back to article How Sinclair's QL computer outshined Apple's Macintosh against all odds

Two weeks before Apple launched the Macintosh, Sir Clive Sinclair launched his unprecedentedly powerful yet affordable Motorola-powered SOHO computer – starting a line of hardware and software that, remarkably, is still going. The QL remains a much-misunderstood computer. For its time, it was just as radical as the closely …

  1. Martin hepworth


    Lets not forget the heat issues and he numbers of us that replaced the heat sink and dropped in a larger transisor

    1. jpennycook

      Re: Overheating

      or replaced the underspec-ed voltage regulator!

  2. Big_Boomer Silver badge


    I remember my Dad's QL with equal doses of love and hate. I loved the fact that it was so fast and allowed me to do so much more than my much upgraded 64Kb ZX81, but I hated those infernal pieces of sh!t,... the Microdrives. After I left home Dad replaced them with external 3½" floppy drives which worked much better and could store 360Kb each. We both migrated to X86 machines in the late 80s and never looked back, but the QL was a very capable machine in it's day and 18 months earlier than the Atari/Amiga that killed the QL off.

    1. Doctor Syntax Silver badge

      Re: Love/Hate

      "those infernal pieces of sh!t,... the Microdrives"

      Fitting something so far from wat was already a standard storage technology was one example of what I think of as the 90% approach ot the UK PC vendors at the time. It's not a variation on the 80/20 rule, it's that they included 90% of what was needed to make the kit worth buying.

      Why? Was it really cheaper to design and build the drive and its media rather than just buy in something that already existed?

      1. juice

        Re: Love/Hate

        > Why? Was it really cheaper to design and build the drive and its media rather than just buy in something that already existed?

        Yep. To get up and running with the original ZX Spectrum Microdrive, you had to fork out £79.95 for the interface 1 unit and a microdrive; you could then chain additional microdrives on at £49.95 apiece. And the microdrive cassettes initially cost £4.95 apiece.

        Conversely, looking at Your Computer from July 1983 (the month that the Microdrive launched), an Atari or C64 disk drive cost around £300.

        And a quick check of the BoE's inflation calculator indicates that the microdrive cost the modern-day equivalent of £260, whereas a disk drive alone would set you back £960.

        Admittedly, even at launch, floppy disks were both cheaper at £2.50 apiece and offered double the storage. But this was an era where technology was so expensive that people were willing to compromise on both features and reliability.

        ... or at least, that's what Sir Clive presumably thought, based on his experience in the 70s when it came to calculators and watches [*].

        However, Moores Law was starting to kick in for the home-computer industry - aided by a bitter price war between Commodore and Atari - and once prices started to drop on standardised floppy disk units and media, Sinclair's "quirky" technologies simply didn't have the capacity, reliability or economies of scale to compete.

        [*] even if the failure rate on his watches did pretty much bankrupt Sinclair Radionics, forcing Sir Clive to spin off a new company which would then become Sinclair Research...

        1. Jan 0 Silver badge

          Re: Love/Hate

          Were looped tape drives really new? I think I remember them featuring either in an article, or adverts, in Scientific American long before the QL arrived.

          1. steelpillow Silver badge

            Re: Love/Hate

            They were not new. This was the era of CB radio, and along with it in any self-respecting Ford Cortina went the 8-track cartridge tape player and massive speakers.

            The microdrive was a tad smaller, is all. About the size of a large postage stamp. ISTR it was just cheap cassette tape slit down the middle. All done with the usual Sinclair cavalier attitude to cheapskate design and build quality (knowing them they bought cassette tape and slit it themselves). Evil little beggars, no two ways about it (the microdrives that is, not the staffers!). Rumour had it that the later OPD ones were adequately engineered, but I never met anybody who had one. Later on the world introduced genuine mini two-spool cassettes for mass storage, using rather more expensive tape of the same width. By then I was incapable of trusting them.

            1. drgeoff

              Re: Love/Hate

              It was videotape, not audio cassette tape.

              1. anthonyhegedus Silver badge

                Re: Love/Hate

                It was videotape grade but it was very thin. VHS Video tapes are 12.7mm wide and Microdrive tape was 1.9mm wide.

                1. Mage Silver badge

                  Re: Love/Hate

                  Some video and cassette tapes may have only varied in thickness of polyester (mylar aka PET) base and width. I'd bet it was cassette tape.

                  1. Dan 55 Silver badge

                    Re: Love/Hate

                    Sinclair claimed it was videotape:

                    Infinite loop: the Sinclair ZX Microdrive story

                    It was confirmed that each 43 x 30 x 5mm Microdrive cartridge contained not a disk but a loop of tape, 2mm wide and claimed by Sinclair to be made of same same material as high-quality videotape and not what you’d find in an ordinary audio cassette.

                    Towards the end of the article it mentions how unsold Spectrum microdrives were repurposed and manually calibrated for the QL case... very Sinclair.

            2. Alan Brown Silver badge

              Re: Love/Hate

              I recall them being sold as "stringy floppies" and saw them fail more times than they worked

          2. John Brown (no body) Silver badge

            Re: Love/Hate

            "Were looped tape drives really new?"

            No. The larger and more expensive Exatron Stringy Floppy pre-dated it. Something I considered for my Video Genie (TRS-80 clone) at the time, but lack of funds meant it was not to be. By the time I'd saved up enough beer tokens, the price of a floppy drive interface + floppy drive was almost within reach so I put off the purchase a little longer and went what was fast becoming mainstream :-)

      2. AlgernonFlowers4

        Re: Love/Hate

        If you can run with Microdrives and keep your data,

        Or walk with Microdrives —nor lose the database;

        If neither load errors nor successful updates can hurt you;

        If all data bytes count, but none too much;

        If you can recover the unforgiving minute data lost

        With sixty seconds’ worth of tape looped run—

        Yours is the Earth and everything that’s in it,

        And—which is more—you’ll be a Backup Man, my son!

    2. Tron Silver badge

      Re: Love/Hate

      Uncle Clive's desire to hit a low price point was a double-edged sword. Without it, I would probably be a poet now, and incapable of ordering from Amazon. But like every innovator, he would sometimes get key decisions wrong. The QL needed to be more reliable (although the Apple III was worse). It also needed an industry standard drive, even if it was in a different model (you could add them, but it needed to be standard). Then, it may have actually challenged the Mac, as it was massively cheaper. But this is with the benefit of hindsight. Sinclair and Acorn produced tech that was as good as anything anyone else produced. And Sinclair were producing first generation electric vehicles decades before anyone else considered it worthwhile to even think about. Being a kid in 80s Britain was brilliant. The contrast with today is just shocking.

      1. Mage Silver badge

        Re: Sinclair were producing first generation electric vehicles decades before


        Victorians had electric cars (Nickel Iron and Lead Acid) and even battery powered trains as well as the better known external power electric trains

        Some early battery trains used Nickel Iron rather than Lead Acid, but of note was the Drumm nickel-zinc battery used on four 2-car sets between 1932 and 1946 on the Harcourt Street Line in Ireland (Dublin - Bray). It charged at each end.

        (Electric trams before WWI and Electric trolley buses from WWII but no battery)

        Electric milk floats, before 1960s

        Electric fork lift trucks.

        Mobility vehicles (some trikes, some like motorised wheelchairs from 1930s

        The first electric golf cart 1932, More popular from 1950s

        The problem was battery weight.

        The C5 wasn't exactly practical and too low to the ground. It used a washing machine motor to give more power than the motors used in mobility / electric chairs

        1. steelpillow Silver badge

          Re: Sinclair were producing first generation electric vehicles decades before

          Sir Clive later said that he knew the C5 was shafted by its own useless batteries, he just wanted to goad an over-complacent battery industry into stepping up to the mark.

          Milk floats - humbug! One pulled out right in front of my motor bike and I had nowhere else to turn. Lucky the crates were plastic and overhung the tailboard...

        2. John Brown (no body) Silver badge

          Re: Sinclair were producing first generation electric vehicles decades before

          "It used a washing machine motor"

          The C5 and motor was made by Hoover, but it wasn't an off the shelf washing machine motor.

          See the fine wikipedia article.

          "Although it was later said that the C5 was powered by a washing machine motor,[32] the motor was in fact developed from a design produced to drive a truck cooling fan."

  3. Dan 55 Silver badge

    My Dad got one in Dixons' fire sale when Amstrad was sold to Sinclair. I think the Psion office software tipped the balance otherwise he would have bought an Amstrad PCW.

    The BASIC was nice and I managed to mess around in BASIC creating windows, but I couldn't find the rumoured multitasking anywhere. As it's Sinclair I guess he cut development short before the release and the programmer had to finish off the job themselves in 68K assembly language. :)

    As for the microdrives... well, they taught me the importance of keeping backups.

    1. Liam Proven (Written by Reg staff) Silver badge

      [Author here]

      > but I couldn't find the rumoured multitasking anywhere

      SuperBASIC on QDOS couldn't and didn't multitask.

      SBASIC on SMSQ/E does.

      1. Dan 55 Silver badge

        I felt the multitask claim was overstated if you had to compile your own software to make it actually multitask.

        If Quill and Abacus ran split-screen in two different windows it would have blown everyone's socks off. Maybe.

        1. DaemonProcess


          You are correct, the 68000 was an upgrade from the old Spectrum's Z80 in that it had a supervisor mode. The o/s simply wasn't pre-emptive in that way. I do know someone who programmed scheduling directly in assembler, with simulated trains going round a track with points.

          I had an OPD at ICL. In parts of the company they were hot property and coveted, because they looked cool. In other parts, just meh, we've got better stuff to do all that, like a PC with MS Windows 1 or AT&T system 4 Unix.

          Plus the OPD Microdrives were still bad. More reliable means they would break, tangle or stretch a tape after a month rather than a week.

          Sir Clive's weakness was that he wanted to miniaturise everything on a tight budget, when it simply didn't need to be done in many cases.

        2. Liam Proven (Written by Reg staff) Silver badge

          [Author here]

          > If Quill and Abacus ran split-screen in two different windows

          You know what, that's a very good point.

          QDOS had the basic concept of text-only windows, but AFAIK the OS didn't have any ability on its own to put whole tasks/jobs into separate windows. Oberon-style text-mode tiling would have been a substantial improvement.

          OTOH other early multitasking OSes back then, such as SCO Xenix and DR Concurrent CP/M and Concurrent DOS, had no windowing of any kind either. You could hop between different full-screen sessions but everything was full-screen all the time.

          I think it was in part an idea that nobody had come up with yet. Even the Blit terminal used graphics mode to draw overlapping windows:

          And that was just 2Y earlier... and I am willing to bet much more expensive. As in, probably add a zero onto the price: the DMD 5620 was better-specced and more powerful than the QL (full 68000 for example), and it was just a terminal.

          QDOS was very early. It was before IBM TopView, let alone DESQview, before the X window system, before almost anything else within an order of magnitude of its price point.

    2. Anonymous Custard

      Amstrad was sold to Sinclair

      I think Lord Yer-Fired might have something to say about that?

      1. Dan 55 Silver badge


        If he does I'll ask him what on earth he was thinking when decided the PC200 was a good idea.

      2. katrinab Silver badge

        That's what I thought too. Amstrad was sold to Sky, and lives on, I think, in the form of the Sky Box you get if you subscribe to their service, though maybe they now use Roku stuff? They were one of the early investors in it, though they have since offloaded their stake.

        1. Dan 55 Silver badge

          Brain failure when transcribing, as much as I'd have liked it to be a reverse takeover with Sinclair pulling Sugar's strings behind the scenes.

          Perhaps the Spectrum ROM belongs to Comcast now. What a horrible thought.

        2. jpennycook

          RE: Amstrad was sold to Sky

          AMSTrad were nearly sold to PSION, until someone thought they were over-valued

          “We are not buying Amstrad as perceived by its brand and name. Amstrad is in ashes. We are buying the phoenix in those ashes.”

    3. This post has been deleted by its author

  4. Anonymous Coward
    Anonymous Coward


    ..."shined" is for shoes!

    Otherwise fascinating article. I got the impression when the Mac was transitioning from 68k that the Motorola processors were "running out of steam" and that there wasn't room for a CISC competitor to x86, the competition was going RISCy (PowerPC, ARM, MIPS, Sparc...), and this rather cramped the potential of 68k successors.

    1. CrazyOldCatMan Silver badge

      Re: outSHONE

      Sadly, the US lack of spelling and grammaer is becoming ever more present in the English language. At times I wonder if I'm the only one left who knows how to conjugate verbs..

      1. Neil Barnes Silver badge

        Re: outSHONE

        I conjugate

        You conjugate

        He is American...

        But still, it's better to conjugate than never!

        1. CrazyOldCatMan Silver badge

          Re: outSHONE

          But still, it's better to conjugate than never!

          Oo-er missus!

      2. An_Old_Dog Silver badge

        Re: outSHONE

        Other increasingly-common mis-conjugations:

        He speeded up the car. (Correct: He sped up the car.)

        He dived into the the water. (Correct: He dove into the water.)

        He lighted the torch. (Correct: He lit the torch.)

        He shooted the gun. (Correct: He shot the gun.) -- just kidding; I haven't seen this one, yet.

        1. Neil Barnes Silver badge

          Re: outSHONE

          'gifted' as a verb. It's perfectly correct, if until recently archaic, but it makes me cringe every time I hear/see it. They gave it, damnit!

          1. jpennycook

            Re: outSHONE

            "gifted" always seems to be to celebrate the giver rather than the gift or the receiver.

        2. Julian Bradfield

          Re: outSHONE

          "dived" is correct in proper English. Historically "dive" had both weak and strong past tenses (Old English déaf, dyfde, Middle English def, defde), but only the weak survived. "dove" is probably a modern reinvention of a strong past tense.

          "speeded" has been around for a while.

          "lighted" is also a weak past going all the way back to Old English (from the Anglo-Saxon Chronicles: "Me lihtede candles to æten bi.")

          "shined" as a weak past goes back to Middle English, but it didn't survive after 1800 (as intransitive) in modern British English.

          1. Kristian Walsh Silver badge

            Re: outSHONE

            The “-ed” also functions to distinguish between different meanings of what looks like the same verb.

            “Let us hang the Constable in the drawing room” is either a incitement to murder of a serving police officer or a suggestion for decorating a wealthy home. When you review the day’s work using “Hung in the drawing room” versus “Hanged in the drawing room” you are removing all doubt.

            1. Prst. V.Jeltz Silver badge

              Re: outSHONE

              It may be too late at that point!

          2. John Brown (no body) Silver badge

            Re: outSHONE

            So, what you are in fact saying is that a lot of Americanisms are actually quaint old English retained in the colonies after the insurrection while English in it's homeland moved on an evolved? :-)

        3. Ian Johnston Silver badge

          Re: outSHONE

          He speeded up the car (transitive) as a result of which the car sped up (intransitive).

          She shined the shoes (transitive) as a result of which the shoes shone (intransitive).

          "Dived" is standard UK, "dove" is standard US.

          "Lighted" and "lit" are equally fine in the UK; "lighted" being a little more formal.

      3. Alumoi Silver badge

        Re: outSHONE

        ...lack of spelling and grammaer...


      4. Dr Kerfuffle

        Re: outSHONE

        The word I hate most is 'burglarized'. As in ' I came home and found my house had been burglarized'. NO NO NO, it is 'Burgled' ! It always has been. Why make up longer words unnecessarily?!!!

        1. mdubash

          Re: outSHONE

          Ditto 'obligated' when 'obliged' does the job.

    2. Charlie Clark Silver badge

      Re: outSHONE

      Already by then the CISC/RISC distinction was less and less clear. Intel was moving everything that wasn't x86 legacy away from pure CISC but had to keep the core. The PowerPCs were by far the better chips, but Intel simply had better process engineers and money for fabs, and with Andy Grove ("only the paranoid survive"), the right boss to drive for success using all means necessary, legal or otherwise.

    3. Doctor Syntax Silver badge

      Re: outSHONE

      Those irregular verbs are a bitch.

    4. Liam Proven (Written by Reg staff) Silver badge

      Re: outSHONE

      [Author here]


      > I got the impression when the Mac was transitioning from 68k that the Motorola processors were "running out of steam"

      Yes, that's fair.

      And just as AmigaOS could not move on and adopt the memory management hardware of the 68030, because it would break inter-app and app-to-OS communications and thus totally destroy backwards software-compatibility, the QL OS was even more hamstrung, because in order to fit into 48kB of ROM, both QDOS and QBASIC were hand-coded in 68000 assembly language.

      SMSQ/E remains as trapped on 68000 -- not even 68040 or 68060; it uses those as fast 68000s, like AmigaOS -- in the 21st century as it was in 1986.

      That's why Tony Tebby's Stella never happened.

      That, indirectly, is also why Apple's AU/X never made the transition from 68030 to PowerPC.

      That is also why the successful move of Classic MacOS to PowerPC was so amazing, and Gary Davidian's nanokernel, which enabled it, was such an important bit of software.

      1. Kristian Walsh Silver badge

        Re: outSHONE

        The reason there were no 68060 Macs is that Apple, Motorola and IBM had formed the PowerPC Alliance in 1991 to promote that new family of microprocessors, and Apple’s engineering staff was busy developing PowerPC chipsets (yes, Apple used to do their own support ICs back then). The 68060 would have been well underway by this point, and I suspect Motorola let it run to completion as a hedge against PPC being a total flop (the 601 was not a Motorola chip, but IBM’s design).

        68060 was a faster CPU than the P5 Pentium that had come out six months before it, and it was at least as fast, MHz-for-MHz as the first-generation PPC, the 601. A rumour at the time said Mac hardware upgrade specialist DayStar had planned a 66MHz 68060 upgrade card for older Macs with Apple’s “Processor Direct Slot”, but Apple refused to licence the necessary ROM changes because they were afraid that the combination was fast enough to steal sales from the new PowerMac line.

        I would 100% believe that story - the PPC601 machines ran an OS that was about 75% emulated code. Given that the 060 could match the 601 on native code, it would have absolutely smoked the PowerMac, and it would also have run QuarkXPress properly – the only application on Apple’s “must-support” list that broke the 680x0-to-PPC code-translator!

  5. ICL1900-G3


    I thought the QL looked such fun, I sold my Sirius and bought one. When someone eventually brought out a floppy interface, it became usable - sort of. There was also a rom add-on called ICE which gave a primitive gui. Eventually, I gave up the struggle and bought an Atari ST with a 20mb hard disk.

  6. dak


    1. TheFifth

  7. Anonymous Coward
    Anonymous Coward

    QL was more than a failure. It was a total joke.

    Back in 1984 I had a few 128K Macs and a Lisa on my desk. For software development. On a large table at the other side of the office was a heap of QL internals that had been cannibalized to make a few QLs that actually worked long enough to test software on. It was nt just the Stringy Floppy that was garbage. QA was basically non-existent at Sinclair.

    As for software by late 1984 there was already a wide selection of MacOS software available once the 512K Mac was shipped. Word-processors. Spreadsheets. Databases. You name it. By the time of the first MacWorld at the Moscone in San Francisco in early 1985 there was already a rich and deep MacOS software ecosystem. And that included both MS Word and MS Multiplan. And the single best word-processor at the time, WriteNow, had no problems running on even a 128K Mac. A quick look at a MacWorld from late 1984 / early 1985 will give some idea of just how mature the MacOS software ecosystem already was. By 1986 (my first show) it was fully evolved to the form it kept for the next decade.

    And in QL land? Crickets.Dead in the water even before the official (very rushed) launch.

    I can go out and buy an old 128K Mac and as long as the capacitors and the fly back transformer are checked carefully before I turn it on ( capacitors are a problem with all 40 year old electronics) the Mac will work. And all the software written for it at the time will work. With QL's turning one on even in 1984 was always Russian roulette. With the one in the chamber being the QL actual would power up and work without obvious problems. At least for a while.

    Those who had the bad luck to work on QL's back then always considered them a joke. Which is what they were. And by the time the QA issues was mostly addressed it was - who cares. You only get one chance to make a good impression.

    Does n't time fly. It was almost 40 years ago exactly when I got to use a Mac for the very first time. At an official Apple launch party. There was an open bar so I had the two Macs on site pretty much to myself. But after the first few minutes I knew this was going to be the future. My future. And so it proved for the next 15 years. Although I did ship my last product that ran on a Mac 27 years later. A fine innings on a fun platform.. Or at least for the first 13 years it was fun. As were the Apple people. Then it became no fun at all in Cupertino. And the party ended.

    Oh well.

    1. Jurassic.Hermit

      Re: QL was more than a failure. It was a total joke.

      "Those who had the bad luck to work on QL's back then always considered them a joke. Which is what they were. And by the time the QA issues was mostly addressed it was - who cares. You only get one chance to make a good impression."

      I don't recognise this view at all. I bought a QL plus a daisywheel printer and proceeded to launch and successfully run a small business writing CVs and providing career development services. The QL worked a treat, although the microdrives were very slow. I still have it in a box in my cellar, no idea if it works though.

  8. F. Frederick Skitty Silver badge

    Fantastic article - this kind of content is one of the main reasons I still read El Reg. Another reason is that you rarely (never?) switch commenting off on articles, and the signal to noise ratio remains remarkably high. I dislike the DevClass sister site, as although it has some interesting articles I feel that the lack of a comment section means the opportunity for more insights is lost.

    1. irrelevant

      This is the only site I regularly read the comments on.. Indeed, I come here for them, because I learn so much more!

  9. Jason Hindle

    The right stuff; the wrong side of history - never made sense to me

    Even by the time of the fire sales (and it was tempting then), the QL never made sense to me*. It was clear it was never going to get the software (especially games). It was all a bit too grown-up at the time! My move to 16-bit didn't come until the Amiga 500, which ended up making a great deal of sense given the many similarities to the grown-up OSs we used at University. That said, I'm rather surprised at the following the QL has to this day.

    * I'm the classic 80s child. Got a ZX Spectrum for "homework".

  10. Anonymous Coward
    Anonymous Coward

    "Outshined" ?


    1. Liam Proven (Written by Reg staff) Silver badge

      Re: "Outshined" ?

      Got to admit, I twitched, hard. Still am a bit.

      I only write the words, not the headlines...

      1. David 132 Silver badge

        Re: "Outshined" ?

        Don't loose your temper, Liam. The Internet's full of rouge spelling mistakes.

  11. Mage Silver badge

    Fatal flaw?

    A friend got a QL and one evening he attempted to show it off to me.

    Let down by the awful stringy-drive If only it had been even like the 3″ PCW drive, which was actually better than an Apple II floppy.

    Almost every Sinclair product (Audio IC and Radio kits, Black Watch, calculator, ZX expansion) had some near fatal flaw due to overly optimistic cost cutting and adverts that were of dubious accuracy, The IC10 and IC12 appeared to have better specs than the Plessey and Texas parts. but were in fact re-badged sub-spec but functional parts. The two TVs and C5 trike overly ambitious. He did start by reselling failed functional transistors with his opwn code by mail order.

  12. 45RPM Silver badge

    This is a great article. I thoroughly enjoyed reading it. I love a bit of revisionist fiction, I do.

    The truth is that the QL was rubbish, and comparisons with the Mac - and particularly its GUI - are irrelevant. But okay, let’s compare.

    In many ways, the Mac was a statement of intent - and, for all that the original was very limited and more than a bit rubbish, and despite Steve Jobs wishes, expansion was built in. Or, at least, anticipated.

    Networking was standard, albeit using LocalTalk - and whilst LocalTalk is derided as slow now, it wasn’t slow by the standards of the time. And it paved the way for Ethernet to be included easily later, and with a large library of software already available.

    Expansion cards were anticipated - effectively stubbed in the memory map as virtual slots. Sure, nothing could be done with them until later iterations of the machine appeared - but it made it easy to include them when they did.

    Even support for multiple monitors was included from the start, with a virtual desktop which spanned all of them, and third parties took advantage of it within a couple of years of the Mac’s launch. Polyphonic sound, similarly, was catered for - as, indeed was cooperative multi-tasking (although, for the first few versions of the OS, this was limited to support for desk accessories)

    So yes, the original Mac was rubbish - but without the foresight of its developers, without that eye on the future, it would have been nothing more than a footnote in the history of computing.

    The developers of the QL may have had foresight as well, but there’s nothing in the QL to persuade me that they were able to deliver on it. For all the fine words written here about the QL, it was still very much a yesterdays machine - and many other machines of the time were including similar functionalities - Commodores Plus/4, the Elan Enterprise and so forth.

    A better comparison to the QL might be the Amiga or the ST. Those machines show exactly what the QL could have been.

    The QL was interesting. The jingoist in me would love to praise it. But the best I can say about it was that it was interesting, and Linus developed fledgling versions of Linux on it. But that isn’t enough to say that it was a good machine in its own right, or anything more than a technological cul-de-sac.

    1. Liam Proven (Written by Reg staff) Silver badge

      [Author here]

      > The developers of the QL may have had foresight as well, but there’s nothing in the QL to persuade me that they were able to deliver on it.

      And yet, and yet...

      The thing is that the QL, as it shipped, was a fairly usable box. The Mac, as it shipped, really wasn't.

      Yes, Microdrives were not great, but you got two of them. A Mac with dual drives was *way* more usable, as was an Amiga with two drives. Yes, the Microdrives only held 100kB on a good day, but the Mac shipped with single-sided drives, so they only held 400kB. When you only have 128kB of RAM, copying a file to another medium with one drive meant lots of swaps... but with 2 drives, it was easy, even if the OS only had a kilobyte free for buffering.

      Sinclair's gamble was that two small cheap drives were better than one.

      I don't know if Sinclair knew that the Mac was coming, but I don't think it did. I don't think anyone did. I think a just-about-affordable GUI computer shocked the industry.

      Yes, the QL was horribly flawed. Yes, it was overall a worse computer. No argument there.

      But I had a ZX Spectrum because it was the only way I could get a home computer with sound and colour graphics for £80 (second hand) in 1983 or so.

      I knew quite a few people who had QLs because the machine was the only way you could get a multitasking 16-bit workstation-class computer for less than several thousand pounds -- until the Amiga and so on came along, and don't forget, the Amiga 1000 was 18 months later and at nearly $1,500.

      (The ST was one month earlier than the A1000, but didn't have multitasking until years later.)

      I am not trying to defend the QL or say is was better. It wasn't. But it was unique. There was nothing like it for a long time afterwards. It launched 7 months before the IBM PC AT, which the late great Guy Kewney called "my first taste of Raw Computer Power" (his capitals).

      And remember, the AT also didn't have multitasking. That only came much later, with SCO Xenix, because Intel removed a feature from the final version of the 80286 CPU which DR Concurrent DOS 286 needed to multitask DOS apps.

      The fact that there were over a dozen QL compatibles says something too. The early ones were maxed-out QLs, with all the RAM it could take, proper floppy drives, and better keyboards... but they cost 3-4x as much. There was real genius in Sinclair's ability to make something viable for so little money, just as it did with the ZX80, ZX81 and ZX Spectrum.

      I think that's why even now, you can still buy a new QL compatible. I think that's amazing, and I think it says important things about staying power, about what it offered.

      Sinclair bet wrongly. It bet that multitasking was key; it wasn't, that was the GUI. Sinclair bet that two small slow drives were better than a single bigger, faster, but more expensive one. That, in the end, was wrong: they were too small, too slow, too cheap, and most of all, too unreliable. Sinclair bet that getting into the expanding business market was more important than a better games computer. It was wrong, and Amstrad proved that by making money from the ZX Spectrum, something I believe Sinclair never did. (While the ZX81 was hugely profitable, the Spectrum wasn't until the +2 and +3.)

      But it's not wrong or bad to have aspirations. Sinclair wanted to escape from the niche of unbeatably cheap games computers by making an unbeatably cheap business computer. That's a worthy aspiration.

      Yes, the Amiga and ST were vastly better in every way, but they also came along a year or so _after_ the Mac and so they were informed by it. I think the ST was the machine the QL should have been: a cheap, commodity 68000 computer, with slightly-better-than-an-eight-bit graphics and, if I remember correctly, a variant of _the same_ sound chip as the ZX Spectrum 128. It was, as the saying goes, Just Barely Good Enough. A minimal viable 68000 computer, built as much as possible with ordinary off-the-shelf bits and an inspired adaptation of an off-the-shelf OS -- in ROM, so single-disk operation was tolerable.

      Which, I think, is why the QL OS moved to the ST and only later back again.

      Yes, the QL was horribly cheap and constrained by its price point... *But so was the original Mac, even though it cost 5x as much*... And at £400 the QL was *less* constrained than the 128k Mac at £1,840. That, looking back, is IMHO amazing and worthy of respect.

      The QL did not save Sinclair. It was a bold try, though. Sinclair didn't to know the world was about to perform a massive technological pivot mere weeks later.

      But the QL wasn't a flop. It inspired multiple ranges of compatible machines and a line of technological development on its OS that continued for over 20 years, a time span in which...

      1. The Mac moved to a new CPU family (the Power Macs)...

      2. Then moved to a whole new OS (OS X, that is, NeXTstep with a Mac-like facelift)...

      3. Then Apple launched Power Macs that couldn't boot the old OS, to nail the coffin lid shut...

      4. Then Apple moved that new OS to a new CPU family... killing the ability to run the old OS in a VM...

      5. Then removed support for legacy apps via emulation on the new OS.

      In all that time, the last QL OS was still being developed.

      1. steelpillow Silver badge

        "The thing is that the QL, as it shipped, was a fairly usable box."

        Not really. For every lucky user who actually got something done on one, there were half a dozen frustrated lusers pouring lighter fuel on the thing. It wasn't just the microdrives, it was all the usual Sinclair bits of bent brass and out-of-tolerance tooling. Connectors? Think disconnectors.

        "I don't know if Sinclair knew that the Mac was coming, but I don't think it did. I don't think anyone did."

        Everybody did. Lisa's potential had blown minds everywhere, Apple were not shy about it and it was only a question of time. But we all thought it would be called the Lisa II.

        "Sinclair didn't to know the world was about to perform a massive technological pivot mere weeks later."

        I honestly don't think the pivot was the thing. The command shell lived long on CP/M, MS-DOS and UNIX, all quite different price points from the GUI scramble. Rather, my firm impression was the constant refrain "if only it had been properly engineered". And those later developemnts of it were.

        "I am not trying to defend the QL or say is was better. It wasn't. But it was unique."

        Now that, I will drink to. Icon for that and the whole memory dump.

      2. Mage Silver badge

        making an unbeatably cheap business computer

        Yet, really the 1985 PCW8256 / PCW8512 was that even though it only had a Z80.

        It also ran 10 year old CP/M as the alternative to locoscript, so people went mad porting business software to it. Bundling the printer and screen made it a success for those that couldn't afford an IBM PC ( £2,400) or XT (never mind a 1984 AT at US$6,000 equivalent to $17,600 in 2022) as the entry model was only £400. It was easily upgraded with RAM and even 3.25" floppies (though +12 & +5 needed reversed and the ribbon cable adapted). And then the PC needed a screen and printer too.

        By 1987 the PCW9256 was aimed more at wordprocessing with a daisywheel instead. Later models had an inkjet printer.

        When the PCW line was retired in 1998, 8 million machines had been sold, though the last two models were complete failures due to dramatic drop in cost of PC compatibles and the take off of Win3.x from 1992 and they were 1995, madness.Linux Kernel 1993 and NT launched in 1993.

        So you only needed a 68K or a 386 if you were having a GUI and a real multi-tasking OS. Also without a decent screen and GUI it's hard for a single user workstation to take advantage of the multitasking. The QL was unique, but stupid, so was the derived OPD.

        Really the QL was a flop. I met one person, ever, that owned one. Met plenty of people that had the mad early Macs, Tandy TRS80, PET, Atari, Amiga, PCs (and Xenix & Cromix and Cromencos), S100 CP/M boxes, Apples, RM380Z Research machines, later RMs that ran Dos / Win 3.x, MS OS/2 with LAN manager servers, Apple II. BBCs, Electrons, Jupiter Aces.

        Tesco later foisted PC Jr on people.

        The little Epson LCD letterbox portables with a cassette built in (1981 or 1982?) maybe more successful than QL.

        The QL was hobby niche.

        Never saw a Lisa. Maybe saw one Apple III.

        1. Ian Johnston Silver badge

          Re: making an unbeatably cheap business computer

          In the academic world the PCWs were revolutionary. Everybody in the humanities had one.

          1. John Brown (no body) Silver badge

            Re: making an unbeatably cheap business computer

            Probably because they could pretend it wasn't a computer :-)

            Put the LocoScript disk in and bingo, a working word processor. No messing about with DOS boot disks and learning the commands to start programmes, copy files around etc and since the screen and printer were included, they are all the same. No messing with choosing a printer and then fighting with printer drivers. And cheap. Never forget cheap :-)

            It was an amazing innovation to introduce computers to technophobes that could be bought from the "normal" budget, not have to go through capital spends processes :-)

        2. Antony Shepherd

          Re: making an unbeatably cheap business computer

          British Science Fiction Fandom also pretty much ran on Amstrad PCWs. I know people who still rave about them!

      3. Ian Johnston Silver badge

        (The ST was one month earlier than the A1000, but didn't have multitasking until years later.)

        As I recall, it never did, officially, through there was something called multiTOS which sort-of did it, within the limits of a 68000. The later Ataris could, but few of them ever made it into the wild.

        I was in a research group which used a MegaST4 as our main working computer, and jolly good it was too. We had a Spectre GCR box on it which meant we could run Mac software faster - and with a bigger screen - than a Mac could.

      4. ianbetteridge

        “ I don't know if Sinclair knew that the Mac was coming, but I don't think it did. I don't think anyone did. I think a just-about-affordable GUI computer shocked the industry.”

        Hmmm, not sure about that. Development of the Mac proper kicked off in 1981, and Apple was a pretty leaky ship so I doubt Sinclair wouldn’t have caught wind of it. But it’s worth remembering, as I’m sure you do, that at the time the “obvious” superiority of GUIs wasn’t accepted by everyone, and I suspect Sinclair probably didn’t have the resources to build one on a 68000 - no mean feat, especially with the memory limitations that were inevitable at that price point.

      5. d.fawcus

        Multitasking vs dates

        "And remember, the AT also didn't have multitasking. That only came much later, with SCO Xenix, because Intel removed a feature from the final version of the 80286 CPU which DR Concurrent DOS 286 needed to multitask DOS apps."

        I'm not sure what date you have for that version of Xenix [1], but CDOS-286 had protected mode pre-emptive multitasking of its own native apps at release - whenever that was [2]. It was inherent in how the system worked. The 68000 port of it (ver 1.2) also had such support, and it was available at latest by August '86 [3]. Someone could try porting that to the "2 bit" QL, or even an Atari ST. The porting kit is available.

        Anyway, since you brought Xenix in as a valid comparison for multitasking, you should also include M/PM, Concurrent C/PM, and Concurrent DOS. The former would be '81 (on an 8080/Z80/8086), '82 for CCPM, but more practically for DOS apps only from '84.

        (All of these OS dates would have been earlier for various OEM evaluations)

        As to a 68008 based business machine at the time. Well it could have worked - if it had come out first, someone recently retro-fitted a 68000 to an 8088 based IBM PC, and found it gave comparable performance. Just that the QL wasn't really up to snuff, as it'd have to have been one with 16 bit memory, to start to compete. Since 8086 based DOS machines were already available, and a machine with a different CPU would have had to be faster.

        [1] WinworldPC,and Bitsavers both suggest PC Xenix was available in 1984, possibly July.

        [2] Ver 1.3 is dated Dec '86 (Bitsavers). It apparently worked (poorly) for DOS apps with the E-1 Step of the 286 by Aug '85, and better with the E-2 Step. That chip was supposed to be available from "first quarter 1986".

        [3] I've seen references to it first being available in '85, possibly June. However not before Jan '84 when the work was announced. A Usenet post from Jun '86 suggests it had been operation for "about a year and a half", so maybe from Jan '85. Which then also suggests that CDOS-286 v1.2 was around (for OEMs) at that same time (sans DOS apps), as it was renamed FlexOS with the 1.3 version.

        1. Liam Proven (Written by Reg staff) Silver badge

          Re: Multitasking vs dates

          [Author here]

          A great response, and thanks for this!

          > PC Xenix was available in 1984

          I can believe that. I was still at school myself and wouldn't meet Xenix 'til 1988 but it seems plausible that Xenix was the first multitasking PC OS, before IBM TopView.

          But it needed a hard disk and a 286 PC, meaning a $10K PC.

          So, I think, did Concurrent CP/M which was slightly earlier, in about 1983. (Even the single-tasking CP/M-86 didn't appear 'til 1982, a year after the PC, and at about 5x the price of PC DOS.)

          So this sort of underlines my point, I think.

          For its price point, the QL was the first multitasking PC that was aimed at single user desktop use. CCP/M and Xenix were multiuser OSes aimed at lots of dumb terminals and one host machine with a hard disk, a system some 10x the price of a QL. Even MP/M was aimed at multiuser systems.

          > CDOS-286 had protected mode pre-emptive multitasking of its own native apps at release

          Yes it did. That's why DR removed "DOS" from its name and renamed it FlexOS.

          But already by the time CD286 came out, the PC industry was ~5Y old and DOS apps ruled. The OS needed DOS apps, and it couldn't multitask them, which robbed it of its killer feature and probably is the single biggest factor that propelled DR into a death spiral.

          1. d.fawcus

            Re: Multitasking vs dates

            For Xenix, while it needed a hard disk, it didn't need a 286. See page 8:


            Even before that, there was Veniix/86 demo in May '83, also running on an 8086. Again I believe it needed a hard disk.

            (Note that the IBM PC-AT was not released until August '84)

            As to Concurrent CP/M (and hence Concurrent DOS) - they did not need hard disks to perform their multitasking. That multitasking was limited, but OEM tunable, plus it did support more than just full screen windows. It supported multiple overlapping resizable text windows.


            (The rest of your comments stand)

            Elsewhere there is mention of a multitasking version of C/PM-68k and of GEMDOS.

            Well, there sort of was a multitasking C/PM-68k, but only as a side effect of the CDOS-68k effort. Hence too late for the QL. Plus too big. The system image (CDOS.SYS) is 187764 bytes resident text+data+bss. The text could have been ROMed, but that alone is 161622 bytes.

            In addition to its native API, whereas CDOS-286 had an DOS front end, CDOS-68k also came with a C/PM-68k front end. The compiler supplied with its development kit was the C/PM-68k version of the Alycon compiler. Since CDOS-68k used a FAT filesystem, so does the C/PM-68k front end.

            As to how well it works, I don't know as I've not yet got it running. Lacking a VME/10 system, and not wanting to write an emulator for one, I'll have to port it to a virtual 68k environment.

            As to if a concurrent C/PM-68k could have been produced sooner; given a customer demand, probably - and with lower resource requirements.

            Concurrent C/PM-86 was an outgrowth of M/PM-86. Now M/PM existed for 8080/Z80 and 8086, and the v1 spec is quite small. So the core modules could probably have been ported to the 68k (since they're so simple), and the early C/PM-68k source added on top. I believe the first version of C/PM-68k was in assembly (or maybe Pascal), later versions were in C.

            As to GEMDOS, yes it had a CLI; it can be seen in its porting kit. I don't know if the ST was supplied with one, but I doubt it - given that a bloke I knew in Uni wrote his own one for it.

  13. Charlie Clark Silver badge

    Engineered down to a price

    Sinclair was a clever guy but a poor businessman and this was obvious in all the products: they were too obviously designed down to a price, as you might expect in what was "the Far East" (Japan, Hong Kong, Taiwan, Korea) at the time but not possible in the UK any more. In the absence of sufficient investment capital, time-to-market and refinancing via cashflow were key and this meant that corners had to be cut at almost every opportunity. But at the same time, production could never be really scaled up for worldwide demand, because cashflow was being funnelled directly back into production.

    The sweetspot was probably with the Spectrum, which ironed out most of the awfulness of the ZX80 and ZX81 predecessors and, for a while, became the dominant small computer, at least in the UK. And the competition wasn't really with Apple: both Commdore and Atari used the 68000 but produced machines and software that didn't look like it had been knocked up over the weekend, which the Sinclairs always did: great for hobbyists but a real killer for business. They were also keen in getting as much out of the 68000's ISA advantage over the crippled x86 ISA and, by doing so, like Apple, were able to address markets that the PCs at the time simply couldn't serve, such as video production. For business, by 1984 the PC was becoming established and the sunk invesment of the "platform" meant that alternatives to MS-Intel would always have it hard. Even now, 40 years on, the Wintel still dominates business computing.

    1. steelpillow Silver badge

      Re: Engineered down to a price

      Sinclair came from the world of hobby electronics: magazine journo, inventor of A, B, C grading for cheap transistors, pioneer of the op-amp and negative feedback for homebuild hi-fi amps, he cut his digital teeth on possibly the world's first digital watch - another kit. Throughout, we all happily bent bits of brass, soldered in over-short wires and painted over little embarrassments rather than pay four times the price for a Heathkit full of his graded transistors. He took that mindset into the Speccy, which promptly outgrew his hobbyist world. He never really learned, or understood that he needed to. Not so much engineered down to a price, as bodged down to a hacker's discount.

  14. Jeremy Allison

    Linus Torvalds and I both enjoyed the QL

    I offered to go with Linus to Sao Paulo zoo once to help him avoid having to meet Lula, the president of Brasil which he really didn't want to do :-). I did so only on the condition he do an interview with me. I was fed up of people asking Linus about Linux, so I only asked him questions about the Sinclair QL, which both he and I enjoyed. Interview is still available on youtube here:

    1. David 132 Silver badge
      Thumb Up

      Re: Linus Torvalds and I both enjoyed the QL

      Namedropper :)

      My only connection with Linus, as I learned from an article here a day or two ago, is that he too lives just outside Portland Oregon and has been affected just as I have by our widespread power outages and sub-freezing temperatures. You have me well and truly beaten.

      1. Jeremy Allison

        Re: Linus Torvalds and I both enjoyed the QL

        Nah, I just wanted to push the view count for this 13 year-old video :-).

  15. RobDog

    Apple rubbish

    Everything of that time outshone Apple II. But people bought them because then, as now, expensive seemed to mean better. Even a branch of NASA used Amigas believe it or not because the engineers found that for the cost they were far better than Apple or PC but the mgrs of the lab that was reprimanded them BECAUSE they were much cheaper and therefore couldn’t be as good.

    1. ThomH

      Re: Apple rubbish

      If everything of that time, i.e. 1984, hadn't outshone the Apple II, which was first launched in 1977, then it'd be surprising.

      How do you feel the computers of 1991 compared to those of 1984?

      1. RobDog

        Re: Apple rubbish

        I see what you’re saying, yes true, but they were still told off for buying Amiga vs ‘current’ Apple, not old Apple.

        1. ianbetteridge

          Re: Apple rubbish

          That must have been a fair bit later, as the Amiga didn’t launch till 1985 and production issues meant they were rare in any quantity till 1986.

  16. Anonymous Coward
    Anonymous Coward

    The 128K Mac was not "rubbish"..

    As someone who put in several hundred hours of dev time on 128K's before we got the first 512K swap boards from Apple in late summer 1984 and whose first product shipped for 128K's the Mac that shipped in January 1984 was anything but "rubbish".

    The MacOS in the original 64K ROM's had only one crash bug that could be triggered from the desktop. The 63 region scaneline transitions bug. And one major custom menu (MDEF) bug if a third party application did not check the total menu width / height before redraw. Due to the Menu Manager saving the under menu bitmap and there not being enough memory. There was only 59K free for applications after the System had loaded. And that was it. Both bugs quickly fixed by System patches.

    And the hardware had one major problem (the fly-back transformer) due to a dodgy supplier. Which only started showing up a few years later. That was all.

    The MacPlus and Mac SE's which shipped until 1990 were just incremental upgrades of the Mac shipped in January 1984. The Plus added SCSI. The SE had an internal slot. And both had more memory. A six year run and more than 10 million sold is pretty good for "rubbish"

    Although the MultiFinder System gave the MacOS cooperative multi-tasking in 1987 from the very start in 1984 both VBLTasks and device drivers gave us ways of implementing good enough multi tasking. Especially VBLTasks. In the summer of 1984 on 128K Macs we were shoveling code into VBLTasks to see just how many instructions could be executed until it started having a noticeable effect on the screen redraw. Quite a lot it turned out. But device drivers were usually a better safer option unless you absolutely needed that 1/60 sec "heartbeat". And even then it made sense to put most of the heavy lifting code in a custom device driver.

    Later I got to work on a bunch of Apple prototype Macs and the prototypes Apple shipped to developers were far more stable and of far higher quality than the QL's that Sinclair sold to paying customers. Even the Apple prototypes without proper cases were rock solid in comparison to the QL's.

    So much for Apple shipping a "rubbish" 128K Mac. It was a fantastic machine. And Burrell Smith with the hardware boards and Andy Hertzfield and the gang with the MacOS software did an amazing job. Never bettered since.

    1. steelpillow Silver badge

      Re: The 128K Mac was not "rubbish"..

      No quibble about the engineering, but the spec was just not up to the dreams yet. It was still too small to do any of the things it inspired you to try and do. Lisa had never been more than a statement of intent. But upgraded Macs soon made the grade and the rest is history.

      1. Anonymous Coward
        Anonymous Coward

        Re: The 128K Mac was not "rubbish"

        Were you actually around in 1984? There was nothing like MacPaint, MacWrite etc on any other platform in 1984. And certainly nothing like MacPascal which MS had shut down. Ever use it? Only in 1987 ThinkC started showing people the same kind of features that Apples MacPascal had at the start of 1984. Alice was fun but MacPascal was a real jaw dropper. Even MacDB the two Mac debugger that shipped in early 1984 had features that only I started to see elsewhere in the late 1980's/ early 1990's. But as so few people had two Macs to develop with it was the single machine Macsbug that won that war. Can you IL, SM, DM? Or even $4E71?

        The first Macs shipped to end users in quantity in late Feb 1984 and the first official 512K's started arriving about 7 months later. But you could already get unofficial piggy-back upgrades. Just like with the very unofficial '020 cards and 21 inch color monitor cards that shipped the following year. In 1985. And you could get external and internal harddrives before the HD20 shipped in late 1985. So to us in the trenches developing shrinkwrap software for the MacOS in 1984 the 512K was nice but we already had products shipping. And had been dealing with the 128K base IBM XT's etc for quiet a while. Which themselves were a step up from the 32K/48K IBM PC's and Apple II's. All 512K Macs meant was that you had to worry far less about the segment size and organization for the final build. Far fewer SegLoads. And less worry about doing LoadResource() / ReleaseResource() etc. Thats all

        You had to be there. Actually working on this stuff to appreciate just how good the original 128K was and just how incremental the 512K was. As was the MacPlus. The Mac II I that arrived on my desk at the end of 1986 was the first - now this is a big improvement - moment. But it was n't until the CX and CI shipped two years later that it made much of an impact of the base target platform you developed for. Which was still the original Mac shipped in 1984. Just with 512K/1M memory rather than 128K. Although we did usually still kept a 128K around to test on. And shipped product worked. Just a lot slower (often a lot lot slower) as it SegLoad'ed and LoadResource'd a lot more. But by that stage MPW made most of this kind of book-keeping a lot simpler.

        1. Dan 55 Silver badge

          Re: The 128K Mac was not "rubbish"

          LoadResource() / ReleaseResource()

          Watching a classic Mac user waxing lyrical about how hardware and software architecture was all so much better than IBM PCs is sort of like watching two bald men fight over a comb.

          Shame Commodore only knew how to market the 8-bits. Apart from the TEDs of course.

        2. steelpillow Silver badge

          Re: The 128K Mac was not "rubbish"

          "The first Macs shipped to end users in quantity in late Feb 1984 and the first official 512K's started arriving about 7 months later. ... just how incremental the 512K was."

          Yep. Seven months with nothing but a feckin' "If only it had a second floppy drive and twice the RAM" frustration machine. You make my case well. But I agree it wasn't "rubbish", it was a brilliant work-in-hand.

          Meanwhile, there was me playing Mike Singleton's Lords of Midnight on my Speccy and mapping out the Land of Midnight all ... er, night long, when I was supposed to be working next day.

          But what I really loved was the series of chimp-with-brand-X-box cover photos that PCW ran at that time, whenever some new groundbreaking machine came out. Still have them around somewhere.

          1. Dan 55 Silver badge

            Re: The 128K Mac was not "rubbish"

            As Douglas Adams said, "But what I (and I think everybody else who bought the machine in the early days) fell in love with was not the machine itself, which was ridiculously slow and underpowered, but a romantic idea of the machine. And that romantic idea had to sustain me through the realities of actually working on the 128K Mac."

    2. RobDog

      Re: The 128K Mac was not "rubbish"..

      Just because you liked them and know lots about them, doesn’t make them good.

      1. Anonymous Coward
        Anonymous Coward

        Re: The 128K Mac was not "rubbish".. really?

        Well maybe would you give us a run down of the various machines / OS's that were kicking around in 1984. Actually shipping. Not vaporware. Like Win 1.0 Or VisiOn. Theirs strengths and weakness. For processional software developers of end user software. In 1984. And how the 128K Mac stacked up against the competition.

        Thats the world I'm talking about. Which I was living in. When the 128K Mac shipped this month 40 years ago in 1984. In the previous year I had worked on products for both MS/DOS and Apple II's on machines as diverse as IBM XT's and the Sharp PC5000. The one with bubble memory and a built in thermal printer. A laptop that weighed 11lbs. The most fun machine I used in 1983 was an Atari 800XL. A review machine. Playing Frogger and Centipede on the 800XL certainly hit productivity hard until the boss locked it up. That was a really nice little machine. With a great keyboard too.

        1. ianbetteridge

          Re: The 128K Mac was not "rubbish".. really?

          Getting a really very good windowing operating system into 64kb ROM and 128Kb RAM is an amazing achievement.

          That it only had 128Kb RAM was down to Jobs’ intransigence.

    3. Colin Bull 1

      Re: The 128K Mac was not "rubbish"..

      A good example of the beauty of the original Mac was a program called Musicworks. No idea of the authors, but it was a lovely piece of software. Compose music, multiple instruments, play back with multiple screens, add to boast of their ability to multitask they had a utility called trails that would leave a ghosts of the mouse movements at the same time as everything else.

      And the best bit was their manglement of the standard apple file menu. Remember Apple introduced a new coherence in the pull down menu options across all software vendors. In musicworks when you clicked on the standard File / Volume button it showed an image of a power plug with disconnected wires. To control volume Musicworks had its own volume control button.


  17. You are standing at the end of a road before a small brick building.

    Unfortunately the QL was not backwards compatible with the ZX Spectrum.

    1. Liam Proven (Written by Reg staff) Silver badge

      [Author here]

      No, it wasn't.

      Firstly, the tech wasn't there yet. Next-gen 16-bit boxes weren't powerful to emulate their predecessor 8-bit ones, and any hybrid design would have been ruinously compromised and vastly expensive, killing any chance it might have in the market.

      For comparison, the Amiga wasn't compatible with the C64 or the Atari 8-bits that were its team's previous design. The ST wasn't compatible with old Ataris or anything, although it could read MS-DOS floppies, which was handy.

      Another short tech generation later came the Acorn Archimedes in 1987, although RISC OS 2 was when it started to work properly and that was 1989. That was a 32-bit RISC chip and it *was* powerful enough to emulate a BBC Micro -- or even an IBM PC, although not very quickly.

      Secondly, though, look at it the other way: the way these 2nd-gen home micros _weren't_ compatible freed up the designers to do clean-sheet designs, unencumbered with backwards compatibility, and they were better machines for it, I think.

      This too is why it took Microsoft OSes on PC compatibles until about a decade later, in the mid-1990s, to catch up and offer _both_ a decent GUI with usable multitasking _and_ the ability to run DOS apps.

      (Yes, I know, NT 3.x was a better OS, but not a better GUI; OS/2 2.x had both, but it was substantially less easy.)

      1. Charlie Clark Silver badge

        Yes, but there wasn't even any hardware compatibility which made exising peripherals useless. No attempt at a plarform or ISA.

      2. Uncle Slacky Silver badge

        > Next-gen 16-bit boxes weren't powerful to emulate their predecessor 8-bit ones

        I recall a ZX-81 emulator for the Atari ST, though it may not have run at full speed - it's on this page:

        1. Dan 55 Silver badge

          The Amiga had a paid-for C64 emulator in 1988 by Readysoft but it was a bit slow. Nowadays of course we have Vice and PiStorms.

      3. Colin Bull 1
        Thumb Up

        IBM PC compatible with Apple II

        I had an IBM portable with a Quadram 6502 card that I used to demonstrate apple IIe software. It was also better than than IIe itself in that it would run colour IIe games with an external monitor that was difficult and expensive with the IIe

      4. steelpillow Silver badge

        "That was a 32-bit RISC chip"

        ISTR it was a 28-bit chip or something weird like that. 32-bit ARM took too long to arrive.

        1. Liam Proven (Written by Reg staff) Silver badge

          [Author here]

          > ISTR it was a 28-bit chip or something weird like that.

          A true 32-bit RISC, but the top 6 bits of the program counter were flags, so in effect, it was running a sort of 26-bit instruction set.

          In 1987, the theoretical memory space of a 32-bit chip -- 4GB of RAM -- was unimaginable. 4GB would have been impossibly big for a hard disk -- my Archimedes had 20MB -- let alone RAM. It would have been millions of pounds' worth of RAM chips. That wasn't even mainframe or supercomputer stuff yet when the ARM2 shipped.

          So lopping a few bits off the top for flags seemed like a very reasonable compromise. 26 bits of address space gave you a theoretical 64MB of RAM and even a decade later that would have cost tens of thousands of pounds.

          The first 32-bit RISC OS desktop computer was the Castle Iyonix in 2002, a decade and a half later. That is about how long it took to become a problem.

  18. ThomH

    In some regards the Macintosh almost feels like a Sinclair

    It has a single video mode, which is a plain frame buffer. It routes two out of four of its mouse direction inputs through unused inputs on its serial chip, and has one half of each quadrature signal trigger a CPU interrupt, the CPU doing all incrementing and decrementing of mouse position.

    All audio output is just buffers that the CPU computed in advance, collected as an adjunct to video. Even drive rotation speed — it has zoned approximate constant angular rotation — is controlled by pulse width modulation that the CPU has filled into the same buffer as the direct audio output.

    Summary: a very low-component design, where the CPU is deputised to do almost everything.

    1. Anonymous Coward
      Anonymous Coward

      Re: In some regards the Macintosh almost feels like a Sinclair...???

      Single video mode? Wrong.

      Just because you saw a 9 inch screen with square pixels actually meant nothing when it came to what the OS was doing. And could do.

      Quickdraw 1.0 allowed remarkable flexibility as to both resolutions, color etc. Which is why if you had gone to a Mac trade show in 1985 in the US (or UK) there were a bunch of companies showing various color monitors hanging off expansion cards for the Mac. Now Color Quickdraw in 1987 was the first to give us full component pixmaps but Quickdraw 1.0 was anything but single color single mode. Just like with the AT if you wanted color it would cost you. I even saw some pretty nice color hardcopy output from a CopyBits and some fancy QD footwork. Then the LaserWriter shipped in 1985 and the real fun started.

      The Serial Mouse? All mice were hanging off serial ports until ADB came along. And later USB in PC-land. And the thing about the serial port is - sooner or later there is an interrupt - somewhere or other. To see where the little bastard has moved to now. If you had connected a mouse to a PC running Window 3.0 circa 1990 guess which port you connected it to. And guess what the mouse driver did on install. Register a serial port interrupt handler. Well the more responsive ones did. At least on the Mac with non ADB mice the most serious problem was a fluffy mouse ball. One of the great problems of life solved by optical mice.

      The audio. Having PCM buffers as an option was a joy to work with even if the early Toolbox trap support was rather sparse. Took the PC world years to catch up. And the basic sound synthesizer types supported by the first driver was low stress to use. The reason why it was shoehorned in the way it was simply there is only so much you can do given the hardware you have to work with. Nothing unusual with that. You should see what some of the 16 bit / 32 bit game console hardware does. I've got both Sega Gen and PSX 1 stories.

      The 3.5 floppy drive? Well there was that IWM chip on the logic board. The Integrated Woz Machine. That did most of the floppy drive heavy lifting. The only bit the CPU was really involved with could be found in SonyVars in the Low Memory Globals. Which was just feeding the IWM. To control a feature the Sony drives shipped with. The way that Sony told people to. I think I still have the Sony docs for that drive somewhere. Probably in MacWrite format.

      Have a look at the logic board schematics . There are quite a lot of chips on there. Not the just the CPU, the RAM, and the usual jelly bean 74LS chips tying it all together. And unlike the original Apple II motherboard schematics (from 1977) its actual pretty easy to work out what does what.

      Its a very efficient and very creative logic board design. Which is always the sign of a really first rate hardware designer.

      1. ThomH

        Re: In some regards the Macintosh almost feels like a Sinclair...???

        > Single video mode? Wrong.

        Here's the emulator I wrote and maintain of the earliest Macintoshes. Feel free to submit bug reports for those other video modes that I've somehow overlooked.

        > The Serial Mouse? All mice were hanging off serial ports until ADB came along.

        Suggest you brush up on your comprehension. The issue discussed was triggering an interrupt upon every tick of one of the quadrature signals and having the CPU do the corresponding arithmetic. The Amiga does 8-bit accumulation in hardware. The ST has a dedicated microcontroller for mouse and keyboard.

        Neither forces a CPU interrupt and requires corresponding hardware polling on every single pixel moved.

        Both of those machines are also implemented by me in the emulator linked above if you want to check it out.

        > The 3.5 floppy drive? Well there was that IWM chip on the logic board.

        Each byte of floppy data is individually read by the CPU, and GCR decoding occurs on the CPU. Even the ST manages to place fully-decoded sectors into memory without any CPU intervention; the Amiga DMAs undecoded sectors into memory, but then uses the blitter to decode them. That's partly why it uses non-IBM formatting.

        > Have a look at the logic board schematics .

        I'll grab my copy of Inside Macintosh off the shelf then, shall I?

        1. Anonymous Coward
          Anonymous Coward

          Re: In some regards the Macintosh almost ... so never actual worked on MacOS 1.0 then?

          So based on your comments you were not developing shrink wrap software on the MacOS in 1984. Or ever worked professionally on any of those early Macs. Or shipped anything for this MacOS.

          So you wrote an "emulator" well whoop de doo. For what looks like some kiddie project on github. Big deal. Ever had to write an emulator for a commercial product dev team because the actual hardware was still not available yet? Now thats an interesting technical challenge. After the first few you get the hang of it. But not as much a challenge as reverse engineering an OS with just a few undocumented header files to work off. I actually reversed engineered the Mac ROMs in 1984. Started because you could only find out so much from the three binder loose leaf (not even telephone book) Inside Mac when you are trying to fit a 140K program into 59K of available RAM. And for that project any tech support was 8 time zones away. So which A trap actually crashed. Who knows. We have no ROM memory map. Lets find out.

          You really have nt a clue what you are talking about. When it comes to how the first Macs worked. As computers. Just know some arcane trivia which was totally irrelevant at the time. To professional MacOS software developers.

          FYI you wont find the schematics for the 128k/512k logic board anywhere in Inside Mac. Any edition. It was in the hardware maintenance manual. Which was not available to the public but which people like me had at the time. For developing product. So not surprised you haven't a clue how the IWM / SonyVars stuff worked. On real machines. In 1984. Off the top of my head SonyVars was around the $280's in Low Mem Globals. Could be wrong. Last time I had to look at them for a project was in 1987.

          Here is some documentation I made of disassembly (IL'ed in Macsbug though a serial port onto a VAX) in mid 1984. Recognize it? Got another 64K where that came from. That particular interrupt was not second pass single stepped to fill out a final version because it was irrelevant to the project. And every other protect I shipped. I think I chucked out the 8 inch thick line-printer hard-copy of those docs years ago . It must have weighted 10 or 12 pounds. I really hated typing on the VT100's but the early release copy of MacTerminal I had was still too flaky talking to the VAX.

          Either way. Happy days. Even when I had cramps in my hands from typing for hours on those horrible DEC keyboards.

          ; =============================================================================



          ; (mouse vertical)


          ; Interrupt Level 2 Trap #1


          ; =============================================================================

          00400BEE: LEA $02CF,A3 ; Move SCCBSts to A3

          00400BF2: LEA $02BE,A2 ; Move ExtStsDT to A2

          00400BF6: BRA.S *+$08 [00400C00] ; Branch continue

          ; =============================================================================



          ; (mouse horisontal)


          ; Interrupt Level 2 Trap #5


          ; =============================================================================

          00400BF8: LEA $02CE,A3 ; Move SCCASts to A3

          00400BFC: LEA $02C6,A2 ; A2 = ChannelACommInt Ptr

          00400C00: MOVE.B (A0),D0 ; D0 = ( A0 ) ??

          00400C02: MOVE.B (A3),D1 ; D1 = ChannelACommInt Ptr

          00400C04: EOR.B D0,D1 ;

          00400C06: MOVE.B D0,(A3) ;

          00400C08: MOVE.B #$0010,(A1) ; (A1) = 16

          00400C0C: CMPI.B #$08,D1 ; Is D1 <> 8?

          00400C10: BNE.S *+$02 [00400C14] ; CBranch continue

          00400C12: ADDQ.W #$4,A2 ; A2 + 4

          00400C14: MOVE.L (A2),A2 ; Extract ptr

          00400C16: JMP (A2) ;

          ; -----------------------------------------------------------------------------

          00400C18: ADDQ.L #$1,$020C ; Time + 1

          00400C1C: MOVE.B #$0001,$1A00(A1) ; vIFR = 1

          ; =============================================================================


          ; NOP TRAP


          ; Level 1 Trap #3,#4,#6,#7 - Level 2 Trap #0,#2,#3

          ; =============================================================================

          00400C22: RTS ; RETURN FROM TRAP

          1. ThomH

            Re: In some regards the Macintosh almost ... so never actual worked on MacOS 1.0 then?

            You said a bunch of things that aren’t true and were called on it. Grow up and stop trying to tear everyone else down.

  19. Anal Leakage

    Classic Reg

    No one else in history would position the QL as upstaging the Mac… but that’s one of the things that makes El Reg so… special. Let’s go with special.

    For sure, no one envied the QL spaghetti 8-track—er—I mean, ‘floppy’. No one. But let’s not let reality get in the way of our revisionist fantasia, shall we?

    1. steelpillow Silver badge

      Re: Classic Reg

      "For sure, no one envied the QL spaghetti 8-track"

      Oh, I knew a few. Model railway enthusiatsts. One of 'em "borrowed" my Speccy's Interface 1 and a couple of microdrives after they had been discontinued, and I never saw them again (the drives or the enthusiasts).

  20. bregister

    OPD in a bingo hall

    Years ago, when UK bingo halls were large going concerns there was a special game of Bingo called the National Game.

    Each bingo hall had an OPD to link into the nationwide network and the first prize was a million pounds.

    The was the late 1980's when a million pounds bought you more than a cup of coffee!

    Happy memories of the Mecca Bingo Hall in Islington, London.

    I will also never forget seeing Renee and Renato(!) live on stage at that same bingo hall one Christmas.

  21. Pete Sdev Bronze badge

    In an alternative reality...

    the QL had

    * A 3.5" floppy instead of the damm microdrives

    * A CP/M-86 compatible OS

    * The full 68000 not the 68008

    <sigh>. We might then have avoided the US beige box + Micro$soft take over.

    Externalising the keyboard might have made entry to European markets easier too, though mind you even today I still like the aesthetic of keyboard built in to the machine.

    1. Peter X

      Re: In an alternative reality...

      I guess Apple got the chance to have a second go at building the machine they intended to build (Mac vs Lisa) so it's a shame Sir Clive Sinclair didn't manage to *properly* fix the issues with the QL and build a successor. In particular, if ICL had managed to re-engineer the micro-drives to make them somewhat reliable, I am left wondering *WHY* Sinclair didn't do that themselves or at least licence back the fixes? I *can* see the logic of going with cheaper microdrives over floppy-drives since (going on other comments) that reduced costs considerably. Although the target business customers were likely slightly less price sensitive than regular consumers, so the QL probably could've been another £100 more (in 1984 money) without adversely affecting sales. I think Sinclair just needed another C-level person to sanity check products before they finalised stuff.

      1. Dan 55 Silver badge

        Re: In an alternative reality...

        Sinclair seemed to hate anything NIH. He didn't even incorporate Texas Instruments' new graphics modes used on the Timex Sinclair for the 128K.

        1. Dan 55 Silver badge

          Re: In an alternative reality...

          Oops, don't know where I got TI from. I meant NCR.

          1. Liam Proven (Written by Reg staff) Silver badge

            Re: In an alternative reality...

            [Author here]

            I did a bit of digging, and you are right! NCR fabbed the SCLD for the TS2068. I never knew that!


            I felt the Spectrum 128 was a misstep. If it had incorporated those 2 extra video modes, and the ability to page out the ROM and replace it with RAM, it would have been a much more capable computer, able to run better-looking games, and hi-res text, and run CP/M. Built-in Interface 2 as well -- joystick ports and a ROM cartridge slot -- would have been more use than a numeric keypad socket.

            I speak as the owner of one of those numeric keypads!

            *That* was the machine Sinclair needed. I think it would have sold better, and been a better games computer, and might have prevented the Amstrad buy-out.

            Also, let us not forget the next computer from Sir Clive, the Z88. That is a great little machine in a lot of ways, and that too should have done better than it did.

            1. Dan 55 Silver badge

              Re: In an alternative reality...

              I think Investronica tried to design something which they thought would please Sinclair who was known to believe his computers were suitable for business, but wouldn't be too expensive as really they were just the IT department from a Spanish department store, so they came up with the number pad.

              Also there was the Inves Spectrum+ which was designed in that year when Spain had a tax on computers with less than 64K RAM which weren't localised into Spanish. This had the 48K ROM from the Spanish Spectrum 128K, different RAM with no memory contention, and a TAHC10 from TI (I knew I got TI from somewhere) instead of the normal ULA. Unfortunately because of these changes some games didn't work.

              Then in the Timex branch of the family tree there was the Timex FDD which was an external disk drive similar in design to the Commodore 1541, i.e. almost another computer only it was three separate boxes (power supply, disk controller, 3" disk drive) released in 1985, a year after Microdrives came out.

              It was all there, it just needed putting together but Sinclair didn't seem bothered about the product that was paying all their wages. It didn't even need to be Clive taking an interest, just the company but they were all running round the office with their hair on fire.

              In the end Amstrad gave you everything you wanted two years later in the +3 except the video modes and the ROM cartridge slot, which I think wouldn't have been successful as the Interface 2 wasn't a success.

              1. Liam Proven (Written by Reg staff) Silver badge

                Re: In an alternative reality...

                > Also there was the Inves Spectrum+

                Fascinating -- I'd never heard of that before. Thank you!

                > in the Timex branch of the family tree there was the Timex FDD

                Yes, that was a fascinating device. I wish someone would clone it using a Raspberry Pi. The idea of a cheap "disk" interface which can run its own programs faster than the host pleases me. :-) It could emulate a fast CP/M box, or a QL. How about a plug in device that turns a Spectrum into a QL? How cool would that be?

                > It was all there, it just needed putting together but Sinclair didn't seem bothered about the product that was paying all their wages.

                I can't substantiate this right now but I have read that while the ZX81 was immensely profitable and made SR millions in profit, the Spectrum never was and absorbed that cash and more. That's why SR was desperately trying to crack the business market with the QL.

                In hindsight, I think the obvious path was to "lean in" to the Spectrum as a games machine: do a cheap upgrade with Investronica and Timex parts, and first offer more colours and a joystick port, like this Inves + machine or the Portuguese TC2048.

                Then, later, as an upgrade to that, 128kB of RAM, and an optional proper disk interface -- as you say, Amstrad got it right. (But with no improved graphics, sadly.)

      2. Liam Proven (Written by Reg staff) Silver badge

        Re: In an alternative reality...

        [Author here]

        > *WHY* Sinclair didn't do that themselves or at least licence back the fixes?

        I think OPD microdrive cartridges were not readable in QL microdrives. Lack of backwards compatibility.

        ICL did later offer a disk interface, and I have read that someone is currently trying to restore one and get it working.

        But the fact that ICL did this is indicative.

        Microdrives were that much cheaper that even a more expensive, more business-focused machine retained them. ICL didn't equip the OPD with a floppy drive instead.

        5.25" drives were bigger, and though the media were cheaper than 3.5" this was a small machine.

        The QL was being designed right around the time Sony's existing but pre-standard 3.5" disk was being replaced by a cross-vendor standardised version.

        The latter is what the Macintosh used, but I suspect it might have added 50% or more to the price of the QL, as well as physically not fitting, so it would have required a case redesign too.

        Like picking multitasking over a GUI, it's an obvious mistake in hindsight, but not so at the time.

        About the only cheaper alternative around that time was the Triton QuickDisk, which a few devices did use.

        TBH and again in hindsight, Amstrad's move to buy up all the remaining stock of 3" drives, equip their machines with this non-standard size and then make a killing selling customers very expensive media was a shrewd commercial strategy.

        Another boon of this format is that the budget models (CPC, early PCW 8-series) came with single-sided drives, but you could turn the disk over and use the other side. That was impossible with Sony's design. Mac users just lost the capacity of the other side of every diskette.

        The standardisation hadn't settled yet. The rivals were much bigger and significantly more expensive. Sinclair had its own in-house tech which it invented, owned and did work.

        I owned Sinclair Microdrives: an Interface 1 and a single drive at first. I still have the cartridges, and some replacement drives. They were slow but they did work and they were *so much* better than using the Spectrum with audio cassettes and a tape recorder.

        Flawed, yes. Fatally flawed, no.

        Sure, as soon as I could afford it, I bought an MGT disk interface and I loved it. So fast, so reliable. But Microdrives tided me over for a year or two and were vastly better, for the price, than any alternative.

    2. Liam Proven (Written by Reg staff) Silver badge

      Re: In an alternative reality...

      [Author here]

      > * A CP/M-86 compatible OS

      OK, two questions.

      [1] You meant CP/M-68K, right? Again, that's kinda sorta what the Atari ST had.

      [2] OMG, *why* would you want that?!

      1. Pete Sdev Bronze badge

        Re: In an alternative reality...

        You meant CP/M-68K, right?

        Yes indeed, my bad.

        OMG, *why* would you want that?!

        My premise (feel free to disagree) is a that a level of CP/M compatibility would have helped against the known vicious circle of the time of:

        * Consumers won't buy a new system if there isn't enough software available

        * Software companies won't bother writing software for or porting to a system unless there's a large enough user-base to make it worthwhile.

        It would have eased porting existing software, particularly business software which would have helped give the QL a more 'business orientated' look.

        If the QL had had more business uptake, that would have then had a knock-on effect. After the glory of the Acorn/Beeb machines, most schools decided, not unreasonably, to go for biege boxes because that's what became dominant in businesses and it makes sense for kids to have the same computers at school as they'll later be using in the "real world".

        Regarding the floppy drive: if they'd waited a couple of months (and I think we all agree that the QL was released months too early), the mostly-Sony standard could have been used.

        The microdrives were a case of NIH as someone else mentioned. They might have saved a few quid per unit, but that didn't make up for the development costs, in money as well as crucial engineer-time. The fact that they were (originally) unreliable (very bad for business) meant their true cost was the entire company...

        The 68008 was down to poor business negotiation skills by all accounts. The too-early release being a classic marketing over reality.

        Funnily enough I had an Atari STE (early 90s IIRC) after my +2 bit the dust, the +2 being the successor to my 2nd hand 48K.

        So yeah, I guess I want the QL to have been more like the ST but in black and with the Sinclair logo :-) .

        Thanks for the article by the way.

        1. Liam Proven (Written by Reg staff) Silver badge

          Re: In an alternative reality...

          I do see what you mean, but I think that's all kinds of different improbabilities, stacking up far into the territory of "never was gonna happen".

          1. CP/M -- I don't think CP/M-68K ever sold much, did it? Except the hybrid of CP/M-68K and DR-DOS called GEMDOS in the ST, which didn't even come with a CLI? I think the very small degree of compatibility with CP/M-80 wouldn't have sold it enough. CP/M was a bit too primitive for the 1980s and it didn't even have a standard disk format. ST TOS got that right: use the MS-DOS FAT format.

          2. I could be wrong but I think CP/M-68K had no multitasking. Sinclair was going for this big new step: a multitasking personal computer. And it did deliver, in a limited way. So it wouldn't have gone for CP/M-68K -- just as it didn't go for GST's rival offering, 68K/OS:

          3. The QL was too early? Well maybe. But if it had launched after the Mac, then it would have needed to be a whole different machine: a budget Sinclair GUI computer. And while that is a fascinating line of thought, it's pure speculation, because Sinclair never did anything like that. The _next_ Sinclair computer, the Z88 -- because remember the QL was not the last -- was text-only too.

          As I said, the QL done right with an awareness of the market was basically the Atari ST, and that was a very different and very un-Sinclair-like machine.

          4. I think the fact that the first round of QL clones, the up-specced OPD/Tonto etc., didn't have floppy drives shows that floppies were still just too expensive in '84. This was a 16-bit priced like an 8-bit.

          5. The 68008 wasn't so awful. There's an amazing homebrew based around it:

          The 68008 compares closely to the Intel 8088: an 8-bit bus version of a more successful 16-bit chip, allowing for much cheaper motherboards and much cheaper RAM.

          In my recollection, in the very early PC clone era, 8088s far outnumbered 8086 machines. Indeed in Europe it was Amstrad's affordable PC clones that drove the DOS PC platform and one of their *unusual* features was that they used a full 8086.


          And of course that leads us to another alternate reality: the last Sinclair *branded* computer, the Amstrad PC-200.

          I always thought that was a dismal effort. A few years ago, I specced out what I thought it should have been... to compete with and out the Amiga 500 and ST520, I proposed:

          • 80286

          • VGA

          • Built in Soundblaster or Adlib sound chip and joystick port

          That was doable, but the PC200 launched in 1988. It was early. Before the 386SX existed.

          A VGA 286 in 1988 would have cost £2000 or so.

          Maybe Amstrad could have done it for £1500. I guess it would sort of work with twin 1.4MB floppies and 1MB of RAM, for a bit of disk cache, but it would not have been fun to use.

          But a floppy-based 286 would have been fairly pointless. So what more Amiga-competitive OS could Amstrad have offered in 1988, with a built in GUI and multitasking? OS/2 1.1!

          But that needed 4MB of RAM and a hard disk, so that would have forced the price up to about £4000.

          The PC-200 cost £300.

          That is ridiculous. It would have collapsed so fast and so hard there's a risk of forming a black hole.

          So, no, let's be vaguely realistic: in 1988 Amstrad also launched the PC 2086:

          It was £600 though.

          But Amstrad *could* have offered a PC200 that didn't suck so badly. VGA graphics and twin 720kB floppies. Throw in a sound chip and that wouldn't have been _so_ awful...

          1. Pete Sdev Bronze badge
            Thumb Up

            Re: In an alternative reality...

            CP/M was a bit too primitive for the 1980s and it didn't even have a standard disk format.

            True true, though I was fantasising about an OS that had compatibility not just a pure CP/M OS.

            I don't know how useful it would have been (obviously), although even today companies are running old versions of Windows and old web browsers because they need compatibility with older systems.<shrug>.

            I could be wrong but I think CP/M-68K had no multitasking.

            There was I think a Concurrent CP/M-86 since 1981(?) that did, so would've have needed something like a Concurrent CP/M-68 :-)

            The QL was too early? Well maybe

            The release announcement (with "the first orders will be shipped in 28 days") was blatently too early. Leading to delays, dongles, ASA. complaints, and generally stuffing the reputation before the machine had even been tried. That's obviously a marketing FU though.

            But if it had launched after the Mac, then it would have needed to be a whole different machine: a budget Sinclair GUI computer.

            I don't think the Mac would have had that much influence simply because it was in a totally different price class. It wasn't AFAIK particularly aimed at business either. I was also thinking about a few more months development, not a year.

            A Sinclair QL with something like GEM is an interesting thought though.

            The 68008 wasn't so awful.

            No but from what I've read Sinclair could have got the 68000 for the same price as they ended up paying for the 68008 if they'd negotiated better with Motorola. Which meant it was a pointless performance reduction.

            In the real world I think the best thing about the QL and its derth of software was that it later forced a certain Finn to take up programming :-).

            I'm grateful in general to Sinclair for, quality problems aside, making computers low priced and thus accessible to a lot more households than otherwise. Like many of my generation, I owe my current career to that fact. I still find it a shame that it got sold to Amstrad, and x86+Microsoft (retrograde in many ways) ended up dominating for years.

  22. John Savard

    The One Flaw

    The Sinclair QL was an amazing computer for the price.

    However, I live in North America. I did pick up a brochure for one at a computer store, but I don't know if they actually brought any in. I do know that I never saw any Microdrive cassettes on sale anywhere, but then perhaps they just had to be ordered direct from Sinclair.

    And, instead of a hard disk, it was to offer a module based on wafer-scale technology - which did not materialize.

    Had it offered a floppy disk drive, it would have been more expensive, but it would have still been cheaper than the competition, and it would have had a real chance to succeed.

    However, there's no need to turn the Sinclair QL into an Atari ST, since we already had - several years later - the Atari ST. And that, despite having a successful run, failed too. More ignominiously than the QL, given that the QL, like the Amiga, still lives on today among retrocomputing enthusiasts. (No doubt the Atari ST does too, but to a lesser extent.)

    Apple was one of the partners in developing the PowerPC, and the Amiga, the Atari ST, and the Sinclair QL together weren't successful in creating a market that would lead to 68020-compatible processors, with improvements and advancements, being available right up to the present day in the same way that 8086-compatible and 80386-compatible processors, with improvements and advancements, such as x86-64, AVX-512, and so on, continue to be available to this very day.

    That's "failure". It may not be a very fair standard, but in practical terms, it's the one that counts. Offices everywhere aren't, at least a sizable fraction of them, browsing the Internet, drafting letters on word processors, or calculating with spreadsheets on descendants of the Sinclair QL, they're all using descendants of the IBM PC or maybe the Macintosh. Even those who are using those descendants of the IBM PC to run Linux instead of Windows.

    1. Liam Proven (Written by Reg staff) Silver badge

      Re: The One Flaw

      > 68020-compatible processors, with improvements and advancements, being available right up to the present day

      One small point... I believe you *can* still buy new 68020 chips today.


      It is interesting to speculate what a Sinclair could have done with a 2nd gen QL. For instance, could it have had a single 3.5" floppy drive?

      But the thing is how much would that cost?

      For comparison, the ICL OPD was £1200 with a mono screen and nearly £1700 with a colour screen.

      OK, no screen, no phone, no modem... but still, a floppy-based QL would have been a _lot_ more expensive.

  23. jollyboyspecial


    I think one thing that didn't help the QL was the perception of the Sinclair Spectrum as a toy. So people who wanted a SOHO focused machine didn't buy the QL because they considered it a toy. People who bought the QL wanting an upgrade path from their Speccy were disappointed that what they'd got wasn't a games machine.

  24. fromxyzzy

    Q68 owner

    Hah, interesting that you found the Q68. I purchased one from Derek a bit more than a year ago after a few years in the queue, it's presumably one of the only ones in the US. He's a gent, takes time to build batches but he does it for the community. The Q68 is a solid, extremely compact little black metal box, uses PS2 ports for keyboard/mouse and VGA, however it plays happily with a good HDMI converter.

    I purchased it out of curiosity, since the QL never made it to the US. I haven't delved too deeply into it but it's fairly obvious that it just didn't have a specialty that would have commanded the loyalty of the Amiga's multimedia capabilities or the ST's rock solid MIDI timing. Add that to the flaws with the original hardware design and if it weren't for the extremely basic design it really didn't have much of a reason to be kept alive by the community. Since it was a simple enough design with off the shelf parts, it made it easy to clone it and re-implement it in ways that weren't really possible with Amigas based on specialty chips.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like