Discussion:
Read SCO Unix file system
(too old to reply)
philo
2019-12-22 04:12:00 UTC
Permalink
I was just given a P-1 from the wife a deceased computer science college
professor.

She wanted me to recover any data on the machine.

She of course does not have the password, so I popped the drive into on
of my Linux machines.


There is one Windows partition and I was able to copy off some photos.

Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.

Of course I'd still like to have a look on the SCO partition.


Gparted sees it as an unknown file system.


Wondering what the file system is and if there is a package avail for
gparted that could read it.


Thanks
Melzzzzz
2019-12-22 04:22:44 UTC
Permalink
Post by philo
I was just given a P-1 from the wife a deceased computer science college
professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive into on
of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail for
gparted that could read it.
https://cateee.net/lkddb/web-lkddb/SYSV_FS.html
Post by philo
Thanks
--
press any key to continue or any other to quit...
U ničemu ja ne uživam kao u svom statusu INVALIDA -- Zli Zec
Svi smo svedoci - oko 3 godine intenzivne propagande je dovoljno da jedan narod poludi -- Zli Zec
Na divljem zapadu i nije bilo tako puno nasilja, upravo zato jer su svi
bili naoruzani. -- Mladen Gogala
Bud Frede
2019-12-22 16:57:38 UTC
Permalink
Post by Melzzzzz
Post by philo
I was just given a P-1 from the wife a deceased computer science college
professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive into on
of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail for
gparted that could read it.
https://cateee.net/lkddb/web-lkddb/SYSV_FS.html
Openserver used this horrible thing called htfs. It was slow, fragile,
and got fragmented after very little use. SCO said that there were no
working defrag tools, so you had to backup the partition to tape and
then re-divvy the disk and finally restore from the backup. This had to
be done every 2 or 3 months.

I don't know if anything other than Openserver supported htfs. Maybe
Unixware after SCO acquired it? I remember that they were making a big,
big push to migrate their customers to Unixware, and I think they got
Unixware to support Openserver binaries and disk volumes.
Paul
2019-12-22 06:05:10 UTC
Permalink
Post by philo
I was just given a P-1 from the wife a deceased computer science college
professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive into on
of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail for
gparted that could read it.
Thanks
https://www.win.tue.nl/~aeb/partitions/partition_types-1.html

(type) 63 Unix System V (SCO, ISC Unix, UnixWare, ...), Mach, GNU Hurd

A Unixware 7.1 partition must start below the 4GB limit.
(If the /stand/stage3.blm is located past this limit,
booting will fail with "FATAL BOOT ERROR: Can't load stage3".)

https://www.cgsecurity.org/wiki/TestDisk

Recognize, not read.

*******

This isn't much of a breadcrumb.

https://www.aplawrence.com/scofaq/FAQ_scotec1linuxfs.html

I would think this would be a "fun" project.

I thought I could make headway on a project like this,
but once the tools made it clear they weren't going
to work with a foreign filesystem, it's pretty hard
to make progress.

You can find references to data recovery for SCO disks
in Google, but when I go to a web page, the product
is missing from the page. Probably too old to continue
selling or supporting or something.

*******

https://en.freedownloadmanager.org/Windows-PC/Stellar-Phoenix-SCO-Open-Server.html

http://download.stellarinfo.com/stellar/StellarPhoenixSCOOpenserverRecovery.exe

Clean scan, unusual for such a thing.

https://www.virustotal.com/gui/file/dbb1a2e53627e381210e73953d71356b858147e07df698b97f1e9ed1d9822b27/details

It's going to tease you, if you're lucky, show some filenames, then ask for...

$400.00

at a guess, since SCO is likely considered a business system
rather than a consumer one, so soaking the customer goes
with the territory. It's not likely to be $39.95.

This definitely looks like fun. Here you can see it's
started and sitting there in WINE (of course, this is
not a workable environment for it, merely to show
what would happen if you run the installer in Windows7).

Loading Image...

Since I have no sample filesystem to test with, it's
pretty hard to say how it works.

HTH,
Paul
dillinger
2019-12-22 22:04:24 UTC
Permalink
Post by philo
I was just given a P-1 from the wife a deceased computer science college
professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive into on
of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail for
gparted that could read it.
Thanks
There are SCO and UnixWare isos available at archive.org.
You may be able to install a matching version on another disk and access
it from there.
Unless it is encrypted, of course.
I haven't tried any of this.
philo
2019-12-23 01:47:36 UTC
Permalink
Post by dillinger
Post by philo
I was just given a P-1 from the wife a deceased computer science
college professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive into
on of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail for
gparted that could read it.
Thanks
There are SCO and UnixWare isos available at archive.org.
You may be able to install a matching version on another disk and access
it from there.
Unless it is encrypted, of course.
I haven't tried any of this.
Thanks. I just may do that.
philo
2019-12-23 04:14:04 UTC
Permalink
On 12/22/19 7:47 PM, philo wrote:
<snip>
Post by philo
Post by dillinger
Post by philo
Wondering what the file system is and if there is a package avail for
gparted that could read it.
Thanks
There are SCO and UnixWare isos available at archive.org.
You may be able to install a matching version on another disk and
access it from there.
Unless it is encrypted, of course.
I haven't tried any of this.
Thanks. I just may do that.
The ISO at archive.org was corrupted but I found one at tachytelic.net

I could not install it as it's a licensed product.

Oddly when I Googled for tech help for SCO, the name that came up often,
is Bela Lubkin who


get this...is the son of the professor who's computer this is.


I think I will just return the HD to the woman I got it from and let her
son deal with it.


I find this hilarious
D***@decadence.org
2019-12-23 06:11:34 UTC
Permalink
Post by philo
<snip>
Post by philo
Post by dillinger
Post by philo
Wondering what the file system is and if there is a package
avail for gparted that could read it.
Thanks
There are SCO and UnixWare isos available at archive.org.
You may be able to install a matching version on another disk
and access it from there.
Unless it is encrypted, of course.
I haven't tried any of this.
Thanks. I just may do that.
The ISO at archive.org was corrupted but I found one at
tachytelic.net
I could not install it as it's a licensed product.
Oddly when I Googled for tech help for SCO, the name that came up
often, is Bela Lubkin who
get this...is the son of the professor who's computer this is.
I think I will just return the HD to the woman I got it from and
let her son deal with it.
I find this hilarious
Pretty cool. Think about the odds.

You have come fool circle. :-)
philo
2019-12-23 16:31:31 UTC
Permalink
On 12/23/19 12:11 AM, ***@decadence.org wrote:

<snip>
Post by D***@decadence.org
Post by philo
The ISO at archive.org was corrupted but I found one at
tachytelic.net
I could not install it as it's a licensed product.
Oddly when I Googled for tech help for SCO, the name that came up
often, is Bela Lubkin who
get this...is the son of the professor who's computer this is.
I think I will just return the HD to the woman I got it from and
let her son deal with it.
I find this hilarious
Pretty cool. Think about the odds.
You have come fool circle. :-)
Really interesting.
I'd still like to come up with a way to read the drive.

Anyone know if it could be done from Solaris or BSD?

I have the install media somewhere
dillinger
2019-12-23 17:31:55 UTC
Permalink
Post by philo
<snip>
Post by philo
The ISO at archive.org was corrupted but I found one at
tachytelic.net
I could not install it as it's a licensed product.
Oddly when I Googled for tech help for SCO, the name that came up
often, is Bela Lubkin who
get this...is the son of the professor who's computer this is.
I think I will just return the HD to the woman I got it from and
let her son deal with it.
I find this hilarious
   Pretty cool.  Think about the odds.
   You have come fool circle.  :-)
Really interesting.
I'd still like to come up with a way to read the drive.
Anyone know if it could be done from Solaris or BSD?
I have the install media somewhere
I don't know, but if you can get/make a SCO emergency boot disk you may
be able to reset the root password:
https://www.tachytelic.net/2019/01/reset-root-password-sco-openserver/
philo
2019-12-23 18:23:58 UTC
Permalink
Post by dillinger
SCO emergency boot disk
Thanks


When I Googled for an emergency boot disk, one of the links referred to
that Lubkin guy.

I hope to soon contact him in person one of these days
Paul
2019-12-23 18:02:05 UTC
Permalink
Post by philo
<snip>
Post by D***@decadence.org
Post by philo
The ISO at archive.org was corrupted but I found one at
tachytelic.net
I could not install it as it's a licensed product.
Oddly when I Googled for tech help for SCO, the name that came up
often, is Bela Lubkin who
get this...is the son of the professor who's computer this is.
I think I will just return the HD to the woman I got it from and
let her son deal with it.
I find this hilarious
Pretty cool. Think about the odds.
You have come fool circle. :-)
Really interesting.
I'd still like to come up with a way to read the drive.
Anyone know if it could be done from Solaris or BSD?
I have the install media somewhere
You can see in the picture I posted, when the tools
are installed, there is a file "Crypserv.exe" which
tells you the file system supports some kind of
encryption. A tool like this likely supports multiple
file systems, so the evidence doesn't tell us which
supported file system needs a capability like that.

https://i.postimg.cc/DwFc1xft/stellar-teaser.gif

And if there isn't a lot of open source or forensic activity
with regard to a sorta-proprietary format like that,
that could spell trouble for recovery. They reverse
engineered NTFS, after a fashion, but imagine how
crusty that format could be (not enough old farts
to help you). To reverse engineer, it really
helps to have running systems, with debug tools
loaded so you can capture runtime details.

I hope you kept an image of the SCO disk, for your
experiments...

https://www.unix.com/unix-for-dummies-questions-and-answers/146023-htfs-file-system.html

mount -F htfs /dev/sda6 /myscodisk

https://sourceforge.net/projects/htfs/

"The htfs is a fuse (userspace) file system. It lives
entirely in ram. Its size is dynamically expanded as needed."

https://ayera.dl.sourceforge.net/project/htfs/htfs/htfs-0.1.1/htfs-0.1.1.tar.bz2

I somehow doubt it is that easy. The SCO disk must
have some notion of slices or something. You could attempt
to use disktype

sudo disktype /dev/sdb

and see what the setup looks like.

Paul
philo
2019-12-23 18:24:44 UTC
Permalink
Post by Paul
Post by philo
<snip>
Post by philo
The ISO at archive.org was corrupted but I found one at
tachytelic.net
I could not install it as it's a licensed product.
Oddly when I Googled for tech help for SCO, the name that came up
often, is Bela Lubkin who
get this...is the son of the professor who's computer this is.
I think I will just return the HD to the woman I got it from and
let her son deal with it.
I find this hilarious
   Pretty cool.  Think about the odds.
   You have come fool circle.  :-)
Really interesting.
I'd still like to come up with a way to read the drive.
Anyone know if it could be done from Solaris or BSD?
I have the install media somewhere
You can see in the picture I posted, when the tools
are installed, there is a file "Crypserv.exe" which
tells you the file system supports some kind of
encryption. A tool like this likely supports multiple
file systems, so the evidence doesn't tell us which
supported file system needs a capability like that.
https://i.postimg.cc/DwFc1xft/stellar-teaser.gif
And if there isn't a lot of open source or forensic activity
with regard to a sorta-proprietary format like that,
that could spell trouble for recovery. They reverse
engineered NTFS, after a fashion, but imagine how
crusty that format could be (not enough old farts
to help you). To reverse engineer, it really
helps to have running systems, with debug tools
loaded so you can capture runtime details.
I hope you kept an image of the SCO disk, for your
experiments...
https://www.unix.com/unix-for-dummies-questions-and-answers/146023-htfs-file-system.html
   mount -F htfs /dev/sda6 /myscodisk
   https://sourceforge.net/projects/htfs/
     "The htfs is a fuse (userspace) file system. It lives
      entirely in ram. Its size is dynamically expanded as needed."
https://ayera.dl.sourceforge.net/project/htfs/htfs/htfs-0.1.1/htfs-0.1.1.tar.bz2
I somehow doubt it is that easy. The SCO disk must
have some notion of slices or something. You could attempt
to use disktype
   sudo disktype /dev/sdb
and see what the setup looks like.
   Paul
Thanks Paul
philo
2019-12-23 19:39:46 UTC
Permalink
I am now in contact with Bela Lubkin and will proceed from here.
philo
2019-12-24 19:17:37 UTC
Permalink
After working with Bela Lubkin I managed to reset the password. He have
me about 75% of the info I needed. Managed to figure out the rest myself.


Now I need to get the data off the drive.


Since there is a fat16 partition all I have to do is mount it and copy
the data over. I can retrieve the data by putting the drive in my Linux
machine.


Problem is I cannot figure out how to mount the drive.

Will keep working with him though
Paul
2019-12-24 20:07:32 UTC
Permalink
Post by philo
After working with Bela Lubkin I managed to reset the password. He have
me about 75% of the info I needed. Managed to figure out the rest myself.
Now I need to get the data off the drive.
Since there is a fat16 partition all I have to do is mount it and copy
the data over. I can retrieve the data by putting the drive in my Linux
machine.
Problem is I cannot figure out how to mount the drive.
Will keep working with him though
Is there a dd (disk dump) available on that system ?

Can you dd the entire slice, into a file, then loopback
mount the file on one of your other systems ?

It helps to know the exact size of the partition and
its offset, if working in the dark. If the disk designators
are working, then maybe "dd" will "just work" to copy the
partition precisely from end to end, where ever you want to put it.

*******

Note that Linux supports loopback mount, using an identifier,
plus a byte offset. In other words, if you know the precise
sector where the FAT starts, you can actually use a loopback
mount plus offset, and get the thing that way. This allows
"stranger" format disks, where you don't understand how the
partition table works, to *still* be mounted in Linux. I
did this by dead reckoning, in an Acronis Capacity Manager
setup, typing in a 64 bit offset in hex, and precisely hitting
the partition I wanted. Only took me about an hour to figure out.
Note that the offset can also be done in hex (because sometimes,
the offset desired is a "nice round number"). Here, I'm
measuring an offset from the beginning of SDC, and expecting a
FAT file system header sector to be sitting at that offset.

sudo mount -t VFAT -o loop,ro,offset=1184440320 /dev/sdc /mnt
cd /mnt
ls -R
cp -Rp .....
sudo umount /mnt

*******

I can show you a little trick I learned. This is how
I captured my entire Macintosh G4 drive, without opening
the computer case. For this to work, you must have a good
understanding of how physical namespace and slices work
on the SCO box (good luck with that!). It's definitely
not going to be /dev/sda, because SCO is not Linux.
It'll be something else. The last time I was inside
FreeBSD, I couldn't figure out theirs well enough to
do anything... :-/ I wouldn't expect the SCO to go
nearly as well.

1) Set up an FTP server on your Linux (or Windows) box.
On my Mac G4 (MacOSX), this was a single tick box! (Unbelievable!)

2) Start an FTP session on the SCO box.
Set transfer type to binary, before doing this.
The pipe symbol is the "streaming transfer source"
for the put command. From a security perspective, there
can be barriers to this working at the source end. The
destination (within reason) doesn't care nearly as much.

ftp> put "|dd if=/dev/sda bs=73728" sda

What that is doing, is transferring some partition (or a whole drive),
to a file called "sda" on your Linux box and FTP server. Whatever
the directory is on the FTP server, the file is likely then
sitting in there. You can then loopback mount the partition.

sudo mount -t VFAT -o loop sda /mnt

Arguments that go after loop, are comma separated, as in

sudo mount -t VFAT -o loop,option1,option2 sda /mnt

Then,

cd /mnt
ls -R

*******

That should whet your appetite for trouble.
Good luck, Sherlock Ohmes.

HTH,
Paul
philo
2019-12-25 00:22:39 UTC
Permalink
Post by Paul
Post by philo
After working with Bela Lubkin I managed to reset the password. He
have me about 75% of the info I needed. Managed to figure out the rest
myself.
Now I need to get the data off the drive.
Since there is a fat16 partition all I have to do is mount it and copy
the data over. I can retrieve the data by putting the drive in my
Linux machine.
Problem is I cannot figure out how to mount the drive.
Will keep working with him though
Is there a dd (disk dump) available on that system ?
Can you dd the entire slice, into a file, then loopback
mount the file on one of your other systems ?
It helps to know the exact size of the partition and
its offset, if working in the dark. If the disk designators
are working, then maybe "dd" will "just work" to copy the
partition precisely from end to end, where ever you want to put it.
*******
Note that Linux supports loopback mount, using an identifier,
plus a byte offset. In other words, if you know the precise
sector where the FAT starts, you can actually use a loopback
mount plus offset, and get the thing that way. This allows
"stranger" format disks, where you don't understand how the
partition table works, to *still* be mounted in Linux. I
did this by dead reckoning, in an Acronis Capacity Manager
setup, typing in a 64 bit offset in hex, and precisely hitting
the partition I wanted. Only took me about an hour to figure out.
Note that the offset can also be done in hex (because sometimes,
the offset desired is a "nice round number"). Here, I'm
measuring an offset from the beginning of SDC, and expecting a
FAT file system header sector to be sitting at that offset.
   sudo mount -t VFAT -o loop,ro,offset=1184440320 /dev/sdc /mnt
   cd /mnt
   ls -R
   cp -Rp .....
   sudo umount /mnt
*******
I can show you a little trick I learned. This is how
I captured my entire Macintosh G4 drive, without opening
the computer case. For this to work, you must have a good
understanding of how physical namespace and slices work
on the SCO box (good luck with that!). It's definitely
not going to be /dev/sda, because SCO is not Linux.
It'll be something else. The last time I was inside
FreeBSD, I couldn't figure out theirs well enough to
do anything... :-/ I wouldn't expect the SCO to go
nearly as well.
1) Set up an FTP server on your Linux (or Windows) box.
   On my Mac G4 (MacOSX), this was a single tick box! (Unbelievable!)
2) Start an FTP session on the SCO box.
   Set transfer type to binary, before doing this.
   The pipe symbol is the "streaming transfer source"
   for the put command. From a security perspective, there
   can be barriers to this working at the source end. The
   destination (within reason) doesn't care nearly as much.
   ftp> put "|dd if=/dev/sda bs=73728" sda
What that is doing, is transferring some partition (or a whole drive),
to a file called "sda" on your Linux box and FTP server. Whatever
the directory is on the FTP server, the file is likely then
sitting in there. You can then loopback mount the partition.
   sudo mount -t VFAT -o loop   sda /mnt
Arguments that go after loop, are comma separated, as in
   sudo mount -t VFAT -o loop,option1,option2   sda /mnt
Then,
   cd /mnt
   ls -R
*******
That should whet your appetite for trouble.
Good luck, Sherlock Ohmes.
HTH,
   Paul
I cannot mount the partition

if I try

mount /dev/hd0d /dos

I get the error

invalid file system type DOS


OTOH it I simply list file system type for that partition,

fstype /dev/hd0d

I get <DOS>


I even reformatted the partion as fat16 just to make sure


Anyway Bela told me that I can use it as raw tar / cpio storage


but he will need to give me exact instructions but that should work.

Too tired to do any more this evening, I am more of a morning person
Paul
2019-12-25 01:47:59 UTC
Permalink
Post by philo
Post by Paul
Post by philo
After working with Bela Lubkin I managed to reset the password. He
have me about 75% of the info I needed. Managed to figure out the
rest myself.
Now I need to get the data off the drive.
Since there is a fat16 partition all I have to do is mount it and
copy the data over. I can retrieve the data by putting the drive in
my Linux machine.
Problem is I cannot figure out how to mount the drive.
Will keep working with him though
Is there a dd (disk dump) available on that system ?
Can you dd the entire slice, into a file, then loopback
mount the file on one of your other systems ?
It helps to know the exact size of the partition and
its offset, if working in the dark. If the disk designators
are working, then maybe "dd" will "just work" to copy the
partition precisely from end to end, where ever you want to put it.
*******
Note that Linux supports loopback mount, using an identifier,
plus a byte offset. In other words, if you know the precise
sector where the FAT starts, you can actually use a loopback
mount plus offset, and get the thing that way. This allows
"stranger" format disks, where you don't understand how the
partition table works, to *still* be mounted in Linux. I
did this by dead reckoning, in an Acronis Capacity Manager
setup, typing in a 64 bit offset in hex, and precisely hitting
the partition I wanted. Only took me about an hour to figure out.
Note that the offset can also be done in hex (because sometimes,
the offset desired is a "nice round number"). Here, I'm
measuring an offset from the beginning of SDC, and expecting a
FAT file system header sector to be sitting at that offset.
sudo mount -t VFAT -o loop,ro,offset=1184440320 /dev/sdc /mnt
cd /mnt
ls -R
cp -Rp .....
sudo umount /mnt
*******
I can show you a little trick I learned. This is how
I captured my entire Macintosh G4 drive, without opening
the computer case. For this to work, you must have a good
understanding of how physical namespace and slices work
on the SCO box (good luck with that!). It's definitely
not going to be /dev/sda, because SCO is not Linux.
It'll be something else. The last time I was inside
FreeBSD, I couldn't figure out theirs well enough to
do anything... :-/ I wouldn't expect the SCO to go
nearly as well.
1) Set up an FTP server on your Linux (or Windows) box.
On my Mac G4 (MacOSX), this was a single tick box! (Unbelievable!)
2) Start an FTP session on the SCO box.
Set transfer type to binary, before doing this.
The pipe symbol is the "streaming transfer source"
for the put command. From a security perspective, there
can be barriers to this working at the source end. The
destination (within reason) doesn't care nearly as much.
ftp> put "|dd if=/dev/sda bs=73728" sda
What that is doing, is transferring some partition (or a whole drive),
to a file called "sda" on your Linux box and FTP server. Whatever
the directory is on the FTP server, the file is likely then
sitting in there. You can then loopback mount the partition.
sudo mount -t VFAT -o loop sda /mnt
Arguments that go after loop, are comma separated, as in
sudo mount -t VFAT -o loop,option1,option2 sda /mnt
Then,
cd /mnt
ls -R
*******
That should whet your appetite for trouble.
Good luck, Sherlock Ohmes.
HTH,
Paul
I cannot mount the partition
if I try
mount /dev/hd0d /dos
I get the error
invalid file system type DOS
OTOH it I simply list file system type for that partition,
fstype /dev/hd0d
I get <DOS>
I even reformatted the partion as fat16 just to make sure
Anyway Bela told me that I can use it as raw tar / cpio storage
but he will need to give me exact instructions but that should work.
Too tired to do any more this evening, I am more of a morning person
There is a manual page here :-)

The example mounts a CD drive. And seems to use /mnt as a mount point
(if it's not already in use).

http://osr600doc.sco.com/en/man/html.ADM/mount.ADM.html

mount -f HS -o lower,write,devmap=./mapfile,defuid=root /dev/cd0 /mnt

"For example, a DOS filesystem understands options such as ro, remount,
and exec but it does not understand options such as trunc and suid"

So maybe it's something like

mount -f DOS -o ro /dev/hd0d /mnt

even though DOS doesn't get mentioned elsewhere in the manual page.

Paul
philo
2019-12-25 02:33:35 UTC
Permalink
Post by Paul
Post by philo
Post by Paul
I can show you a little trick I learned. This is how
I captured my entire Macintosh G4 drive, without opening
the computer case. For this to work, you must have a good
understanding of how physical namespace and slices work
on the SCO box (good luck with that!). It's definitely
not going to be /dev/sda, because SCO is not Linux.
It'll be something else. The last time I was inside
FreeBSD, I couldn't figure out theirs well enough to
do anything... :-/ I wouldn't expect the SCO to go
nearly as well.
1) Set up an FTP server on your Linux (or Windows) box.
    On my Mac G4 (MacOSX), this was a single tick box! (Unbelievable!)
2) Start an FTP session on the SCO box.
    Set transfer type to binary, before doing this.
    The pipe symbol is the "streaming transfer source"
    for the put command. From a security perspective, there
    can be barriers to this working at the source end. The
    destination (within reason) doesn't care nearly as much.
    ftp> put "|dd if=/dev/sda bs=73728" sda
What that is doing, is transferring some partition (or a whole drive),
to a file called "sda" on your Linux box and FTP server. Whatever
the directory is on the FTP server, the file is likely then
sitting in there. You can then loopback mount the partition.
    sudo mount -t VFAT -o loop   sda /mnt
Arguments that go after loop, are comma separated, as in
    sudo mount -t VFAT -o loop,option1,option2   sda /mnt
Then,
    cd /mnt
    ls -R
*******
That should whet your appetite for trouble.
Good luck, Sherlock Ohmes.
HTH,
    Paul
I cannot mount the partition
if I try
mount /dev/hd0d /dos
I get the error
invalid file system type  DOS
OTOH it I simply list file system type for that partition,
fstype  /dev/hd0d
 I get <DOS>
I even reformatted the partion as fat16 just to make sure
Anyway Bela told me that I can use it as raw tar  / cpio   storage
but he will need to give me exact instructions but that should work.
Too tired to do any more this evening, I am more of a morning person
There is a manual page here :-)
The example mounts a CD drive. And seems to use /mnt as a mount point
(if it's not already in use).
http://osr600doc.sco.com/en/man/html.ADM/mount.ADM.html
   mount -f HS -o lower,write,devmap=./mapfile,defuid=root /dev/cd0 /mnt
   "For example, a DOS filesystem understands options such as ro, remount,
    and exec but it does not understand options such as trunc and suid"
So maybe it's something like
   mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
   Paul
Thanks Paul, I will give that a try


in the mean time I decided to see if I can at least view the photos. The
machine has Netscape Navigator, so I was able to view some the the jpg's
then get a pretty good photo of the screen.


There really are that many images on the drive and most of them are of
no photographic value.


Since the family only wanted to view a handful of old memories, If I
even up not being to get the actual date off the machine, I can give
them a few family photos which they will undoubtedly like


Well back down into my basement workshop

This is proving a be a great Christmas eve :)
Wes Newell
2019-12-25 14:01:23 UTC
Permalink
Post by philo
Thanks Paul, I will give that a try
If you're trying to mount a fat16 partition on a device, it's pretty
simple.
Make your mount point with mkdir and then mount it using mount. I had to
create fat16 partitions, format them, and copy a file to one, but here's
all that required to do it.
***@mythfe0:/home/wes# mkdir /dossdb1
***@mythfe0:/home/wes# mkdir /dossdb2
***@mythfe0:/home/wes# mount /dev/sdb1 /dossdb1
***@mythfe0:/home/wes# mount /dev/sdb2 /dossdb2
***@mythfe0:/home/wes# mc

***@mythfe0:/home/wes# ls /dossdb1
myremotes.nec
***@mythfe0:/home/wes# ls /dossdb1

The mc in there is what I used to copy the file. Now to delete all that
crap.
Starting point is fdsik -l to tell you what you have.
philo
2019-12-26 03:53:50 UTC
Permalink
Post by Wes Newell
Post by philo
Thanks Paul, I will give that a try
If you're trying to mount a fat16 partition on a device, it's pretty
simple.
Make your mount point with mkdir and then mount it using mount. I had to
create fat16 partitions, format them, and copy a file to one, but here's
all that required to do it.
myremotes.nec
The mc in there is what I used to copy the file. Now to delete all that
crap.
Starting point is fdsik -l to tell you what you have.
I know it is simple but it refused to mount the fat16 partition

If I could have, it would have saved a lot of work
philo
2019-12-25 02:43:43 UTC
Permalink
Post by Paul
So maybe it's something like
   mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
   Paul
That did not work.

BTW: The first thing Bela suggested was setting up an ftp server.

Unfortunately the machine will not boot up with a net card attached as
it hangs when it gets to the (evidently network) printer
Paul
2019-12-25 03:11:24 UTC
Permalink
Post by philo
Post by Paul
So maybe it's something like
mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
Paul
That did not work.
BTW: The first thing Bela suggested was setting up an ftp server.
Unfortunately the machine will not boot up with a net card attached as
it hangs when it gets to the (evidently network) printer
Does the machine have Kermit ? :-)

Smokin...

http://www.columbia.edu/kermit/current.html

Kermit and a serial port.

A serial port with hardware flow control would go faster,
but that's expecting a lot. For example, I run 57600 between
the Test Machine and the Typing Machine, and I use

console=ttyS0,57600n8

a lot on the Linux boot lines on the Test Machine, and then
I can sit at the Typing machine and watch why Xorg won't start
and so on.

In the old days, gutless computers (like the ones we built at
work), could do 4800 baud. Because there was no hardware flow
control that I can remember. No XON/XOFF either. I remember the
look of glee, when our explosive genius managed to optimize
the driver and get the serial port to run at 9600 baud.
But the computer couldn't do much else but use the serial
port at that rate.

And I remember doing a project, where networking was broken,
and I exchanged files between Mac and PC for *3 weeks* using
crap like that. Serial ports, serial transfer of design files
for engineering. It was my way of saying to the boss
"from crap materials come crap ideas" :-) Because the small
files were source files and the like, the usage of a serial port
wasn't a major impediment. If it had been CAD files, I would
have gone on permanent vacation.

Paul
Paul
2019-12-25 03:25:41 UTC
Permalink
Post by philo
Post by Paul
So maybe it's something like
mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
Paul
That did not work.
BTW: The first thing Bela suggested was setting up an ftp server.
Unfortunately the machine will not boot up with a net card attached as
it hangs when it gets to the (evidently network) printer
They appear to have gone through an Mtools era.
This looks similar to some of the Mtools stuff we used to use.

http://osr507doc.sco.com/en/man/html.C/doscmd.C.html

Yikes!

I was hoping there was a mount command.

But the mount command doesn't show a value of DOS for type,
so I don't know what the deal is there.

The mount command might require that the equivalent of fstab
be populated first. You could benefit from finding whatever
/etc file controls the other mounts on the box, and see
if the DOS partition happens to have a mount set up for it.
Then the syntax could be more compact.

If you decide to clone the disk you're working on, and
trying to boot the cloned copy, the drive may need to be
limited in size. Using the CLIP jumper can limit an IDE
drive to 33.7GB or 2GB, and it depends on the "interpretation"
of the machine at BIOS level, as to how the system comes up
afterwards. You could also use an HPA (Host Protected Area)
to trim down a drive. I have one port on my Typing machine
that accepts HPA commands, and I can "make drives smaller",
like SATA drives, using that capability. As otherwise, some
of those older OSes have heart failure caused by the
CHS geometry. I've clipped down IDE drives before for
experiments, but they never come up showing 2GB, and they
always managed to be 33.7GB or so.

Paul
philo
2019-12-25 03:47:23 UTC
Permalink
Post by Paul
Post by philo
Post by Paul
So maybe it's something like
    mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
    Paul
That did not work.
BTW: The first thing Bela suggested was setting up an ftp server.
Unfortunately the machine will not boot up with a net card attached as
it hangs when it gets to the (evidently network) printer
They appear to have gone through an Mtools era.
This looks similar to some of the Mtools stuff we used to use.
http://osr507doc.sco.com/en/man/html.C/doscmd.C.html
Yikes!
I was hoping there was a mount command.
But the mount command doesn't show a value of DOS for type,
so I don't know what the deal is there.
The mount command might require that the equivalent of fstab
be populated first. You could benefit from finding whatever
/etc file controls the other mounts on the box, and see
if the DOS partition happens to have a mount set up for it.
Then the syntax could be more compact.
If you decide to clone the disk you're working on, and
trying to boot the cloned copy, the drive may need to be
limited in size. Using the CLIP jumper can limit an IDE
drive to 33.7GB or 2GB, and it depends on the "interpretation"
of the machine at BIOS level, as to how the system comes up
afterwards. You could also use an HPA (Host Protected Area)
to trim down a drive. I have one port on my Typing machine
that accepts HPA commands, and I can "make drives smaller",
like SATA drives, using that capability. As otherwise, some
of those older OSes have heart failure caused by the
CHS geometry. I've clipped down IDE drives before for
experiments, but they never come up showing 2GB, and they
always managed to be 33.7GB or so.
   Paul
Bela gave me something else to try viz: writing to standard output


zip -r > /dev/hd0d

It is now adding to archive but I don't know how long it will take. If
it's not done in half an hour I'm going to bed and will check in the
morning.

I guess Bela used to work for SCO, so he seems to know his stuff


As to using a serial port.

Sheesh,. the last time I did a serial port transfer was almost 20 years
ago when I got my first digital camera...

I was using Win95 which I was happy with but upgraded to Win98 for USB

Then the upgrade game began :)
Paul
2019-12-25 03:55:32 UTC
Permalink
Post by philo
Post by Paul
Post by philo
Post by Paul
So maybe it's something like
mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
Paul
That did not work.
BTW: The first thing Bela suggested was setting up an ftp server.
Unfortunately the machine will not boot up with a net card attached
as it hangs when it gets to the (evidently network) printer
They appear to have gone through an Mtools era.
This looks similar to some of the Mtools stuff we used to use.
http://osr507doc.sco.com/en/man/html.C/doscmd.C.html
Yikes!
I was hoping there was a mount command.
But the mount command doesn't show a value of DOS for type,
so I don't know what the deal is there.
The mount command might require that the equivalent of fstab
be populated first. You could benefit from finding whatever
/etc file controls the other mounts on the box, and see
if the DOS partition happens to have a mount set up for it.
Then the syntax could be more compact.
If you decide to clone the disk you're working on, and
trying to boot the cloned copy, the drive may need to be
limited in size. Using the CLIP jumper can limit an IDE
drive to 33.7GB or 2GB, and it depends on the "interpretation"
of the machine at BIOS level, as to how the system comes up
afterwards. You could also use an HPA (Host Protected Area)
to trim down a drive. I have one port on my Typing machine
that accepts HPA commands, and I can "make drives smaller",
like SATA drives, using that capability. As otherwise, some
of those older OSes have heart failure caused by the
CHS geometry. I've clipped down IDE drives before for
experiments, but they never come up showing 2GB, and they
always managed to be 33.7GB or so.
Paul
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will take. If
it's not done in half an hour I'm going to bed and will check in the
morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???

Is that trying to erase hd0d ???

Paul
Bobbie Sellers
2019-12-25 04:46:06 UTC
Permalink
Post by Paul
Post by philo
Post by Paul
Post by philo
Post by Paul
So maybe it's something like
    mount -f DOS -o ro /dev/hd0d /mnt
even though DOS doesn't get mentioned elsewhere in the manual page.
    Paul
That did not work.
BTW: The first thing Bela suggested was setting up an ftp server.
Unfortunately the machine will not boot up with a net card attached
as it hangs when it gets to the (evidently network) printer
They appear to have gone through an Mtools era.
This looks similar to some of the Mtools stuff we used to use.
http://osr507doc.sco.com/en/man/html.C/doscmd.C.html
Yikes!
I was hoping there was a mount command.
But the mount command doesn't show a value of DOS for type,
so I don't know what the deal is there.
The mount command might require that the equivalent of fstab
be populated first. You could benefit from finding whatever
/etc file controls the other mounts on the box, and see
if the DOS partition happens to have a mount set up for it.
Then the syntax could be more compact.
If you decide to clone the disk you're working on, and
trying to boot the cloned copy, the drive may need to be
limited in size. Using the CLIP jumper can limit an IDE
drive to 33.7GB or 2GB, and it depends on the "interpretation"
of the machine at BIOS level, as to how the system comes up
afterwards. You could also use an HPA (Host Protected Area)
to trim down a drive. I have one port on my Typing machine
that accepts HPA commands, and I can "make drives smaller",
like SATA drives, using that capability. As otherwise, some
of those older OSes have heart failure caused by the
CHS geometry. I've clipped down IDE drives before for
experiments, but they never come up showing 2GB, and they
always managed to be 33.7GB or so.
    Paul
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will take. If
it's not done in half an hour I'm going to bed and will check in the
morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???
Is that trying to erase hd0d ???
   Paul
zip is archive, -r I will bet is recursive letting the zip access the
sub-directories, etc.

But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!

bliss
--
bliss dash SF 4 ever at dslextreme dot com
Aragorn
2019-12-25 11:41:55 UTC
Permalink
Post by Bobbie Sellers
Post by Paul
Post by philo
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will
take. If it's not done in half an hour I'm going to bed and will
check in the morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???
Is that trying to erase hd0d ???
zip is archive, -r I will bet is recursive letting the zip access the
sub-directories, etc.
But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!
Syntactically, that would not be correct. Whatever "zip" does, the
output redirection ">" means that it's going to write its output to
/dev/hd0d, which is not a directory but a block device.

So Paul is correct: that command — at least, as Philo typed it here —
is going to overwrite the block device with data from "zip", and I
don't know where "zip" gets is data from when not used with a file, so
I'm guessing it'll be from stdin, i.e. the console.
--
With respect,
= Aragorn =
philo
2019-12-25 16:11:11 UTC
Permalink
Post by Aragorn
Post by Bobbie Sellers
Post by Paul
Post by philo
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will
take. If it's not done in half an hour I'm going to bed and will
check in the morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???
Is that trying to erase hd0d ???
zip is archive, -r I will bet is recursive letting the zip access the
sub-directories, etc.
But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!
Syntactically, that would not be correct. Whatever "zip" does, the
output redirection ">" means that it's going to write its output to
/dev/hd0d, which is not a directory but a block device.
So Paul is correct: that command — at least, as Philo typed it here —
is going to overwrite the block device with data from "zip", and I
don't know where "zip" gets is data from when not used with a file, so
I'm guessing it'll be from stdin, i.e. the console.
I tried to recover just a single file and it worked.


He had me use "zip" because the recovery partition is only one gig and
we not sure if there was enough room. As it turned out "zip" was not needed.


anyway what I did was use

cpio -ivdm > /dev/hd0d


It wrote raw data to hd0d (which of course destroyed the file system)


To recover, from my Linux machine I reversed the process and from my
"recovery" folder used the command (from a root terminal)


cpio -ivdm < /dev/sdb1

(sdb1 was how Linux sees the drive)


This is kind of exciting because not only am I now recovering data,
after all the years that I've been using Linux, this is the first time
I've ever used actual Unix


Oddly, my wife does not share my enthusiasm and keeps telling me:
You know we are having guests today>


sheesh
D***@decadence.org
2019-12-25 16:27:35 UTC
Permalink
Post by philo
Post by Aragorn
Post by Bobbie Sellers
Post by Paul
Post by philo
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will
take. If it's not done in half an hour I'm going to bed and
will check in the morning.
I guess Bela used to work for SCO, so he seems to know his
stuff
What does that command do again ???
Is that trying to erase hd0d ???
zip is archive, -r I will bet is recursive letting the zip
access the sub-directories, etc.
But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!
Syntactically, that would not be correct. Whatever "zip" does,
the output redirection ">" means that it's going to write its
output to /dev/hd0d, which is not a directory but a block device.
So Paul is correct: that command — at least, as Philo typed it
here — is going to overwrite the block device with data from
"zip", and I don't know where "zip" gets is data from when not
used with a file, so I'm guessing it'll be from stdin, i.e. the
console.
I tried to recover just a single file and it worked.
He had me use "zip" because the recovery partition is only one gig
and we not sure if there was enough room. As it turned out "zip"
was not needed.
anyway what I did was use
cpio -ivdm > /dev/hd0d
It wrote raw data to hd0d (which of course destroyed the file
system)
To recover, from my Linux machine I reversed the process and from
my "recovery" folder used the command (from a root terminal)
cpio -ivdm < /dev/sdb1
(sdb1 was how Linux sees the drive)
This is kind of exciting because not only am I now recovering
data, after all the years that I've been using Linux, this is the
first time I've ever used actual Unix
You know we are having guests today>
sheesh
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great labor
and many pots on the stove. ;-)

Tell her she is lucky times are so much more automated. (for some
things).

Back then, you would have to wait to get time on the mainframe for
your workstation/program run.
Aragorn
2019-12-25 16:53:11 UTC
Permalink
Post by D***@decadence.org
Post by philo
You know we are having guests today>
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great labor
and many pots on the stove. ;-)
Tell her she is lucky times are so much more automated. (for some
things).
Back then, you would have to wait to get time on the mainframe for
your workstation/program run.
You say that, but look over there in the corner. That PDP-7 hasn't
been used in ages. Are you thinking what I'm thinking?

Gaming time! :)
--
With respect,
= Aragorn =
philo
2019-12-25 17:50:27 UTC
Permalink
Post by Aragorn
Post by D***@decadence.org
Post by philo
You know we are having guests today>
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great labor
and many pots on the stove. ;-)
Tell her she is lucky times are so much more automated. (for some
things).
Back then, you would have to wait to get time on the mainframe for
your workstation/program run.
You say that, but look over there in the corner. That PDP-7 hasn't
been used in ages. Are you thinking what I'm thinking?
Gaming time! :)
No PDP-7 here but I played my first game on a PDP-8

Text based Moon Lander


BTW: The Unix machine I'm working on has quite a few text based games on it
D***@decadence.org
2019-12-25 19:11:29 UTC
Permalink
Post by philo
Post by Aragorn
Post by D***@decadence.org
Post by philo
Oddly, my wife does not share my enthusiasm and keeps telling
me: You know we are having guests today>
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great
labor and many pots on the stove. ;-)
Tell her she is lucky times are so much more automated. (for some
things).
Back then, you would have to wait to get time on the
mainframe for
your workstation/program run.
You say that, but look over there in the corner. That PDP-7
hasn't been used in ages. Are you thinking what I'm thinking?
Gaming time! :)
No PDP-7 here but I played my first game on a PDP-8
Text based Moon Lander
BTW: The Unix machine I'm working on has quite a few text based games on it
NetHack Just released 3.6.4 on the 21st. Near impssible for a guy
like me to win. Too impatient, man.

<https://www.nethack.org/>
D***@decadence.org
2019-12-25 19:08:07 UTC
Permalink
Post by Aragorn
Post by D***@decadence.org
Post by philo
Oddly, my wife does not share my enthusiasm and keeps telling
me: You know we are having guests today>
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great
labor and many pots on the stove. ;-)
Tell her she is lucky times are so much more automated. (for some
things).
Back then, you would have to wait to get time on the mainframe for
your workstation/program run.
You say that, but look over there in the corner. That PDP-7
hasn't been used in ages. Are you thinking what I'm thinking?
Gaming time! :)
Distributed processing!
Aragorn
2019-12-25 20:13:39 UTC
Permalink
Post by D***@decadence.org
Post by Aragorn
Post by D***@decadence.org
Post by philo
Oddly, my wife does not share my enthusiasm and keeps telling
me: You know we are having guests today>
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great
labor and many pots on the stove. ;-)
Tell her she is lucky times are so much more automated. (for some
things).
Back then, you would have to wait to get time on the mainframe for
your workstation/program run.
You say that, but look over there in the corner. That PDP-7
hasn't been used in ages. Are you thinking what I'm thinking?
Gaming time! :)
Distributed processing!
Well, no, not really, but that's how AT&T Unix was born. Thompson,
Ritchie, Osanna et al were playing games on the Multics mainframe, but
it took away costly CPU cycles, and they had an unused PDP-7 [*] sitting
in a corner, so they wrote a scaled-down version of Multics as the OS
for that mini, so that they could play their games on that instead. ;)


[*] Or was it a PDP-8? I'm not sure anymore. They either way rewrote
it in C later — it was originally written in assembler — for a
PDP-11.
--
With respect,
= Aragorn =
philo
2019-12-25 17:48:47 UTC
Permalink
On 12/25/19 10:27 AM, ***@decadence.org wrote:
<snip>
Post by D***@decadence.org
Post by philo
It wrote raw data to hd0d (which of course destroyed the file system)
To recover, from my Linux machine I reversed the process and from
my "recovery" folder used the command (from a root terminal)
cpio -ivdm < /dev/sdb1
(sdb1 was how Linux sees the drive)
This is kind of exciting because not only am I now recovering
data, after all the years that I've been using Linux, this is the
first time I've ever used actual Unix
You know we are having guests today>
sheesh
At least it is not 1969 and you are Ken Thompson. Or 'the old
lady' would be cooking up that dinner for those guests at great labor
and many pots on the stove. ;-)
Tell her she is lucky times are so much more automated. (for some
things).
Back then, you would have to wait to get time on the mainframe for
your workstation/program run.
I remember those days.
Took FORTRAN-IV back I 1968 and learned to hate punch cards

With all the waiting in line , it took a week to get a program running.

By 1982 I decided never to touch a computer again in my life and stayed
away until 1999.
Paul
2019-12-25 17:05:55 UTC
Permalink
Post by philo
Post by Aragorn
Post by Bobbie Sellers
Post by Paul
Post by philo
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will
take. If it's not done in half an hour I'm going to bed and will
check in the morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???
Is that trying to erase hd0d ???
zip is archive, -r I will bet is recursive letting the zip access the
sub-directories, etc.
But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!
Syntactically, that would not be correct. Whatever "zip" does, the
output redirection ">" means that it's going to write its output to
/dev/hd0d, which is not a directory but a block device.
So Paul is correct: that command — at least, as Philo typed it here —
is going to overwrite the block device with data from "zip", and I
don't know where "zip" gets is data from when not used with a file, so
I'm guessing it'll be from stdin, i.e. the console.
I tried to recover just a single file and it worked.
He had me use "zip" because the recovery partition is only one gig and
we not sure if there was enough room. As it turned out "zip" was not needed.
anyway what I did was use
cpio -ivdm > /dev/hd0d
It wrote raw data to hd0d (which of course destroyed the file system)
To recover, from my Linux machine I reversed the process and from my
"recovery" folder used the command (from a root terminal)
cpio -ivdm < /dev/sdb1
(sdb1 was how Linux sees the drive)
This is kind of exciting because not only am I now recovering data,
after all the years that I've been using Linux, this is the first time
I've ever used actual Unix
You know we are having guests today>
sheesh
So /dev/hd0d was your "tape drive".

That's one way to do it. Cool.

*******

And "my wife does not share my enthusiasm"...

"You could at least stir the gravy pan." They give the important
jobs to those standing around, in the way. Preventing gravy from
burning, by using cheap human labor, is important. Stirring gravy is
one of my specialties (since I have no other kitchen skills).

If we had real automation, your Christmas dinner would look like this.
They never seemed to fight over seconds with these plates.

https://scifi.stackexchange.com/questions/169923/what-is-being-eaten-by-kirk-and-company-in-this-scene?rq=1

Paul
philo
2019-12-25 17:55:18 UTC
Permalink
Post by Paul
Post by philo
Post by Paul
Post by philo
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will
take. If it's not done in half an hour I'm going to bed and will
check in the morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???
Is that trying to erase hd0d ???
zip is archive,  -r I will bet is recursive letting the zip access the
sub-directories, etc.
    But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!
Syntactically, that would not be correct.  Whatever "zip" does, the
output redirection ">" means that it's going to write its output to
/dev/hd0d, which is not a directory but a block device.
So Paul is correct: that command — at least, as Philo typed it here —
is going to overwrite the block device with data from "zip", and I
don't know where "zip" gets is data from when not used with a file, so
I'm guessing it'll be from stdin, i.e. the console.
I tried to recover just a single file and it worked.
He had me use "zip" because the recovery partition is only one gig and
we not sure if there was enough room. As it turned out "zip" was not needed.
anyway what I did was use
cpio -ivdm > /dev/hd0d
It wrote raw data to hd0d  (which of course destroyed the file system)
To recover, from my Linux machine I reversed the process and from my
"recovery" folder used the command (from a root terminal)
cpio -ivdm < /dev/sdb1
(sdb1 was how Linux sees the drive)
This is kind of exciting because not only am I now recovering data,
after all the years that I've been using Linux, this is the first time
I've ever used actual Unix
You know we are having guests today>
sheesh
So /dev/hd0d was your "tape drive".
That's one way to do it. Cool.
*******
And "my wife does not share my enthusiasm"...
"You could at least stir the gravy pan." They give the important
jobs to those standing around, in the way. Preventing gravy from
burning, by using cheap human labor, is important. Stirring gravy is
one of my specialties (since I have no other kitchen skills).
If we had real automation, your Christmas dinner would look like this.
They never seemed to fight over seconds with these plates.
https://scifi.stackexchange.com/questions/169923/what-is-being-eaten-by-kirk-and-company-in-this-scene?rq=1
LOL on Start Trek


My wife is an expert cook and usually wants me out of the kitchen.

Now the work of cooking has been handed down to my step-daughter

I have to watch my manners though.


My wife was warned early on:

First date, when she got up to use the ladies room, she returned to an
empty plate. I ate what was left of her meal!
Ken Hart
2019-12-25 22:06:28 UTC
Permalink
Post by Paul
Post by philo
Post by Paul
Post by philo
Bela gave me something else to try viz: writing to standard output
zip -r > /dev/hd0d
It is now adding to archive but I don't know how long it will
take. If it's not done in half an hour I'm going to bed and will
check in the morning.
I guess Bela used to work for SCO, so he seems to know his stuff
What does that command do again ???
Is that trying to erase hd0d ???
zip is archive,  -r I will bet is recursive letting the zip access the
sub-directories, etc.
    But bliss is not an expert on SCO or even GNU/Linux.
So looked it up and I was correct!
Syntactically, that would not be correct.  Whatever "zip" does, the
output redirection ">" means that it's going to write its output to
/dev/hd0d, which is not a directory but a block device.
So Paul is correct: that command — at least, as Philo typed it here —
is going to overwrite the block device with data from "zip", and I
don't know where "zip" gets is data from when not used with a file, so
I'm guessing it'll be from stdin, i.e. the console.
I tried to recover just a single file and it worked.
He had me use "zip" because the recovery partition is only one gig and
we not sure if there was enough room. As it turned out "zip" was not needed.
anyway what I did was use
cpio -ivdm > /dev/hd0d
It wrote raw data to hd0d  (which of course destroyed the file system)
To recover, from my Linux machine I reversed the process and from my
"recovery" folder used the command (from a root terminal)
cpio -ivdm < /dev/sdb1
(sdb1 was how Linux sees the drive)
This is kind of exciting because not only am I now recovering data,
after all the years that I've been using Linux, this is the first time
I've ever used actual Unix
You know we are having guests today>
sheesh
So /dev/hd0d was your "tape drive".
That's one way to do it. Cool.
*******
And "my wife does not share my enthusiasm"...
"You could at least stir the gravy pan." They give the important
jobs to those standing around, in the way. Preventing gravy from
burning, by using cheap human labor, is important. Stirring gravy is
one of my specialties (since I have no other kitchen skills).
If we had real automation, your Christmas dinner would look like this.
They never seemed to fight over seconds with these plates.
https://scifi.stackexchange.com/questions/169923/what-is-being-eaten-by-kirk-and-company-in-this-scene?rq=1
   Paul
At least it's not Soylent Green.
--
Ken Hart
***@frontier.com
philo
2019-12-26 12:02:15 UTC
Permalink
Post by philo
I was just given a P-1 from the wife a deceased computer science college
professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive into on
of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her the
images and tell her I got everything and I'd sure she'd be happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail for
gparted that could read it.
Thanks
All data now backed up.

I needed extensive help as the Unix commands work a bit differently that
Linux.


a fun excursion!
D***@decadence.org
2019-12-26 12:52:29 UTC
Permalink
Post by philo
Post by philo
I was just given a P-1 from the wife a deceased computer science
college professor.
She wanted me to recover any data on the machine.
She of course does not have the password, so I popped the drive
into on of my Linux machines.
There is one Windows partition and I was able to copy off some photos.
Since she has no idea what is on the machine, I could give her
the images and tell her I got everything and I'd sure she'd be
happy.
Of course I'd still like to have a look on the SCO partition.
Gparted sees it as an unknown file system.
Wondering what the file system is and if there is a package avail
for gparted that could read it.
Thanks
All data now backed up.
I needed extensive help as the Unix commands work a bit
differently that Linux.
a fun excursion!
IIRC, UNIX will process c script right from the console prompt (not
that that tid bit is pertanent here).
Coders back then were a lot smarter than now.
philo
2019-12-27 16:20:25 UTC
Permalink
On 12/26/19 6:52 AM, ***@decadence.org wrote:
<snip>
Post by D***@decadence.org
Post by philo
All data now backed up.
I needed extensive help as the Unix commands work a bit
differently that Linux.
a fun excursion!
IIRC, UNIX will process c script right from the console prompt (not
that that tid bit is pertanent here).
Coders back then were a lot smarter than now.
I was not much of a coder anyway.


BTW: There was one thing I thought funny.

The login screen for entering user and password had a "help" tab.

It explained that you put user name where the "user name" box is and
password where the "password" box is.
Paul
2019-12-27 18:05:57 UTC
Permalink
Post by philo
<snip>
Post by D***@decadence.org
Post by philo
All data now backed up.
I needed extensive help as the Unix commands work a bit
differently that Linux.
a fun excursion!
IIRC, UNIX will process c script right from the console prompt (not
that that tid bit is pertanent here).
Coders back then were a lot smarter than now.
I was not much of a coder anyway.
BTW: There was one thing I thought funny.
The login screen for entering user and password had a "help" tab.
It explained that you put user name where the "user name" box is and
password where the "password" box is.
These Computer Scientists think of (almost) everything.

Now, if they could only set the clocks properly, on space ships.

Paul
D***@decadence.org
2019-12-27 18:22:01 UTC
Permalink
Post by Paul
Post by philo
<snip>
Post by D***@decadence.org
Post by philo
All data now backed up.
I needed extensive help as the Unix commands work a bit
differently that Linux.
a fun excursion!
IIRC, UNIX will process c script right from the console
prompt (not
that that tid bit is pertanent here).
Coders back then were a lot smarter than now.
I was not much of a coder anyway.
BTW: There was one thing I thought funny.
The login screen for entering user and password had a "help" tab.
It explained that you put user name where the "user name" box is
and password where the "password" box is.
These Computer Scientists think of (almost) everything.
Now, if they could only set the clocks properly, on space ships.
Paul
I always thought the term "space ship" was kind of funny.
With such small crew counts, it seems more like an inflatable raft.
Probably why none have been dubbed "SS" anything yet. I dunno.

Maybe fix the guidance on rendezvous craft too. ;-)

Might be more inclined to go.

Things as they are down here though, y'all could just stick me in a
can and kick it off toward Mars. I'd be a perfect test ride for
determining vulnerabilities. Not waste all those millions they put
into the crew they want to send.

Some fucker up there would probably call me a troll too. In a
language I do not even know.
philo
2019-12-27 23:40:47 UTC
Permalink
<snip>
Post by Paul
Post by philo
BTW: There was one thing I thought funny.
The login screen for entering user and password had a "help" tab.
It explained that you put user name where the "user name" box is and
password where the "password" box is.
These Computer Scientists think of (almost) everything.
Now, if they could only set the clocks properly, on space ships.
   Paul
Speaking of dates, I recall the big W2K scare.
Though I assumed nothing would happen, I set the dates on a few of my
machines ahead and all was OK.


Got a big laugh when the years 2000 arrived, another machine that I had
not pre-tested came up with the year 19100

Continue reading on narkive:
Loading...