Now that Twitter has effectively become “the Daily Horror app”, I’ve gone back to reading conscientiously, and the book I am currently engrossed by is Electronic Dreams: How 1980s Britain learned to love the computer by Tom Lean. This is one of the few popular history books I’ve read where I can say “I was there”, and the resultant flood of memories has prompted this post that will probably be the nerd equivalent of Jumpers For Goalposts. (But it definitely won’t be about football, so it will be an improvement in one respect at least.)
The Sinclair ZX Spectrum came out in 1982, just as I turned 12. I’d been aware of home computers since at least the appearance of the ZX80 two years earlier, and can recall playing about with the ZX81 and similar microcomputers in stores: “Playtime’s over, son,” rumbled the salesman in Dixons, flipping the power switch as I sat crosslegged in front of their demo VIC-20 about to start my second round of Galaxians. A week previously, he’d been happy enough to give me his full patter on it, but he was clearly now aggrieved that I hadn’t yet nagged a parent into buying one - for the very good reason that they simply dismissed it as too expensive, which it was at £200 (and it needed its own proprietary tape deck). But yes, I wanted one. How badly? I’d already hung around with some of the geekier lads in the year above, playing with the school’s Commodore PET 2001 in the classroom during breaks. One of the Chemistry teachers owned a Sharp microcomputer, which was a rare possession at the time, and he let a select few of us tentatively prod at it during a couple of lunch hours, although we had to endure some formalised teaching of programming from him (while we mainly wanted to make these new wonder machines print silly messages or rude words). Those early encounters had led to borrowing a paperback that covered basic BASIC programming from my local library, which I studied closely. Following on from an existing interest in electronics, this was a natural progression and yet pleasantly diverting: the application of rigorous logic within an entirely conceptual frame. But I still lacked a ready means of trying any of it out.
Feeding the fever
In the meantime, I obsessively devoured the burgeoning hobby press for this growing interest - like most hobbies, debating and comparing the various ways to spend money, even if only in your own mind, was half the fun. The magazines were odd compendia of product news, reviews, tips and tutorial articles at the time, each covering a smorgasbord of micro models and companies ranging from well-known international brands to local independent retailers and one man band software sellers. As no one system had yet gained significant market share, they couldn’t favour particular products so had to write about the entire gallimaufry of competing brands and every quirky little plastic box on sale (and quite a few that, while well-known, weren’t available yet and actually would never appear). Your Computer magazine, an entirely representative example of the milieu (you can find scans of old editions online), ran a series on porting programs between various BASIC dialects, alongside pieces on ZX81 machine code, the month-by-month development of a database program for the BBC Micro and several pages of type-in programs, including reams of hex numbers for the machine code ones. It seems faintly incredible now that people used to sit at their cherished devices tapping in streams of opaque hieroglyphs, and further hours trying fruitlessly to identify the mistakes in them (some of which could be down to not merely a miskeyed entry but actual printing errors, meaning they had almost no hope of ever making it work). One featured program was for a flight simulator which, over the course of a very long evening, my friend, his father and I could not get to work on their borrowed Acorn Atom - but we stuck at it because we really wanted to see that flight sim.
There were multipage promotions from the major makers, extolling the virtues of their particular models, and sincere price promises from the newly minted Thatcherite wide boys hastily opening slick retail premises to shift boxes (although if you didn’t have the readies to hand then of course, “playtime’s over, son”), alongside quarter page ads placed by the archetypal lone bedroom ‘entrepreneurs’ bashing out ten “puzzle games” on a C30 cassette that faithfully promised a world of entertainment from some fairly rudimentary text handling. One company urged the new ZX81 owner to “stop playing boring games” and use their advanced new technology to run a “database filing system” for sheer unadulterated thrills instead. But part of what the magazines had uncovered was the enormous latent demand by ordinary people to own multifunctional devices that they could direct to their own purposes - to do things that only the malleability of software allows (today we have vast app stores and composable web services, and you don’t need to learn how to code). Clearly, nobody in their right mind would want to run a database off cassette-based storage, yet individuals were beginning to do that in preference to tedious or unworkable manual solutions. This was a step beyond previous consumer product booms such as hifi, although it had something in common with the DIY craze that arose in the early seventies.
‘I just had this instinctive feeling that nothing this cool could be useless.’ (Guy Kewney)
Over time, this charming amateurishness gave way to a wide range of titles each targeting both a specific system and a particular readership. The market for ‘thinly aimed at teenage boys’ was especially large; of magazines such as “Load Runner”, “Big K” and “Crash”, only the first named was upfront enough to call itself a computer comic. Those of us who read too many of these publications had a distressing tendency to believe that talking like a computer - dropping words like ‘input’ and ‘output’ into normal conversation - did not somehow beg for derision. Even the prevailing use of BASIC (boo, hiss) in the program listings and tutorials lent an undercurrent of exotic Americanisation and an unfamiliar entrepreneurial spirit with the liberal use of the dollar sign and ruthless categorisation of data as ‘strings’ or ‘integers’.
Yet still the entry price for hopping onboard this craze appeared to start
upwards of two hundred pounds (at least if you wanted to play - or, being
generous, write - games in colour). Until May 1982, that is,
which happily coincided with my twelfth birthday.
A 16K ZX Spectrum at £125 was just about within the realms of
parental indulgence, if I played the usual ploys of “I’ll have it for
birthday and Christmas!” and “I’ll never ask for anything again!”
[Narrator: “Neither of these were true.”] It took seemingly forever to
arrive, though it might actually have been anything from six weeks to
three months (Sinclair had massive shipping delays at first, and I suspect
it also took a month or two initially to reassure my dad that his money
wouldn’t just disappear).
I filled the time covering pages of notepads with my own fledgling BASIC
programs (partly because I still hadn’t grasped the use of
was thus unrolling all my loops in long hand) and then, when the coveted
Spectrum still hadn’t arrived, I resorted to ‘entering’ them on the full
page photo in the Sinclair brochure “to practice my typing”. I was
jonesing for my fix bad. Around this time, I recall buying the
August 1982 issue of Your Computer as “beach reading” (sic) for a day out
at Rhyl, to satiate my gnawing impatience at the continued wait. (Why yes,
I was a pale and slight adolescent forever getting sand kicked in my
While I was waiting for my personal holy grail to arrive, an entire panoply of low priced, ‘fully-featured’ competing microcomputer models flooded on to the market to exploit the demand newly exposed by Sinclair’s latest success. I’m sure that at least a few owners of the Video Genie, Newbrain, Oric-1, Jupiter Ace or Atari 400 were pleased with their purchases, but the companies behind them might as well have farted into a paper bag for all the lasting impact they would have on the home computing scene in the UK. By the end of the year, it was clear that there were only three main players in town - Sinclair, Commodore and Acorn - and the nascent software industry necessarily coalesced around them. (Jupiter devoted most of their adverts to justifying the novel use of FORTH in the Ace, as if writing code rather than running it was the main selling point. Nevertheless, the impression persisted that if your parents bought you one of these esoteric outliers, they probably also ate muesli and banned you from watching ITV.)
Home computing madness
Eventually, my ZX Spectrum arrived. I can’t recall exactly when it came or how it felt, but I suspect no first time heroin shooter ever matched the rush of that day. I wrote my little programs and I played the sample “Breakout” game supplied by Sinclair on their demo cassette endlessly. (I once wrote an entire database program in a morning - no idea why other than because I could - and then couldn’t save it because the tape recorder was faulty. And my mother agonised and then decided not to tell me about the brand new cassette deck hidden upstairs for Xmas.) A teacher and fellow enthusiast at school arranged a trip to a regional computer show for some of us, and for the princely fiver I had scraped together (or possibly cadged), I bought my first game, Artic’s Gobbleman, an early Pacman clone that at least hadn’t been written in BASIC and was thus a ‘proper’ program. But in slowly collecting more and more games, I rapidly discovered the limitations of my basic Spectrum.
‘When you’ve got a computer and you want to make something, you can do it. You’ve got everything there. And I loved the fact that I could have an idea and I could immediately put it into action.’
Tom Lean makes the excellent point that computing offered a hobby which was largely self-contained and theoretically less prone to ‘acquisition syndrome’ than most others. Once you had the computer, almost everything you could want to do with it was immediately possible, only requiring your imagination, time and intellectual nous. I was a failed veteran of both model railwaying and DIY electronics already (shockingly predictable, I know), pursuits that each demanded endless supplies of ‘stuff you don’t already own’ before you could achieve much.1 But here at last was a gizmo that was complete in and of itself, that only needed the application of logic and creative insight to conjure anything I could think of into existence. Oh, and some extra RAM, because it quickly turned out there weren’t many games that worked in only 16K. Oh, and a joystick. And a joystick interface (because the Spectrum hardware was really minimal). Actually, an endless succession of joysticks (surely there must be one that will let me win at games?!). And a Microdrive now they’re available. And maybe a proper keyboard. Or a printer. And an Assembler program, because I’m bored with BASIC after writing half a dozen ropey games and I need to learn Z80 machine code to be a games developer. Although actually, Z80 machine code is reeelly hard, but it’ll be much easier on an entirely new computer with better graphics and sound, Dad. (I’ve no idea how I ever perpetrated this last outrageous con to the tune of a £200 Commodore 64 - rising Eighties middle class affluence innit - but I went on to repeat it twice more, although at least I used my own wages and only fooled myself on those later occasions.)
A technical digression
In today’s era of compiled high level languages, Object Orientation, IDEs, frameworks, toolkits and modern 64 bit platforms, we’ve largely forgotten how compromised and arcane early microcomputer architectures were. Clever hacks were creatively employed, and I’m sure for solid reasons, to produce the capabilities that the market demanded from low end hardware, but none of them made life easy for assembly language programmers looking to hit the bare metal to turn out arcade quality games. The ZX Spectrum’s display memory map was laid out in three sections of eight character rows, each comprising the top row of pixels for each successive row of characters in that section of screen, then the next row of pixels for each character row, etc. (This was quite apart from the colour layout, which was a sequential array of character squares superimposed on the hi-res screen - hence the famous ‘attribute clash’ when objects of differing colours met onscreen, since the character space they shared could only have one assigned foreground colour.) So that’s a screen split into thirds, each of which is split again into segmented rows of characters. Imagine trying to animate a game object moving vertically through that space, one pixel row at a time - in Z80 assembly! In fact, there’s some nifty hexadecimal maths you can use to do this, but there’s no doubt that the various code explanations are frankly convoluted on first glance - as well as subsequent horrified glances. (I tried to find a link that illustrated the screen layout to explain it better, but drew a blank. Not surprised really. The best way to understand it is to observe a loading screen in any emulated Spectrum game.)
Screw this, thinks the aspiring games maker. I’ll get a C64 instead, they’re meant for gaming. And it’s true, the Commodore 64 had hardware sprites and a proper sound chip, and you could even program them from BASIC by POKEing values into memory addresses - in fact, that was the only way to program them in any language. C64 programmers usually graduated to assembly language in a very short space of time, because Commodore BASIC was so awful as to be little different from it. But hold on, the C64 uses banks of memory and so-called ‘shadow RAM’ - its 20K of ROM is actually overlaid in 4K blocks on the 64K RAM. This is a plus in one sense, in that you can copy the ROM to the underlying RAM, then switch out that bank and modify the copy or use the shadow RAM for something else entirely, but it also adds the complications of managing all these disconnected bits of memory, not all of which is accessible at once by the various custom chips. Add to which, it turns out that 6502 assembly is even more primitive than Z80.2 Good luck, kid! Oh and by the way, the Internet isn’t generally available yet. You can’t google the answers. You’ll have to buy these huge thick programmers reference guides and figure it out. (Until machine-specific coding books appeared, I regularly saw recommendations to use generic Z80 or 6502 programmer texts, such as those by Rodney Zaks, which is a bit like telling someone to learn how to drive by studying Haynes manuals.)
Needless to say, I ended up owning Toni Baker’s classic Mastering Machine Code on the ZX Spectrum, the Commodore 64 Programmer’s Reference Guide and at least two thick system programming references for the Amiga. (You can find scans of all of these online now, btw, in case you run out of insomnia meds.) I knew the lore of programming these systems at a low level almost off by heart by dint of hours spent poring over such tomes, but it was entirely theoretical - absolutely none of it ever came to be applied on a practical level. I didn’t even get round to buying and then never using an assembler program for the C64. Simply reading about it became sufficient. Such is the curse of being bright but staggeringly lazy. (This is the point where most functioning people can write, “However, at about this time I discovered girls instead.” Yeah, that didn’t happen either.)
How then, did all this useless esoterica and unapplied erudition benefit us? Simply, unlike the previous generation, we were the first to generally understand that computers were not the all-powerful, all-knowing, ‘intelligent’ entities of popular sci-fi - that they were in fact dumb as rocks. But if you could command the rocks into life and make them perform the most basic tasks, they would be able to do them much faster and more accurately than any person. And if you could break any complex task down into its most elemental steps, it could do that too. The earliest home computer games were based entirely on the simplest possible manipulation of screen characters - moving single ASCII symbols around one block at a time in response to either player command or some inner brutal logic:
410 IF player_position>alien_position THEN 420 LET alien_position = alien_position-1 430 END IF
Within two or three years, game sprites were obeying the laws of mass, inertia, gravity and thermodynamics - simulated entirely by self-taught teenagers working at the lowest level of CPU programming (though I wasn’t one of them, obviously). I’m not going to point at modern teens and sneer about vlogging by comparison, as I think that also demands a similar level of skill in psychology, presentation and narrative, but this was probably the only time in history at which ‘deep science’ subjects were broadly comprehended by a significant part of a generation. This stuff was quite literally rocket science.
Moreover, for me personally, my maths finally started to improve. Content to bump along with the minimal adequate effort in a middle set, I suddenly decided maths might actually be useful in future and began to pay attention. (I learnt trig from the ZX Spectrum manual before the class covered it.) Miss B, bless her soul, pushed my lazy arse with extra tuition to get me moved up to the top set and eventually, O and A Level grades. (And yes, I work in IT now, although the gulf between what I did on my ZX Spectrum and what I do now is incomprehensibly vast.)
As the magazines began to target specific platforms, I lapped up all their technical articles and avidly read the installments of their “programmers diary” features (a genre surely rivalled only by “Journal of a Chartered Accountant” and “The Librarian’s Day” for exotic thrills), such as Andrew Braybrook’s wonderful Birth of a Paradroid in the otherwise risible Zzap!64. The newly launched Your Spectrum mag set out their stall early on with a series of tutorials by Toni Baker on writing machine code for the Spectrum; prior to this, we’d had to endure the excruciatingly worthy Sinclair User with their ‘User of the month’ feature, about people who’d use their ZX printer to print their shopping lists.3 (Worse, they often put the user on their cover, which was frequently embarrassing when classmates found it amusing to identify me with the person in question. The toddler was bad enough, but the stick I got for the Morris Dancer one - the resemblance was plain, apart from the costume - was mortifying. But then I was naïve enough to be seen with it openly in school.)
“People could program any of the new machines, but the expectation was that typical users would just use them as software players.”
The Amiga, when it eventually became generally available, was potentially easier to develop software on, as it had custom chips that could perform a lot of the work independently, a proper 16 bit architecture and a well-developed set of OS libraries to handle the basic tasks - but by that point, the complexity and expectations of the typical game were almost beyond what one person could hope to satisfy. Although this generation of hardware would be the peak of the home computing market, it also marked the slow death of the amateur ethos that had lingered from the early days. From here on, the market split into gaming consoles and desktop PCs, both of which were basically appliances for running purchased software. (Those who still enjoyed typing stuff in were by now discovering UNIX at university, which would lead them to Linux so they could run it at home.) In the face of the new 16 bit machines, the makers of the older 8 bit computers attempted to launch enhanced versions - the Commodore 128, Spectrum+, etc - but found themselves hampered by the petrified architecture of their ancestors and the need to maintain compatibility with the existing software base. This was before any thought of “graceful degradation” or API versioning (or, indeed, APIs), and so the legacy system was usually bolted on as a ‘compatibility mode’ in a way that even Dr Frankenstein would think inelegant.
Popularity, or not
To read Lean’s history, one might think that home computing was a pervasive craze among all youngsters in the early 1980s. It was significantly popular - mainly for the opportunity to play arcade games and the easy way one could copy and swap those games with others - but to think this held true across the entire playground would be hugely mistaken. In my experience, the kids who were most obsessive about these new devices were the ones who would have have been obsessed with any highly technical, esoteric and largely insular hobby. Those users who were actually doing any serious programming were a select few; those working commercially even fewer; and of those, probably only a handful writing code to the level of the top games. I don’t recall much interest in programming amongst my peers, but then others sharing that interest in my year may also have been similarly introverted and quiet, and if so, they were probably wiser than me in keeping it to themselves (whereas I might as well have been wearing a “Geek, kick me” t-shirt). A little later, when the school computer room opened, offering a new source of breaktime respite to those who had previously concealed themselves in the library, I was the only member of my year regularly in attendance. There were some younger boys, who seemed most interested in provoking irritation, and a group of older know-alls who would pronounce with authority on the subject, some of whom were even mature enough to talk to girls with passable confidence, which to me was akin to having a superpower. Not that we saw many girls in the computer room during breaks - if it hadn’t been for formal lessons, almost none of them would ever have got near a BBC Micro. (The lessons weren’t available to the upper years, probably because the school didn’t offer O level Computer Studies and nobody could be bothered to integrate IT into their traditional exam curriculae.)
It was through this lab that I first encountered computer networking, which was initially seen within the industry as a way to share expensive printers and disk drives across several machines rather than to allow user communication. Once again, I hit the system programming guides to see if it was possible to hack into remote BBCs using Acorn’s ‘Econet’ protocol and make them misbehave, but it seemed it was insufficiently vulnerable to allow that without collusion. We went back to distracting Mr M., who was nominally ‘in charge’ of the room, so we could liberate the floppy disks holding his games collection (a session on Elite being the main bribe used to get kids to complete their computer work). The worst that could happen was being identified by a teacher as ‘someone good with computers’ and therefore liable to being pressed into service on whatever IT project they had in mind, most of which were viewed as cringeworthy and twee by those of us who found ourselves volunteered. It may seem strange and hypocritical that a bunch of computer geeks could develop any concept of what was ‘uncool’, but there were definite lines drawn within our little sect. Publishing a school newsletter: yuk. Hacking the caveman game FRAK! so that he uttered a similar but different four-letter word: cool.
Weirdly though, a shared interest in computer games, even though I wasn’t particularly good at or even especially keen on playing them, finally kickstarted some sort of social life outside of school. Friends would come over for late evening sessions of Revenge of the Mutant Camels or, even better, 2-up Double Dragon. (Jetpac, in which the aliens exploded with short farty noises, was a particular favourite when we discovered that hammering the pause key during an explosion could draw this out into a prolonged raspberry, to great hilarity.) This would eventually lead in sixth form, when we were old enough to get away with underage drinking, to nights out at the pub instead. (It was at this point that I discovered the undemanding joys of ‘having a laugh with your mates’, and computers finally began to take a healthier backseat to everything else in adolescence - except girls. Plus ça change.)
My last ‘home’ computer was an Amiga A1200, with an internal hard drive, bought with the seeming riches earned from my first post-graduation job as the closest thing I could get to the UNIX workstations I used at work - you could even download a version of C-shell for it (although it wouldn’t run Minix due to the lack of a Memory Management Unit in the pared-down CPU used, and believe me I tried). Unfortunately, the hard drive - an inconceivably expansive 80MB Western Digital model - had an infuriating habit of locking up when it became warm, making it very frustrating to use. There was still no widely available Internet access outside of academia (not that the Amiga supported TCP/IP either, and even on Windows it was a dreadful hack), and I still didn’t find the time to do any proper programming. When I moved on to a non-academic job, I bought a cheap Macintosh and I don’t now recall what happened to that Amiga.
You can always go back
One of the most pleasant surprises of writing this post was the discovery that it’s (almost) all still around. There are images of all the old games across various sites that you can download to run under faithful emulation on the platform of your choice. (Manic Miner on a tiny smartphone screen? Seems crazy but you can do it. Even Gobbleman is out there.) There are native cloned versions of the most popular ones if you prefer that. There are lots of scanned magazines going back to the earliest days of the hobby, and PDF scans of complete books. There are even people still writing games for these systems today. I suppose working in a modern development environment must feel a bit like skimming the outermost atmosphere of a particularly large, murky planet with an unfathomable ecology beneath, whereas an 8 bit computer is like a decently sized asteroid whose entire structure you can understand inside and out within a reasonable time. At this degree of enlargement, one realises fundamentally that everything a computer does only ever involves, at the most basic level, adding small numbers together.
Tom Lean’s laudable book brought all this back to me. I pulled up a few emulators and tried out the old games again. And I’m still useless at them, and I still can’t be bothered to persist with any but the very easiest for more than ten minutes. I read a few modern blog posts on writing machine code games for these ancient platforms - and quickly realised that I still have neither the time nor the inclination to go very far down that rabbit hole. I even finally tried a bit of online 6502 assembly programming. But what an agreeable Proustian rush it’s been to look back on those pioneer days. Thanks, Tom.
(All highlighted quotes taken from the book.)
Top Five Games
- Revenge of the Mutant Camels (C64): Although I was a huge fan of Jeff Minter - whose role in shepherding teenage boys to their destiny of becoming Pink Floyd fans should not be underestimated - this was the only Llamasoft game I didn’t find too complex to master. Press fire button, direct bullets, destroy weird alien attack waves (or leap gingerly over exploding sheep).
- Paradroid (C64): A beautifully simple concept but with enough going on to keep you playing, and probably the most elegantly designed game I ever saw.
- Scuba Dive (Spectrum): Beautiful marine animation, plenty to see and a slightly gentler pace, involving a lot of carefully timed underwater maneuvring.
- JetPac (Spectrum): Once the control method is mastered, the straightforward gameplay of this early Ultimate release made it far more enjoyable than anything they did later (including the sequel, Lunar Jetman, which by stark contrast was impossibly hard).
- Double Dragon (Amiga): Laughably poor beat ‘em up with highly silly sound samples turned into hilarious fun with the addition of a second player and several piss-taking spectators. Even better, you could - accidentally or intentionally - twat the other player’s character with a baseball bat.
Honorable mentions: Impossible Mission, Beam Rider, Elite.
- Extract from the book, on how the Spectrum inspired creative games.
- Spectrum vs. C64: let’s settle this - you all smell.
I’ve still yet to see the fabled ‘pattern makers dowels’ referenced in every guide to joining model railway baseboards together, and I carted two such boards around for several years, fated never to be linked in the ongoing absence of these items and my apparent inability to consider alternatives.↩
There’s nothing like LDIR on the 6502, for example, which can shift 64K chunks of data in one instruction.↩
…He said sneeringly. Actually, I just dug out the article I was thinking of. It’s from p.66 of Issue 16 of Sinclair User from July 1983: an interview with a lady called Mrs Celia Sims who has a 1K ZX81 for her and her two sons. Although the journalist appears to be forcing a story out of not-very-much, on rereading I find Mrs Sims to be one of the most progressive and insightful computer users I’ve encountered, and a rare voice in computing both then and now. But while SU’s selected users showed an overlooked diversity within the hobby, they were typically also middle class parents and carers keen to give their kids a further leg up from their already privileged positions.↩