Why the UK's IBM Failed
Vložit
- čas přidán 15. 06. 2024
- In 1968, the British government arranged a shotgun marriage between three computer manufacturers to create International Computers Limited, or ICL.
The government envisioned ICL as the United Kingdom’s answer to IBM - a national champion in computers. But the company could not compete against American and East Asian competition and eventually sold to Fujitsu.
The United Kingdom pioneered computing technology. So why did ICL fail?
In this video, we are going to trace the history of the British computing giant from its punched-card days to its final demise in 2000. Sit back and relax. This is going to be a long one.
Hat tip to the National Museum of Computing (TNMOC) for some imagery
Links:
- The Asianometry Newsletter: asianometry.com
- Patreon: / asianometry
- The Podcast: anchor.fm/asianometry
- Twitter: / asianometry
This was a long one. Remember to check out the newsletter: asianometry.substack.com/
Why not tell the truth about IBM and the Holocaust? Literally the title of the book and historical records show that they killed millions.
Conveniently left that out? Do you deny it? Because denying that is illegal in most places.
You forgot IBM helped the NAZIS!!!!
Quality doesn't need to be compressed always in short videos (although I can understand that one wants viewers). One possibility is to make "long" (if at all) videos in parts.
I enjoyed this video but you took way too long on the historical context before the company even existed (25 minutes or 70% of the video!), before doing 10 mins actually on the title of "Why the UK's IBM failed". The history is important and interesting, but it should be more like 30% rather than 70%. Also, there must be some video clips you can use to make it a bit less static. Even slowly zooming and panning can help, or gradually building up a slide with other images rather than just a static slide. Otherwise it's like a PowerPoint presentation with the text removed, which is a bit weird. Also, you could narrate more dynamically as if talking rather than monotonically as if reading a script. I sound like I'm being harsh, but these are a few easy ways you can make your content much more engaging! Thanks for making the video.
@@ytubeleo I'd much rather static slides than panning over the image. I eventually gave up on The History Guy channel just because he pans so much. Sometimes I want to look at an image for its details without pausing the narration.
The narration is one of the attractions here. I know I'm getting real information, not just a show. Dynamic animated narration is fine for kitten or puppy videos.
I joined ICL in 1982, was still there in 2000 when it was bought out by Fujitsu, and I'm still there today! That tall building you show at 25:58 is the one I am nominally based in (WFH nowadays), which looks much the same today, just with a big Fujitsu sign on the roof. Starting off as a programmer I worked on 2900 and ME29 systems at first, then the first ICL PC systems (they licensed the Rair Black Box PC), then a bit of everything, the ICL-built CDOS based PCs, ICL One-Per-Desk, ICL DRS 20, early Unix systems such as Perq and DRS 600, Series 39 mainframes both large and small, I even did one project porting stuff over from a very old System 4 (that was a seriously weird architecture), later ICL DRS 3000 Unix and DRS 6000 Unix (first System V Unix to run on Sparc chips). After that it was predominantly Fujitsu servers, and today it's mostly VMs and cloud, but occasionally I have to go into the office to swap out a dead disk, or give some recalcitrant piece of hardware a kick. I really enjoyed this video.
Thanks for listing so many of the amazing projects and achievements that are glossed over in the video... but we were treated to a long sequence about punched cards, so that's okay then ;-)
@@G7LWT Don't knock the punched cards! In my first project for ICL,we still needed to occasionally punch job cards (the 2 or 3 cards needed to start a program) but we didn't have a card punch, we had to borrow one from a team on the other side of a very large building complex. So I rebuilt one from 2 broken hand punches. One was Hollerith, one was ICT. Must have been 40-50 years apart in age, but completely parts compatible. A childhood spent playing with Meccano was not wasted.
@@EddieCochrane Please don't misconstrue my reply - punch cards and tape are dear to me too :-) The video skipped over Quattro, Series 39 and CAFS, Clan series Unix machines (OfficePower anyone?), CDOS/Unix on DRS300. Even the DRS PWS (DRS60/80 workstation) deserves a mention as the world's first multitasking PC supporting MS-DOS 4.01 (and the funky memory cards). Most platforms interworked with everything else - it was all quite elegant. The open systems networking fundamentals that I learned in STE04 and WIN01 are still in use every day. There are plenty of videos about punched cards - not so many about the topics mentioned above :-(
Fujitsu YIKES - I heard their name mentioned in the Lords - I interviewed at Lloyds Bank and had to ask the 3 interviewers why their managers were in UK prisons 37 years - eventually started working in IT for Porsche JCB and others and _Feb 2020_ they terminated me with pneumonia - I was in ICU on oxygen 8 days
Thanks for sharing your experience! Did you ever consider making videos to share this experience? That will be really exciting to watch!
Part of the reason for stagnant innovation in the UK was cultural. If you graduated with a degree in engineering in 1975 you could either go to a UK university and continue a living wage in academia, go and work for a British company in Bedfordshire where your perpetual boss would be a self-serving ignoramus in - crucially - a public school tie or you could bugger off to the USA where you'd have an office in California, Ford Mustang, a McMansion and trophy spouse by the time you hit 30. Hence the fruits of the UK's excellent academia often flowed away the moment the very moment they became productive.
It is still largely true today. Many graduates immediately head overseas for more money and better prospects than a UK company will offer them.
The Brits are very creative and are up to speed in aviation, computing, electronics and medicine to name a few. The problem is marketing - UK companies love putting accountants in charge of big business. It's very difficult sell an accountant a new idea and fill them with enthusiasm.
Ha ha what a myth. British management at ICL epitomise their failings. Why are so many British companies owned by and run by non British corporations. Rolls Royce (cars) , Bentley, JLR, British Airways, Morrisons, etc etc. We are shite at marketing too.
My (software) company tried to do business with ICL in the mid 1980s. It was almost impossible. Scientists and academics everywhere instead of product managers.
One could argue it is also difficult to compete when the US government openly medals in the private sector under the guise of national security, when in reality they are lobbied to line their own pockets and that of US interests.
But the technology came from an American.
@@brooklynknite Did it?
Great Video! I learned to program in FORTRAN on an IBM 1130 in 1967 while I was in high school. Among my earliest professional IT adventures was sorting a file containing approximately 6 million cards; it took weeks, simultaneously running 3 IBM sorters. The 1130 and its 16K of RAM stood me in good stead for the microcomputer revolution. I was invited to Silicon valley in 1980 after winning a software competition. (While there I had dinner and drinks with Steve Jobs and Steve Wozniak) Alas I never became a Silicon Valley billionaire, but did have a relatively good life writing software. As I approach my 70th birthday, I still get the occasional small contract to supplement my pension. My cousin Bob worked for ICL in Manchester as a machine "commissioner." Essentially he ran tests on all of the mainframes before they went out to a customer. My father worked in Liverpool at English Electric in the 50's but he was making jet engines not computers. Many of your presented images brought back memories from a much simpler time. Well done!
Great story!! When I saw ‘sorters,’ my mind instantly to Unisys check sorters, where I got my start in IT, but as a lowly maintenance tech; I wish Imhad taken an interest in software as you did. Anyhow, it paid the bills, and got me a job at IBM (where I avoided the check sorter team :)) which eventually sprung me into sysadmin work. Things changed fast - I started out monkeying w a Unisys sorter (DP700??) that used core memory and 8.5” floppies. My favorite was the 9TB IBM Shark, a disk subsystem about 3 racks wide of SSA hard drives. And now? Maybe I should apply at a computer history museum. Anyway, thanks for sharing your story, brought back some decent memories.
My dad was one of the team that developed Fortran. He also worked at IBM on the STRETCH project with the idea to run more than one program on a computer at a time.
Later I started work using an ICL 2950 using DME and George 2 with the promise that we would move to the new fangled VME very soon.
VME was way ahead of the game, but possibly too clever.
Most companies just wanted something that would do the job at a good price and not spent millions converting to the 'better' VME and many jumped ship to IBM at that point.
I always felt that ICL was an example of missed opportunities.
To say that they ignored the small computer market would be incorrect and seemed to be well sorted until taken over by the total dinosaur that was STC.
ICL had the DRS range, ME29 mini computers plus the Quarto PC's.
While you were in high school learning Fortran, I was in graduate school (University of Chicago) dutifully submitting my punch-card deck in the morning and waiting for the print-out results (usually an error) in the afternoon. In those days university research computer centers were big/bureaucratic deals...no longer, thank goodness.
@@johnweiner
My father took me to work with him before I attended my first school.
He then taught me very basic FORTRAN.
As he was on the team that developed the language, he knew a thing or two.
My first program was made on punched cards and so was the output, but being next to the computer meant that the turn around was quick.
At the time I had no idea how lucky I was to get to play with real computers back in the 1960's.
There was still segregation at that time? Lol Martin Luther king?
As someone who is old enough to remember working with mainframes and punch cards and saw the rise of the personal computer I found this backstory video fascinating and enjoyable thanks!
Punched card, what did you do? Voting systems, bank?
i only like coding, this is all hardware here...
Hmmm! Lucas, computer software needs an understanding of hardware and older hardware was very close to software.
I learned about punch cards from CNC machining in the 90's. I guess they were slow to catch up too.
@@ryanmartin4602 CNC will always bit a bit slow. lots of bespoke hardware on its digital side, the machines as a whole are huge CapEx to update, and when maintained they keep chugging along for a very long time with the precision required by the clients of the shop.
I worked for ICL in the late eighties/early nineties. The reason they failed was the inability to see the rise of distributed computing and personal computers. They were slow to embrace open standards and then to realise that open standards would lead to a computer on each desk rather than a terminal linked directly to a mainframe.
Several mis-steps followed with the introduction of the likes of the OPD and the over-engineered early DRS models. But in the end the products ICL were offering were nowhere near as easy to integrate as cheaper PC clones running MS-DOS and windows. When they did jump on the PC desktop bandwagon they had to buy Nokia Data to buy-in PC manufacturing capability rather than develop and manufacture their own products.
I was working in the manufacturing side, but when it became obvious that manufacturing was going to be hived off, I jumped ship at the end of 1993.
By 1999 I was working as an IT contractor and was surprised to see a bank still using their 2900 VME system even that late. It brought back memories. lol.
I was working for DEC then and they made exactly the same mistakes.
IBM was lucky that their PC became the platform for Novell Netware, but they messed up on Unix. ICL was onto the Unix platform, but hammered by the idea of VME. By the time ICL had PC's (8088's at first) it was already being bypassed by Dell and Compaq. It's Unix solution was not enough and it's stable of VME and S25's was far too old.
I think some companies were emulating ICL 2900s well beyond 2000.
Kodak was yet another who guessed the wrong way thinking they knew better. They looked up, saw the future then stuck their heads back in the sand.
@@LeslieGilpinRailways This was a proper, original 2900 system, with 7500 consoles, the whole shebang. In a proper nuclear hardened data centre too. Brought a tear to my eye, I tell ya! But yes, VME lived on after the mainframes were pensioned off.
When I started high school in 1992 in Lincoln, England. I walked in to discover rooms full of brand new shiny ICL desktop computers - enough to have one for every student. I may not have gotten to see ICL become a dominant world leader in mainframes, but I did get to see their involvement in our school change our world as students.
Thank you ICL.
Completely agree we should of gone with working with local government and local education yet many beloved local government should not be funded looking at you Thatcher 😊
When I was working in IT in the '70s and '80's there was the saying that 'Nobody ever got fired for buying IBM'.
And that attitude has transferred to Microsoft, while Microsoft has since completely reversed it's business principles from working for customers to trying to exploit them at any cost. Yet blind trust in Microsoft makes it increasingly difficult to convince others to be wary of the current Microsoft.
@@johndododoe1411 Well, we're a Microsoft-free zone - Linux on all PC and Laptops. Android phones (As it runs on a Linux kernel).
@@johndododoe1411 but to certain degree, Microsoft had to change. Once you get too much market share you have no choice but to ‘eat your young’. Shareholders only care about dividends, so you have to make that money. If there is really not much market share to be able to gain, you have to look to your users to get those additional profits from.
Well I am not surprised that ICL failed when you consider the calibre of the employees ! My neighbour an ex RAF pen pusher was recruted in about 1981 and was hailed as a computer wizz. I at the time was running a machine code Nascom 1 and my neighbour didn't have a clue about programming or electronics. No surprise there then. Needless to say he had the obligatory company car.
Without doubt, they employed a good many people with far better minds, than your arrogant one.
Actually, I noticed the same military "tie" brotherhood in ICL and decades later in Vodaphone, usually in non-jobs.
This is called network marketing, he was probably hired for his contacts in the RAF, given that ICL were the Prime supplier to the armed forces and the biggest single system was the RAF logistics application run by pen pushers meant he was far more valuable to them than some one tooling around with a Nascom. Remember the saying. "He who does not play around with assembler as a teenager has no heart, he who does it as an adult has no head " Z
Ikr
Communists supporters
Nascom compared to a large business computer is like a bicycle compared to a train.
Ferranti Mk1 was built in conjunction with Manchester University who had built the Manchester M1 .. mostly by a load of ex-government workers who could not say what they had worked on during the war ... they had been working at Bletchley Park on Colossus, i.e. Farranti had computer engineers who had been computer engineers longer then ENIAC had existed ...
I think you might have focused on LEO a bit more. Granted they were tiny, but for a fleeting moment they did hold a technological lead, as it ran it's 1st business applications in 1951. Lyons, the company that made LEO did something that would be unimaginable today, they recognised very early on how useful computers would be to business, and without any prior involvement in electronics, decided to make their own. They were also very British, in as much as their core business was operating tea shops, they were the equivalent of today's fast food, with a Lyons Corner Houses as ubiquitous as McD or Star Bucks are today, having at least 1 on every high street in the UK.
I haven't finished viewing the video yet but was really surprised that LEO wasn't mentioned when the development of Computers was discussed. The LEO was the first computer to be used for Business and so deserves a mention. Also the work of Manchester University seems to have been omitted with no mention of the MU Baby which ran the first stored program. I will probably have more comments when I watch the rest!
@@chutalotr Fully agree that LEO (Lyons Electronic Office) Computers Ltd deserve a mention. At a time when most other computer installations were concentrating on 'number crunching' large mathematical problems (artillery trajectories, weather prediction, Manhattan project, and other scientific tasks), J Lyons & Co Ltd. realized the possibility of processing large amounts of business data very quickly. In building a computer for their own needs, they eventually (almost by accident) created a computer manufacturing subsidiary. The company did not have a strong electronics background and sold only one product (computers). In April 1963, LEO merged with English Electric to become English Electric-Leo Ltd in an effort to stave off the growing IBM threat.
@@greybeardza9197 I think the potoffice bought a LWO machine for telephone billing. I also understand that LEOs were emulated on some of the microcoded machines - the 2960 I think. Where I worked bought a couple of 2960s to start the move to VME but landed up running George 3 on them using DME.
@@chutalotr All told, I believe that the Post Office bought 14 of the LEOIII machines. I was told that the Post Office requirements to print monthly invoices was the main motivator in Introducing the Xerographic printer that just about burned down LEO's computer room. Previously, using teletype machines, the GPO could not keep up with printing invoices for London alone - let alone the rest of the country. If the account I was told is to be believed, forever after the fire, when the Xerographic printer was being used, there was always someone on standby with a fire extinguisher! In the absence of suitable lasers at the time, the Xerographic printer used powerful arc lamps as a light source to de-ionise the printing drum - but they could just as well set fire to the printer paper (rolls of newsprint) if there was a paper jamb!
Somebody at J. Lyons & Co had incredible foresight. They deliberately set out to determine how some of the advanced technology developed during WW2 could be applied in their business in the immediate post-war period. The LEO I was a direct development of the Cambridge EDSAC which first ran in 1949 and several years ahead of UNIVAC in the U.S.
Considering the importance of computers in everyone's lives and work today, this stuff should be taught in school, in history classes. Great video, as usual, thank you !
Ironically this stuff isn't taught even in computing science "schools", aka, graduation courses.
It's something you have to chase yourself really. There was a really good PBS series about early computer history called, The Machine That Changed the World.
@@monad_tcp Yeah, because you better be studying your maths not worrying about niche history.
If everything that was taught that every dumb ass on the webernets should be taught, the brats would never get out of school. For the uninitiated school seems to be there to teach you everything you will need in life, but it isn't! It is there to teach you how to learn! For example, how important is it to learn how to win the Victoria Cross? Some people want you to learn that, this guy thinks it important to include his poor Harry Potter humour. It isn't funny. I am an adult, I never read Potter, it is for children, it is like Dr. Suess, without the skill in writing or observation.
@@minirock000 school isn't there to teach anything besides compliance.
And universities almost seems like a MLM pyramid when you look at it from the commercial pov.
Great video, brought back lots of memories for me. In 1966 I joined BMC at Longbridge as a trainee computer operator, working on an I.C.T. 1904 system. In due course I rose up to become a shift controller, overseeing the 1904 and an IBM System 360/40 in the same computer room. I left in 1970 because I could see that the British Leyland Motor Corporation (as it had become by then) was failing rapidly. I went to work for IBM Laboratories Ltd. at their Hursley plant (near Winchester). IBM was a great company to work for, the best I ever had - night and day compared to BLMC.
I'm old enough that one computing lesson at school was the teacher showing us how punch cards worked using index cards, a hole punch and a knitting needle.
They made really nice flash cards and Christmas wreaths.
We had four actual IBM 029 keypunch machines along with other "antique" electro-mechanical and electronic business machines in the classroom. A basic add/subtract/multiply/divide solid state noiseless electronic calculator that was the size of a large toaster oven.
A little further back and you might have had a school visit from the ingenious Mr Babbage! :)
The knitting needle is apt as those punch cards originated to program control Jacquard looms
You had a computer class? You were lucky.
If the UK hadn't been so concerned with keeping everything secret (colossus etc) and had given the people who had built those early computing devices during the war more involvement in the industry, maybe the UK could have had a better outcome.
That’s what I also assume(d).
I think it is a cultural thing.
It blew my mind that Tommy Flowers ended up in debt after building the world's first programmable electronic computer. Plus he even went back to his old job. He did help create the first all-electronic telephone exchange to be used in a public network. So his skills weren't completely wasted.
UK did that with rocket engines as well made them top secret just like computers after WW2 thus stopping UK progress :(
@@favesongslist It seems like the US put a lot of pressure on the UK to not develop all sorts of tech after the war and to instead leave it to the US to develop (supersonic tech being another big one where the US took what the UK had and refused to share in the other direction)
I really like that you always jump straight into the topic
In the mid-1990s I worked for ICL repairing computer equipment (mainly computer monitors). Most of our work was contract repairs for commercial clients and some military organisations. Then ICL / Fujitsu introduced the PCTV (worth a Google, that). It was so badly designed (both electrically and mechanically) it would often fail before it reached the customer because a slight knock would bend the aluminium frame in such a way that the CRT's tube base PCB would collide with the line-output transformer.
Other wonderful faults were that the main tuner PCBs were located above the PSU, which got hot. This regularly cooked the EEPROM that stored user settings such as channel tunings, volume and screen brightness. When the EEPROMs went bad they set everything to maximum, so switching PCTVs on with that fault was fun - you'd be met with a very bright, snowy image with white noise from the speakers at full volume.
You could also format the hard drive from the remote control if you knew your way around WIndows 3.1.
Despite us being the same company, we had one set of schematics, and we weren't allowed to have more sets, we had no help or time to figure out some of the more interesting faults, no access to unit-specific spares and management under-estimated the time to service them, so we lost money on every unit we fixed.
I can't understand why the product never became a success...
Interesting. I was working for Cogar/Singer in the mid-1970s when ICL bought Singer. At the time, I was working with a small team developing software for the Singer 1500 desktop device that was made in Utica, NY. The 1500 line was originally designed for doing "Intelligent Data Entry" and had a max of 16K of semiconductor memory. It was fully programmable and featured a 256 character CRT (8 lines of 32 characters) , a keypunch style keyboard, and two cartridge tape drives. The 1500 could function standalone or up to 16 could be networked together and share a 20 MB disk drive.
The team, of which I was a part, developed a set of data entry programs for use by Chevrolet dealers to enter and submit vehicle order, delivery notices, warranty claims and financial statements. Chevrolet had opened their program to eliminate paper to several vendors - Singer/ICL being one. Communications with GM was via dial-up asynchronous comm via 202 modems.
The product was successful and over 500 were sold to GM dealers.
It failed because of absolutely crap senior management. The same as other big British companies such as British Leyland. It is so depressing that the British have a proud history of innovation, yet we consistently fail because of inept management, risk averse investors and lack of Government support for small businesses. There is so much the UK Government could do to improve things, but it never will because it is only interested in big companies.
STC did not help, as they had even worse management.
I think it is more than that. British MPs which become government ministers are dominated by lawyers, accountants, people with degrees in history, geography PPE - politics, philosophy, economics.
They don't have a clue about technology and engineering.
@@deang5622 Exactly...as per usual...The British are led into commerce (& war) by donkey's....Been that way since 1945.---IMHO.
And there is no sign the situation will improve.
What Mr Thomas Watson Sr of IBM said at the beginning...still holds true...basically: The Brits never get their priorities right.
And, sadly, there is no indication they will ever do so.
British Leyland cars were crap.
ICL were a joke. In the 1970's I worked for a company who tried to put an ICL 100A 5V Switched Mode PSU into production. It had been designed by university students with little or no experience of real electronic engineering and was without doubt the worst designed pile of crap I have ever the misfortune to work on. The design was quietly withdrawn.
Any details? It sounds interesting, what was fucky with it?
Please do share some details!
I worked for ICL Mainframe Division for 8 years. I remember that PSU well. Broke frequently. Only the 2 Power Transistors went. In those days switched mode power supplies
were brand new technology. Not surprising that there were problems. The Power transistors were BJT types (Mosfets did not exist then). So they had to be a matched pair
due to stored charge delays when turning off. The other switched mode psu's in service were made by Gould and Advance if I remember correctly. Still have a few of those
in my possession when they scrapped the mainframes. Including one of the 5V 100A type. It is about 50cm wide, 30 cm deep and 10cm high, more ore less. If that is the
one you are referring to. Today I manufacture switched mode psu's in my own company for use in our in house products. But those days were never forgotten.
RIP ICL. Had good times there, met many smart people. Very glad to have been part of the heydays. Left in 1983 just as Fujitsu took over. Saw the writing on the wall.
@@esoterex what was it like working in the industry at that point in time? Did you expect the world of computing to change so radically to what it is now? Thank you so much for sharing your story!
@@helpme8993 It was fairly demanding work.
Especially as most large systems ran 24/7.
In my case, as I was responsible for a large
dual Mainframe installation with 50 Disc Drives, lots of Mag Tapes, Printers etc. we had do do what we called standby duty after hours. Although the actual Mainframes
were fairly reliable but the other kit needed a lot of work. Had to manage a dozen 'Engineers' which was more of a problem than the actual hardware. But looking back
all those years I enjoyed my work. Did a lot
of advanced troubleshooting when the others got stuck on a problem, devised
numerous modifications which improved
performance. Spent a lot of time at the local
pub with a 2 way radio as cell phones did not exists waiting for a breakdown. So it was not all work and no play.
Today, after running my own electronic design and manufacturing outfit for over 30
years I often think back to the good old times, and yes they were the good old times.
Am I surprised by the incredible rise of computing? Not really. Once the underlying
technologies exist progress is fairly rapid.
Not just in computing. When I look at my Porsche and Helicopter, how they have become so advanced so quickly its because due to the rapid rise of electronics and computing power at low cost. It is just sad that the wet ware has not kept pace.
Disappointed to see you didn't mention, in detail, the Lyons Electronic Office. The LEO Mk.1 was the first stored program computer to be put to work in a private commercial setting in 1951.
It’s a 34 minute video. Had to draw the line somewhere
@@Asianometry another video about that would be good
This is one of many serious errors in this video. I had a lecturer who worked on the Leo, he told us the same thing. I worked as a student at ICL in the early 1990’s in the processor chip design section and in the mainframe OS section, so your characterisation at the end is totally wrong.
As a former owner of a MGA roadster and a BSA motorcycle, I can only assume that ICL went under because of massive oil leaks.
Did you ever hear of the bizarre electrical problems MG and Austin were know for? Those kept many car mechanics very well paid for decades.
I have seen two Nortons broken down at the same intersection at the same time. One stopped to help the other and then couldn't get started himself.
No no... the joke goes: "Why don't the Brits make computers? Because they couldn't find a way to make them leak oil."
I worked for both ICL and Fujitsu (before and after the sale).
One problem was that management were dumbfounded by the transition to distributed computing. When we were making desktop PCs, they were great quality but we were losing money on every one we sold.
Also, people who were using centralised processing started to move from mainframes to minis. IBM already had an offering, but it was a market that ICL never made a dent in.
But the main problem was that management never made the mental transition from being a 'protected' British manufacturer (by the Government) to a genuine free market. They were so used to operating in splendid isolation that, when they were forced to compete in a free market, they had no idea how to do so.
Agreed
Gee, its amazing how many commenters on here worked for ICL.
I thought I was old as I remember DOS and bat files.
I was working in a computer repair store in 1993, and the most common computer brand to need fixing was the ICL PC clones. We had stacks of them, all with different intermittent issues, non of them reported on the report card, so they went back into the our junk room stores unfixed.
The Brits were terrible at making technology if you had to fix that many.
Nah. Sounds more like lack of mantainance / incorrect use by the end customer, which still affects tons of electronics today.
I recall when the Sony PS3 was new, lots of owners where complaining of overheating... when in most cases the console had been set up incorrectly by them, and was not able to draw enough air to cool itself... or worse, was drawing in carpet fibers which swiftly broke the fan and caused the circuit board connections to melt... complete idiocy on the part of the end user, as it said in the manual [to paraphrase] *do not obstruct air flow around the unit, nor place on a carpeted/rugged floor.*
Similar daftness happened in the car sector, with the *Triumph Stag* getting a bad rep' for blown engine head gaskets, which often was actually due to the end user not having read the manual on how often to get the timing belts changed / service garages not being appropriately aware of it.
Product failure is often blamed on the manufacturer in this country [often by a comically underinformed media carping upon matters for which they know nothing, nor bother to], yet later turns out to be someone else not using the thing as directed in the first place.
What a great video. I well remember the IBM 360 from my grad school days (1973 - 76), and also of carrying around those boxes of punch cards for my fortran program. The university was very proud to have that technology. IBM was king of the hill in those days, but today seem to be only a shadow of their former selves. Thank you for the work that was put into this video.
Did you have a hand punch tablet for quick corrections like I did?
Maybe you should look up IBMs cooperation with the Nazis
@@rogerhudson2814 , I had one for punch tape too, actually I still have it as a memory from 1968 forward. My first computer was an Elliot 803, and for Cobol and punch cards a ICL.
@@slcpunk2740 , and the royal family's too perhaps if you want to live in the past.
@@hurri7720 IBM doesn't get to erase their past regardless of your whataboutism
Coincidentally the UK’s General Electric also fell apart. Check out the history of GEC (General Electric Company) for a similar story
Well if it's any consolation both IBM and General Electric are kind of shadows of themselves even if they still exist.
I always thought General Electric was American. TIL
@@monad_tcp General Electric is American, its just there was another (UK) company called General Electric Company
@@monad_tcp The British General Electric Company plc should not be confused with the American General Electric Corporation which still exists today. The American GE was in the mainframe business, but sold it to Honeywell who later sold it to Groupe Bull.
Well large parts of it ultimately became BAe via Marconi
Imagine watching a video where time is lavished on the punched card era but huge swathes of detail are missing from the 80s and 90s - that's exactly what I just experienced. There are so many chapters glossed over, including all the technology innovations and so many partnerships. Hopefully, a good video about ICL will come along soon; it's a company history that deserves to be preserved.
Amen to that, and not a mention of George 3 or VME, both streets ahead of IBM's offerings.
Make the video then. This isn't his expert topic, it's an insight. All of the whining from ex ICL employees betrays the absolute arrogance that brought the company to its knees. Should he have done a 4 hr video? Be respectful. We're supposed to be a nation that prides itself on manners.
@@teebonesteak8015 Speaking of manners... Best not to leave it too late to find out if there's a refund policy on that "Charm and Written Etiquette for Novices" correspondence course. ;-)
Excellent structure and pacing to an excellent topic... as always :)
In addition to being an excellent overview of the UK market, your information about the US, Japanese, and French firms is extremely helpful.
¶
An additional brief note about the many US firms may be helpful. In the early decades the US had IBM, Burroughs, Univac, NCR, CDC, GE, RCA and Honeywell. While these firms did compete, to the outside world they also appeared to be monolithic. This was long before highspeed data networks and de-mountable disk drives. There were two key interchange standards: punch cards and magnetic tape. The US DoD required tape standards from all of its vendors, first 7- and later 9-track. The US already had strong standards -- setting organizations. This helped data tape interchange in DoD (and other government agencies, too) and for industrial, commercial, and (higher) educational users, too.
It is hard to overstate how important these standards groups were _and_ still are today! They include the IRE (now the IEEE), ACM, ASME, SMPTE, AES, and numerous others. The close cooperation amongst these organizations greatly helped US member firms and also impede market entry to non-US companies.
¶
This is a topic which needs a fuller portrayal.
IBM, Burroughs, Univac, NCR, CDC, GE, RCA, and Honeywell were nicknamed Snow White and the Seven Dwarfs. After GE and RCA exited the computer business, the remaining Dwarfs were referred to as the BUNCH (their initials).
hi, this video was in my recommended section today and I just wanted you to know that I'll be binge watching a bunch of your videos from now on. thank you for the work you put into these.
Just subscribed to the newsletter, it’s been too long and I’m thrilled to hear from you!!
You forgot almost completely the world's first commercial computer, LEO - Lyons Electronic Office, produced not by a punched-card manufacturer, but by a company better known for cakes and tea shops.
And it became part of ICL. Who until the introduction of the 2900 series running under VME, for a while ran computers under 2 architectures the ICT 1900 under George and the Leo series 4
That last sentence of the video: IBM's decline into irrelevance? They last posted quarterly sales of 16.7 billion USD and a quarterly net profit of 2.33 billion USD. IBM. Not the undisputed market leader of the old days but still quite relevant I'd say.
Great video. Brought back a few memories of at first dialing into the Oxford council 1904 via an acoustic coupler from our school and taking my first steps programming in BASIC. Then going to Uni and using a punch card system on an IBM 370 before it was upgraded to a 4341 programmed in Pascal and COBOL. Then worked for ICL and worked on System 10 and 25 and also 2900 series with GEORGE and VME. Also worked on DME on ME29s. I stayed with them and bought my own PCs from them including an M30, a low end first venture into desktop personal computers for ICL. Then I got an NB386 - ICLs first notebook/laptop. I stayed with "the company" until 2019, though they morphed several times over the years such as acquiring CFM as an entry into the outsourcing business. I also still have the vinyl album called A New Way Of Seeing with music written for the launch of the 7500s and ME29 (cat.no. ICL001).
IBM was a beast in this period..the S/360 ecosystem was a brilliant concept
S/360 was, in the opinion of many, primitive and crude, but it was marketed very well. I've developed systems on Burroughs (from L/TC to B7700), IBM from (360 to 3090), Univac, Digital (VAX).
Probably your best one yet. I absolutely loved it. Not only did you give a very detailed historical review of the history of the different companies. You also showed hints of where these machines were used for during the different era's of the time. For it is really hard to imagine what these machines were used for in the early 20th Century. I absolutely loved this video. Now I'm still searching for a good video of the evolution of pattern weaving machines to punchcard machines. And I'm still a bit at loss on how the products of IBM machines were used by small and large businesses on a daily basis.
The IBM 30XX series (303X, 308X, 3090) mainframes were targeted to large corporations and enterprises. the 43XX series mainframes were targeted to medium size corporations. These families of mainframes were advanced to the System/390, then the eServer zSeries (900, 800; 990, 890), then the System z9, then System z10, then zEnterprise System (z196, zEC12), and now the IBM Z (z13, z14, z15). Programs written for the instruction set of the System/360 will still work on today's z15.
Meanwhile, IBM also created a successful line of midrange systems (then called minicomputers) that were targeted to small businesses that do not have their own IT departments. These midrange systems were turnkey systems with no need for programming personnel and were incompatible with the mainframe because of the different instruction set. The IBM System/3 was the first of IBM's midrange systems released in 1969 and used punch cards that were 96 columns across (instead of 80 for the mainframe punch cards) and the holes were circular instead of rectangular, but the cards were smaller than their 80-column counterpart. The System/3 advanced to the System/32, then System/34, then System/38, then System/36, then AS/400, then today as IBM Power Systems running IBM System i. The mainframes are System z. The RISC-based computers were System p which later converged with System i.
IBM's shrinking personal computer business and Intel-based server business were sold to China-based company Lenovo.
@@RaymondHng thank you for that extensive overview.
@@RaymondHng thanks 😊 this extensive overview. I find it amazing that these old programs were able to run on much newer hardware.
This is full on documentary quality condensed into a short you.t video, great job man, well impressed on the quality. Wasnt even interested at first till you grabbed my attention : )
I worked for STC during the early 80s when it took-over ICL. At the time STC were heavily invested in their Honeywell mainframe and mini-computers (all before PCs appeared on the scene). It was pointed out that it was bad-form for a computer company to have a machine from their competitors, so STC bought some small ICL mainframes and we all went on various training courses how to use them. It was a bit of a disaster, and the ICL system ran slower than our H'well ...one week the weekly payroll system didn't run quick enough and all the payment deadlines were missed! There were also a large number of systems which would require a complete re-design and re-write to make them work on the ICL. The ICL system was pushed into a corner; it still ran, but never did anything important of time critical.
It took ICL 2 years to buy their way out of STC. Honeywell machines were IBM clones. ICL had an operating system light years ahead of IBM's . IBM killed themselves with their PS2 and the toxic relationship with Microsoft and early Windows. Honeywell was just another cashed-up US defence company with no real skin in the game. ICL were victims of the Thatcher govt and too well-entrenched management who would not alter their thinking to modernise. STC went belly up in 1991.
@@fordyootbling2189 Thatcher's no lame duck policies destroyed so much UK strategic industry which was also infested with complacent, ossified, old school tie management. Contrast that with the huge federal subsidies for R&D in tech institutions, which was handed to American corporations with large tax breaks. In the last few decades the laissez faire, "free trade" mantra has damaged R&D for a quick buck mentality while countries like China did the same as the US formerly did. It is staggering that Western governments never wake up to the fact that free trade is an illusion because the countries whose economies have done the best do not practice free trade at all. Add to that the massive technological and commercial IP theft and illicit copying by China and you have a strategic error of epic proportions.
Wow, this brings back memories! 3 shift patterns operating an IBM 370/135 mainframe in 1977 followed by a 370/138 and a 4341 before transferring to Cobol and PL1 programming/systems analysis in 1980. I can even remember the finger blistering disc drive numbers (3330/3340), tape drive numbers (3420) and chucking frisbee tape case lids on night shift. How sad. Chewing write protect rings was bad for your teeth and dropping printer chain assemblies got you one hell of a bollocking...
Well out of it all by 1985 thank goodness.
Reminds me of another shotgun marriage that also failed. The forced marriage of British Motor Holdings and Leyland Motors into British Leyland.
And the perpetrator of both: HM Government
Excellent content! Hope you continue make more stuff like these
I started on ICL-1904 in school in 1977 - BASIC & FORTRAN on punch paper tape. Then ICL-ME29 mini in Local Government and 2966 VME/B in the civil service. then we wne ICL-VME Series 39 which was the BEST corporate server - f/c connected peripherals, virtual machine environment, easy DR and sophisticated volume / partition mirroring and the catalogue for machine management (bit like device manager). superb. I then went all-windows in 2000/2003
I have a colleague who frequently stated that his two biggest career fears were 'British management' and 'visiting American expert'.
My dad was a data processing engineer with ICL, starting at Powers-Samas after leaving the RAF in 1955. He often talks about going to faulty tape machines, printers and punched card machines. All to often a thump on the side of a cabinet would get them working again. The company knew what the faults were but, time and again wouldn't fix the problem permanently.
What a great informative video, thank you. In the punch card and paper tape days I worked with hardware from Burroughs Machines.. I changed direction when microcomputers took off. I remember someone saying "I can't see the day when everyone will have their own computer on their desk" Now everyone carries a computer with previously unheard of power in their pockets!
Just found your channel. Excellent essay, lots of stuff I didn’t know. Subscribed!
In the very late 1980s I remember going to the government funded site in Southampton to work on a small micro/mini computer my company had sold them (MIPs based) and to my suprise being ushered into a computer room which had our little computer in the corner and the rest of this huge room taken up with endless washing machine sized disk drives and rows of RTR tape drives. Even then, it felt a little bit like having stepped back in time by a fair few years. In conversation one of their IT guys told me that our little MIPS R2000 system probably had about as much processing power as everything else in the room put together.
I started working for ICL in 1976. The machines were not the best but I did like their operating systems. I worked on all their ranges as well as PDP 11/34, 11/45 then also IbBM 360 and 370. I could go on , but there is too much to tell.
One more interesting point I think the king of the vacuum tube computers was I believe called SAGE which ran US air defense. I knew someone who worked on the hardware and he said the first troubleshooting step was to turn out all the lights and look for dark vacuum tubes (burned out filaments). The term computer repair people used for that was Easter Egging (looking for telltale symptoms) and I fixed a few myself that way. The first major multimillion dollar mainframe computer I fixed that way by seeing a small smoke trail coming off a set of cards and followed it to the offending logic card. Lots of training backed up by a keen sense of smell and sight!! haha
Watching this on the day of the Lunar New Year's eve, I can sense what your trying to picture. Thanks Jon and enjoy the New Year holidays!
My mom retired like 10 years ago, but she was in some form of computing industry for her whole career. She was working for ICL's local software service company ICL Invia in Finland when that Fujitsu merger happened, but as far as I can remember it took a while for the name change to happen. Apparently they had 2300 people on the payroll in Finland and the Nordics at the time (2002). I think she ended up at ICL from a sale of the computing department of local milk producer Valio. Nice video, keep up the great work!
Fishing, hunting and shooting. You forgot drinking which comes first
That's why other name for Britons is Inglish.
By the way,... I reckon ... f...ing is the first 😏.
I worked in the industry back in the 80s and 90s, I clouding with some ICL folks. I think another factor was the gap between ICL's corporate approach and the sheer bloodthirsty competitiveness of the hundreds or thousands of startups that were arising in the beginnings of Silicon Valley. IBM's own rather bloodthirsty sales machine kept them alive longer than any other big player but even they eventually fell away as the democratization that microprocessors brought to the industry allowed fast, low overhead, antibureaucratic upstarts who (for example) could sell an entire machine for less than the monthly rent on an IBM mainframe, priced below the max budget of a department head, and eliminated the wait for the IT department to schedule your job.
Essentially all the mainframe manufacturers fell afoul of a new expression of the economies of scale.
Great video. In the early 1970s I worked at GEC who did their computing on an ICT mainframe. In the mid 1970s, I programmed (in COBOL and Fortran) a Univac 9700, which at that time was the most powerful computer in the UK. It had 256k of memory and used punched cards and tapes for storage. Univac was an early American competitor to IBM.
I appreciate the straightforward nature of your videos. Cheers asianometry
I'd also add, the final incarnation of S39 I worked on was the Trimetra - able to run VME, Unixware & Win-NT all on the same intel hardware - it was partitioned. After Fujitsu took over ICL, we continued to work with the new company and eventually, ended up with a DC made up of fujitsu blades, physical & virtual network switches, f/c running VMWare - which basically was the Trimetra in modern form. ICL as a company might have 'died' but the ethos - especially in the quality of the Fujitsu intel servers and services - lived on.
"The United Kingdom pioneered computing technology. So why did ICL fail?" Britain pioneered just about everything and then it eventually failed, not always at the same time or rate or for the exact same reasons, but lack of innovation, short sighted management, underestimating rivals, poor labour relations [heavily caused by poor management but not only them], and the economic dominance of the financial sector over manufacturing, eventually.
Nothing changes, Look at ARM sold to Softbank or Deep Mind sold to the USA for just £400m; politicians said that was a good deal for the UK :(
No Britain did not pioneer almost everything that is just an inbuilt tomfoolery with the British, and it's deep and old and mostly total rubbish.
@@hurri7720 Ooh...GB welcomed in talented immigrants like Marconi (and Marx), offered a stable society and in Victorian times a population taught to read properly unlike today, it acquired territories which spearheaded advances, like the first mechanical clock to work on board a ship. We should teach this in schools..Britain would be first in many sectors - until other nations also stabilised and caught up, GB was the best overall.
@@hurri7720 I remember an American bighead asking what Britain had developed, he challenged me to name three. I quickly said, the jet engine, stainless steel, first nuclear power station. He went quiet after that.
@@francishunt562 ,
I wonder if you are joking or serious.
The first patent for a jet engine was by a French guy, but it was never built.
The first jet engines in the air were German about 2 years ahead of the British.
The mother of all jet engines and the first axial flow engine is the Jumo004.
World War Two JET POWER
czcams.com/video/4N5lNOYlOb8/video.html
Junkers Jumo 004
en.wikipedia.org/wiki/Junkers_Jumo_004
Jet engine
en.wikipedia.org/wiki/Jet_engine
The two first nuclear powers stations built producing electricity where American and Russian.
en.wikipedia.org/wiki/Nuclear_power
Electricity was generated for the first time by a nuclear reactor on December 20, 1951, at the EBR-I experimental station near Arco, Idaho, which initially produced about 100 kW.
On June 27, 1954, the Obninsk Nuclear Power Plant in the USSR became the world's first nuclear power plant to generate electricity for a power grid, producing around 5 megawatts of electric power.
What the English have invented is the "silly" word to insert in a sentence to enable the use of the word "the first".
So now we find this designed to fool the British who are very willing to be fooled.
The world's first commercial nuclear power station, Calder Hall at Windscale, England was connected to the national power grid on 27 August 1956.
The amazing thing about this is that the British never seem to grasp their fraud and never question anything.
A very popular English claim is that the steam engine is an English invention, which it is not but again we find the use of the silly word - commercial there too.
A few more, The oldest (silly word "sitting") parliament.
The oldest (silly word "independent") air force.
I could go on and it's absolutely amazing how widespread and how deep it is in the British.
One of my favorites from last year was a women from the BBC who on the Telly pulled this - "the British parliamentary system is the envy of the whole world."
Amazing.
Regarding stainless steel:
The invention of stainless steel followed a series of scientific developments, starting in 1798 when chromium was first shown to the French Academy by Louis Vauquelin. In the early 1800s, British scientists James Stoddart, Michael Faraday, and Robert Mallet observed the resistance of chromium-iron alloys ("chromium steels") to oxidizing agents. Robert Bunsen discovered chromium's resistance to strong acids. The corrosion resistance of iron-chromium alloys may have been first recognized in 1821 by Pierre Berthier, who noted their resistance against attack by some acids and suggested their use in cutlery.[15]
In the 1840s, both of Britain's Sheffield steelmakers and then Krupp of Germany were producing chromium steel with the latter employing it for cannons in the 1850s.
en.wikipedia.org
/wiki/Stainless_steel
Fujitsu and IBM still competing in making fastest supercomputer. Your channel is really interesting, full of information and original compared to many other tech youtuber that only make trendy video with fancy editing. I want to request US-Japan 1980's tech war and capacitor plague in early 2000's.
Thanks very much, this was informative. I love your historical progressions
I recall, back in the 1970s, one Hollerith card (80 chars) can hold the whole student record.
Nakata ask you Jonnie Walker to stop taking all the cats.
High school class registration was done with punch cards in my school days.
2:51 There is a phrase from this era, when companies would routinely send you these cards for you to send back to them: “do not fold, spindle or mutilate”. It was a reference to the fact that your very identity was tied up in these coded cards.
What a way to protest by putting a big fat crease in the punch card and jam up the company's card readers. My teacher once punched an extra hole in his punch card from AT&T and turned the debit into a credit on the card and returned it with his check payment to AT&T. He got a letter from AT&T sternly telling him not to do it again.
This was a phenomenal effort. Really excellent.
It's always been fun seeing pictures of a tape drive labeled "a computer".
I worked at the ICL software house, ICL Dataskil, for 2 years, 1977 to 1979, and then left to go freelance. Rather a lot glossed over in this video. As I recall, the Thatcher government came to power in June 1979. They deliberately created a scenario of high interest rates, that brought in a huge amount of 'hot' foreign money. That raised the value of the £ to a ridiculous level, which crippled exports for many companies. As I recall, a year or so previously, ICL had received the Queens Award to Industry for its exports. On a wider scale in the UK, government policy destroyed 10% of manufacturing companies in the space of 18 months.
My first contract, in early 1980 took me to Dublin, where I and couple of other contractors got a new mini-computer system, an ICL 2904, up and running. My next, in autumn 1980, took me to the London suburbs to a software house which had a contract to translate a German language software to a workable version for their client, ICL. A few months in, the word went round that ICL was going bust. I ended up, round the turn of 1981, working on the basis of one week's notice at a time. I couldn't run my life that way, and started asking round. Within a few weeks, I was in Salisbury, Wilts., on contract with a life insurance company. Pay was great, the work mind numbingly boring. I stayed there for almost 2 years, until I could get out, and find contracts that used my specialist knowledge of manufacturing systems
Basically, Thatcher and Co were completely uninterested in UK manufacturing. The only thing they were interested in was financial services. In the end though, the advent of the IBM PC, and the PC clones in a relatively few short years consigned the mainframe , and mini computer manufacturers to history. Not just ICL, but IBM ( as a manufacturer ), and the grouping known as the BUNCH..., the nickname for the group of mainframe computer competitors of IBM in the 1970s. The name is derived from the names of the five companies: Burroughs, UNIVAC, NCR, Control Data Corporation (CDC), and Honeywell.
As a final comment, on something that the video creator here briefly mentions.. Japan. For decades after WW2, Japan was in effect run by and for the benefit of a few huge industrial companies. They and the government were, in effect, one and the same. In the US, major computer companies were greatly aided by Department of Defence monies, what used to be called the military/industrial complex. In the UK, however, the Ministry of Defence were always somewhat reluctant customers of ICL, though public services, eg The GPO ( General Post Office ) were large customers. An illustration of the Thatcher government's attitude, when in 1980/81, ICL struggled, and approached the government for help, I understand that Thatcher's initial response was something similar to ... "F*ck them...". At which point, someone pointed out that most UK councils used ICL computers, and if ICL went bust, there would be chaos in local public services.
" in June 1979. They deliberately created a scenario of high interest rates, that brought in a huge amount of 'hot' foreign money." Nope. All over the world in 1979 the economy was in rough shape, the rate of annual inflation was 11.3% and US mortgage rates were at an all-time high rate of ~12.5%, companies everywhere were in poor financial shape. I'm defiantly not a fan of Thatcher for many reasons but she had little to do with world economics.
@@davidnull5590 What I’ve stated is correct. She chose the policy of Friedman economics.
@@andrewpreston4127 --- Do you remember Friedman believed that government intervention did harm, and that the best economy would be a liberal free-market economy? Thatcher had her own brand of misguided, she believed in trying to target the money supply to reduce inflation. It involved: Higher interest rates. Higher taxes and spending cuts. I was the recipient of economic policy from the US and world economic policies in 1979, inflation worldwide was very high and business interest rates for my business in the US were crippling.
Rolls Royce at the time had a new type of engine on the drawing boards which was far more fuel efficient than any jet engine on the market, but they had put so much R & D in it they had gone bust before it was ready, so you can understand the rationale. These same engines are still in production today and are world leaders. So it was a case of punt on the technology, not the "all companies are equal" view. There were a lot of protests as well. The government was pressured.
The UK government was pressured to save Rolls Royce because they were costing Lockheed a lot of money by failing to deliver the promised engines for the L-1011. In the end Lockheed chose to never again participate in the civil aviation market.
Rolls Royce should have been bailed out before the engine development slowed.
My father worked for Westinghouse in the US in the defense industry. They designed a jet engine for the government to government specs. This, as it turned out, required using a lot of titanium, and the cost overruns related to figuring out how to properly tool that was the main problem. In the end, the government took all the technical expertise learned and told Westinghouse to forget about the incurred costs and so gave Westinghouse the shaft. Westinghouse got out of the Defense Industry never to return. It was, after all, the coldest part of the cold war. My father made the shift to doing missile guidance systems for GE and had a long and stable career with that. If the planet blows itself all to hell with his missile guidance systems, he might still save himself for history as he is on the gold disk in the Voyagers.
@@thomasjamison2050 I think we have something in common. My father designed the aerial for the beacon so the first moon landing module could be recovered once it splashed down.
@@Andrew-rc3vh It would seem to be the case. My father only had a few things to do with the space moon programs. He did work on the Mercury heat shield, and he handled a contract for a tile patch kit for the shuttle. They did have one with them when they went down, but as you well know, once the stuff under the tiles starts to burn, it's too late to do any patching. And neither can one do a space walk during launch. Pleasure to say hi!
Curious Droid did an excellent video on the Rolls Royce RB211 engine.
It would be awesome to see a History of EPSON video too
No, no, no! Not 'Bomber aircraft frames'! BTM was the main manufacturing partner for Bletchley Park codebreaking machines. They made BOMBES (and possibly the 'frames' for them). Bombes were the machines that cracked the Enigma machines' rotor starting positions allowing the day's Enigma traffic (for that particular German cryptosystem) to be read. That huge cube machine you show - but don't comment on - is an actual Bombe (one of the early ones). BTM were much less involved in the Colossus computer which attacked the 'Tunny Code' - actually the super-secret Lorenz S-40 cipher machine for top level Nazi comms. Colossus used VACUUM TUBES not the mechanical switches which BTM excelled at. BTM might have made precision sub assemblies for wartime aircraft gun and bomb sights but they had ZERO involvement in aircraft construction in WW2.
I couldn’t actually confirm the extent of their involvement with that. Do you have a source?
@@Asianometry Wikipedia: en.m.wikipedia.org/wiki/Bombe
U mad bro?
@@Asianometry I was at Bletchley Park last week. He's right.
Colossus was constructed at Post Office research property in Dollis Hill, London. Project led by Tommy Flowers.
Mainframe dinosaur is still going very strong in banking, insurance and other finance industries, COBOL is still running hundred of millions of lines of code......
26:10 I was wondering when you were going to mention minicomputers. But I was thinking of the original minicomputers, mainly from the 1960s through to the 1980s, offered by companies like DEC, DG, HP, Perkin-Elmer, Prime and others. They showed that small computers, while less powerful, could be a lot more versatile than big ones.
Minicomputers deserve a separate video of their own. IBM also created a successful line of midrange systems (then called minicomputers) that were targeted to small businesses that do not have their own IT departments. These midrange systems were turnkey systems with no need for programming personnel and were incompatible with the mainframe because of the different instruction set. The IBM System/3 was the first of IBM's midrange systems released in 1969 and used punch cards that were 96 columns across (instead of 80 for the mainframe punch cards) and the holes were circular instead of rectangular, but the cards were smaller than their 80-column counterpart. The System/3 advanced to the System/32, then System/34, then System/38, then System/36, then AS/400, then today as IBM Power Systems running IBM System i. The mainframes are System z. The RISC-based computers were System p which later converged with System i.
Olivetti was actually the first producing a what we call a personal computer now.
Yes, what he calls minicomputers are actually called microcomputers, and the real minicomputers were the ones you mention.
Please also focus on other industries apart from semi conductors. Your analysis is very interesting to hear. A lot of countries, particularly Asia and Africa have tried their luck with several other manufacturing industry and while some have emerged successful, some have not. Documenting these will be useful for sectorial analysis of industries in different countries. Keen to watch a wider variety of topics from your channel !!
I joined British Coal as an mainframe operator in 1974. Due to being a nationalised industry there was pressure to have ICL. What actually happened is that of the 6 UK data centres 3 were mainly ICL & the other 3 were mainly IBM. That meant that all our software had to be coded for both IBM & ICL and then additional programs written to allow data to be transferred & shared.
I started with 1900 series before move to 2900 series and finally an ME29. At that point, with Thatcher shrinking the coal industry, we went all IBM with 4300 series and series 370 for the main centres.
I can still remember the vast rooms full of metal cabinets that contained used punch cards by the million.
that's a great logo for English electric
26:10 "the mini-computer boom of the 90s"
Wouldn't that rather be the 70s?
This video, and many of the comments, brought back so many memories. I have had a few contacts with ICL during my adult life including working for them at Stevenage early in my career (beginning of 1983 to early 1984). I also studied comp sci at Manchester where studying the MU5 architecture was a core part of the course and that architecture heavily influenced the 2900 series architecture. I was also one of a 7 (I think) person joint AT&T/ICL team who, in 1988, went out to AT&T Bell labs in New Jersey to port Unix SVR4 onto SPARC for ICL's upcoming SVR4 based systems although that initial port was actually done on Sun workstations with another team of Sun staff also doing porting work on the west coast.
My earliest encounter with ICL however was during the 1977 summer holidays just after taking my A levels and before going to university when I got a summer job as a computer operator at a fairly big ICL installation. (I now realise I can't remember if it was a System 4 or a 1900 series.) What I remember most was the ICL support engineers who were supposed to be in the on-site engineers' office at all times (it was a 3 shift 24 hour operation). During the evening shift the on-call engineer used to come in on time at 4pm but at about 6pm when all the regular office-hours staff had left he would go out again leaving us with a list of telephone numbers for all of the local pubs that he was going to be at that evening. If there was a hardware fault we had to start working our way through the list to try and track him down and if the problem occurred in the latter part of the evening shift we were pretty much doomed because once we had tracked him down and he got back to site he would invariably be so drunk as to be pretty much incapable of fixing anything!
It was a very different world then but if that sort of service and support was replicated at other sites then I am not at all surprised that ICL lost customers to other suppliers. (That site that I was working at started transitioning to Honeywell a year later).
Do us a favour, get real. You think ICL employed known alcoholics.
Nice presentation! My college used a 4341 in the 1980s, and we all wrote assembler, cobol, fortran, PL1, Basic, RPG II, and more for our class projects. I later went to work for a company that, after a year or so, consolidated a 370 and a CDC machine to a single IBM 4341, just like the one we had in college. The 4341 did not look anything like the 4381 at 29:41 in this video, but instead was a low, wide freezer-chest arrangement, upon which we would put the main consoles. "Kaboomed the market" is right. The processor itself could be leased for a few thousand a month or bought outright for under $300K. That's about $830K in today's dollars. But by 1985, you could buy a 4341 for about $50,000 or lease it for a small fraction of that. For small companies with fewer than 500 users, this was doable. The 4341, mated to fast disk (Direct Access Storage Devices were called "DASD" in the old days, and pronounced "DAZDEE") could serve hundreds of people concurrently, could run manufacturing lines, could support CAD design and manufacturing long before people could put Bridgeports, 3D printers, or CNC machines in their workshops, inventory control, and component replenishment on factory assembly lines. It could run your finances and print your company's payroll checks; complete with mailing address on the envelope exteriors! Oh, and you could use it dirt cheap for 10 years before your coding boffins in the basement would be able to write super complex business systems that would require a processor upgrade, and sometimes you could just slap more CPU modules into it without buying a whole new box! And by 10-15 years later, you could go to a 98xx model and keep that for ANOTHER 10 years. It wasn't until the late 1990s that operating systems such as Windows were beginning to employ techniques that had been in use by the mainframe OSs since the 1970s or earlier. Things like using virtual storage, multi-processing, the use of multiple CPUs, large memory management, dedicated I/O buses, and stuff like that. The IBM mainframe was the woodshedding of the modern computer era.
And I heard that IBM rejected the use of the screen and keyboard at some stage.
Big companies are sometimes just too big.
It's a little more complicated. The IBM SNA communication architecture introduced in 1971, is Block Mode, it waits for the user on a terminal to press the "enter" key before the data is sent as one packet. Most non-IBM terminals from the 1970s and 1980s send one character at a time, like TCP/IP. IBM's communication strategy and their hardware were based on block mode. Parts of IBM wanted to switch (my group there) to TCP/IP. It was a battle inside IBM, it came down to the profits of the existing communications products being too important, this was decided at the company's highest level. This decision left a huge opportunity for other companies, enter Cisco and others - they made massive profits.
I was the sysadmin of a Fujitsu system running DRS-NX back in the 90's and I recall the documentation having ICL badges on it. We replaced it with a Sun machine running Solaris that was considerably more powerful and I've seldom seen the Fujitsu systems again. I think this whole business of governments trying to make national champion companies is fraught with danger and very often doesn't work. Had the UK govt stayed out of it the situation may have been different.
They never learn. The UK government is doing it again, with satellite based internet this time. Another billion pound hole to fill and then abandon.
A great and interesting video history. I enjoyed an ICL 2960 at The University of Kent. Did the final years of my computing degree with it, writing an Editor as a group project, that was used on the system for a while. Also operated the system over the University holidays. A great memory and experience.
Matt Lee BSc hons.
I remember working at ICLs headquarters in West Gorton Manchester in the late 80s early 90s, I was a printer service engineer working for the Dutch company Oce at the time, I have very fond memories of the place.
I did some Oce chip designs
Hi. I totally love your videos. I've watched every one of them. They're very well written and interesting. I hope you don't mind me pointing out a mispronounced word you keep saying. It's the word niche. You say "nitch" when it's pronounced "neesh". Just saying because I speak a foreign language and when learning it no one ever pointed out when I said things incorrectly when I wish they had. Other than that keep up the good work!
History of automobile industry in Japan can be a long video. Can you do it? If you want you can. I got the idea from merger of Japanese automakers in 1960s as you mentioned. Hope you try. While ICL is dead, Fujitsu still chugs along with sizeable IT business, just like IBM today.
Might be a bit off topic but the Aussie government forced a bit of product sharing between Toyota and Holden resulting in the Toyota Lexcen, a rebadged Holden commodore and the Holden Apollo and Nova, rebadged Toyota Camry and Corolla… the idea was to get the Aussie automotive industry to be more globally competitive by reducing tariffs… it lasted from 89 till 96 and replaced an earlier model sharing agreement between Holden and Nissan that ended in 84.
i was an IT manager for a company using ICL system 10 and then system 25 computers. i found this video very interesting and certainly brought back loads of memories.
Those two machines were a best kept secret
Thank you so much. This was really informative and fascinating!
I worked for ICL/Fujitsu on the Post Office Counters (Horizon) project, hundreds of honest lives ruined by anti union government staff and incompetent ICL Managers, and complicit staff hiding the problems with the software and hardware and then lying for years as honest Post Office staff were convicted of non existent theft. No disrespect intended or implied as regards the pioneers and earlier employees. Wrong people in jail.
Please, the history of NORSK DATA.
Thanks
Thanks a lot - this is the first time I really had the comprehensive story of ICL's rise and fall. One day, maybe you could take a look at Univac's similar rise and demise?
The first and last company I worked for had 3 ICL 1900s when I joined them straight from school. They were in the process of migrating to an IBM 3032 but they still had to get a fourth 1900 before they eventually were decommissioned. Some of my first work was coding programs on the IBM to replace ICL programs written in Plan. A tool called Protean was used to convert ICL COBOL source code for the IBM.
Good video. In 1949, the BTM decision to go alone probably had more to do with the UK's balance of payments crisis and lack of being able to pay in USD.
Very educational, I didn’t realize punch cards were so old
They go back to the 1700’s.
Yep. This video doesn’t even mention that punch cards had been in use for a long time before ever being applied to tabulation and computing.
1979 80 I installed many air-conditioning systems to some of these machines in Sydney Australia.
Specialised work in the industry at the time...
Interesting work and payed well.
I do remember (as others have noted the punch card (heck-even the 'interpreter' machine that created the punch cards). I worked as a computer operator in the mid 80's - to early 90's.
I do remember reading many magazine articles that were touting 'this and that' about what could be the next big thing.
This look back was fascinating. It still never ceases to amaze me how companies that seemingly have a monopoly or near monopoly in their respective industries can see their 'fortunes' fall and in some cases fall very quickly because of being wedded to the idea that they 'own' that industry. The computer industry is just full of these stories. Just as in other industries....
Many thanks for this historical tour de force!!
A very good little documentary. It really does show the pit falls that do happen, when politicians, interfere and mismanage an industry.
The UK politicians have done the same to our AI research, could not believe they thought that the selling of Deep Mind to the USA for £400m was a good deal for the UK :( Same with ARM.
I'm a retired IBM mainframe programmer. Floating around the internet, I apparently have resumes of my experiences. I haven't worked for 20 years. I'm getting daily job offers. As the amount of data in the world multiples, I wonder how non mainframes handle the data explosion. I wonder what Google and Amazon use. At the end of my career, companies had me run their business on their mainframe, while their newly hired college graduate programmers, installed dozens of small computers to take over the work load. It got very complicated. Every application had it's own computer and software package.. I retired from IBM.
I started my IT career in 1976 as an operator of Unigate's twin ICL 1904a's, at it's computer centre at Bellefield house in Trowbridge, Wilts. I had the dubious privileged of closing down the final ICL machine in 1980, as the conversion to a Sperry Univac 1100/12 was complete by then.
27:38 The LM324 was an analog quad op amp released in 1972 and they produced it for decades. You can still get a new LM324 today. It's probably not a good representation of the "tightening cycles of computer semi-conductor production". Interesting video overall apart from a few minor issues.
Also the LM324 is still being used as a standard component in new designs 50 years later!
29:10 Some semiconductor devices (with big heatsinks) from the late 80s, beginning 90s are shown: not a good representation either.
@@AndyFletcherX31 I've got a tube of them (and another of LM358s) sitting a couple of feet away from me now :)
No mention of the 2900 series and the VME operating system which were highly successful in the 1970s. Back in 1980, ICL had developed IC Design Automation (DA) software and methodologies that were at least 8-10 years ahead of any commercial vendor. This meant that they were quick to embrace increasing levels of silicon integration. From an engineering perspective, ICL was a leader. It certainly gave my career a fantastic start. I'll leave it to others to comment on the true nature of its decline.
Having worked at ICL in the 70's I must agree, The development of 2900 based on the Manchester MU5 machine was a very bold step to try and bind the conglomerate together and focus on a single product line. Not only was the DA side extremely advanced, but the whole concepts of VME itself, being written in the S3 language (an in-house developed high-level procedural language based on Algol 68). There was a lot of innovation going on but the company was not always the best to exploit it in commercially successful products, e.g. in-house chip plant in Manchester, project "Little" single board, networked desktop computer prototypes built before ethernet was a standard, the Distributed Array Processor etc. etc
I learned COBOL on an ICL mainframe in 2000, I remember going on IDMS training to an institute in a gorgeous building in Windsor.
The punched card shown at 2:22 has round holes, which probably means it is a Univac card, not IBM, although they were the same size, which happens to be the same size as pre-1934 dollar bills because Herman Hollerith re-used money sorting machines for his punched card equipment. The pattern makes clear it is a sample card. IBM cards had 80 columns of 12 rows of rectangular holes. Univac had 45 columns of 12 rows of round holes, but called it a 90 column card because they used 6 holes per character, not the 12 IBM used. The price paid for those extra 10 characters was flimsy cards more prone to jamming.
One problem holding back British electronic computing was the government's extreme secrecy over Enigma decoding; they sold those machines to other countries and wanted to read the traffic. They forbade Turing and everybody else from using anything related to Colossus to keep it out of the public eye. I think I remember reading some estimate of holding back the British computer industry by 5 years, an eternity then which gave IBM time to wake up and take the lead. Whether that would have overcome the lackluster British sales efforts is debatable.
It's hard now to imagine the shock of mini-computers on the main frame market, and then the PC revolution. I liken it to the change from slide rules to calculators, or the later change from telephones and bulletin board forums to WiFi and the Weird Wobbly Web; by the time everyone realized it had happened, it was hard to even remember the old technology.
The takeover by STC happened as ICL was cash strapped because of the development of the 3900 series. When the profits from this new range rolled in, they were used by STC to support its core business instead of being reinvested in R& D for ICL.