The New ROM Antics – building the ZX Spectrum 128 Following an unlikely series of events involving British Telecom, Prince Philip and a VTX-5000 modem, your teenage protagonist found himself at a drunken dinner party in 1984's West London. Across the table, excitingly, sat my boyhood hero Sir Clive Sinclair, and he seemed genuinely interested in what I had to say about the …
Re: " their ego crashes the company (you don't hear about those in the news)."
I think Gerald Ratner counts here.. Another I can think of (although this wasn't in the mainstream news) is Adam Osborne, who developed the Osborne 1. For those who don't know, this was the first "portable" (put in quotes as the thing was the size of a large sewing machine, required a mains supply and weighed 11 kilos, so more luggable). While the sales were already dropping thanks to competition, The sales of the 1 were not helped by the fact that 2 years after the launch, he started talking about how their next computer, the "Executive" was so much better.
I think Gerald Ratner counts here..
To be fair to Ratner, that was more the media than him.
The joke he made in the context of the wider speech he was giving was amusing, and went down well at the time. What he didn't appreciate was there were journalists in the audience who (rightly) saw that single joke as a great story.
Out of context the comments were ruinous. Tim Harford does a good bit on it here.
While Mr Ratner resigned from his position, Ratner group simply rebranded as Signet Jewelers and Ratners shops became H Samuel and Ernest Jones (among others) and it is currently the largest retailer of diamond jewelry in the world with annual revenue of $7.8bn (2022).
That is the popular story but reality was significantly different.
Osborne appointed a manager to handle the Osborne 1 business while he focused on completing the Osborne 2. That went really well until this caretaker found a stack of unpopulated Osborne 1 main boards and decided to build them up into complete computers. All the tooling was worn out and had to be replaced. Many of the parts had long lead times. All the company's cash was tied up for months with no new Osborne 1's to sell. There was no money to complete version 2 and put it into production.
This was at a time when clock speeds and memory sizes doubled every 18 months at constant or falling prices. The competition moved on while Osborne was stuck with a potential pile of old computers that would not sell and a new design he could not afford to build.
In the early 1980s I was at Inspex, the big UK Inspection Equipment show. This was at the time when electronics were coming in to the main stream* instead of manual equipment, we looked at one very expensive (as in 5 figure) machine and I immediately commented. "Interesting use of a ZX81!" They were a bit taken aback that I'd recognised it from the screen but admitted that yes, that was at the heart of it.
The reality was you could by a ZX81 for far less than any other Z80 based system board and of course they'd taken all the control lines neatly out to the edge of the board.
*There were earlier machines, the company I was working for at the time inherited a 3D measuring machine from the head office driven by a PDP-8.
This post has been deleted by its author
When computers came with manuals just as cool as Cloud City. Nice to see the Next carries on the tradition.
...and here it is:
http://www.retro8bitcomputers.co.uk/Content/downloads/manuals/zx81-basic-manual.pdf
(Incidentally, the author of that manual - Steven Vickers - had a very quirky sense of humour. Why do computer manuals nowadays not have lines like "Suddenly, your housekeeper rushes in to tell you that eggs are now 61p a dozen"? Oh, yes... there are no manuals with computers any more. Bah.)
[Shrug] Stranger things have happened. I saw a Commodore VIC-20 running a substantial chunk of Chemical engineering at an ICI site, and spent a year working on a modular interface system to allow a Commodore 64, a BBC Micro, or potentially a humble Speccy to do the sort of industrial control work previously considered the domain of DEC PDP 11s and the like…
Most of the Hong Hong Stock Exchange was run on Beebs, and the Hong Kong Harbour Authority modelled water flows using a network of Beebs connected to monitoring equipment. And then all those factories where, under half an inch of oil, grime, and metal shavings, was a Beeb controlling the processes.
I did my apprenticeship back at a popular camera company back in the day, and one of the things they got me to do in the second year was rewrite the power control system for two paper coating tracks on a BBC Micro (a prior apprentice had written the original system as part of his degree). The bosses wanted quite a complex level of integration and alarm management and so I set out to write something from scratch. The hardware was driven via the printer port using all 8-bits for addressing and the bidirectional lines for strobe and readback. I left that company in 1988 during my third year, and was told by others who remained that my software (baked into an autostarting ROM) was still in use way into the 20's.
The first PC I owned was a Spectrum but after the second one failed, I returned it to Dixons and exchanged for an Oric 1, which sits on my desk and still works today. We had a lot of fun as devs back in those days as you really could create anything if you had a computing device, a bit of time and some electronics experience.
"It might have even been a power station". At the time of the ZX81 we had a manager at out Power Station that thought it would be a good idea to have a Mill simulator for training operators running on a ZX81.
The same guy thought there was no need for a Word Processor since he could do all his writing on a spreadsheet!
Anyhow said manager recruited an operator with zero programming experience to write the code. It took months and never did work.
Excellent article. It's amazing to think that the Spectrum could have been a business machine, but thinking about it, it wasn't that much less powerful than the first generation IBM PCs, but it was about a 20th of the price.
Of course, those IBM PCs were so well built that I always thought if you threw one at a brick wall, the wall would come off worse, and they had the advantage that at the time there was a perception that no one was fired for buying from IBM. I think if Sinclair had developed a standard floppy drive (I know other companies did have floppies for the Spectrum, but they had all sorts of formats), and possibly even a hard drive, with a reasonable DOS, the Spectrum could have been a contender business wise. The problem is, both would have cost too much.
I love both my Speccy 48K and my ZX81 (with the wobbly RAM pack and ZX Printer), but being honest, Sinclair didn't have a great rep for reliability.
It's amazing to think that the Spectrum could have been a business machine,
I think it would never have been one, because it didn't look serious. Imagine if the Spectrum said no to someone. They would have asked the computer operator of this oversized TV remote to bring in a real computer to compute the answer instead.
Not in the original case, and certainly not with the rubber keyboard. But given a decent case, and a better keyboard, it could have. Of course, a better case and keyboard would have raised the price, but Amstrad managed to build the PC1512 and 1640 for a lot less than the IBM cost, so there is likely no reason Sinclair could not have done it.. Cases and keyboards can be changed. The question is, whether it is feasible to do so.
> So did Beta BASIC, the third-party upgraded programming language for it.
I was such a big fan of Beta BASIC that I wrote the original Wikipedia article about it.
https://en.wikipedia.org/wiki/Beta_BASIC
The SAM Coupé's SAM BASIC was by the same author and even better. Sadly, though, Dr Andy Wright did not open up the source code and the ongoing SE BASIC does not inherit most of its niftiness.
Liam is wrong about Andy Wright and the SAM BASIC source. It was freely released on disc by the author and (mislabelled as an anotated disassembly) is available in PDF format at http://sam.speccy.cz/rom/sam_rom_v3-0_disasm_annotated.pdf . That shows how very much better use could have been made of 32K ROM than the Sinclair/Investronica lash-up, given sufficient time and talent.
For almost no functional benefit (a sluggish editor with no key roll-over, prone to losing characters or whole lines, buggy RAM disc and a pointless hardware music player copied from Microsoft, ignorant of bars and blocking anything else from running as it bleeped) the Spectrum 128 ROM adds more than twice as many bugs as the original 16K ROM had (released without the knowledge of the developers at Nine Tiles, who hadn't finished it) and it fixed none of the earlier errors well-documented in the Logan book Goodwins understandably cribbed from.
They had an upgraded graphics chip in the form of the SCLD developed by Timex in 1983 (which Sinclair was planning to use in the Pandora). This had a 512x192 pixel mode that was capable of 80x24 text using a 6-pixel wide font. But the 128 development was driven more by Investronica than Sinclair and it didn't know about the existence of the SCLD. In recent times it's become possible to retrofit the Timex modes to the 128 using replacement ULAs. But there wasn't enough interest to sustain production.
But the 128 development was driven more by Investronica than Sinclair and it didn't know about the existence of the SCLD.
It's crazy that Sinclair took so little interest in the one product that was paying all their wages that he left it to US and Spanish distributors to lay out their own individual upgrade path, both incompatible with each other.
I have vague memories of reading about the PC200 at the time in my Spectrum-owning friend's Sinclair magazine. And the one thing I remember is that even *they* were critical and dismissive of it.
But I don't think that one was meant to be a cheap office/serious machine.
I assume- in hindsight and with the benefit of having read more about it since- it flopped because it was obviously trying to target the 520ST and Amiga 500 with its very similar all-in-one case, but with far weaker graphics, sound and general spec that would have turned off home/low-end users who cared more about that than compatibility with some dull, text-based IBM PC apps. (And probably too crap for those who *did* want a half-decent PC compatible).
Amstrad had already targeted the word processing and "serious" market elsewhere, and they'd been very successful in *that* attempt, which was the non-IBM-PC-compatible Z80-based PCW series (mentioned elsewhere in this thread).
You are so right. I was a business planner at ICL and was given the use of a 1901A to run the modelling. Except as I recall the only software available was PROSPER and there wasn't even a FORTRAN compiler available. My enlightened boss allowed me to fiddle my expenses to acquire a surreptitious TRS-80. Knocking stuff up on BASIC was trivial and I was knocking out results to go senior management. I can't take that to the board said my manager. 80pp matrix wasn't ICL and use of alien kit was forboden.
So lashed up a Termiprinter with a fancy bit of cabling. Results on standard 132pp green/white lined paper. RESULT! I got a commendation for getting their modellng software to work so well - and so FAST!
The Termiprinter cost rather more than the Trash-80 & Matrix printer combined but no one but me and my boss were any the wiser.
Huh? I used a Mannesmann-Tally MT81 dot matrix printer with my Speccy (va a Romantic Robot Multiprint Centronics interface). The output was rendered by the printer using its built-in fonts, and so looked the same regardless of the computer the text came from. I used Softechnics' "The Writer" word processor, which rendered to the screen at 40 characters per line, and scrolled horizontally so you could see the entire line. The printer was £99, and the Multiprint about £40.
ZX81 , Serial I/O card (Maplin) and a Brother thermal typewriter all hooked up together to do printouts.
I even did a simple database with the ZX81 and a fair bit of machine code work, it was a pain having to save everything to tape....and one bad tape...
Oh they were fun days.
Yep, much the same setup here over at the Reg FOSS desk.
Quickly replaced the +D with a DISCiPLE, which was much more reliable.
I had a Panasonic KXP 1081 and the Last Word, with a "huge" 64-column display. Best looking essays in the department.
A bit later I ended up losing one of them to a disk crash and pulled an all nighter writing my own disk sector recovery tool to retrieve the thing.
I think it would never have been one, because it didn't look serious. Imagine if the Spectrum said no to someone. They would have asked the computer operator of this oversized TV remote to bring in a real computer to compute the answer instead.
Sinclair should have realised you can have more than one compatible model at the same time, like Jobs but Jobs was cutting down models, Sinclair needed one more, then he could have pushed the boat out with a proper keyboard, real floppy disks, and shamelessly copying and pasting Timex's hi-res screen mode for the pro desktop model.
> Sinclair should have realised you can have more than one compatible model at the same time
TBF, they did release the 16k and 48k models at the same time ;)
For me, the reason the IBM PC came to dominate things, was because of two interlinked things:
a) It was made by IBM and "no-one ever got fired for buying IBM"
b) It used standardised hardware and had an openly published spec for the BIOS
And once the BIOS was reverse engineered to remove the IBM tax from "compatible" machines, the economies of scale truly kicked in, which meant that even relatively niche hardware (e.g. graphics accelerators, sound cards) could sell enough units to be profitable.
At least until the hardware became fully commoditised and the various companies found themselves stuck in a merger-or-die loop...
As such, I occasionally harbour a fantasy of Sinclair doing similar, and releasing a set of standardised designs for ZX Spectrum peripherals - here's the sound-chip we recommend, here's BIOS support for up to 128k of RAM, here's a standardised API for mass storage devices, etc.
Alas, in the real world, Sir Clive was obsessed with reaching a specific price point, and didn't really see computers as anything other than just another electronic device with limited long term potential.
So what we got was dozens of independently produced and only occasionally compatible peripherals. Kempston (aka Atari), Fuller and SJ2 joystick interfaces, the Currah Speech Box, light guns, printer interfaces - all these and more, each competing with each other and the few peripherals which Sinclair did release, and which came with their own "designed to a budget" quirks.
C'est la vie...
"As such, I occasionally harbour a fantasy of Sinclair doing similar, and releasing a set of standardised designs for ZX Spectrum peripherals - here's the sound-chip we recommend, here's BIOS support for up to 128k of RAM, here's a standardised API for mass storage devices, etc."
Gawd, even just "Here's a standard API for *ANYTHING*...."
I want to output to the printer stream. Ok, roll dice and call....... 1601. I want to load a file. 4D16 gives me.... 0556, but even that's only for tape, no way to say "no, not tape, some other device", you have to lovingly hand-craft your own code to do something else to just. save. to. somewhere. else. And if there's an error, BANG! right back into Basic, so not even a roll of 4D16 to help you, you have to envelope that in yet more code.
Any.
Other.
System.
LD registers
CALL entryblock[n]
Half my coding on the Spectrum was just coding *around* the Spectrum. Oh well, I spent Christmas redocumenting and repackaging my Spectrum libraries to lay a solid path through the swamp for other people.
Along with a lot of other people outside the UK and EU I’m still waiting for my Spectrum Next from the Kickstarter campaign. The latest hurdle for the project team is UPS’s insistence that from a shipping point of view, there’s no difference between Li-Ion laptop batteries and the CR2032 coin-cells that the Next uses.
But when I do receive it I’m looking forward to rediscovering the joys of Speccy programming!
>there’s no difference between Li-Ion laptop batteries and the CR2032 coin-cells t
<rant mode> There is an exception in ISO13485 for medical device electronics that allow a computer to be shipped with a single coin cell battery.
Intel's marketing dept (a bunch of Golgafrinchans who couldn't invent fire or decide if people wanted fire that could be fitted nasally) decided to add one of those little greeting card chips to the box to play an Intel Jingle when you opened it. = Now it's 2batteries and the same shipping requirement as 2Ton of Tesla batteries wrapped in Semtex and drizzled with Nitro-Glycerine
A Z80 based business machine? Banked memory really is not as much fun as it sounds. At least the 8088 in the PC could keep its BIOS and interrupt table visible while you fought the segmented memory model and gave a fairly straightforward route to increase the amount of RAM available to any program (except GW-BASIC that was hardcoded for 64K).
Not on the Spectrum but I built a microspectrophotometer controlled by a Z80. Another lab apparently built one controlled by a PET. Microsoft (Or was it two words or one word camel case back then?) had a FORTRAN compiler for CP/M and UCSD p-Stytem also ran on it. Given the layout of memory on the kit we had there was only the lower 48k available plus a bit of high memory sitting above the BIOS & video - or possibly squeezed in between.
Without bloat it's surprising what you could do with 8 bits.
My first real job, apprentice at Ferranti, I had to use an impressive automated PCB testing machine. Massive thing, size of an upright piano with a desk in front, onto which you wired up the PCB in question. So big it had a room of its own. Whole lot driven by an early PET which just sat on top of it, along with a printer to output the results to.
First computer I had was the Sharp MZ-80, which, if you think about it, wasn't all that far removed from the ZX81 in terms of capability. It came with a serious case capable of doubling up as a wrecking ball, a monochrome display and a tape deck. The keyboard was like that from a cash register of the period; which is to say, bloody awful.
A disk interface was available, and I think there were some useful business software of the day out there. I only remember tape basic for that machine though.
A major headache of the system is that it's BIOS was absolutely negligible. Barely enough to start a tape loader to perhaps load a basic interpreter or other environment. This made booting up to a point of usefulness a little slow. And a hard reset a major pain.
ZX81 had a little more ROM available to it, and so, a hell of a lot nicer to use. But it did come a few years after the Sharp did.
Both systems did what academics hate - taught a whole generation of kids, including myself, programming by GOTO. One must unlearn what you have learned...
The IBM-PC was relatively expensive because it was well built -- IBM, made stuff to a specification rather than a price.
They obviously attracted a lot of lower price clones, partly because of their practice of supplying a full set of manuals with the machine, manuals that included schematics and a BIOS listing. This meant that practically anyone could build a PC, and they did. The limitations were merely the mad scramble for parts (especially RAM) and the tendency to try to wring that extra few cents out of the design. (The company I worked for, for example, screwed themselves because of the founder's insistence that you didn't need a multilayer circuit board, you could get everything on a two sided board.)(That, and a tendency to slap small capacitors on the logic lines to tweak timings and suppress transients......)
I have no idea why people were still building budget priced home computers and trying to pass them off as professional units. Its one thing to have a ZX81 to play with at home, something else to use one in a business situation, its just not going to work. Also, by 1984-5 computers were starting to get hard disks, networking and then the 80286 -- a whole different world from those of a few years previous. (OK -- the hard disk might have been 5MBytes, the network was coax and the 80286 still had to pretend it was an 8086/8088 but it didn't take much imagination to see where things were going. ROM BASIC had also passed into irrelevance as well.)
"It's amazing to think that the Spectrum could have been a business machine, but thinking about it, it wasn't that much less powerful than the first generation IBM PCs, but it was about a 20th of the price."
And was easily accessible to very small businesses. I used to know a guy in the building trade, now sadly no longer with us, who used one to revolutionise his business.
As The Boss, much of his days were spent going out to look at / measure up new jobs, and his evenings were spent preparing estimates for the prospective clients, based on his visits. All written up by hand, as was the norm for all the local builders, it took ages to do.
He saw the potential in home computers, bought himself a speccy, and learned how to use it. He learned enough BASIC to write simple programs, knocked up something to work out his estimates for him and print them using the thermal printer.
Suddenly, he was streets ahead of his competitors. His estimates were done and presented next day, in a format that was much easier to read than the traditional scruffy handwritten scrawl. His clients loved it, he got more work, his business grew. He was able to pitch for local authority jobs, which needed to be presented in printed form and were beyond the reach of the handwritten estimate brigade. He was unsure that the little thermal printer output would be good enough, but the authority was happy with it and lots more work came in.
Not bad, for the cost of a speccy + printer and the time spent learning a little BASIC.
Some family of ours were in a similar position in the early 80s, they invested very early in CAD for a construction business. The edge it gave them being able to outpace paper draughtsmen by orders of magnitude (once the initial proposals were done) won them an awful lot of work.
The firm was later swallowed up by some anonymous corporation, one of the ones that now forms the cartel operating the UK Construction scene. Exorbitant rates and poor service being the default; with few fall backs as the red-tape requirements are very clearly designed to shut down the little guy from operating. Bloody bureaucrats!
A lovely vignette of the time..
As far as hiring practises go, the biggest problem was (and amazingly still is) knowing how to decide if someone was good for the job. Outside of academia (which was largely above the squalid battles of home computers), there were no qualifications to check, and no easy way to discern if people were capable of doing something that next to no-one had done before. You hired people, put them to the fire and saw whether they melted, exploded or achieved something remarkable. Companies weren't just figuring out how to build computers for the masses, they were figuring out how to be companies.
And these computers are still living on - there are some cracking Z80 kits (and, hiss 6502) out there, that recreate the experience of having a machine over which you have complete and utter control. Some allow you to build out interesting systems to your specific needs, and some allow you to (ahem) play games. Remarkably, quite a bit of new software is being written, from entirely new operating systems to.. yes, ok, games.
> It's too bad the 6809 fell out of production, as I think it would have made a better retro computing platform than the 65C02.
The 6809 begat the 68000 which did great. The 6502, more indirectly, begat the ARM, which did even better.
> Still, I find the Commander X16 appealing.
It's a weird, compromised design, and it is *WAY* too expensive for what it is.
https://www.thebyteattic.com/2021/09/engineering-for-purpose-silliness-of.html
The author of that piece designed the Agon Light, which is in its way a much much nicer 8-bit home computer, and it's £50.
CP/M was pretty portable CP/M-86, CP/M-68k,CP/M-8000 but booting a VAX into a hypothetical native CP/M-VAX as there wasn't even a CP/M-11 is Munchausen worthy. ;)
I would imagine the writer meant there was a VAX/VMS hosted PL/M-80 cross-development toolchain.VAXes running VMS or Ultrix (or even BSD) were "commonish" back then before the more affordable Sun kit became de rigeur although VAX cluster were used by Sun itself for its financials until quite late I believe.
Sinclair gear wasn't all that common in AU and the Amstrad models CPC464,664,6128 were a bit more common or visible. Having never seen one in real life, I always thought the Spectrum was M68K based (68008?)
The CPC6128 came with both CP/M-2.2 and CP/M-3 (plus) - the latter had a relocatable 8080 assembler and bank switching. For my sins I could probably even now translate the Z80 ED and FD prefixed instructions in my head.;)
《Insert: "The Four Yorkshiremen sketch" here》
Side note: I had a microvax 2000 on my desk, running Ultrix, in my first full-time job. A wonderful bit of kit. My colleague, on the desk that backed up to mine, had one running VMS. We called them 'bonnie' (the Ultrix machine) and 'clyde' (VMS). Fond memories.
I'm late to the party here but for posterity this was implemented by plugging a real Z80 card into the VAX11/780 backplane and there was some magic to route disk and console support out to the VAX and a users terminal. As I recall we could afford 2 of them and they had to be shared.
I can only assume we did that because the Tektronix cross assemblers that we used on the VAX for the QL development didn't support the correct dialect to be able to assemble the original source code. At the time Tektronix did a range of cross assemblers but they all shared a common set of directives that likely as not were not the same as the original processor manufactures assembler.
The only other site in the UK that was using these boards at the time was some defense contractor down Southampton way, that was just far enough to justify the use of the SRL twin engine turbo prop for the day trip from Cambridge to go and see them in action. Normally that plane was used for the monthly (possibly weekly) trip to Timex in Dundee to try and resolve manufacturing issues.
Different times.
I used to repair Spectrums as a part time job after school.
I still remember the day we had one come in that smelt funny. Opened it up, and the ULA was a charred mess, along with the motherboard beneath it. Turned out the son had plugged in the wrong power supply to the wrong socket at the back, at which point it was said the Spectrum started blowing out grey smoke....
Still managed to get it working again, though the board looked like a cat's cradle after I replaced all the burnt out tracks!
Well, according to the "many worlds" cosmological theory, there will exist an infinite number of parallel universes where you did this. Thankfully we don't live in one of those, where everybody's walking round using Sinclair smartphones with 2 colour lcd screens and squishy rubber keyboards with 15 functions on each key.
It had the Spanish enhanced ROM, pretty different to the UK version. In this video it shows editing a string with the editor and using the calculator pad to enter numbers. Then it shows the calculator pad doesn't work in 48K mode. Finally it shows strings can't be edited on the UK version and the calculator pad edits in BASIC mode but enters numbers in Calculator mode.
...my home computer was a QL, and my work computer was a Merlin Tonto (the OPD version of a QL with a modem build in). I only used the Tonto as a dumb terminal to talk to the actual computer I used for work back then though, an IBM mainframe (via some sort of SNA/3270 gateway/emulator). Of course, I worked for BT then - I don't think BT ever actually sold many of them to real customers.
Could the Speccy have grown up? There were so many reasons why the original was not a serious business machine. (I should know, I tried using one for stock control). Yet it went through endless upgrades, first with Sinclair, then with Amstrad, and eventually as Miles Gordon Technology's SAM Coupé, which itself was pushed forward in various ways after the fall of MGT. The Coupé had more memory, better display, extended procedural BASIC, bays for two floppy drives, and a keyboard that felt a bit clacky but worked okay. But it looked like a toy and lost out to the beigely-suave 16-bit Amigataris.
Yes, beige was suave in those days. Sheesh! Black was for boys' toys, and the Speccy proved it. It was many a long year before offices could allow black kit to be seen in there. And its dead-flesh membrane keyboard was the ancestor of 99% of today's office keyboards. So the Speccy did get a few bits of the future right, even if the world was not ready for them at the time.
> Could the Speccy have grown up?
It did.
http://www.nedopc.org/nedopc/Sprinter/About
https://en.wikipedia.org/wiki/Sprinter_(computer)
And a host of others...
https://en.wikipedia.org/wiki/List_of_ZX_Spectrum_clones
For enhanced clones, for example, look at the Scorpion and Pentagon.
Well as someone who was filling out a reverse engineered disassembly of the MacOS 64K ROMs on a VAX 11/780 that very year (eventually got everything but the low level PASCAL QuickDraw code) I found two statements a little bit odd.
CP/M was just a monitor and the internals were readily available in places like early issues of BYTE etc from the late 1970's. So mapping CP/M calls onto VMS or even just rolling your own (which many did) was really not a big deal. Maybe 10K/15K of exe. There were even a couple of books kicking around with fully commented listings of full featured monitor / OS's. In Z80, 8080 and 6800. A quick visit to Foyles on Charing Cross Road would usually turn up something interesting. Once you could actually find where they had hidden the books. Mostly organized by publisher back then. And really old books might still have L/S/D prices on them. More than a decade after decimalization. Dillons had a much smaller selection of tech books but you could actually find books quickly.
And bank switching for Z-80 boards that addressed more than 64K had been around since a few months after the first Z-80A shipped in 1977. For system software bank switching and supporting code overlays were part of the same problem. Running a program that was much bigger than available RAM and 64K. X86 16/32 segment registers were basically just bank switching supported on CPU rather than off CPU and overlay were a fact of life in MS/DOS and MacOS applications (the Segment Loader) until dependable VM came alone with the i486 and the 68020 a few years later.
But hey, we we all just hacking around back then. Which is why it was so much fun. Not knowing just how little we actually knew. Which is why we got so much done. With so little memory and disk space. With no processing power. Counting instruction clock cycles. And knowing how many clocks it took to execute an instruction. If you dont know something is "impossible" you can usually get it to work.
That weird way the books were organised by publisher was something that Christina Foyle, the daughter of one of the founders, insisted on. Her Wikipedia page describes some of her other obnoxious business practices, but doesn't mention her admiration for Hitler and the Nazis. A friend who worked at the sprawling old store back in the 1990s said the basement was full of long out of print books, which would occasionally be sold in one of the obscure corner rooms. Since moving to the new shop earlier this century, the amount of shelving space is a fraction of the old store and it's a shadow of its former self where browsing often turned up some really interesting books.
If you had wandered into a big bookshop in the 1910's or 1920's in London it would have been organized pretty much like Foyles in the 1970's and 1980's. Which if my memory serves me right you still had the bring the book to a desk to write up the invoice, then pay at a cash desk, then take the stamped invoice back to collect the book. Just like in Victorian and Edwardian shops.
As for the "Nazi stuff". She was in good literary company. Like George Bernard Shaw. Who never met a dictator he did not like. At least for a while. And if you want real cringe-worthy stuff from that era try the books by the Webbs (founders of the Fabian Society). Simpering over Comrade Stalin's building of a Soviet Utopia. So more than enough mud to go around. Only people like Robert Byron saw through all the guff. By both sides.
The great thing about Foyles was that if you willing to mount an expedition into the darkest recesses of the buildings there was no knowing what treasures you might find. I knew someone who struck gold when he found a large cache of 1960's classical Greek and Latin books (he was studying classics) at late 1960's prices. Would have been 1981. When he went up to pay for them he was expecting the prices to be marked up at the latest prices but the person at the counter just added up all the prices as marked in pounds, shillings and pence and thats what he paid. With the shillings and pence converted to New P. One book I remember he paid 27/6 for a book that was almost 20 quid in the reprint..
No matter how frustrating Foyles was I miss it. Although I miss the old Dillons much more. Once HMV/Waterstones took it over it was never the same again.
"The great thing about Foyles was that if you willing to mount an expedition into the darkest recesses of the buildings there was no knowing what treasures you might find"
Many decades ago, on a school trip to London, I spent an hour or so wandering around the old Foyles and got myself a few books, including one on analytical chemistry which is still in the bookcase behind me and still useful. A bookcase which now holds more technical technical books than the entire engineering department of Foyles when I last visited them - which made me feel very sad indeed, old fossil that I am.
I propose a toast to all the technical bookshops some of us had the privilege of growing up with, which often guided our careers.
For me the biggest loss when it came to academic bookshops was when the Stanford University Bookstore was handed over to management by some bookstore chain making big promises in the late 1990's. Who soon gutted it. Sorry. Rationalized its stock. And my annual purchases of technical books went from multi $K to almost nill in a few years.
The Stanford bookshop before that was the best anywhere in the world at the time. By a wide margin. I actually did a direct comparison with Blackwell's in Oxford in the early 1990's and apart from certain areas of British history and the Classics the Stanford Bookstore had a larger and better chosen selection of titles on pretty much ever subject I looked at. I spent so much time (and so much money ) in the decade before its destruction that I can still wander round the floors of the building (now full of sweatshirts and other tat - very few books) and still tell you where the shelves for every section was. And downstairs was the biggest most comprehensive computer science / electrical engineering section of any bookshop I'd ever seen. And that was back in the days of Stacys, Codys and the Computer Literacy at its biggest. And certainly nowhere in the UK came even close back then. Not even Foyles labyrinthine floor maze.
Well at least I still have the books I bough back then. And the herd of Stanford Cow cuddly toys I acquired over the years which they gave out around Christmas for every $100 spent are still in use all around the world having been given as children's Christmas presents to family and friends. I still have one of the cows somewhere.
There are many reasons to hate Amazon, I lived in Seattle for a while so heard all the stories, but Jeff Bezos's destruction of bookshops is one of his greatest crimes against humanity. Just so people can buy cheap junk at what they think is a bargain price while in fact its little different from how we used to buy stuff mail-order from all over the world decades before Amazon. At pretty much the same relative prices too.
For me the idiosyncratic organisation of the floorplan was less frustrating than the wilful unhelpfulness of the staff*. One time, having managed to locate the sought tome and waiting to pay for it, the person ahead of me in the queue (there was always a queue) asked the cashier where he might find the Peter Norton Programmer's Guide to the IBM PC**. Without hesitation the response came that they didn't have it. I pointed at the shelf behind and observed "it's right there".
-A.
* A friend worked there. Apparently they all hated the place.
** An IT bestseller at the time.
Foyles was one of those places where back in the day I dont think I ever asked any of the staff anything. Not a single question. It always had the vibe of a Soviet department store. In somewhere like Novosibirsk. And on more than one occasion I was able to point a bemused customer in the right direction. There again I have had that happen in several bookshops over the decades. More than once it was only after I had directed the customer to the correct area that they realized I did not actually work there.
Ah, Peter Norton. Who ran that rather "odd" software company in Santa Monica before it was acquired by Symantec. Lets just say he had a reputation in the business at the time. As a "difficult" person. And no one was too surprised when he disappeared without trace from the biz after being bough out. Very much a one trick pony branding exercise.
"Foyled again? Try Dillons."
When I was a kid my mum used to do home help visits to an old lady who was a member of the Foyle clan (far enough away so that she didn't have any roles at the bookshop) and I sometimes went with her (lack of babysitting capability mostly!). Nice old lady - gave me a Victorian card-case (silver and tortoiseshell - modern me cringes at the latter) because I was admiring it in her cabinet.
The Osborne 1 came with a substantial software set.....for around £1500:
- All the Digital Research tools
- Wordstar
- Supercalc
- MS BASIC and CBASIC
For a few quid more you could get dBASE-II.
......when a basic (no software) IBM PC cost twice as much.
Interesting too that much of the first generation of IBM PC software was CP/M application software simply ported from CP/M -- 8080 (or Z80) assembler just translated to 8086 assembler.
Adam Osborne's mantra was "Adequacy Is Enough"...........and the Osborne 1 was both cheap (for its time), and perfectly adaquate as well!
I rememeber when Amstrad took over Sinclair and you could pick up so much old Sinclair kit of "pennies" I bought about 10 Sinclair machines of various types for about £600, drives and disks. I'd just started my first job, living at home and had money to burn!
My parents refused to buy Sinclair micros when i was a kid they'd both bought the original Sinclair odds and sods electronic kits and hated them, so we had to have a "Proper home computer, not that cheap Sinclair shit!", so we had two Dragon computers, then various Amstrads, Amstrad PCs and then Tandem XTs before me and my dad started building our own PCs. My Dad is now 82 and still build his own PCs and does stuff with Arduino, and he still hates Sinclair with a passion! LOL!
The retro community are brilliant at reverse engineering hardware, scanning schematics and preserving software. Stories like this are often not that well known. I've often felt we need to get better at recording details like this now - none of my generation are getting any younger, and they may soon be lost forever. They may seem inconsequential, but they do give context.
There's an excellent episode in the Cautionary Tales podcast, 'laser v parchment', about the BBC's reboot of the Domesday Book. It was distributed as a laserdisc.
Fast forward a few decades and they were virtually unreadable as the required a BBC micro with Phillips videodisc player, a combination as rare as hen's teeth.
A guy took on the task of deciding the custom format (a mix of analogue video plus digital data) and after years of work finally managed to publish it online
Unfortunately he has subsequently died and no one was able to keep the site up and running, so the info is lost for a second time!
Of a mis-spent youth... well actually... a well spent youth bashing away at sinclairs latest and greatest(sinclair's because we could barely afford the 175 quid for a 48K one, even the bbc A was 300 quid and well beyond a lot of people's finances (a model B was 400 quid).... but sir clive coming up with a price , then forcing his staff to build it to that price.
Glad to see a mention of the ROM disassambly, still got that book knocking around somewhere amongst my junk, most of it well beyond my programming abilities until I saw the floating point calculator and figured out how that bit worked.. then the rest kinda fell into place and then never forgave clive for using interupt mode 1 on the Z80 instead of the full on interupt table method of mode 2. made it much harder to do timer circuits/PIO interupt routines.
But a spectrum/QL as a business machine.... was not going to happen
No decent keyboard
No floppy drive/HDD capability
No IBM logo
However the old story about the ZX81 controlling a factory? well a PLC with a 16 bit input port and a 16 bit output port and a serial port to upload the ladder is what a lot of people use.. and a Z80 chip with 2 PIO chips and a Dart chip and a 5v power pack would do that very easily... and that would give you 16 control lines to whatever valves/relays you wanted to fire off in any order.
And in industrial control you want rugged equipment with a long life span(10 yrs plus)
sinclairs latest and greatest(sinclair's because we could barely afford the 175 quid for a 48K one, even the bbc A was 300 quid and well beyond a lot of people's finances (a model B was 400 quid)
My parent's budget stretched to a 16K Spectrum, I had to wait until the following year for the 32K RAM upgrade.
It gave me a career I guess. If the BBC had been the only thing available I'm not sure what I would have ended up doing.
And most importantly I know when to use GO TOs.
Likewise.
A day or two after Christmas, I was showing a "horse racing" game (i.e. pick number 1-3 and see if it matched the one randomly picked by the computer) to my uncle that I had typed in from a magzine. It said either "You won!" or "You lost!" and he said, "There must be a way that you can make it tell you which horse of the three won". He looked at the code, saw the variable, worked out how to print it to the screen and I thought "wow". From then on, I tweaked code, added things to it, and discovered that this wasn't a fancy, unreachable "thing" - I was in complete control of it.
I also learned that writing system software for a paged memory map and no MMU is as much fun as root canal work via nostril. Don’t ask.
But it's still the reality for anyone who's ever worked with a Microchip PIC. It's principal 8-bit competitor these days is the 8051 copies. They are arguably even odder with no less than four address spaces, but a lot more comfortable to work with in practice.
Well the startup menu did serve to cause me much consternation for a few days when I first got a 128. I was upgrading one of my games to take advantage of the extra hardware. No matter what I tried I just couldn't get the memory page to change.
Turned out I was starting it up in 48K mode which locked out the additional hardware features. I think the reason I did that was that 48k BASIC got you a whole keyword per keypress as opposed to a letter per keypress in 128 mode so it was faster to type in the LOAD command that booted up the environment.
ISTR we had a small stub program on microdrive that then ran a loader to bring the rest of the game image over from a PC running a Z80 cross compiler. I probably still have the hardware for that somewhere... I reverse engineered the PC (ISA?) card on wirewrap boards as originals were a bit pricey.
Rupert, this brings back so many foundational memories for me and so many others who begged our parents for a ZX81 which then became the genesis of some great careers. Those early days poking our noses into places that we should not go and making mischief while we were minors and the laws were rather ineffective became crazy stories that I shared with my kids and, hopefully will, with my grandkids. Thanks for the write up, brother!
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
