 
Beautiful Sunny Winfield




[no subject]

2004-11-30 Thread Philip Warrick
My /etc/apt/sources.list in the chroot looks like this:

deb ftp://ftp.debian.org/debian/ testing main
deb-src ftp://ftp.debian.org/debian/ testing main
# deb http://security.debian.org/ testing/updates main

as created by apt-setup (I said no the the security option, not knowing if I 
wanted this).

These files do exist in the chroot: /etc/hosts, /etc/passwd, /etc/group (not 
/etc/groups), /etc/shadow,
/proc (as described in the AMD-64 How-to).  I had done nothing with /etc/mtab 
and its entries were:

proc /proc proc rw 0 0

whereas my chroot /etc/mtab was simply

/dev/sda7 / ext3 rw,errors=remount-ro 0 0
proc /proc proc rw 0 0
devpts /dev/pts devpts rw,gid=5,mode=620 0 0
proc /var/chroot/sid-ia32/proc proc rw 0 0
/dev/sda1 /mnt/winC ntfs ro,noexec,nosuid,nodev,umask=0222 0 0
/dev/sda5 /mnt/winD ntfs ro,noexec,nosuid,nodev,umask=0222 0 0
/home /var/chroot/sid-ia32/home none rw,bind 0 0
/tmp /var/chroot/sid-ia32/tmp none rw,bind 0 0
//lmsrd01/DataDrive /mnt/lmsrd01 smbfs rw 0 0
/dev/sda6 /mnt/winE vfat rw,noexec,nosuid,gid=100,umask=000 0 0

Following ?good? advice elsewhere 
(http://hacktavista.com/howto/chroot_slackware.html) I did the following for 
the mtab file:
rm -f /etc/mtab
ln -s /proc/mounts /etc/mtab

the contents of /proc/mounts in the chroot are:

rootfs / rootfs rw 0 0
/dev2/root2 / ext3 rw 0 0
proc /proc proc rw,nodiratime 0 0
devpts /dev/pts devpts rw 0 0
proc /var/chroot/sid-ia32/proc proc rw,nodiratime 0 0
/dev/sda1 /mnt/winC ntfs 
ro,noatime,nodiratime,nosuid,nodev,noexec,uid=0,gid=0,umask=0222,nls=cp437,errors=continue,mft_zone_multiplier=1
 0 0
/dev/sda5 /mnt/winD ntfs 
ro,noatime,nodiratime,nosuid,nodev,noexec,uid=0,gid=0,umask=0222,nls=cp437,errors=continue,mft_zone_multiplier=1
 0 0
/dev2/root2 /var/chroot/sid-ia32/home ext3 rw 0 0
/dev2/root2 /var/chroot/sid-ia32/tmp ext3 rw 0 0
//lmsrd01/DataDrive /mnt/lmsrd01 smbfs 
rw,nodiratime,nosuid,nodev,file_mode=0744,dir_mode=0755 0 0
/dev/sda6 /mnt/winE vfat 
rw,nodiratime,nosuid,noexec,gid=100,fmask=,dmask= 0 0

After all that I still get the same errors.  apt is definitely able to connect 
(apt-get update downloaded some packages successfully).

Any ideas?

Thanks,

Phil


[EMAIL PROTECTED] (Bob Proulx) writes:

 Phil Warrick wrote:
 I'm going with the chroot approach then.
 
 Following the AMD64 HOWTO in section Running applications inside the 
 chroot,  I followed the instructions and then I wanted to try to run a 
 32-bit program.
 
 First I wanted to try to run some X-based program so I tried openoffice, 
 which was not yet installed.  When I tried to install it in the chroot, 
 I got the following errors below.
 [...]
 Can someone indicate what might be wrong?  I'm using kernel 
 2.6.8-9-amd-k8-smp on a dual Opteron system.

 Check your /etc/apt/sources.list file *in the chroot* and make sure it
 can get to the 32-bit archive.  Make sure that 'apt-get update' works
 in the chroot.  It looks to me like apt has no package lists at all.

 If you have 'dchroot' configured you can run the commands like this:

   dchroot editor /etc/apt/sources.list

   dchroot apt-get update
   dchroot apt-get install openoffice.org

 Bob

Don't forget /etc/hosts, /etc/passwd, /etc/groups, /etc/shadow,
/etc/mtab, /proc in the chroot.

MfG
Goswin


ia32-libs vs. ia32 chroot

2004-11-30 Thread Philip Warrick
(Sorry for the no-subject post)

My /etc/apt/sources.list in the chroot looks like this:

deb ftp://ftp.debian.org/debian/ testing main
deb-src ftp://ftp.debian.org/debian/ testing main
# deb http://security.debian.org/ testing/updates main

as created by apt-setup (I said no the the security option, not knowing if I 
wanted this).

These files do exist in the chroot: /etc/hosts, /etc/passwd, /etc/group (not 
/etc/groups), /etc/shadow,
/proc (as described in the AMD-64 How-to).  I had done nothing with /etc/mtab 
and its entries were:

proc /proc proc rw 0 0

whereas my chroot /etc/mtab was simply

/dev/sda7 / ext3 rw,errors=remount-ro 0 0
proc /proc proc rw 0 0
devpts /dev/pts devpts rw,gid=5,mode=620 0 0
proc /var/chroot/sid-ia32/proc proc rw 0 0
/dev/sda1 /mnt/winC ntfs ro,noexec,nosuid,nodev,umask=0222 0 0
/dev/sda5 /mnt/winD ntfs ro,noexec,nosuid,nodev,umask=0222 0 0
/home /var/chroot/sid-ia32/home none rw,bind 0 0
/tmp /var/chroot/sid-ia32/tmp none rw,bind 0 0
//lmsrd01/DataDrive /mnt/lmsrd01 smbfs rw 0 0
/dev/sda6 /mnt/winE vfat rw,noexec,nosuid,gid=100,umask=000 0 0

Following ?good? advice elsewhere 
(http://hacktavista.com/howto/chroot_slackware.html) I did the following for 
the mtab file:
rm -f /etc/mtab
ln -s /proc/mounts /etc/mtab

the contents of /proc/mounts in the chroot are:

rootfs / rootfs rw 0 0
/dev2/root2 / ext3 rw 0 0
proc /proc proc rw,nodiratime 0 0
devpts /dev/pts devpts rw 0 0
proc /var/chroot/sid-ia32/proc proc rw,nodiratime 0 0
/dev/sda1 /mnt/winC ntfs 
ro,noatime,nodiratime,nosuid,nodev,noexec,uid=0,gid=0,umask=0222,nls=cp437,errors=continue,mft_zone_multiplier=1
 0 0
/dev/sda5 /mnt/winD ntfs 
ro,noatime,nodiratime,nosuid,nodev,noexec,uid=0,gid=0,umask=0222,nls=cp437,errors=continue,mft_zone_multiplier=1
 0 0
/dev2/root2 /var/chroot/sid-ia32/home ext3 rw 0 0
/dev2/root2 /var/chroot/sid-ia32/tmp ext3 rw 0 0
//lmsrd01/DataDrive /mnt/lmsrd01 smbfs 
rw,nodiratime,nosuid,nodev,file_mode=0744,dir_mode=0755 0 0
/dev/sda6 /mnt/winE vfat 
rw,nodiratime,nosuid,noexec,gid=100,fmask=,dmask= 0 0

After all that I still get the same errors.  apt is definitely able to connect 
(apt-get update downloaded some packages successfully).

Any ideas?

Thanks,

Phil


[EMAIL PROTECTED] (Bob Proulx) writes:

 Phil Warrick wrote:
 I'm going with the chroot approach then.
 
 Following the AMD64 HOWTO in section Running applications inside the 
 chroot,  I followed the instructions and then I wanted to try to run a 
 32-bit program.
 
 First I wanted to try to run some X-based program so I tried openoffice, 
 which was not yet installed.  When I tried to install it in the chroot, 
 I got the following errors below.
 [...]
 Can someone indicate what might be wrong?  I'm using kernel 
 2.6.8-9-amd-k8-smp on a dual Opteron system.

 Check your /etc/apt/sources.list file *in the chroot* and make sure it
 can get to the 32-bit archive.  Make sure that 'apt-get update' works
 in the chroot.  It looks to me like apt has no package lists at all.

 If you have 'dchroot' configured you can run the commands like this:

   dchroot editor /etc/apt/sources.list

   dchroot apt-get update
   dchroot apt-get install openoffice.org

 Bob

Don't forget /etc/hosts, /etc/passwd, /etc/groups, /etc/shadow,
/etc/mtab, /proc in the chroot.

MfG
Goswin


net install image seems not be updated for a while.

2004-11-30 Thread Jin Zhao
Just had a look and found it still timestamped as 26-Oct-2004 05:56, 
but other images are updated daily.

  http://debian-amd64.alioth.debian.org/install-images/
Does everybody use net boot image to test amd64?
Jin



Re: ia32-libs vs. ia32 chroot

2004-11-30 Thread Sven Krahn
Philip Warrick wrote:
[...]
After all that I still get the same errors.  apt is definitely able to connect 
(apt-get update downloaded some packages successfully).
Any ideas?
 

In your chroot, modify your /etc/apt/sources.list:
deb ftp://ftp.debian.org/debian/ sid main contrib non-free
deb-src ftp://ftp.debian.org/debian/ sid main contrib non-free
From your earlier posts I think you have forgotten the 'contrib' and 
'non-free' entries.
Hope this helps.

--
Best regards
Sven



You are Approved

2004-11-30 Thread Tonko Michael

Dear Applicant,

Your application was processed and approved. You are eligible for $ 400,000 
with a 2.1 % rate.

Please verify your information here: http://www.ibestrate.com/?partid=aaks9

We look forward to hearing from you.

Tonko Michael, Account Manager
Joset Marketing

not interested - http://ibestrate.com/st.html


Re: Tyan Tiger in a 2U

2004-11-30 Thread Patrick Flaherty
In case anyone was wondering in ended up just using a pci graphics card
and customized (mutilated) it's back planeworks good. so if anyone
wants to put a tiger in a cheap case, get a different powersupply and
cut your own back plane plate. 

patrick
On Mon, 2004-11-29 at 13:58, Patrick Flaherty wrote:
 Has anyone found a moderatly cheap 2u rack to put a tyan tiger k8w
 (S2875) in? i've been looking around, but it seems the only way i can
 get one to fit properly is to use an agp riser and block all the pci
 slots. they fit into 3U's pretty easily, but i'd prefer to only use 3u
 cases when i'm going to need to put lots of drives in them.
 
 currently i've been beating it into a
 http://rackmountpro.com/productpage.php?prodid=1953 with a different
 power supply. running in console only mode. while this solution works it
 sorta sucks, esp if i want want to put any sort of expansion card in it.
 in addition to the riser issue, the 5/12volt connector next to the ram
 get's blocked by any power supplies with the input below the switch, as
 well as having an unfiltered 120v line running past my drives and ram.
 AWESOME. 
 
 also, if anyone knows of different server hardware oriented lists i
 could send this too i'd be happy to try them too.
 
 thanks
 patrick
 --
 ---
 Patrick Flaherty
-- 
---
Patrick Flaherty