Recommended archive for LKG pristine emufiles

[Adam Billyard]

Hi

I've now got my Tektronix 4404 up and running OS 2.0 (as well as older OS 1.5).  Of course I have backups of these emufile images but in the unlikely event somebody wanted a known good image in the future,  is there a recommended place to post them?

Adam

[David Gesswein]

If you don't know of a Tek community location I can put a copy on my site.
Upload to http://www.pdp8online.com/upload/abuploadab.html

> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/aa55cc99-6f90-40a5-9b04-0a7ea0f682b8n%40googlegroups.com.

[David Gesswein]

Put the file online. See last entry in table.
http://www.pdp8online.com/mfm/status.shtml

On Fri, Apr 28, 2023 at 11:38:37PM -0700, 'Adam Billyard' via MFM Discuss wrote:

[Al Kossow]

>> I've now got my Tektronix 4404 up and running OS 2.0 (as well as older OS
>> 1.5). Of course I have backups of these emufile images but in the unlikely
>> event somebody wanted a known good image in the future, is there a
>> recommended place to post them?

my site
bitsavers.org

[Adam Billyard]

Nice.

I had been using a Uniflex precursor of fsck called “diskrepair”  that showed no errors in FDN etc.  Having run your mfm_util and seen many blocks with missing LBAs I found a util on the Tek4404 that checks low level formatting: “devcheck /dev/disk”  and it comes out with the same blocks missing LBAs.  Adding these to the badblocks record does appear to improve stability.

I just dont understand how *some* blocks can be missing LBA info and most are fine.  Any ideas?

To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/ZFGnAiUFzkwHjOha%40hugin3.

[David Gesswein]

To understand how those tracks get wrong format we would need to capture
the data when formatting those tracks. That would require a scope or logic
analyzer that can capture at least 100 MS/s for at least 17 ms. We would need
to verify if they occur at the same location each format and if they do I can
program a pin on the beaglebone to generate a trigger when the correct
cylinder and head is being written.

Normally when formatting a track its done at one contiguous write. Don't
see how the emulator could be messing up the headers the way we see. If they
are trying to rewrite the header for one sector its possible it gets
inserted into the wrong location but seems like data writes would also
get corrupted if this was happening.

We can also fix the disk image. I can add ext2emu support for this format
and we can extract the data and recreate the emulator file with correct
headers. Any file that was on the messed up sectors will still be bad. Do you
know of any way to figure out what file is on a LBA? Could also
intentionally make data errors in the emulator file and try to read every
file on the disk to find the bad ones. If we can find them I assume you can
then replace them with good copy from the install floppies.

> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/35FF18CA-DCE4-48A8-A5DD-F27479D0E59C%40googlemail.com.

[Al Kossow]

Did you ever put the 4404 emulator files up anywhere?

[David Gesswein]

See last entry for 2.0. Not sure if this was the final image or more work
was done after I got this copy. We still don't know why their are errors in
the formatting.

http://www.pdp8online.com/mfm/status.shtml

> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/e83fe817-0d8f-42f5-9942-554e2deeb190n%40googlegroups.com.

[Adam Billyard]

[Apologies,  day job got in the way.]

I've spend a few hours building a fresh OS2.0 + compiler + UNIX utils but on a disk image formatted as 20MB and have yet to see any read/write errors.   Whereas a 40MB image had occasional disk corruption which meant having to roll back etc.

So I may start using this image and assume I'm not going to run out of space - my plan is to get uemacs3.9 working and then implement a window system so I can run multiple shell windows.

[Adam Billyard]

Hi Al,

Is there somewhere I can upload an mfm image to?  I have shared it with DavidG.

best

  Adam

[David Gesswein]

On Sun, Jun 25, 2023 at 10:54:38AM -0700, 'Adam Billyard' via MFM Discuss wrote:
> [Apologies, day job got in the way.]
>
> I've spend a few hours building a fresh OS2.0 + compiler + UNIX utils but

> on a disk image formatted as *20MB *and have yet to see any read/write
> errors.
>

The emulator container file is 6 heads 830 cylinders so 43 MB. It looks
like it only used heads 0-2 when you formatted at 20 MB and the upper
heads have data from previous usage. That makes mfm_util generate a lot
of errors. If you want I can see if I can remove the other 3 heads from
the file which will make a less confusing image. You'll need to test and
see if it still works. Let me know.

[Adam Billyard]

Yes, that would be great and much appreciated.

Annoying not to understand why 6 head setup is so unreliable. Stillnot had any errors with 3 head image.

Adam

> On 26 Jun 2023, at 16:45, David Gesswein <d...@pdp8online.com> wrote:

> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/ZJmyhKGjVgvb2HmY%40hugin3.

[d...@pdp8online.com]

Finally got time. Image is up at http://www.pdp8online.com/mfm/status.shtml. Let me know if it works for you.

[Adam Billyard]

Oh nice!   Thank you for that - as I started writing more original code for the Tek4404 I’m keen not to lose stuff!

I’m not in the UK with the Tek4404 for a while now but will look forward to giving a whirl and reporting back.

Thanks again David.

  Adam

To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/54a7d0f0-edca-48a5-a165-52d210f31aa6n%40googlegroups.com.

[David Gesswein]

In case you made changes to the images you made you can use the attached
code if you wish to shrink them.

gcc -o copy_emu copy_emu.c mfm/emu_tran_file.c -I mfm/inc mfm/obj/msg.o mfm/obj/crc_ecc.o -lm

run copy_emu and it will fix emufile_20MB_network creating emufile_20MB_network_3head

On Mon, Jul 17, 2023 at 05:40:51PM +0200, 'Adam Billyard' via MFM Discuss wrote:
> Oh nice! Thank you for that - as I started writing more original code for the Tek4404 I’m keen not to lose stuff!
>
> I’m not in the UK with the Tek4404 for a while now but will look forward to giving a whirl and reporting back.
>
> Thanks again David.
>
> Adam
>
> > On 16 Jul 2023, at 17:50, d...@pdp8online.com <d...@pdp8online.com> wrote:
> >

> > Finally got time. Image is up at http://www.pdp8online.com/mfm/status.shtml <http://www.pdp8online.com/mfm/status.shtml>. Let me know if it works for you.

> >
> >
> > On Tuesday, June 27, 2023 at 5:58:42 AM UTC-4 adambi...@googlemail.com wrote:
> > Yes, that would be great and much appreciated.
> >
> > Annoying not to understand why 6 head setup is so unreliable. Stillnot had any errors with 3 head image.
> >
> > Adam
> >
> > > On 26 Jun 2023, at 16:45, David Gesswein <d...@pdp8online.com <applewebdata://926720AD-613E-4D7E-AC76-0C15806F3BD0>> wrote:
> > >
> > > On Sun, Jun 25, 2023 at 10:54:38AM -0700, 'Adam Billyard' via MFM Discuss wrote:
> > >> [Apologies, day job got in the way.]
> > >>
> > >> I've spend a few hours building a fresh OS2.0 + compiler + UNIX utils but
> > >> on a disk image formatted as *20MB *and have yet to see any read/write
> > >> errors.
> > >>
> > >
> > > The emulator container file is 6 heads 830 cylinders so 43 MB. It looks
> > > like it only used heads 0-2 when you formatted at 20 MB and the upper
> > > heads have data from previous usage. That makes mfm_util generate a lot
> > > of errors. If you want I can see if I can remove the other 3 heads from
> > > the file which will make a less confusing image. You'll need to test and
> > > see if it still works. Let me know.
> > >
> > > --
> > > You received this message because you are subscribed to the Google Groups "MFM Discuss" group.

> > > To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com <applewebdata://926720AD-613E-4D7E-AC76-0C15806F3BD0>.
> > > To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/ZJmyhKGjVgvb2HmY%40hugin3 <https://groups.google.com/d/msgid/mfm-discuss/ZJmyhKGjVgvb2HmY%40hugin3>.

> >
> > --
> > You received this message because you are subscribed to the Google Groups "MFM Discuss" group.

> > To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com <mailto:mfm-discuss...@googlegroups.com>.
> > To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/54a7d0f0-edca-48a5-a165-52d210f31aa6n%40googlegroups.com <https://groups.google.com/d/msgid/mfm-discuss/54a7d0f0-edca-48a5-a165-52d210f31aa6n%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/7C9363F2-C589-4D2C-BC66-FAB23FB264B2%40googlemail.com.

[Adam Billyard]

Just to report back on this.   This appears have totally resolved the unreliable disk issues I was seeing.  I spent all weekend writing C utlities for the Tektronix and compiling/linking etc with zero issues. 

Thanks again for this David.   Now I have a 20MB disk, I really am going to have to implement NFS on Uniflex :-)

[Al Kossow]

On 9/12/23 6:58 AM, 'Adam Billyard' via MFM Discuss wrote:
> Just to report back on this.   This appears have totally resolved the unreliable disk issues I was seeing.  I spent all weekend writing C
> utlities for the Tektronix and compiling/linking etc with zero issues.

it would be nice to archive these images on bitsavers to try on my machine.

[Adam Billyard]

Definitely. *almost* got my implementation of telnetd working (can connect from a Mac to the Tektronix 4404, but some issues). Uniflex sockets differ subtly from modern day sockets - who knew..

My implementation of DHCP client is unfortunately not working yet - its just a bunch of DGRAM messages - and the same code works on my M1 Mac - so bit stuck.

Its on my github repo if you want to browse.

> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/4e24ca93-cf6c-eb6e-a35c-46521501cd7d%40bitsavers.org.

[David Gesswein]

The working image without the additinal network programs is linked to from
the last entry in this page.
http://www.pdp8online.com/mfm/status.shtml

Didn't find the github repo in a quick search. Would be good to post a link.

> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/4e24ca93-cf6c-eb6e-a35c-46521501cd7d%40bitsavers.org.

[Adam Billyard]

https://github.com/Elektraglide/tek4404

You received this message because you are subscribed to a topic in the Google Groups "MFM Discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mfm-discuss/nhlWdf23NFw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mfm-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/ZQD4eKSQ%2BhsIc2ai%40hugin3.

[nosn...@gmail.com]

              OMG: Here's what I did:

 

1.                   On the PC: copied the Tektronix_4404_OS2.0.emu file to the BeagleBone with pscp Tektronix_4404_OS2.0.emu ro...@10.0.0.64:Tektronix_4404_OS2.0.emu

2.                   on the BeagleBone: cd mfm

3.                   on the BeagleBone: ./mfm_util --ana --emu Tektronix_4404_OS2.0.emu --ext Tektronix_4404_OS2.0.dsk

4.                   on the PC: copied the Tektronix_4404_OS2.0.dsk file to the PC

5.                   on the PC: used EmEditor to trim off the sectors that UniFLEX does not know about (using the System Information record to see what the file size should be)

6.                   on the PC: loaded the Tektronix_4404_OS2.0.dsk file in FloppyMaintenance

 

This is what I got:

 

 

              And all of the files seem to be intact. I can read them and extract them. I AM STOKED.

 

Michael Evenson

http://www.swtpcemu.com/swtpc/Downloads.htm

http://www.flexusergroup.com:8080     <- http link to the sftp site

To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/f2fd9d22-ffe8-45f3-ab3c-a2fffb9c7a0an%40googlegroups.com.

[Adam Billyard]

Very cool.  I have been ftp-ing the file system to my PC so I can run better reverse engineering tools on the binary files etc

[nosn...@gmail.com]

Adam,

 

              I tried to do an extract of all the files with FloppyMaintenance, and it gets hung up on extracting the tail end of smalltalk/demo/demoChanges with the error message: An attempt was made to move the file pointer before the beginning of the file. This would indicate that there is something wrong with the fdns for this file. Do you get an error when you try to cat the file? If I skip this file, all of the rest of them extract perfectly.

 

Michael Evenson

http://www.swtpcemu.com/swtpc/Downloads.htm

http://www.flexusergroup.com:8080     <- http link to the sftp site

 

 

 

 

From: 'Adam Billyard' via MFM Discuss <mfm-d...@googlegroups.com>
Sent: Wednesday, September 13, 2023 10:08 AM
To: MFM Discuss <mfm-d...@googlegroups.com>
Subject: Re: Recommended archive for LKG pristine emufiles

 

Very cool.  I have been ftp-ing the file system to my PC so I can run better reverse engineering tools on the binary files etc

 

On Wednesday, 13 September 2023 at 14:46:05 UTC+1 nosn...@gmail.com wrote:

              OMG: Here's what I did:

 

1.                   On the PC: copied the Tektronix_4404_OS2.0.emu file to the BeagleBone with pscp Tektronix_4404_OS2.0.emu ro...@10.0.0.64:Tektronix_4404_OS2.0.emu

2.                   on the BeagleBone: cd mfm

3.                   on the BeagleBone: ./mfm_util --ana --emu Tektronix_4404_OS2.0.emu --ext Tektronix_4404_OS2.0.dsk

4.                   on the PC: copied the Tektronix_4404_OS2.0.dsk file to the PC

5.                   on the PC: used EmEditor to trim off the sectors that UniFLEX does not know about (using the System Information record to see what the file size should be)

6.                   on the PC: loaded the Tektronix_4404_OS2.0.dsk file in FloppyMaintenance

 

This is what I got:

 

To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/c13b86b2-9e99-4384-9662-3f4451e21fa2n%40googlegroups.com.

[nosn...@gmail.com]

Adam,

 

              I have posted an updated version of FloppyMaintenance to the web site and the SFTP site that allows you to skip a file during export if an error is encountered. If you have a Windows machine, you should download it and give it a try. I can also put your .dsk image of the Tektronix file on the sftp site as well if you like. The login to the sftp site is:

 

user: flexuser

password: flex

 

Enjoy.

 

Michael Evenson

http://www.swtpcemu.com/swtpc/Downloads.htm

http://www.flexusergroup.com:8080     <- http link to the sftp site

 

 

 

 

From: 'Adam Billyard' via MFM Discuss <mfm-d...@googlegroups.com>
Sent: Wednesday, September 13, 2023 10:08 AM
To: MFM Discuss <mfm-d...@googlegroups.com>
Subject: Re: Recommended archive for LKG pristine emufiles

 

Very cool.  I have been ftp-ing the file system to my PC so I can run better reverse engineering tools on the binary files etc

 

On Wednesday, 13 September 2023 at 14:46:05 UTC+1 nosn...@gmail.com wrote:

              OMG: Here's what I did:

 

1.                   On the PC: copied the Tektronix_4404_OS2.0.emu file to the BeagleBone with pscp Tektronix_4404_OS2.0.emu ro...@10.0.0.64:Tektronix_4404_OS2.0.emu

2.                   on the BeagleBone: cd mfm

3.                   on the BeagleBone: ./mfm_util --ana --emu Tektronix_4404_OS2.0.emu --ext Tektronix_4404_OS2.0.dsk

4.                   on the PC: copied the Tektronix_4404_OS2.0.dsk file to the PC

5.                   on the PC: used EmEditor to trim off the sectors that UniFLEX does not know about (using the System Information record to see what the file size should be)

6.                   on the PC: loaded the Tektronix_4404_OS2.0.dsk file in FloppyMaintenance

 

This is what I got:

 

To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/c13b86b2-9e99-4384-9662-3f4451e21fa2n%40googlegroups.com.