Collected Works
Laid-back Llamasoft founder Jeff Minter reflects on a decades-long career making tripped-out games
CENTIPEDE
Developer/publisher dK’Tronics Format ZX81 Release 1982
GRIDRUNNER
Developer Llamasoft Publisher HESware, Quicksilva Format Atari 8bit, C64, VIC-20, ZX Spectrum Release 1982
HOVER BOVVER
Developer/publisher Llamasoft Format Atari 8bit , Commodore 64, Intellivision, PC Release 1983
LLAMATRON: 2112
Developer/publisher Llamatron Format Amiga, ST, PC Release 1991
TEMPEST 2000
Developer Llamasoft Publisher Atari, Interplay Format Jaguar, PlayStation, PC, Saturn Release 1994
TEMPEST 3000
Developer Llamasoft Publisher Hasbro Interactive Format Nuon Release 2000
TXK
Developer/publisher Llamasoft Format Vita Release 2014
SPACE GIRAFFE
Developer/publisher Llamasoft Format 360, PC, Xbox One Release 2007
MINOTAUR RESCUE
Developer/publisher Llamasoft Format iOS Release 2011
POLYBIUS
Developer/publisher Llamasoft Format PS4, PSVR Release 2017
Back in 1978, a 16-year-old Jeff Minter wandered into a room at his sixth-form college and found a fellow student sitting in front of a device which looked something like a cross between a television and a calculator. On the screen was a crude mass of pixels the person was controlling. Minter had encountered computers in the past, and had already been exposed to a few arcade games by this point, but when he asked how this particular game had found its way into the college device, the boy’s answer changed his life: “I typed it in,” he replied.
As a result, Minter – who nowadays lives in Wales, where he has a flock of sheep and a variety of other livestock, many of which have made their way into his games – has spent the best part of four decades typing things into computers, carving out a charismatic style that is instantly recognisable. His latest venture, PSVR shooter Polybius, represents something of a zenith for his quest to induce a zen state in anyone who plays his games. But we’ll start 35 years ago, when Minter was busy dreaming up his own version of Atari’s Centipede.
CENTIPEDE Developer/publisher dK’Tronics Format ZX81 Release 1982
“I’d never actually played arcade Centipede. I’d seen it from a distance on the other side of a pub, so I knew what the idea was. But I didn’t realise that the player craft went up and down, because I hadn’t played it. So I just made a
Centipede- style game on the ZX81. It was the first time that I ever thought there might be some degree of fame attached to making games. I remember I was queuing up to get into one of the ZX Microfairs, and there were these two lads in front of me talking about the games they’d been playing the other night. One of them mentioned playing Centipede, and I said, “Do you mean dK’Tronics Centipede? Because I made that.” And he said, “You must be Jeff Minter!”
Looking back on the game now, it’s well ropey. It really is quite rough. But at the time, the reviewers seemed to really like it. They liked the presentation I did with the big centipede that waggled its antennas. It was written in hybrid BASIC and assembler – well, machine code, because you didn’t have an assembler back then. That’s what got me started.
We also did a version for dK’Tronics’ Graphics ROM. It was a little ROM that you put in the ZX81 which gave you access
to loads of alternate character sets. But I actually fell out with Dave Heelas, who ran the company, over that. It was also a funny time in my career because dK’Tronics and I fell out quite a lot in the early days – there were a lot of dodgy arrangements going on, and often I seemed to end up on the wrong end of them.
I designed all the graphics for this ROM – five complete sets, all plotted out on graph paper, before working out the hex [values] and handing it in. But then he didn’t cut me in with anything at all. He was selling that for £30 for several years, so he must have done all right out of it. In the end we managed to get him to cough up £500, very grudgingly. I just assumed that if I was working with him on something that I would get at least some sliver of it. I did do a version of Centipede on that, and a version of Space Invaders, actually – which was the only version of
Space Invaders I’ve ever done. But it was that kind of stuff that prompted us to start Llamasoft.”
GRIDRUNNER Developer Llamasoft Publisher HESware, Quicksilva Format Atari 8bit, C64, VIC-20, ZX Spectrum Release 1982
“Gridrunner is when I really started to realise that I’d have a lot more fun if I started trying to put my own stuff in there, rather than just porting things. There were loads and loads of Centipede games on the VIC-20 at the time, so while I quite fancied doing a Centipede- style game, I didn’t want to just make another one. Also, at that time Atari were just starting to throw their weight around and beginning to hassle people. So I started to add things to it – it had the whole thing with the bombs and the cross lasers. But I took some things away, as well – I didn’t have as many mushrooms. It worked out really nicely. It was hand-assembled on a VIC-20 and made in a week, start to finish. I started on a Monday morning and finished on the Sunday afternoon – I remember Wednesday night was Basingstoke Computer Club, and I took in a half-finished version to show people at the club! They seemed to quite like it. I had some good times at that club and made some good friends there.
The shop was run by Basil de Ferranti, and he was a lovely guy, but not the most technical bloke in the world. One time I was in the shop and I’d written a little program on the VIC-20 that copied the character set down into RAM, but copied it inverted, and then switched it so that basically the computer was working but all the characters were upside down. I called him over and said, “You might have to send this one back – I think they might have plugged the character-generator chip in upside down!” He was at the point of unplugging it and sending it back when I told him I’d just been messing with his head [laughs].
But Gridrunner was probably the best week’s work I ever did – the game ended up doing really well on cartridge in the US, and it’s one of the few times in my career that we actually made some decent money. But money is an annoyance, really – I think the more things are motivated by it, the more shit they tend to be. In a way, I really wish that I could have had one big hit that would give me enough in the background so I wouldn’t have to worry about it and scrabble around trying to get money all the time – it pisses me off no end. It gets in the way of being able to do what you want to do. What I do, I do mostly for the love of it. I’m obviously not aiming for the massmarket, because I find that quite boring. I make games that I want to play, and I hope that there will be enough of a market for them to at least fund me to get to the next one.
I don’t really want that much money, to be honest with you – there’s not a lot I would do with it. I suppose I could buy myself a new car or whatever, but I’ve got one that works at the moment. I can’t really go on massive holidays because I’ve got the flock to look after. I don’t really want many things – I’m not really a thing kind of person. All my favourite toys are the ones which I use for development anyway, and I tend to get given those. I just don’t like the way money changes
“MONEY IS AN ANNOYANCE – THE MORE THINGS ARE MOTIVATED BY IT, THE MORE SHIT THEY TEND TO BE”
people. If I’d made a game like Angry Birds, and ended up with millions and millions of quid, I’d be thinking, ‘I can do whatever the fuck I want for the rest of my life!’ And I’d be coming up with all sorts of mad stuff. But they’re just going, Angry Birds,
Angry Birds, Angry Birds… Why? You’ve got freedom – why just sit in a rut producing the same thing again and again? It’s obviously because they’re more interested in making money than they are in making games. The way it should be is that you concentrate on making something nice, and then hopefully the money will follow. Rather than concentrate on making money and then hoping your game’s all right. I think a lot of places have got it the wrong way round, and I try to keep things what I consider to be the right way around.”
HOVER BOVVER Developer/publisher Llamasoft Format Atari 8bit , Commodore 64, Intellivision, PC Release 1983
“Hover Bovver isn’t in my usual shoot-’em-up genre, and it was co-designed between me and my dad. We went to a computer show in Birmingham, I think, and we were staying in this quite posh bed-and-breakfast farmhouse in Solihull somewhere. The place had lovely grounds, and one morning we were getting ready to go to the show, having a full English breakfast – I remember the scrambled eggs were glorious – looking out of the window, and there was the groundsman mowing the lawn outside. I remember saying to my dad: “We could make a game about that”.
So we started tossing the idea back and forth between the two of us, and all these simple rules – the dog chasing you, borrowing the lawnmover from a neighbour – emerged and sounded like they would come together into a very British, almost Terry & June-style, game. The only thing we couldn’t decide straight away was what we should call it. But in the car on the way to the show, I remember hearing one of the old Qualcast adverts where they kept going on about, “It’s a lot less bovver than a hover,” and so I went, “It’s got to be called Hover Bovver, really”.
We designed it verbally, and when I got home I turned it into actual code. Those ideas ended up being just as fun in the game as I imagined they would be, but I’m particularly fond of this one because my dad did love his games. I remember getting him into games with my first Atari VCS that I brought home from university. He protested that I was too old to play games, so I gave him the joystick and got him to play Space Invaders, and then I didn’t get it back for several hours. From that point on he was completely into games for the rest of his life – he’d always have the latest consoles, and my mum used to say she was a computer-game widow. All the family were quite involved with Llamasoft – they all helped out when we did exhibitions and stuff like that. My mum did a lot of the business side of things because she was far better at that than I am – I’m a rubbish businessperson. But it was particularly nice to have my dad actually co-design a game with me.”
LLAMATRON: 2112 Developer/publisher Llamasoft Format Amiga, Atari ST, PC Release 1991
“I love Robotron, but Llamatron illustrates quite nicely the difference between programming for the arcade and programming for the home. Llamatron looks like Robotron, initially it plays like Robotron, too, but it’s a lot more gentle. Robotron’s job is to kick your arse as quickly as possible, to get you to put another coin in the machine, whereas in
Llamatron I wanted there to be a bit of a journey, and so there are 100 levels and it’s much more lenient. It won’t kill you so quickly, it’s got loads of powerups, and there’s stuff to discover on every level. I mean, I get quite annoyed about stuff like free-to-play, and people will say, “It’s only like an arcade game where you have to put in 10p for each go”. But I’m not into designing that kind of experience – I don’t want to design a game whose main objective is to get you to put money into it constantly. I want to design something that’s a nice, well-rounded whole that you’ll enjoy sitting down to. I’ll make my money, hopefully, when you buy my next game because you liked the first one.
Nobody seemed to be interested in that kind of game at the time, so we
decided to do it as shareware. And this was proper shareware where you give the whole game out and you only ask people to pay if they like it. It was a time when we were running out of money, but we had it distributed on the cover of ST Format and we did really well. But that also caused some controversy: some of the software houses didn’t like the idea of magazines distributing entire games on the cover disk. They felt threatened by that, because they thought people might not go out and spend money on their games. But for me it was a really heartwarming experience – we desperately needed to get some cashflow going. We had people who weren’t only sending £5, but also letters saying how much they appreciated it, and how much they enjoyed the game. We even had people sending us more money than we asked for. It was really pure, and I don’t think I’ve ever seen anything like that since.”
TEMPEST 2000 Developer Llamasoft Publisher Atari, Interplay Format Jaguar, PlayStation, PC, Saturn Release 1994
“I’d always loved Tempest. The first time I saw it in an arcade, I was just entranced by it. It was just so differentlooking, very abstract, and that really fast, shooty gameplay was right up my street. You can see right there the basis for my love of that kind of style that’s gone on throughout the years. But getting to do a version surprised me. I was at a Jaguar development convention and wasn’t really sure what I was going to do, if anything, on the Jaguar. I’d done some work on the Atari ST in the UK, I’d also worked on the cancelled Panther system briefly, and Atari had published a couple of my games and were interested in showing me the Jaguar. Towards the end of the conference they asked if anybody was interested in doing some conversions of Atari arcade titles, and started calling out game names. When they said “Tempest”, I was like, “Fucking hell – yes”. I was a bit scared because I hadn’t done any 3D at that point. Everything I’d done had been sprites and tile maps, and I’m not that good at maths. So I was thinking, ‘Oh, no, I’m going to have to learn some shit if I’m going to do this’. It was a bit of a struggle at first, but after a couple of months I had a very reasonable version of Tempest going. I sent that in, but then I started thinking, ‘How am I going to make this my own?’ So I started filling in the polygons and added the things that shattered. Actually, I got told off by John Mathieson, the guy that designed the Jaguar, for that. I had a few run-ins with him. The effect was actually running something that was supposed to be used the other way around to try to simulate texture mapping under certain conditions, but it proved quite effective my way.
I discovered another thing at that time, which absolutely blew my mind. I sat down and had my morning spliff, and when I finished it I thought, ‘I wonder what happens if I make a sprite be the image of the previous screen?’ So I put that onscreen and started to spin it and got all that feedback, and I was like, “Fucking hell! That’s amazing!” And so I hooked up the joypad to some of the parameters and sat there the whole day going, “Woah – this is great!” That’s lasted me for so many years – I’ve used it in so many different things.
When the Jaguar launched, I got invited out to New York to go to the launch party, and John was there, too, of course. We’d both been drinking, and he
“IT WAS REALLY PURE, AND I DON’T THINK I’VE EVER SEEN ANYTHING LIKE THAT SINCE”
came up to me and said, “I’ve seen Tempest
2000. It really doesn’t use my machine that well at all, and really, for Atari, it’s a makeweight game – nobody thinks that much of it.” And so I was like, “Oh, thanks, John”. But then, of course, when it came out he had to eat his words! We went on to become good friends after that, and we laughed and laughed about it later on. I think we were both a bit tired and emotional at the time [laughs].
TEMPEST 3000 Developer Llamasoft Publisher Hasbro Interactive Format Nuon Release 2000 TXK Developer/publisher Llamasoft Platform Vita Release 2014
“I was very happy with 2000, but there were certain things I didn’t like about it – the framework could get choppy when it’s under load, and these days it looks a bit blocky. I think with TxK I’m satisfied that that’s pretty much the definitive version – as far as I want to go in that direction. And I also had the opportunity to do
Tempest 3000 on VM Labs’ Nuon. That was a fucking challenge, I can tell you. It was a 54Mhz chip, it doesn’t have any hardware assist for graphics, and every pixel has got a significant calculation going on. So it’s effectively doing the same kind of stuff that you do with a shader these days. The chip has four different processing elements, and you had to program them in parallel. I had three and a half when I first started, but gradually more and more system resources got taken up and so I got shoved off one, and shoved off another until there wasn’t a lot left. Each of these CPUs had 4K memory, so you had to page stuff in and out constantly, and it was just a tremendous dance to get anything out of it. When I look at it now, I know that the framerate is too choppy, and the resolution isn’t what it should be, but considering the resources I had to do that in, I think I did bloody well.”
SPACE GIRAFFE Developer/publisher Llamasoft Format 360, PC, Xbox One Release 2007
“Even though it looked Tempesty, there were quite a lot of differences in the gameplay – if you played it like Tempest, then you wouldn’t do very well. In a way that worked against it, though, and I wish I’d made it look less Tempesty because then people might have found it easier to access the new play style. People took a look at the web and thought, “Oh, it’s
Tempest.” So I did myself a disservice there. But that game was intended to be the confluence of my particular style of videogames, and the lighting stuff I did – it was built right on top of the Neon engine which we’d just used for making the Xbox 360 music visualiser.
I had the idea that I wanted to use the environment as part of the difficulty, but some people really objected to that. They thought it was their god-given right to see everything in perfect clarity and not have any psychedelic warping going on because that was against the rules. But to me, there were no rules. Part of the skill of the game was learning how to read the stuff. All the cues were still in there: if you look at level 64, the first time you see it, you just think, ‘How the fuck can anybody see anything in that?’ But when you learn it, you can play that level through without losing a life. A lot of cues were moved out to the audio. Yes, it was experimental, but I’m very proud of it, even though it was sharply divisive and quite a lot of people hated it.
I think it’s remembered a lot more fondly now, though, and people are
“THEY THOUGHT IT WAS THEIR GOD-GIVEN RIGHT TO SEE EVERYTHING IN PERFECT CLARITY”
realising that it’s actually a bit of a classic of its kind. I think it’s far better to make something which causes people to have a strong opinion than it is to create something that’s forgotten years later. So I stand by Space Giraffe, and I still play it to this day. To play it well you almost have to let your defences down and let all that stuff wash over you, and then you find you’ve achieved this kind of zen state. I’m always looking for that in the games I design. I want to get players into this kind of trance state because for me that’s when gaming feels best – when you’re really locked into it and you’re in flow.”
MINOTAUR RESCUE Developer/publisher Llamasoft Format iOS Release 2011
“I really thought there might be a way we could survive doing nice, fun little games on mobile devices. There are so many of them out there that, even if software is being hopelessly devalued by the horrible 69p price point, you’d think there would be enough to sustain them. But unfortunately it didn’t happen because discoverability is awful on mobile. If I’d been able to charge the same price as my VIC-20 games – a fiver – then I could have happily stayed there doing those games even with the numbers I was getting. But no, the moment you start to put the price up above a quid, people think it’s too expensive. I mean, fuck off.
But I’m really proud of the games I made on mobile devices. Each of them was a really nice game, and Minotaur Rescue is a little gem. I was fascinated by the idea of how to get good controls out of a touchscreen device, and I ended up getting that to feel pretty nice. People think that I hate mobile gaming per se, but it’s not that at all – I just hate the fact that the ecosystem is incredibly toxic to developers. I really enjoyed the short-form development – spending a month making one thing, and then moving on to another thing. I mean, I love Polybius but it’s been a long slog and has taken me much longer than I hoped it would. So it was quite a joy doing those little games.
But with each one, I’d release it and hope it would get somewhere, and I’d end up getting 50p back. Then I’d be crushed, and have to build up my enthusiasm again. Then the same thing would happen. All the while I was getting excellent user reviews, and in all the places where you’d think it would be important to get good reviews, like Touch Arcade. But nothing made the slightest bit of difference. After two years of that, I was worn out – I was absolutely emotionally crushed, exhausted, and I was out of money. It actually ended up costing me money to make the games. In a way, that two years was the worst time in my career, I think. I really ended up quite broken up about it.
Then I did TxK on Vita, supposedly a niche system, and it’s so much better than ten games that I’ve done on this device that’s supposedly got billions of users. That’s why I don’t do mobile.”
POLYBIUS Developer/publisher Llamasoft Format PS4, PSVR Release 2017
“There will be a non-VR version of Polybius, but one of the things that I learned with the Oculus Rift version of
TxK is that it just feels really good being in there. It adds to the experience a lot, so I thought it would really nice to put players inside an arcade game. And, of course, the inspiration was this legend of
Polybius which was supposed to be this game which put people in trances and did things to their heads. It just seemed to be the obvious theme for it.
We spent ages at the start building the engine, and that’s a lot of work. The payoff is that we can do things like 2x oversampling and native 120fps, which most people can’t do at the moment. But this time last year we had some of the engine built, but I didn’t have any game at all. I thought, ‘Fucking hell, I need to sit down and get something down, because if I don’t get the game going soon, I’ll run out of money again’. So I put together a demo in the space of about a week and started to play that. I put on some Prodigy and started rushing through the gates, and it all felt really good.
While it looks really hairy, we haven’t had a single bad reaction to the classic virtual-reality motion sickness. I want to take people to trip-out city, but in a really safe way.”