rogues / Rare european machines

Tore Sinding Bekkedal toresbe at ifi.uio.no
Sat Dec 2 18:57:55 CST 2006


On Tue, 2006-11-28 at 00:28 -0800, Brent Hilpert wrote:
> Jos Dreesen / Marian Capel wrote:
> > 
> > Chris M wrote:
> > 
> > >  So what do some of the starrier UK and mainland
> > > European machines go for? Heck which are they?
> 
> Just curious, where do NorskData minis fit in this?:

Well, OK, here goes my second Norwegian tech company history rant mail
today. :) Please ask if there was anything I omitted.

ND was founded in 1968, grew faster than Apple did for several years,
suddenly stopped growing in 1987, went bankrupt in 1992, and the stock
was priced at 0 in 1993. Had 4500 employees at its peak.

Interesting innovations which I cannot pinpoint having been done
anywhere else:
1968: First minicomputer with hardware floating point standard?
1969: First minicomputer with paging? (optional)
1973: First superminicomputer? (NORD-5)
1979: First single-board 16-bit minicomputer? (NORD-100)

I'm sure there are a lot more, but it's 2am, and those are the ones that
spring to a tired mind.

The seed for ND was sown in the NDRE, Norwegian Defence Research
Establishment. There, the SAM and SAM 2 (Which gained the nickname
"FLINK" - roughly translatable to "dilligent") computers were developed
by an ambitious gang of engineers. After the SAM 2 was completed, they
decided to start their own computer company. Many of the designers had
spent time at MIT, and the SAM shows clear inheritance from the PDP-1
and Whirlwind computers. This was to fade away as you progressed to the
NORD-1, however.

I myself have four ND's. One of them is one of their Motorola 88Ks,
however. The one to the far left is on 24-7, fighting a winning battle
with the winter cold. *With the window open.* Accounts to any SINTRAN
nostalgics that may be about are certainly available (Though you'll have
to telnet through my PDP-11!).

http://toresbe.at.ifi.uio.no/fourcompacts.jpeg

>    How prevalent were they in the European marketplace compared to, for
> example, PDP-11s?

Not extremely. The PDP-11s dwarfed Norsk Data in many areas, but Norsk
Data had a very respectable market share with their modular, integrated
office automation system NOTIS, and the NORTEXT integrated typesetting
system was extremely popular. Their SIBAS II databases were also neato.
These three systems were often combined together with rather neat
results - one would input an advertisement in NOTIS-WP, store it in the
ADBAS database, have the typesetting system typeset it, and use the
report generator NOTIS-RG to automatically charge for the
advertisements, for example. Later in the eighties, their Technostation
CAD workstation would also gain a very large market share. 

They also had substantial sales to the Soviet Union, due to some rather
silly behaviour on the side of CoCom - their machines were allowed in,
even though they dwarfed the banned VAXen in performance! 

>    Were they not used much in Britain (why doesn't Tony have any)?, more
> popular on the continent perhaps?

I don't know why Tony doesn't have one. They *were* quite common, though
possibly less so in Britain than other countries.

> Never see them here in Canada. 

A Canadian in the #classiccmp IRC channel passed up a ND-500 a while
ago. They tend to surface in surprising places. Last I heard, the US and
Israeli air force still actively use them as an F-16 flight simulator.
Norway might too.

> I recall seeing a bank of (some model) in the
> beam-control room at CERN, in 1985,

I think I have may have a picture of that room somewhere in a marketing
brochure. Was there a voltage control console with a graphics CRT
outside of the room? They were the orange generation of machines, right?
(NORD-10, NORD-50, NORD-10/S)

>  but I don't know whether that was an
> indication of their popularity/success or part of a bureacratic mandate to
> spread the purchasing money around to member countries. 

CERN had an excellent relationship with Norsk Data. In fact, Tim
Berners-Lee wrote his first prototype of the WWW concept, ENQUIRE,
(which was more like a wiki, really) on a SINTRAN ND-500! The story
behind the first delivery is quite funny, IMHO.

This all got started in 1972. Norsk Data was founded in 1968, and was
still quite a small company at this time. Back then, CERN put out a
tender for the control/acquisition computers for the Synchroton
accelerator. 88 companies replied, including DEC and those giants,
and... Norsk Data. There circulate many stories from the early days, but
these are from credible sources. 

CERN representatives were invited for dinner and a demonstration of the
technology. On the way from the airport, the CERN representatives
debated whether or not to accept their invitation to dinner, so as to
not completely drain the company of money! :) The representatives were
confident that they would give the contract to DEC, and felt like they
were freeloading.

The representatives were seated next to a teletype, and shown a
demonstration of timesharing (Developed by Swedish American Bo
Lewendal), BASIC compiler, FORTRAN compiler, file system, floating point
performance, etc. After around 30 minutes one of the representatives
interrupted - saying that they knew how remote mainframe timesharing
worked, but would rather like to see what their minicomputers could do.
The representatives refused to believe that the minicomputer next to the
teletype had been performing these operations, and had to be shown the
lead going into the machine!

During that demonstration and later dinner conversations, the
representatives grew certain that this was the right system for the job.
The price was the lowest reply to the tender which could feasibly
deliver. Once they got back and announced their intention to use Norsk
Data, all hell broke loose - they'd given the tender to a tiny Norwegian
company which didn't even have a working machine yet! The French embassy
told CERN that when Norsk Data failed to deliver the order, they would
be more than welcome back to Bull.

They had sold CERN a machine which did not yet exist anywhere but on
paper, the NORD-10. Reverse-compatible with the NORD-1 they had been
shown, it was to be a far more compact design. The logic diagrams were
done, and so were most of the microcode. The delivery was made on time -
of an empty rack! They had delivered the first machine, and a couple of
weeks later Harald Eide discreetly came in with a box full of cards, the
solder on which had barely dried. There was no time to debug the cards -
all of this was done on location, using a NORD-1 to spoonfeed the
NORD-10 microcode via an "umbilical cord". When the machine was finally
working, you could barely see the boards for all the wires and tape. But
the delivery was accepted. 

The later machines were driven down by car, by Bård Sørbye, Rolf Skår,
and Trygve Matre. In the back of BS's Renault 4 was a disk drive, TMA
had a couple of teletypes in the back, and RS had his car so full of
cards he could barely see out. Their cars were stopped by customs at the
border - and they almost didn't get out!

The CERN contract was Norsk Data's breakthrough contract, and solidified
their credibility, both nationally and internationally. 

I've translated the Norsk Data history page written by Jonny Oddene, an
ex-ND nostalgic, and I'm working on a more thorough one of my own.
http://toresbe.at.ifi.uio.no/nd/history.html

Here is the story of my own ND-100...
http://toresbe.at.ifi.uio.no/nd/nd100/

WHEW. Hope I didn't bore y'all.

rgds,
-Tore




More information about the cctech mailing list