ICL History.

[Malcolm T]

Yet another sad demise of a British company. Makes one want to weep .

https://www.youtube.com/watch?v=EkTHDgYTh64&t=260s

[Gulliver]

All sad. I grew up in Letchworth Garden City where one of their manufacturing/office buildings was. A good friend's dad worked for them, programming among other things. Circa 1980 he had a wonderful ICL PC with all of a 5 or 10Mb hard drive that didn't have auto parking heads so one had to type a command into CP/M to park the heads before shutting it down. Later I got to try the "One Per Desk" computer with built in modem and telephone....and I think the storage medium for that was a derivative of the Sinclair Microdrive!

When Fujitsu pulled the plug, their building in Letchworth lay dormant for a while before being redeveloped as "The Nexus Building" which seems to be offices. It was only 30% full for several years but finally appears to have been full the last 10 years or so.

[Buzby123]

Back in the 80's I installed an industrial controls project for ICL at Letchworth. At roughly the same time I also did a similar project for IBM in Greenock.

The production lines I saw at both sites were producing similar devices, desk-top terminals, or something like that.

ICL had rows of workstations branched off a long conveyor. Sat at each workstation was a worker who had one job to do, and when that job was done the product moved to the next workstation where another operator did the next part of the process. This huge room had the feel of a Victorian cotton mill, but a bit quieter.

Meanwhile, at IBM, they were having a little celebration. The last worker on their production line had been replaced by a machine, an automatic adjustment of brightness/focus/contrast etc. to get the best picture.

The IBM production line had been designed as a continuous loop, with automatic stations performing the jobs at each station. At the start of the loop the CRT was removed from it's polystyrene packaging. The packaging was sent to the end of the loop, inverted, and then used to hold the finished product. The whole process was automated, with just a couple of maintenance workers overseeing.

I wondered then how long ICL could survive.

Regarding the OPD. It was a product that felt like it was designed by a committee. It had some good features and some bad features, but very little integration between them. And why on Earth did it have removable storage that was compatible with nothing, not even the QL Microdrives on which the OPD drives were based ?

Yes, it is sad to see what happened to ICL, but the writing was on the wall.

[G6Tanuki]

I remember going to ICL in Bracknell to talk to them about their servers/workstations; these used Sun SPARC processors, but they were last-generation ones and the hot-swap disk-drives needed ICL-specific firmware [unlike Sun's current generation which just took any drive of the same size as the others in the array].

We didn't buy.

ICL seemed stuck in the mainframe era, and never really managed the transition to the PC and networked computers [from memory they had their own version of Ethernet too]. This was a big problem or some companies back in the 80s - Digital had their own "PC" [the DEC Rainbow] but it could only run DEC-customised software, Digital's VAX was a workd-beater but they went off-piste with the Alpha chip and nobody was buying minicomputers any more, so DEC were bought out by Compaq [who in turn fell into the hands of HP].

Other 'big' computer companies of the era - Univac, Burroughs, NCR, Honeywell-Bull, also fell by the wayside; the smart ones transitioned away from selling hardware and became 'systems integrators' - Fujitsu took over ICL basically to get hold of their large residual installed user-base in public-sector organisations.

Time moves on. Evolution takes no prisoners. I don't really miss ICL.

[TonyDuell]

I suppose not strictly ICL (in that the original design was done by Three Rivers n the States) but I've always liked the PERQ. The attached photo shows 4 of them, along with other machines. The little thing about the size of a tower PC is an ICL PERQ 3A (AGW3330), a 68020-based machine.

[Bill]

ICL gave us half a dozen PERQ machines , 1980 something in return for some project we completed for them years earlier. Our little unit used them for PCB layouts, IIRC a large floppy was output which then was loaded on the EMMA (Electronic Mask Making Apparatus) photoplotter. The PERQ was a revelation at the time, very fast with a well defined monochrome screen, easy to use compared with what had gone before
I also had a One Per Desk which I found next to useless, but to be fair it was old by then.
I passed it on to a forum member some years ago.

[Bill]

ICLs factories in the Manchester area were mostly inherited from Ferranti. Park Road in Dukinfield was an old cotton mill. West Gorton was a heavy engineering works. Late 70s they opened a brand new assembly plant in Ashton Under Lyne poaching technical staff from other places with offers of good salaries. I think that plant lasted about 5 years and must have cost millions , it was closed with many redundancies, that place is still there but hasn't been involved with computers or electronics since.

[TonyDuell]

The original PERQ1 was close to being the first commecial '3M' machine. That's 'Megabyte, Megapixel, MIPS'. At least one megabyte of RAM, at least one million pixels on the screen, at least the power of a VAX 11/780. It didn't meet the second of those, being only 768*1024 pixels (portrait monitor).

The PERQ 1 processor was a big board of mostly TTL, some programmed PROMs and 4K*48 bits of microcode RAM. The machine loaded the microcode from disk when it booted. PERQ1A and later machines had an upgraded processor using some PAL chips to add a few more features and 16K*48 bits of microcode RAM. Writing PERQ microcode is actually quite enoyable.

There was a simple 'blitter' on the CPU board (the 'RasterOp machine') to handle screen bitmap updating. The display was purely bitmapped, there was no text mode or anything like that, so having fast bitmap updates was important.


It was a spectacular machine when it came out in 1979. But alas the 20 bit (!) processor was limited to addressing 1Mwords (2Mbytes) of memory, which was too little by the mid 1980s. There was the 24 bit PERQ 2T4, but that is so rare you will not see one.

Given the origins of the OPD, I am not surprised it wasn't that useful.

[G6Tanuki]

I remember ICL also had a big operation in Kidsgrove, Staffs. which is where the 2900-series large-machine operating-system VME/B was developed.

VME/K [another OS intended for smaller 2900-series machines] was developed at Bracknell.

So the last letter of the OS names were transposed compared with the places they were developed!

The death of ICL was closely linked to that of Standard Telephones & Cables [STC] - following the failure of STC to win anything significant in the BT 'System X' contract, and with ICL suffering significant losses too, their managements arranged a merger, hoping to benefit from the 'convergence' of voice and data services.

They were a decade or more ahead of this actually happening and the resultant combine went under after a few years.

[I spent some time at STC's Harlow site in times-past, a truly fascinating place where a lot of the fundamental work on fibre-optic cabling was undertaken by Charles K. Kao. What was the site of Standard Telephone Labs in Harlow is now known as Kaopark in his honour - https://kaopark.com/the-park/history/ ]

[wireman]

I once went to a recruitment presentation by ICL (probably mid 80s). I went away with the impression the company was living in the past and was not a place where I would want to work.

[Craig Sawyers]

There was a post about IBM. Back in about 1980 I was a research student in the laser lab at Southampton. IBM were having problems with a new colour monitor - the switch mode supply kept blowing up, and they suspected it might be random flash-over in the picture tube. Could we help?

To trigger flash over, I focused a pulsed laser onto the gun assembly through the envelope. Ping! The power supply died on the first laser pulse. OK, what they suspected was confirmed. A fix was implemented.

Next time - first laser shot, ping! Power supply died.

Then a raft of managers arrived too. Technical guys were sweating, much assurance that the problem was fixed. First laser shot - ping - power supply died.

I had to leave the lab and let them get on with recriminations. They did not come back for another test.

Back on subject. The maths department at Southampton had an ICL mainframe - I seem to recall it was a 1907 or something similar. If you wanted more than 8k of core store for your (punched card) programme you had to get written permission from your supervisor.

A little later we moved to, and lived in Icknield Way in Letchworth - about half a mile from the ICL factory. Sad that they are no more.

Craig

[Jeremy M0RVB]

My first foray into programming was on an ICL 1906A ("By George" I hear! Sorry...) - three huge racks of it plus a room of discs and drums, and a room for engineers who were always on site kicking the thing. Years later I had to oversee the closedown of an ME29 setup including scrapping it - I managed to get a local firm to actually pay money (everyone else wanted to be paid to remove it) - £100 cash (1995 I think) of which I spent £99.99 on a 'fridge for the office and squandered the remaining penny!

[G6Tanuki]

In the early-70s during school/college holidays I 'moonlighted' as a night-shift computer operator on an ICL1901T operated by a local engineering company.

They had normal night-shift-operators doing batch runs to do backups, prepare sales-/financial data and orders/despatch-notes for the next day, but one of the usual two operators was away having her baby; they weren't allowed to have just one operator on-site overnight [health & safety even back then, the operator working alone could have been eaten by the power-operated lid on the line printer] and because I had been trained in first-aid as an Army cadet I got to sit-in and do things like mounting tapes, loading paper into the line-printer, decollating the multi-part forms and suchlike. I even got to type commands into the Teletype-33 that was the console.

The 1901T didn't run George, it ran a thing called Exec, which was rather primitive. I remember you had to manually tell it to load a program, attach the program to the peripherals, and then tell it to begin executing.

The two programs I remember we ran a lot were NITA [NIneteenhundred TAbulator] which took input from a tape-drive and a layout/template-file on disk, and formatted the data before printing it on a horrendously-noisy drum-printer, and XSDC which was a sort-routine.

FI#NITA loaded the program from disk
RU#NITA [or was it GO#NITA?] then ran it.

FILETAB and PLAN were also names I remember.

For an 8PM to 6AM shift I got paid the princely sum of £20 per night! Great money back in the 70s.

[Dave Moll]

When I started with BL in 1971, £20 was my weekly wage as a trainee computer programmer.

[dragonser]

I had many years ago a Rair Black box running CP/M which I think was somehow related to the ICL machine running CP/M ? there was also a version running M/PM I think ?

[G6Tanuki]

Given that the history of ICT/ICL has entwinings with the Canadian Ferranti-Packard computers which ICT adopted in a revised form as the basis of their 1900-series, perhaps a cross-link to this current thread on Ferranti would be relevant: https://www.vintage-radio.net/forum/...d.php?t=196957

[McMurdo]

by coincidence I watched that video last night. It seemed mainly about IBM and how they (and every asian computer firm) trounced ICL. They could easily have substituted ICL for any number of smaller computer companies of the era and the video would've still had a ring of truth. They glossed over the success of Ferranti particularly with defence computers. I thought it was unbalanced, misleading and incomplete.

[Restoration73]

I have a copy of this music LP they commissioned (!)

https://retrocomputingforum.com/t/ic...rvey-1980/2520

[Trevor]

Back in the 70's I knew a guy who had the contract to remove scrap from the ICL computer plant at Kidsgrove
I obtained from him loads of power transistors ADZ11 Germanium units and Silicon MA13.s These I treated as 2n3055 made loads of amplifiers with them
He had loads of caps resistors 8 track heads fans heatsinks and wire
He helped me so much in my early years
Thanks Bill

[hotbulb]

My late father worked for ICT/ICL and was one of the engineers who worked on installing, and maintaining, the ATLAS computer at Harwell. As I child I went into work with him on occasion and marvelled at the place. I still have some bits of it up in the loft - including one of the hard discs, which is about four feet in diameter and will one day make an interesting coffee table...