Need SCO bootdisk

Chip

unread,

Feb 5, 2002, 10:14:24 PM2/5/02

to

Hi,
I have a few servers with SCO OpenServer Release 5
installed on them but don't have the root password or
install CD.

I can't get to single user mode without a boot
disk (or so I've been told).

I have tried to mount one of the disks on several other
Linux machines (Slackware, Red Hat) so I could edit
the password file but can't read the file system.

When I fdisk the drive in Linux I see:
Device Boot Begin Start End Blocks ID System
sda4 * 1 1 1020 4241128 63 GNU HURD

I have looked all over the net for a boot disk image
but have not been able to find one.

I hate to fdisk the drive and waste a good OS just
to reload Linux.

Any help would be greatly appreciated

Chip Carpenter
ccarp...@erols.com

Tony Lawrence

unread,

Feb 6, 2002, 4:53:08 AM2/6/02

to

Chip wrote:

> Hi,
> I have a few servers with SCO OpenServer Release 5
> installed on them but don't have the root password or
> install CD.
>
> I can't get to single user mode without a boot
> disk (or so I've been told).

Well, you were told wrong, but getting to single user mode won't help you.

>
> I have tried to mount one of the disks on several other
> Linux machines (Slackware, Red Hat) so I could edit
> the password file but can't read the file system.

Correct. At least AFAIK right now, there is no Linux support for HTFS
file systems- however, I keep hearing rumors that such does exist- if
anyone does know of this, please let me know.

>
> When I fdisk the drive in Linux I see:
> Device Boot Begin Start End Blocks ID System
> sda4 * 1 1 1020 4241128 63 GNU HURD
>
> I have looked all over the net for a boot disk image
> but have not been able to find one.

Rather you should have looked for "lost root password SCO"

See http://pcunix.com/SCOFAQ/scotec1.html#root_1

--
Tony Lawrence
SCO/Linux Support Tips, How-To's, Tests and more: http://pcunix.com

Chip

unread,

Feb 6, 2002, 7:29:40 AM2/6/02

to

Thanks Tony,

I have searched lots of different sites searching
for "lost root password" but they all start with
the same thing:

"Boot the system from your emergency boot
diskettes (if you didn't make these and keep
them up to date, shame on you, but you should
be able to use N1/N2 instead"

My problem is I have no boot disks of any
kind that will read the file system so I can
mount it.

The pcunix.com site has been very helpful
in the past, and would be exactly what I
would try now but I can't get to the files to
edit them.

Thanks Again
Chip

"Tony Lawrence" <to...@pcunix.com> wrote in message
news:3C60FD70...@pcunix.com...

Tony Lawrence

unread,

Feb 6, 2002, 7:44:01 AM2/6/02

to

Chip wrote:

> Thanks Tony,
>
> I have searched lots of different sites searching
> for "lost root password" but they all start with
> the same thing:
>
> "Boot the system from your emergency boot
> diskettes (if you didn't make these and keep
> them up to date, shame on you, but you should
> be able to use N1/N2 instead"
>
> My problem is I have no boot disks of any
> kind that will read the file system so I can
> mount it.
>
> The pcunix.com site has been very helpful
> in the past, and would be exactly what I
> would try now but I can't get to the files to
> edit them.

Did you even bother to read it or did you just assume your answer wasn't
there?

The faq includes instructions for booting from install media and
crashing out to a shell.

Bill Vermillion

unread,

Feb 6, 2002, 12:53:52 PM2/6/02

to

In article <3C61257D...@pcunix.com>,

Tony Lawrence <to...@pcunix.com> wrote:
>Chip wrote:
>
>> Thanks Tony,
>>
>> I have searched lots of different sites searching
>> for "lost root password" but they all start with
>> the same thing:
>>
>> "Boot the system from your emergency boot
>> diskettes (if you didn't make these and keep
>> them up to date, shame on you, but you should
>> be able to use N1/N2 instead"
>>
>> My problem is I have no boot disks of any
>> kind that will read the file system so I can
>> mount it.
>>
>> The pcunix.com site has been very helpful
>> in the past, and would be exactly what I
>> would try now but I can't get to the files to
>> edit them.
>

>Did you even bother to read it or did you just assume your answer
>wasn't there?

>The faq includes instructions for booting from install media and
>crashing out to a shell.

He did say he had no CDs so I'll assume that he doesn't have the
floppies that go with them. If he had the CD he could make the
floppy.

--
Bill Vermillion - bv @ wjv . com

Tony Lawrence

unread,

Feb 6, 2002, 2:47:02 PM2/6/02

to

Bill Vermillion wrote:

He didn't say he had no CD, just no boot disk.

Bill Vermillion

unread,

Feb 6, 2002, 7:02:02 PM2/6/02

to

In article <3C6188A2...@pcunix.com>,

Tony - you'll have to give up decaf and go with something that has
some real oomph to it.

Here is what I clipeed from the original message.
======================

I have a few servers with SCO OpenServer Release 5
installed on them but don't have the root password or
install CD.

I can't get to single user mode without a boot
disk (or so I've been told).

======================

Unless I mis-interpret "don't" [sounds like a former president]
I interpret this to mean he has nothing but the OS on the system.

And just so you don't think I'm picking on you:

))))
))))
:::: ))))
:::: ))))
---- ))))
---- ))))
:::: ))))
:::: ))))
))))
))))

Must go along with your message about talking to your accounting
system :-)

Bill

Chip

unread,

Feb 6, 2002, 7:16:17 PM2/6/02

to

Thanks Bill,

Yes your right my first post stated:

>I have a few servers with SCO OpenServer Release 5
>installed on them but don't have the root password or
>install CD.

I guess I could have stated that I don't have any CD's
or boot disks instead of install CD, thought that was a
given.

I an not one to jump on a news group and post
a question without looking long and hard for
FAQ's or checking google first.

But I still need help with how to get into this
box.

Thanks
Chip

"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message
news:Gr4Hp...@wjv.com...

Tony Lawrence

unread,

Feb 6, 2002, 7:15:56 PM2/6/02

to

Bill Vermillion wrote:

Well, when I hear that I think "emergency boot disks".

But what difference does it make? He has what he has, no matter what
you or I think. He can buy install media from SCO or on Ebay if he
doesn't have it..

Tony Lawrence

unread,

Feb 6, 2002, 7:21:26 PM2/6/02

to

Chip wrote:

> Thanks Bill,
>
> Yes your right my first post stated:
>
>>I have a few servers with SCO OpenServer Release 5
>>installed on them but don't have the root password or
>>install CD.
>>
>
> I guess I could have stated that I don't have any CD's
> or boot disks instead of install CD, thought that was a
> given.

THEN BUY A CD.

Fer crying out loud :-)

If yiu are running a current release, you can get media kits from
Caldera very inexpensively- if it's 3.2v4.2 then E-bay is full of it.
In fact Ebay is full of 5.0.0 and 5.0.2 CD's as well- of course
sometimes these people are under the impression that the CD has real value..

Chip

unread,

Feb 6, 2002, 7:28:07 PM2/6/02

to

Hi Tony,

Yes I did read the information on the site and
it is as well explained as any I have seen. In fact
I printed and booked marked the page to use as
reference if I can come up with a way to get a boot
disk.

I do not have any CD's of any kind, just the
servers themselves. That is why I had asked
if anyone knew of a site to down load the boot
disk image.

I still need help if you have any suggestions as to how
to get one, or any other ideas.

You guys have far more experience and knowledge
than I and that's why I come here for help.

Again,
Thanks for your time and help
Chip

"Tony Lawrence" <to...@pcunix.com> wrote in message

news:3C6188A2...@pcunix.com...

Chip

unread,

Feb 6, 2002, 7:43:31 PM2/6/02

to

Thanks Tony,

I just checked ebay and they are everywhere
for next to nothing. I had not thought of them.

I was just hoping to not have to wait for ups
and all that stuff if I could find a image some-
where.

Thanks Again.

Chip
"Tony Lawrence" <to...@pcunix.com> wrote in message

news:3C61C8F2...@pcunix.com...

Tony Lawrence

unread,

Feb 6, 2002, 7:51:56 PM2/6/02

to

Chip wrote:

> Thanks Tony,
>
> I just checked ebay and they are everywhere
> for next to nothing. I had not thought of them.
>
> I was just hoping to not have to wait for ups
> and all that stuff if I could find a image some-
> where.

You don't have to wait for ups.

http://pcunix.com/consultants.html is full of people from all over the
world, many of whom probably carry these cd's in their pockets. Hire
one of 'em and you could have one tomorrow..

As someone has said, "Speed costs money- how fast do you want to go?"

Bela Lubkin

unread,

Feb 7, 2002, 5:37:09 AM2/7/02

to sco...@xenitec.on.ca, ccarp...@erols.com

Someone named Chip is looking for a boot disk for some release of SCO
Unix or OpenServer -- I've lost track of the details.

This prompted me to dredge up my old list of FTPable SCO boot disks
(couldn't find it via google, I thought they claimed they now had all of
USENET back to the beginning?) -- but I found a quote of most of it.

I've updated it by adding some new items, and fixing all the URLs so
they point to machines and directories that actually exist as of
2002-02-07. Since this is an external directory of information
maintained by other people, it's entirely subject to change.

Herewith, please find a list of boot disks for "SCO" operating systems.
It attempts to be comprehensive, but I may have missed a few. If your
exact OS flavor is missing, a "nearby" version might work. If multiple
images are listed for a particular version, read their cover letters or
try them serially.

The URLs generally point directly to the file that contains the boot
image; however, you would be wise to look in the same directory for
related materials such as cover letters.

Some of the disks listed may in fact be of such specialized purpose that
nobody could possibly use them any more...

I expect this to be archived by google and thus searchable for many
years to come.

>Bela<

URL Operating system(s)
==================================================== =====================================
N/A Xenix < 2.3.2, Xenix 286 MCA
ftp://stage.caldera.com/SLS/xnx264.n1.Z Xenix 2.3.2 286 ISA
ftp://stage.caldera.com/EFS/efs100.n1.Z Xenix 2.3.2 386 MCA 3.5" (SCSI)
ftp://stage.caldera.com/SLS/xnx232b.n1.Z Xenix 2.3.2 386 MCA 3.5" (ESDI)
ftp://stage.caldera.com/EFS/efs124.n1.Z Xenix 2.3.4 386 MCA 3.5"
ftp://stage.caldera.com/EFS/efs132.n1.Z Xenix 2.3.4 386 ISA 3.5"

ftp://stage.caldera.com/pos/demo/demofloppy/demofloppy.image SCO POS demo

Roberto Zini

unread,

Feb 7, 2002, 6:04:09 AM2/7/02

to

Chip wrote:
>
> Thanks Tony,
>
> I just checked ebay and they are everywhere
> for next to nothing. I had not thought of them.
>
> I was just hoping to not have to wait for ups
> and all that stuff if I could find a image some-
> where.
>

What about your machine ? Is it an ATAPI/EIDE or a SCSI one ?

Please try to describe the disk subsystem (in terms of controller
& disk) the best you can.

Best,
Roberto
--
---------------------------------------------------------------------
Roberto Zini email : r.z...@strhold.it
Technical Support Manager -- Strhold Evolution Division R.E. (ITALY)
---------------------------------------------------------------------
"Has anybody around here seen an aircraft carrier?"
(Pete "Maverick" Mitchell - Top Gun)

Tony Lawrence

unread,

Feb 7, 2002, 6:59:33 AM2/7/02

to

Wonderful - I'll add this to the FAQ

Bela Lubkin wrote:

Chip

unread,

Feb 7, 2002, 7:25:45 AM2/7/02

to

Thanks Bela,

This I just what I needed!!

Chip

"Bela Lubkin" <be...@caldera.com> wrote in message
news:2002020702...@mammoth.ca.caldera.com...

Chip

unread,

Feb 9, 2002, 9:42:01 AM2/9/02

to

Thanks Bela,

I have downloaded the above files and
tried to boot from them with using "tools"
at the boot prompt with no luck.

The demofloppy.image boots to a prompt but does
not have mount or edit programs to make any changes.

The oss604a.boot and others just continue on
with the install process and look for the
install media "which I don't have yet" and
I can't break out of the install to a prompt
without it making me reboot.

I did go out and buy a FreeBSD disk "SCO
coming ups from Ebay, thanks for the idea
Tony" and installed it because of the ufs
support.

Then tried to mount the SCO drive with:

mount /dev/da1 /mnt
# mount: /dev/da1 on /mnt: incorrect superblock

also tried:

mknod /dev/root b 1 42 # this worked
fsck -ofull /dev/root # fsck: illegal option --o
mount /dev/root /mnt # device not configured

When I fdisk with freebsd /dev/da1:

******* Working on device /dev/da1 *******
parameters extracted from in-core disklabel are:
cylinders=1048 heads=127 sectors/track=63 (8001 blks/cyl)

Figures below won't work with BIOS for partitions not in cyl 1
parameters to be used for BIOS calculations are:
cylinders=1048 heads=127 sectors/track=63 (8001 blks/cyl)

Media sector size is 512
Warning: BIOS sector numbering starts with sector 1
Information from DOS bootblock is:
The data for partition 1 is:
<UNUSED>
The data for partition 2 is:
<UNUSED>
The data for partition 3 is:
<UNUSED>
The data for partition 4 is:
sysid 99,(ISC UNIX, other System V/386, GNU HURD or Mach)
start 63, size 8376984 (4090 Meg), flag 80 (active)
beg: cyl 0/ head 1/ sector 1;
end: cyl 1023/ head 126/ sector 63

When I fdisk in Linux:

Device Boot Begin Start End Blocks ID System
sda4 * 1 1 1020 4241128 63 GNU HURD

Again all of your help is appreciated.
Chip

Pat Welch

unread,

Feb 9, 2002, 4:46:58 PM2/9/02

to

Hmmm.

Try following this TA for breaking out of an install 'in progress':

http://stage.caldera.com/cgi-bin/ssl_reference/?105094

See if that works for you.

--
----------------------------------------------------
Pat Welch, UBB Computer Services, a WCS Affiliate
Caldera Authorized Partner
Unix/Linux/Windows/Hardware Sales/Support
(209) 745-1401 Fax: (413) 714-2833
Nationwide pager: (800) 608-7122
E-mail: pat...@inreach.com
----------------------------------------------------

Chip

unread,

Feb 9, 2002, 9:40:29 PM2/9/02

to

Hi Pat,

The following does work fine. The
problem is that you can't get to the
"keyboard type" question without
chosing one of the install media.
(which is what I don't have).

I wonder if the boot disk gets some of
it's programs off of the install media?

I booted from the demofloppy.image disk,
which will boot you to a prompt, but could
not use mount, ect.

I then used the oss604a.boot image and when
breaking out of the install process before selecting an
verifying the install media, could not use mount,
just as the following explains.

<snip>
To break out to a shell press <F8>.
Highlight the Shell escape option and press <Enter>.
The following message is displayed:
"To return to the installation, exit the shell."
followed by the shell prompt: <Installation>
If you break out of the installation prior to the
initialization of the CD-ROM screen, certain
tools won't be available, such as mounting the
hard disk, fsck, file listing, and so on.
Once you get past identifying the installation
media device, and are at the point where you
are asked for the keyboard type, press the
<F8> key. Go to the Shell escape.

In order to access the root filesystem you will have
to re-create the device node:
# mknod /dev/root b 1 42

To run a filesystem check on the root filesystem:
# fsck -o full /dev/root

To mount the root filesystem:
# mount /dev/root /mnt
</snip>

It's a great page and should work but
not without some type of install media.

Thanks for jumping in to lend a hand Pat.

Bela Lubkin

unread,

Feb 10, 2002, 5:13:00 AM2/10/02

to sco...@xenitec.on.ca

Chip wrote:

> The following does work fine. The
> problem is that you can't get to the
> "keyboard type" question without
> chosing one of the install media.
> (which is what I don't have).
>
> I wonder if the boot disk gets some of
> it's programs off of the install media?

Yes, it does. You're going to have to wait for UPS to deliver the
install disks you ordered.

Before OpenServer 5.0.0, the boot disks were a set of N1 (boot and
kernel) and N2 (root filesystem, including things like the mount
command). In 5.0.x, there is a single boot disk (N00) which contains
boot, kernel, and a minimal root filesystem. That root filesystem has
enough code to ask you where to find the install media, mount that, and
proceed.

In fact, it does have a mount command or it wouldn't be able to access
the install media. But the environment is so poor, you would have a
hard time doing anything useful even if you did get to a shell prompt
and had access to mount.

> I booted from the demofloppy.image disk,
> which will boot you to a prompt, but could
> not use mount, ect.

I only included that because I was trying to make a canonical list of
SCO boot disks, for future reference via google.

> I then used the oss604a.boot image and when
> breaking out of the install process before selecting an
> verifying the install media, could not use mount,
> just as the following explains.
>
> <snip>
> To break out to a shell press <F8>.
> Highlight the Shell escape option and press <Enter>.
> The following message is displayed:
> "To return to the installation, exit the shell."
> followed by the shell prompt: <Installation>
> If you break out of the installation prior to the
> initialization of the CD-ROM screen, certain
> tools won't be available, such as mounting the
> hard disk, fsck, file listing, and so on.

When you get to that shell prompt, you _are_ running a Unix shell, which
is a pretty powerful thing. There is no `ls` command, but you can use
"cd" and "echo" to good effect. Try:

cd /
echo *
echo */*
echo */*/*

etc. If any particular list gets to be too much, use more selective
expressions like:

echo etc/*

or whatever.

The environment contained on the boot floppy _does_ have the ability to
mount things, else it would never get access to the install CD. It may
be that this is done through direct system calls, but I'm pretty sure
there _is_ a mount binary out there somewhere.

Once you find any interesting binaries, you can add their locations to
your path, to make life easier. e.g. "PATH=$PATH:/rootFS/etc", if that
turns out to be a place where binaries live.

> In order to access the root filesystem you will have
> to re-create the device node:
> # mknod /dev/root b 1 42
>
> To run a filesystem check on the root filesystem:
> # fsck -o full /dev/root
>
> To mount the root filesystem:
> # mount /dev/root /mnt

The N00 environment ought to have mknod and mount, but not fsck. If you
can mount the root filesystem without the kernel complaining, it doesn't
need an fsck; otherwise, you need that install media.

>Bela<

Bela Lubkin

unread,

Feb 10, 2002, 5:13:44 AM2/10/02

to sco...@xenitec.on.ca

Chip wrote:

> I did go out and buy a FreeBSD disk "SCO
> coming ups from Ebay, thanks for the idea
> Tony" and installed it because of the ufs
> support.
>
> Then tried to mount the SCO drive with:
>
> mount /dev/da1 /mnt
> # mount: /dev/da1 on /mnt: incorrect superblock

I don't know FreeBSD device nodes, but that probably refers to an entire
partition, the equivalent of /dev/hd0a on OSR5. That isn't where the
filesystem lives. You would need a device driver that knew how to see
SCO divisions (which live inside a partition).

> also tried:
>
> mknod /dev/root b 1 42 # this worked
> fsck -ofull /dev/root # fsck: illegal option --o

These are far too OSR5-specific. The mknod worked, but who knows what
major number 1 is in FreeBSD? This is more likely to cause a
spectacular crash than anything useful.

The fsck option exists on a number of OSes and filesystems, but on many
systems you must explicitly state the type of the filesystem to be
checked -- either by running a different binary (`ufs_fsck ...`) or by a
flag (`fsck -F ufs ...`).

> mount /dev/root /mnt # device not configured

that's probably the error message for ENODEV, "no such device or
address" on OSR5. It means that either major #1 doesn't exist, or it
doesn't respond to minor #42. Since block drivers are much rarer than
character, it's probably that major #1 isn't a block driver, so
block-major-1 doesn't exist.

> When I fdisk with freebsd /dev/da1:
>
> ******* Working on device /dev/da1 *******
> parameters extracted from in-core disklabel are:
> cylinders=1048 heads=127 sectors/track=63 (8001 blks/cyl)
>
> Figures below won't work with BIOS for partitions not in cyl 1
> parameters to be used for BIOS calculations are:
> cylinders=1048 heads=127 sectors/track=63 (8001 blks/cyl)
>
> Media sector size is 512
> Warning: BIOS sector numbering starts with sector 1
> Information from DOS bootblock is:
> The data for partition 1 is:
> <UNUSED>
> The data for partition 2 is:
> <UNUSED>
> The data for partition 3 is:
> <UNUSED>
> The data for partition 4 is:
> sysid 99,(ISC UNIX, other System V/386, GNU HURD or Mach)
> start 63, size 8376984 (4090 Meg), flag 80 (active)
> beg: cyl 0/ head 1/ sector 1;
> end: cyl 1023/ head 126/ sector 63

Amazing that after all these years and however many SCO Xenix, Unix and
OpenServer systems in the field (well over a million licenses sold),
they still can't bring themselves to name it. Also amazing the GNU HURD
chose the same ID as SysV Unix -- something like 10 years after SysV
Unix had already claimed it...

>Bela<

Tony Lawrence

unread,

Feb 10, 2002, 7:09:09 AM2/10/02

to

Bela Lubkin wrote:

> Amazing that after all these years and however many SCO Xenix, Unix and
> OpenServer systems in the field (well over a million licenses sold),
> they still can't bring themselves to name it. Also amazing the GNU HURD
> chose the same ID as SysV Unix -- something like 10 years after SysV
> Unix had already claimed it...
>

Yeah, doesn't it just frost you? Far and away SCO had more Unix out
there than anybody else, but it always got ignored- didn't exist as far
as anyone else was concerned.

Now that Caldera owns it, I can ask: did some member of SCO management
do something to really tick everyone else off at some point in the misty
past? Could that be the reason SCO "never got no respect"?

Or was it just the Sysv purists on one side and the BSD bigots on the
other who just wanted to pretend SCO didn't exist?

Chip

unread,

Feb 10, 2002, 7:32:48 AM2/10/02

to

Great information Bela,

I'll spend the afternoon working with
the N00 disk and see what I can learn.

Thanks Again

Chip
"Bela Lubkin" <be...@caldera.com> wrote in message

news:2002021002...@mammoth.ca.caldera.com...

Bill Vermillion

unread,

Feb 10, 2002, 11:30:40 AM2/10/02

to

In article <2002021002...@mammoth.ca.caldera.com>,

Bela Lubkin <be...@caldera.com> wrote:
>Chip wrote:
>
>> I did go out and buy a FreeBSD disk "SCO
>> coming ups from Ebay, thanks for the idea
>> Tony" and installed it because of the ufs
>> support.

>> Then tried to mount the SCO drive with:

>> mount /dev/da1 /mnt # mount: /dev/da1 on /mnt: incorrect
>> superblock
>
>I don't know FreeBSD device nodes, but that probably refers to an
>entire partition, the equivalent of /dev/hd0a on OSR5. That isn't
>where the filesystem lives. You would need a device driver that
>knew how to see SCO divisions (which live inside a partition).

Just an FYI since I run web/mail servers on FreeBSD.

You are right in your assumption that /dev/da0 is the entire
drive. Typically root is mounted on /dev/da0s1a and /usr
is on /dev/da0s1e [s1 - slice 1 - where a slice is a
partition in other environments. Can be confusing]

>> also tried:

>> mknod /dev/root b 1 42 # this worked
>> fsck -ofull /dev/root # fsck: illegal option --o

>These are far too OSR5-specific. The mknod worked, but who knows what
>major number 1 is in FreeBSD? This is more likely to cause a
>spectacular crash than anything useful.

There is only one device with a major 1 and it is 1,0 and is tty.

I've seen other instances where people try to mount drives for
recovery on other systems and wind up destroying any chance they
had for recovery.

>The fsck option exists on a number of OSes and filesystems, but on
>many systems you must explicitly state the type of the filesystem
>to be checked -- either by running a different binary (`ufs_fsck
>...`) or by a flag (`fsck -F ufs ...`).

The fsck in the BSD is pretty close to the V's. It will however
create a lost+found if it does not exist, and if lost+found is not
large enough it expands it. That is counter to the SysV method -
and I have no clue as to how this is done to keep from overwriting
good blocks. The V appoach on the surface appears to be more
conservative - but that's only a guess - and I think most of us who
have used these systems for awhile have cursed [either silently or
out loud] whoever installed the system originally when we get
called to try to fix something.

>> mount /dev/root /mnt # device not configured

No /dev/root in BSD land either.

>that's probably the error message for ENODEV, "no such device or
>address" on OSR5. It means that either major #1 doesn't exist, or it
>doesn't respond to minor #42. Since block drivers are much rarer than
>character, it's probably that major #1 isn't a block driver, so
>block-major-1 doesn't exist.

And there are no block devices in BSD anymore. Everything is
a character device. Another gotcha.

>> When I fdisk with freebsd /dev/da1:

>> ******* Working on device /dev/da1 *******
>> parameters extracted from in-core disklabel are:
>> cylinders=1048 heads=127 sectors/track=63 (8001 blks/cyl)

....

>> The data for partition 4 is:
>> sysid 99,(ISC UNIX, other System V/386, GNU HURD or Mach)
>> start 63, size 8376984 (4090 Meg), flag 80 (active)
>> beg: cyl 0/ head 1/ sector 1;
>> end: cyl 1023/ head 126/ sector 63

>Amazing that after all these years and however many SCO Xenix, Unix
>and OpenServer systems in the field (well over a million licenses
>sold), they still can't bring themselves to name it. Also amazing
>the GNU HURD chose the same ID as SysV Unix -- something like 10
>years after SysV Unix had already claimed it...

Well at least the SCO systems fit in the other System V/386 category
- and for awhile there were a lot of those too. At least six come
to mind at the moment. The GNU/HURD thing also always struck me as
strange.

Bill Vermillion

unread,

Feb 10, 2002, 11:52:52 AM2/10/02

to

In article <3C666355...@pcunix.com>,

Tony Lawrence <to...@pcunix.com> wrote:
>Bela Lubkin wrote:
>
>
>> Amazing that after all these years and however many SCO Xenix, Unix and
>> OpenServer systems in the field (well over a million licenses sold),
>> they still can't bring themselves to name it. Also amazing the GNU HURD
>> chose the same ID as SysV Unix -- something like 10 years after SysV
>> Unix had already claimed it...

>Yeah, doesn't it just frost you? Far and away SCO had more Unix out
>there than anybody else, but it always got ignored- didn't exist as
>far as anyone else was concerned.

I was running SysV systems on iNTEL devices before SCO ever brought
forth their first Unix implementation. The SysV2 was realy raw as it
was one where you had to add all the lines in the /etc/password file
by hand, and make sure you got all the :'s correct, etc. That was
from MicroPort - which I think was the first iNTEL based SysV2. When
I used Esix - which was V.3 - it was about a year before SCO came
out with their V.3 based implementation. So you can't blame someone
for not mentioning something that hadn't yet existed. Of course
Linux came along after that so your point is valid there :-)

>Now that Caldera owns it, I can ask: did some member of SCO
>management do something to really tick everyone else off at some
>point in the misty past? Could that be the reason SCO "never got no
>respect"?

SCO never had a lot of respect in the Unix area. I went to a lot of
local Unix meetings and I think I was probably one of the only
people there who used SCO.

These were people running Vaxes, HPs, DG's, you name it - there
were only 70+ versions of Unix at that time. SCO was a Xenix
system at that time and perhaps the anti-Gates philosophy had
migrated itself to Xenix. The IBM Xenix was about the worst and
buggiest *ix implementation I ever had the displeasure to touch!
And by 1980 - even before the PC - Gates had PO'ed a lot of the
computing world.

>Or was it just the Sysv purists on one side and the BSD bigots on the
>other who just wanted to pretend SCO didn't exist?

The Unix meetings I mention above probably had more SysV people
there. Of course the SunOS crowd was BSD dervived, but I think
there were more SysV people there than BSD. Sometimes we even had
meetings in the AT&T building, the IBM building, the HP building
and we also held one at the AT&T chip building plant and the NCR
manufacturing faciltiy. The map-making part of AAA was also all
SysV based as I recall.

Until SCO brought out the Unix implementation the Xenix would only
support 16 users - so that one reason it was looked down upon from
people in that group. It was thought of more along the lines of a
small car than a large truck.

One of the constant grumblings was that which was forced upon SCO
by licensing issues and that was that only the base system was
there, and you had to purchase the development system and text
processing system separately. I heard that a lot from others
who were using SysV iNTEL based systems, who were independant of
the above group.

But SCO had the most business appications and ran extremely well,
and most of the these people were in engineering environments.
[One person had his pine-board computer. An S-100 on a 1 foot
square piece of wood - naked to the world - an keep your hands away
from the power supply. Another in the group in about 1980
fabricated his own CPU from discrete parts. These were the people
who were anti-SCO. At least those are my observations]

Bill

Bela Lubkin

unread,

Feb 10, 2002, 1:45:39 PM2/10/02

to sco...@xenitec.on.ca

Tony Lawrence wrote:

I've always assumed it was just the Microsoft connection. MS did the
very earliest porting work, the basic hardware drivers that let AT&T 7th
Edition and then System III Unix work on 8088 IBM PC machines. Pretty
early on, they contracted SCO to continue the porting work. SCO got
into the marketing and sales after a while, but Microsoft retained
partial ownership of the company. The Unix community continued to
remember and think of SCO as some sort of Microsoft subsidiary.

>Bela<

Bela Lubkin

unread,

Feb 10, 2002, 1:51:32 PM2/10/02

to sco...@xenitec.on.ca

Bill Vermillion wrote:

> In article <2002021002...@mammoth.ca.caldera.com>,
> Bela Lubkin <be...@caldera.com> wrote:

> >The fsck option exists on a number of OSes and filesystems, but on
> >many systems you must explicitly state the type of the filesystem
> >to be checked -- either by running a different binary (`ufs_fsck
> >...`) or by a flag (`fsck -F ufs ...`).
>
> The fsck in the BSD is pretty close to the V's. It will however
> create a lost+found if it does not exist, and if lost+found is not
> large enough it expands it. That is counter to the SysV method -
> and I have no clue as to how this is done to keep from overwriting
> good blocks. The V appoach on the surface appears to be more
> conservative - but that's only a guess - and I think most of us who
> have used these systems for awhile have cursed [either silently or
> out loud] whoever installed the system originally when we get
> called to try to fix something.

Some of OSR5's current fsck programs _will_ create and expand lost+found
as necessary. It's reall a question of how confident the author of each
fsck is. As long as he's _sure_ he's found unused blocks and space for
the directory entry in the root directory, it's no big deal.

Hmmm, looks like the HTFS fsck will do it; DTFS fsck hasn't even heard
of the name "lost+found"; the other Unix fsck's (S51K, XENIX) will just
complain; the DOS fsck is sort of an archaic monstrosity.

> >> mount /dev/root /mnt # device not configured
>
> No /dev/root in BSD land either.

He had mknod'd it immediately before (1,42).

> >Amazing that after all these years and however many SCO Xenix, Unix
> >and OpenServer systems in the field (well over a million licenses
> >sold), they still can't bring themselves to name it. Also amazing
> >the GNU HURD chose the same ID as SysV Unix -- something like 10
> >years after SysV Unix had already claimed it...
>
> Well at least the SCO systems fit in the other System V/386 category
> - and for awhile there were a lot of those too. At least six come
> to mind at the moment. The GNU/HURD thing also always struck me as
> strange.

As far as I can tell from the mod history of <sys/dio.h>, Xenix first
started using the 0x63 partition ID in 1984. It might not have been
until 1985 that it actually shipped to customers (not sure).

>Bela<

Bela Lubkin

unread,

Feb 10, 2002, 1:57:57 PM2/10/02

to sco...@xenitec.on.ca

Bill Vermillion wrote:

> In article <3C666355...@pcunix.com>,
> Tony Lawrence <to...@pcunix.com> wrote:
> >Bela Lubkin wrote:
> >
> >
> >> Amazing that after all these years and however many SCO Xenix, Unix and
> >> OpenServer systems in the field (well over a million licenses sold),
> >> they still can't bring themselves to name it. Also amazing the GNU HURD
> >> chose the same ID as SysV Unix -- something like 10 years after SysV
> >> Unix had already claimed it...
>
> >Yeah, doesn't it just frost you? Far and away SCO had more Unix out
> >there than anybody else, but it always got ignored- didn't exist as
> >far as anyone else was concerned.
>
> I was running SysV systems on iNTEL devices before SCO ever brought
> forth their first Unix implementation. The SysV2 was realy raw as it
> was one where you had to add all the lines in the /etc/password file
> by hand, and make sure you got all the :'s correct, etc. That was
> from MicroPort - which I think was the first iNTEL based SysV2. When
> I used Esix - which was V.3 - it was about a year before SCO came
> out with their V.3 based implementation. So you can't blame someone
> for not mentioning something that hadn't yet existed. Of course
> Linux came along after that so your point is valid there :-)

The first releases of SCO Xenix were based on AT&T 7th Edition and then
System III, back in 1984. But even then they used the 0x63 partition
ID. MicroPort's '286 SysV was released in early '86, I believe. SCO's
SysV port (Xenix System V 2.1.0 or so) was around the same timeframe,
with the '386 version a year or two later.

> Until SCO brought out the Unix implementation the Xenix would only
> support 16 users - so that one reason it was looked down upon from
> people in that group. It was thought of more along the lines of a
> small car than a large truck.

Was there really a 16-user limitation? I can only think of license
reasons for that, not software...

> One of the constant grumblings was that which was forced upon SCO
> by licensing issues and that was that only the base system was
> there, and you had to purchase the development system and text
> processing system separately. I heard that a lot from others
> who were using SysV iNTEL based systems, who were independant of
> the above group.

SCO was forced to unbundle the devsys and text processing portions due
to unfavorable license agreements with MS and AT&T, respectively. The
royalties due on each of those portions would have made the overall
price of the OS unacceptable. You could argue (and I would agree) that
SCO should have made better royalty agreements with MS & AT&T, initially
_or_ by renegotiating. But it didn't happen.

By the OSR5 timeframe, when we finally got rid of the MS-based
development system, the idea of selling the DS separately was well
entrenched, and persists to date (though you can get really deep
discounts by joining a developer's program, which is either free or
cheap -- I've lost track). And the text processing package had become
almost completely irrelevant.

>Bela<

Tony Lawrence

unread,

Feb 10, 2002, 2:29:55 PM2/10/02

to

Bela Lubkin wrote:

> Bill Vermillion wrote:
>

>
>>Until SCO brought out the Unix implementation the Xenix would only
>>support 16 users - so that one reason it was looked down upon from
>>people in that group. It was thought of more along the lines of a
>>small car than a large truck.
>>
>
> Was there really a 16-user limitation? I can only think of license
> reasons for that, not software...

I don't remember it that way- pretty sure I had 32 user Xenix systems..
it's a long ways to remember though :-)

Bill Vermillion

unread,

Feb 10, 2002, 7:01:39 PM2/10/02

to

In article <3C66CAA3...@pcunix.com>,

Tony Lawrence <to...@pcunix.com> wrote:
>Bela Lubkin wrote:
>
>> Bill Vermillion wrote:
>>
>
>>
>>>Until SCO brought out the Unix implementation the Xenix would only
>>>support 16 users - so that one reason it was looked down upon from
>>>people in that group. It was thought of more along the lines of a
>>>small car than a large truck.
>>>

>> Was there really a 16-user limitation? I can only think of license
>> reasons for that, not software...

>I don't remember it that way- pretty sure I had 32 user Xenix
>systems.. it's a long ways to remember though :-)

You are probably right. I could have been confusing it with the
16MB memory limitation. I never had a Xenix system with more than
about 10 users.

Bill Vermillion

unread,

Feb 10, 2002, 7:00:22 PM2/10/02

to

In article <2002021010...@mammoth.ca.caldera.com>,

Tony's comment was on SCO SysV Unix. I'm well aware of the early
Xenix uses as I maintained several machines with it. The MicroPort
time frame is correct as it was being promoted at the 1986 summer
Usenix conference where I first saw. They were promoting it along
the line of 'buy this hard drive and get Unix free'. This was
also about the same time I'd see your posts on the Dr. Dobbs forum.
[some of us still remember!]

>> Until SCO brought out the Unix implementation the Xenix would only
>> support 16 users - so that one reason it was looked down upon from
>> people in that group. It was thought of more along the lines of a
>> small car than a large truck.

>Was there really a 16-user limitation? I can only think of license
>reasons for that, not software...

That sticks in my memory - but I could be wrong on that one.

>> One of the constant grumblings was that which was forced upon SCO
>> by licensing issues and that was that only the base system was
>> there, and you had to purchase the development system and text
>> processing system separately. I heard that a lot from others
>> who were using SysV iNTEL based systems, who were independant of
>> the above group.

>SCO was forced to unbundle the devsys and text processing
>portions due to unfavorable license agreements with MS and AT&T,
>respectively. The royalties due on each of those portions would
>have made the overall price of the OS unacceptable. You could
>argue (and I would agree) that SCO should have made better royalty
>agreements with MS & AT&T, initially _or_ by renegotiating. But it
>didn't happen.

You notice I did say 'forced upon SCO'. It was the others Unix
users who complained - thinking it was something that SCO did on
purpose. And who of us who was on this list in the early 1990s
will ever forget all of Larry's rants against SCO. SCO's problem
as I see it was that they were about the only pure SW vendor while
others had HW ties.

Intel even had their own brand of Unix for awhile. And maybe you
recall but was that the one that went to Kodak, which then became
Interactive, which then went to Sun. Others came and went.
SCO has always 'been there'. That's more than you can say for
other vendors who championed Unix for awhile and then quit.

Dell comes immdiately to mind - the one Larry championed so
loudly. They pushed it for awhle and then they dropped. Dell
later pushed Linux and then dropped it. Now if they would only
push MS products there might be hope ;-).

The next time I stumble across the SCO price list from 1984/5
that has the pricing for Xenix on the Apple Lisa and the Lyrix word
processing for other platforms [I think the VAX was included] I'll
scan it in. Far too many think of SCO as only working Xenix and
Unix on iNTEL but they were far more than that. Their list of
cross-assemblers for different CPUs/platforms was amazing too.
I had forgotten how broad that field was - had to well over a dozen
at that time.

>By the OSR5 timeframe, when we finally got rid of the MS-based
>development system, the idea of selling the DS separately was well
>entrenched, and persists to date (though you can get really deep
>discounts by joining a developer's program, which is either free or
>cheap -- I've lost track). And the text processing package had become
>almost completely irrelevant.

The EU suit against Micrsoft making them stop forcing the inclusion
of the Xenix code was a good thing. ISTR that it was only about
six months after that when SCO was able to drop that part. People
seem to forget the MS's licensing hurt more than just MS user.
Given the environments where SCO was used I don't know whether
a cheaper or bundled DS would have been beneficial to the business
side or not.

About the only thing the text-processing was being used for by that
time in many was to write/format man pages it seemed. Of course all
of SCO man pages were already formatted, probably because of this.
And writing in troff style was certainly nothing I ever felt I
would like to learn.

The best parts of AT&T text processing never seemed to make it past
AT&T. I was really impressed by the Writers Work Bench. But by
that time serious document production was being done by companies
who speicilized in it - and it really didn't belong in the OS.
FrameMaker comes immeditely to mind. That did some truly amazing
things but it's target customers were HUGE companies. Main users
were places such as drug manufacturers who would generate a
semi-truck full of paper for submission for drug approval, and
automobile manufactures. Unix really shined in those environments.

Bill Vermillion

unread,

Feb 10, 2002, 7:18:29 PM2/10/02

to

In article <2002021010...@mammoth.ca.caldera.com>,
Bela Lubkin <be...@caldera.com> wrote:
>Bill Vermillion wrote:

>> The fsck in the BSD is pretty close to the V's. It will however
>> create a lost+found if it does not exist, and if lost+found is not
>> large enough it expands it. That is counter to the SysV method -
>> and I have no clue as to how this is done to keep from overwriting

>> good blocks. ...

>Some of OSR5's current fsck programs _will_ create and expand
>lost+found as necessary. It's reall a question of how confident the
>author of each fsck is. As long as he's _sure_ he's found unused
>blocks and space for the directory entry in the root directory,
>it's no big deal.

>Hmmm, looks like the HTFS fsck will do it; DTFS fsck hasn't even heard
>of the name "lost+found"; the other Unix fsck's (S51K, XENIX) will just
>complain; the DOS fsck is sort of an archaic monstrosity.

Just shows you how out of touch I can be. I remember times when
lost+found would get full and also going through the rituals of
deciding which pieces to try to save and what to dump. I will say
this though that as the HW and SW gets more robust [and part of it
may be luck on my part] is that I've never seen disks trashed in
the last 10 years like you'd see in the mid-80s.

Thanks for the update on the HTFS.

As I recall the DTFS is quite similar to a log based file system -
a huge rotary file from my impression. An fsck there would just
have to ensure consistancy and then move uknown piece to the end,
would it not? It won't be too much longer before fsck as we have
grown to know it will have to be replaced with backround checks, or
completely new file systems that don't require it. With some
recent HW breakthroughs it looks like 500-700GB drives will be with
us in about 2 years with TB drives after that. You could probably
take your vacation time waiting for a drive of that size to
complete and fsck :-)

>> >Amazing that after all these years and however many SCO Xenix, Unix
>> >and OpenServer systems in the field (well over a million licenses
>> >sold), they still can't bring themselves to name it. Also amazing
>> >the GNU HURD chose the same ID as SysV Unix -- something like 10
>> >years after SysV Unix had already claimed it...

>> Well at least the SCO systems fit in the other System V/386 category
>> - and for awhile there were a lot of those too. At least six come
>> to mind at the moment. The GNU/HURD thing also always struck me as
>> strange.

>As far as I can tell from the mod history of <sys/dio.h>, Xenix first
>started using the 0x63 partition ID in 1984. It might not have been
>until 1985 that it actually shipped to customers (not sure).

Thanks.

Bill

Chip

unread,

Feb 10, 2002, 7:34:55 PM2/10/02

to

Ok Guys,

Your gonna love this one!!

Yes all of the attemps to mount the
drive on a FreeBSD system did screw
up my file system, (as most of you had
guessed I would ;-) but when I restarted
the SCO box it ran a disk recovery
at boot and after fixing about 45 files
owned by root it dumped me into a
shell prompt!!

I don't know how or why but I'm in.

I'll see what I can do from here.

Thanks Again
Chip

"Bela Lubkin" <be...@caldera.com> wrote in message
news:2002021002...@mammoth.ca.caldera.com...

Jean-Pierre Radley

unread,

Feb 10, 2002, 7:41:27 PM2/10/02

to ScoMisc [c.u.s.m]

Bill Vermillion propounded (on Mon, Feb 11, 2002 at 12:01:39AM +0000):

I'm rather sure that filePro for Xenix was sold with licenses for 32, 64
or more users, so the OS had to support that many eyes.minds.bodies in
the first place.

--
JP

Bela Lubkin

unread,

Feb 11, 2002, 4:04:27 AM2/11/02

to sco...@xenitec.on.ca

Bill Vermillion wrote:

> In article <2002021010...@mammoth.ca.caldera.com>,
> Bela Lubkin <be...@caldera.com> wrote:

> >The first releases of SCO Xenix were based on AT&T 7th Edition and then
> >System III, back in 1984. But even then they used the 0x63 partition
> >ID. MicroPort's '286 SysV was released in early '86, I believe. SCO's
> >SysV port (Xenix System V 2.1.0 or so) was around the same timeframe,
> >with the '386 version a year or two later.
>
> Tony's comment was on SCO SysV Unix. I'm well aware of the early
> Xenix uses as I maintained several machines with it. The MicroPort
> time frame is correct as it was being promoted at the 1986 summer
> Usenix conference where I first saw. They were promoting it along
> the line of 'buy this hard drive and get Unix free'. This was
> also about the same time I'd see your posts on the Dr. Dobbs forum.
> [some of us still remember!]

Bet you didn't know this part: I financed MicroPort's '86 trip to
Usenix. For a few thousand dollars I bought 1% of the company and they
got to go to the show and promote their product. This was years before
I got personally involved with Unix -- I'd fooled with BSD for a couple
of days in college, but didn't really start using it until '87 or '88.
Started work at SCO in '89 and have been here ever since (modulo
buyouts). At the time I had just quit a Support job at Borland and was
doing consulting jobs for a barcode software company.

The 1% stake in MicroPort became worthless, of course, but they returned
about 70% of the cash about a year after I bought it (long before they
died), and later I got some hardware (never did anything with it).

>Bela<

Tony Lawrence

unread,

Feb 11, 2002, 5:59:02 AM2/11/02

to

Chip wrote:

> Ok Guys,
>
> Your gonna love this one!!
>
> Yes all of the attemps to mount the
> drive on a FreeBSD system did screw
> up my file system, (as most of you had
> guessed I would ;-) but when I restarted
> the SCO box it ran a disk recovery
> at boot and after fixing about 45 files
> owned by root it dumped me into a
> shell prompt!!
>
> I don't know how or why but I'm in.

Well, apparently it screwed up in a good way :-)

Not sure I weant to keep that one in my bag of tricks though :-)

Fulko Hew

unread,

Feb 11, 2002, 8:31:44 AM2/11/02

to sco...@xenitec.on.ca

Bela Lubkin <be...@caldera.com> replied (with lots of stuff prunned):

> SCO was forced to unbundle the devsys and text processing portions due
> to unfavorable license agreements with MS and AT&T, respectively. The
> royalties due on each of those portions would have made the overall
> price of the OS unacceptable. You could argue (and I would agree) that
> SCO should have made better royalty agreements with MS & AT&T, initially
> _or_ by renegotiating. But it didn't happen.
>

> By the OSR5 timeframe, when we finally got rid of the MS-based
> development system, the idea of selling the DS separately was well
> entrenched, and persists to date (though you can get really deep
> discounts by joining a developer's program, which is either free or
> cheap -- I've lost track). And the text processing package had become
> almost completely irrelevant.

I'd love to get my hands on some of the text processing package again.
(I can't seem to find my original XENIX disks anymore.)
The stuff about style checking, and readability reporting was facinating
and I haven't seen a modern day equivalent yet.

-------------------------------------------------------------------------------
Fulko Hew, Voice: 905-681-5570
Senior Engineering Designer, Fax: 905-681-5556
SITA (Burlington) Email: fu...@wecan.com
777 Walkers Line,
Burlington, Ontario, Canada, L7N 2G1

Ken Wolff

unread,

Feb 11, 2002, 9:00:48 AM2/11/02

to Fulko Hew, sco...@xenitec.on.ca

If you really want it, I have copies of the XENIX Text Processing System,
Rel 2.1 on 4 5 1/2 (real) floppy disks. Let me know if you're
interested. Of course this would assume you had an original license for
this product.

--------------------------------------------------------------
Ken Wolff
Phone: 616-957-4949 Ext: 1111
FAX: 616-957-1614
--------------------------------------------------------------

Bill Vermillion

unread,

Feb 11, 2002, 11:20:45 AM2/11/02

to

In article <2002021101...@mammoth.ca.caldera.com>,

Bela Lubkin <be...@caldera.com> wrote:
>Bill Vermillion wrote:

>> In article <2002021010...@mammoth.ca.caldera.com>,
>> Bela Lubkin <be...@caldera.com> wrote:

>> >The first releases of SCO Xenix were based on AT&T 7th Edition
>> >and then System III, back in 1984. But even then they used the
>> >0x63 partition ID. MicroPort's '286 SysV was released in early
>> >'86, I believe. SCO's SysV port (Xenix System V 2.1.0 or so) was
>> >around the same timeframe, with the '386 version a year or two
>> >later.

>> Tony's comment was on SCO SysV Unix. I'm well aware of the early
>> Xenix uses as I maintained several machines with it. The MicroPort
>> time frame is correct as it was being promoted at the 1986 summer
>> Usenix conference where I first saw. They were promoting it along
>> the line of 'buy this hard drive and get Unix free'. This was
>> also about the same time I'd see your posts on the Dr. Dobbs forum.
>> [some of us still remember!]

>Bet you didn't know this part: I financed MicroPort's '86 trip to
>Usenix. For a few thousand dollars I bought 1% of the company and they
>got to go to the show and promote their product.

Son of a gun! I only got to a couple of the Usenix conferences and
loved every minute of them. '86 was the year that Sun introduced
NFS and AT&T introduced RFS as I recall.

>This was years before I got personally involved with Unix -- I'd
>fooled with BSD for a couple of days in college, but didn't really
>start using it until '87 or '88. Started work at SCO in '89 and
>have been here ever since (modulo buyouts). At the time I had just
>quit a Support job at Borland and was doing consulting jobs for a
>barcode software company.

That was about the time I was seeing you on the Dr.Dobbs forum.

>The 1% stake in MicroPort became worthless, of course, but they
>returned about 70% of the cash about a year after I bought it
>(long before they died), and later I got some hardware (never did
>anything with it).

Well I hope my purchase of the MicroPort helped make some of that
cash come back. I have the four volumes of docs on the shelf
behind me. Never saw binders crammed so full of printing for the
runtime. The 5.25" disks are still in there with the labels
falling off. The Televide Telecat that it was running on [needed
special disks to handle their proprietary drive control that did
1:1 interleave when most of the world was a 6:1].

Thanks for the history update.