• Please review our updated Terms and Rules here

Something's interesting in Denmark

bsdphk

bsdphk

Experienced Member
Joined
Jun 2, 2024
Messages
70
Location
Slagelse, Denmark
Hi gang,

Some of you may know me from other corners: FreeBSD, Vinyl Cache (The FOSS project formerly known as Varnish Cache), Beer-ware, bikeshed.org and so on, but my main hobby is being a volunteer in Datamuseum.dk, focusing on digital preservation.

Datamuseum.dk is private association, funded by member's dues (almost 1000 members) and we have approx 700 EU pallets of physical collection, collected over 37 years.

We are only open to the public on the first Sunday of the month, but if you come from afar and have an couple of hours in Copenhagen, drop me a note and I can try to arrange something.

But you do not have to travel to Denmark, because our "BitArkive" already holds more than 9700 artifacts, and not all of them related to Danish IT-History: https://datamuseum.dk/wiki/Bits:Keyword

Our charter instructs us to "Collect, preserve and present the Danish IT-history," but we are not restricted to that, and in particular we give high priority to preservation of artifacts unknown to the web, and I am creating this thread, so I can post a note here when that happens.

Here are some artifacts recently added:

* LifeBoat's C-compiler for CP/M manual: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/LIFEBOAT_ASSOCIATES
* A pile of Apollo manuals: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/APOLLO
* Mark Williams Coherent 3.2 intall floppies: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/MARK_WILLIAMS_COMPANY
* Firmware for IBM P/370 "Personal Mainframe": https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/IBM/P370
* A tape with a genuine 1584 page C04 ABEND DUMP from an IBM Mainframe: https://datamuseum.dk/wiki/Bits:30009748

And will finish the list with little gem I personally treasure:

We all know about "Backus-Naur notation", and therefore the two gentlemen will forever be tied together with a hyphen.

But here is a telegram to remind us, that of course they did not agree about everything:

* https://datamuseum.dk/wiki/Bits:30009303

Have a nice weekend,

/phk
 
More international artifacts added in the previous two weeks:

* ICL DRS8801 floppies, both native and CP/M: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/ICL/DRS8801
* IBM 3745 Service Processor 90mm MO disk: https://datamuseum.dk/wiki/Bits:30009827
* Install floppies for S/36, maybe Danish Language version: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/IBM/S36
* IBM internal "PCDT PC Debugger Tool", https://datamuseum.dk/wiki/Bits:30009825 and https://datamuseum.dk/wiki/Bits:30009828
* Storagetek 4600 microcode/diag floppies: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/STORAGETEK
* SoftTech's UCSD User's Manual: https://datamuseum.dk/wiki/Bits:30009790

The 3745 (also) contains a lot of OS/2 INF files (in ZIP format) about the 3745 "frontend processor" - also known as "The worlds best engineered and most expensive network bottleneck."

The two PCDT floppies are also interesting, they contain an internal IBM-PC debugger program I have never heard about before, the second disk has the manual.

And no, I had never heard about the ICL DRS8801 either
:)

/phk
 
More international stuff has landed:

* IBM S/36 "DIAG51 Release 3.0": https://datamuseum.dk/wiki/Bits:30009899
* Rudimentary AIX "backup" format support in AutoArchaeologist: https://datamuseum.dk/aa/docs/examiners/ibm/aix_backup
* IBM/RT 6150 Virtual Resource Mgr: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/IBM/AIX
* IBM/RT 6150 AIX V2.2.1 Install/Maintenance: (same)
* IBM/RT 6150 Config & Diagnostics: (same)
* IBM/RT 6150 (vs)Fortran compiler: (same)
* IBM/RT 6150 Fortran77 compiler: (same)
* IBM/RT 6150 Pascal compiler: (same)
* The BD Software C Compiler 1.5 manual: https://datamuseum.dk/wiki/Bits:30009774
* Sykes COMM-STOR Configuration floppy disk: https://datamuseum.dk/wiki/Bits:30009851
* Sykes COMM-STOR support in AutoArchaeologist: https://datamuseum.dk/aa/docs/examiners/sykes/comm_stor/
 
bsdphk said:
* Sykes COMM-STOR Configuration floppy disk: https://datamuseum.dk/wiki/Bits:30009851
* Sykes COMM-STOR support in AutoArchaeologist: https://datamuseum.dk/aa/docs/examiners/sykes/comm_stor/
Click to expand...

If you have a comm-stor, dumping the firmware would be nice.

also, I was the one who won the RCA 301 library tape auction https://www.ebay.com/itm/147242776229, I think you have a Gamma-30.
I haven't dug into what you might have on line from it in detail. I pushed some more info on the 301 and 3301
today. RSS isn't currently working since I'm modifying the script that generates the .rss files to be less resource hungry.
 
Last edited:
Hi Al,

As far as I can tell, we do not have the COMM-STOR itself, only around 75 floppies, which I'm currently curating.

Yes, we have a Bull Gamma 30 - the first computer in the computing centre of the Danish savings banks.

It's huge monster, the printer alone weighs 625kg(!) and we have never really had enough space to take good photos of it.

I know we also have some documentation for it, but it is packed away and I have never gone through it in the first place. Some day...

(Update: We have actually scanned a couple of documents: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/BULL)

(One of the photos in our index shows docs for a "Bull Gamma 10" ?)

As far as I know, we have no intresting tapes for it, but I hope to be wrong about that, and if we find any, I may need your help, because we have no 7-track kit.
 
Last edited:
Well. It has all the electronics needed. You just need to add some microcontroller to operate it. NRZI isn’t exactly rocket science to decode. In this case you might even want to use the analogue signals from the pre-amplifier. I sent the manual. Did you scan it?
 
I didn't say "quite a bit of work" couldn't be done :-)

But if it is for a single tape, it might be more efficient use of time to enlist the help of somebody who already has a setup, than to build a setup from scratch.

As to the manual, no I do not recall scanning it, or even seeing it, but if you gave it to us, we have it, and can find it again.

Anyway, this is a bit theoretical, until we find we have tapes to read.
 
bsdphk said:
And will finish the list with little gem I personally treasure:

We all know about "Backus-Naur notation", and therefore the two gentlemen will forever be tied together with a hyphen.

But here is a telegram to remind us, that of course they did not agree about everything:

* https://datamuseum.dk/wiki/Bits:30009303
Click to expand...
Note the White Plains, NY address for Western Union. I'm guessing John Backus was working IBM Watson Laboratories at the time?
 
shirsch said:
Note the White Plains, NY address for Western Union. I'm guessing John Backus was working IBM Watson Laboratories at the time?
Click to expand...
Yes, he did.
I wonder if there are any telecom nerds left who know how to decode the "vby59/k439 white plains ny 27 26 1257" line at the top ?
vby(59) is obviously "Valby" in Copenhagen, and "white plains ny" is also obvious, but I expect the rest of the line would allow them to pinpoint the originating teleprinter ?
Would be interesting to know if he sent this from IBM or if he went down to the local telegraph nearest his home.
 
It's possible he lived in White Plains at the time as Yorktown Labs are only a 20 minute commute from there. But you can be sure that IBM had an in house teletype service at all major sites.
 
TubeTimeUS (Is he present here ?) stumbled over a HP9133 hard-drive, and since I cannot resist that kind of challenge, the AutoArcheologist now, in addition to LIF support, also has preliminary support for listing HP85-family BASIC programs.

Here's one example output: A Bio-Rhytm program: https://phk.freebsd.dk/misc/LIF9133/14/1446d49f2.html

The code is inspired by and lifted facts from this abandoned github project: https://github.com/vp080808/HP85-Detokenizer but gets a fair bit further.

There is a fair bit more work to do, but a lot of it involves comparing to how a real HP85 LISTs a file and I dont have a real HP85 at hand.

If somebody with a HP85, and preferably Python3 skills, want to help take it further, drop me a note.
 
By a total coincidence Ken Shirriff wrote an article about the SpaceLab computer (https://www.righto.com/2026/05/reverse-engineering-spacelab-computer.html) just as I had started scanning some documents related to SpaceLab.

The documents, including the MITRA 125-S references manuals have now landed in our BitArchive:


A couple of other mildly interesting bits:

* IBM's "suggestion" for a S/360 for Århus Univesity 1967: https://datamuseum.dk/wiki/Bits:Keyword/COMPANY/ÅRHUS_UNI
* Olivetti M20 "OLIWORD2" manual: https://datamuseum.dk/wiki/Bits:30010189
 
PHK,

as you dig through the datamuseum.dk archives, please keep a look out for any Burroughs B6700 materials. As far as I know your museum is the only one on the planet with an intact Burroughs B6700 MDL (Maintenance Diagnostics Logic) panel, otherwise known as blinkenlights / front-panel. I am hoping that whomever saved that MDL might have preserved magnetic tapes or listings of the MCP and compilers/libraries for the machine as well.

Thanks to you and the museum members for your preservation efforts.
 
Noted.
We seem to have B6700 donations from two sources.
The MDL is from "Risø", the Danish Atomic Energy research establishment, I think it is unlikely that they would have donated software.
But we have one (or two?) disk-platters:
I would not be surprised if there are still bits left on those surfaces, but that would be a lot of work for a very long shot...
 
I have one of those platters too (although I have no idea which system they came from, Burroughs used them on their mid-range systems too). I have often wondered whether the bit-density is so low that some sort of Ferrofluid might show the bits enough to read optically.

"B6700 donations from two sources." - I recall that your museum has a B7700 MDL and a B6700 MDL, Finn Verner Nielsen from your museum provided me with some measurements of the metal work that goes with the MDL (I hope one day to make my own replica).

Anything else that is Burroughs B6700 that comes to light I would be glad to learn about. thanks.
 
I've been spending a lot of time in our remote storage, so no new digital artifacts for now.

But we did get the Rational R1000 running again: https://datamuseum.dk/wiki/Rational/R1000s400/Logbook

If anybody else has encountered MB81C71A (64kx1 SRAM) chips which fails, we'd like to hear about it.

We see them failing with power and without power, at a rate of several percent of the chips per year.

We're trying very hard to get to two running R1000 machines, but IOC board number 2 gives parity errors on the Unibus, probably because of a sick SRAM chip.

And yes, the R1000 has a Unibus :-)

The first version was a 4 CPU machine, each CPU had a PDP-11 as I/O processor.

The next version was a single-CPU machine, still with a PDP-11

Then they replaced the PDP-11 with their own I/O processor, flaunting a 68K20, but kept all the Unibus peripherals.

Finally, in the model we have, they also replaced the Unibus peripherals, but didn't want to redesign the I/O processor, so they kept the UniBus.

But it gets weirder: The ethernet is on a separate 68K20, on a CMC ENP100 VME board - on the far side of the UniBus.

And that's not even the weirdest thing about the R1000 :-)
 
This summer Datamuseum.dk received a donation of approx 250 color slides (24x36mm), from the estate of Tage Vejlø, including this wonderful picture of the "patch panel" on the back of the prototype GIER computer:

gier_patch_panel.png

It doesn't get much more vintage than that :-)

The full gallery is here: https://datamuseum.dk/gallery/30010236/

Tage Vejlø was part of the team who built "DASK", the Danish enhanced clone of the IAS computer, via BESK in Stockholm, and he became head of production in Regnecentralen.

The second Danish computer was originally meant to be a one-off prototype, but companies, almost Futurama style, threw money at Regnecentralen, so they started a production, and produced just shy of 50 "GIER" computers in total.

To augment the somewhat deficient GIER prototype documentation, Tage Vejlø photographed the prototype, and at least some of those pictures, like the one above, are in this donation.

Other pictures (in the gallery) show the production, PCB production, thoughtful engineers with pipes, ladies with soldering irons etc. All very "period", which is to say first half of the 1960ies.

The GIER was fundamentally a Von Neumann IAS computer, built with germanium transistors: 40 bit words, 20 bit half word instructions, 1024 words core and drum storage, but it had a number of improvements, including advanced addressing modes suitable for the ALGOL programming language, and sported the world's (still!) fastest and best paper tape reader (a story for another day.)

We have a working GIER computer in datamuseum.dk, serial number 35, probably around 60 years old, but still going strong (except for the IBM typewriter.)

Wiki: https://datamuseum.dk/wiki/GIER

BitArchive: https://datamuseum.dk/wiki/Bits:Keyword/GIER

AutoArchaeologist: https://datamuseum.dk/aa/gier/

PS: If you wonder how an ALGOL-60 compiler looked like in the 1960ies, see artifacts 30000728-30000740 in the AutoArchaeologist: Written in assembler, and commented in pseudo-ALGOL.
 
You must log in or register to reply here.
Back
Top