The "Other" Lord British
Interview with Jez San, OBE
Conducted by Syd Bolton
Jez San may not be a name you instantly recognize, but you’ve probably heard about the games he has been involved with. Argonaut Software and its teams of engineers are also people who you probably have never heard of, and yet, a good number of you have enjoyed technology designed by them. Software and hardware design can often be a thankless job. It’s never been more evident than at a company like Apple. Steve Jobs is the one everyone recognizes, while Steve Wozniak worked away at designing the hardware that gave the company its start in the business.
you’ve actually played on a Commodore Amiga then you might recognize Jez’s
name. He created an incredible game called StarGlider that brought three
dimensional (3D) effects to a computer long before 3D effects were available in
hardware. The game Elite might have ushered in the era of a 3D space flight
simulator, but StarGlider brought the visuals and audio experience to a near
Jez is still in the business. He’s received recognition for his contribution
to the computer game industry by being inducted into the Order of the
you’ve been in this business for a long time. I understand that you used to
program in assembly language (direct machine code) for some of the older
computers. Do you miss that these
days? What was the last game you
contributed actual code to?
Yes, I definitely miss not being a programmer anymore. Especially in assembly language, which was my favourite (the most enjoyable, the most fundamental, the
most difficult, yet the most rewarding of all the computer languages). My last project was an assembler called ArgAsm (one of the cooler names I thought up).
An Assembler is an application that converts assembly source code into object code – that is a stepping stone to producing executable code. It’s like what a C compiler does for C
source code. I wrote ArgAsm because games projects were taking longer and longer to assemble, and so you’d write a piece of code and have to wait many minutes to see your change running in the actual game. I felt that there was a way to make an Assembler run much faster.
I used game programmer technology (optimizations) that are rarely used for serious applications. Things like Global register optimizing, look unrolling, hash and lookup tables… and lots of other game programming tricks to make the assembler tens of times faster than competing software applications. On a standard Amiga A500 (7.16 MHz CPU), ArgAsm was hitting a peak of a million lines per minute. The best that Devpak could do was about 40,000 (and that was way faster than the others). Of course, peak times are nothing, but in reality, the average game was assembled in 100-200,000 lines per minute instead of 10-20,000…
Ultimately, it turned my development cycle down from minutes to seconds , which is great when you're tuning a game to make it the best it can be. The cycle becomes: Tweak code, Assemble, Look at Game,… Tweak more Code, Assemble, Look at Game.. etc… the best games are those that the developers have had plenty of opportunity to tweak them, and the faster the development tools can get you from making a change to seeing the change in the game, the better the chance of making tweaks that improve the game.
The last actual game that my code was in was probably Birds of Prey, or Days of Thunder. But the last game I was directly involved in was StarGlider-2… all the others had vestiges of my code until it eventually got rewritten by even smarter people with even better code optimization skills than me (and I was damn good, but some of the programmers we hired at Argonaut were “Gods” of game optimization).
Argonaut, you guys developed the game “Malice” which was once considered a
possible big hit for the Xbox. In the end, it was released later than expected
and came out on the PS2 as well. The
critics gave it less than stellar reviews, but a quick check of many fan
generated reviews (mine included) concluded otherwise. Although the game
wasn’t perfect, do you feel that the game was judged unfairly because of the
Partly that, but also partly because the game was passed around from publisher to publisher.. and was deeply affected by being tossed in one direction then another and another. First, it was a Microsoft launch title... then a Sierra one... then a Vivendi one (as Vivendi bought Sierra).. each time the “master” changed, so did the game direction. It was wearing the team out, all those constant changes. If it had stayed with Microsoft and come out near Xbox launch, it would’ve been the best game of its genre at that time. The delays meant others caught up in terms of look and feel, so by the time it finally shipped, it wasn’t anywhere near as big a leap as it was intended to be. Showing it off so early at the Xbox launch gave others a big clue how to catch up. Malice had way the best tech around, at the time--but not by the time it shipped.
Argonaut worked closely with Nintendo during the early years of the NES and SNES.
You were fundamental in developing the Super FX Chip for the SNES, which was
used in Star Fox. How did the need
for that chip arise? Would it have
happened if Star Fox didn’t happen?
was in a room with Miyamoto-san and Izushi-san and some other Nintendo people
like Tony Harman and I showed them a prototype of StarFox that we had running,
initially on the NES – codenamed “NesGlider”, and then a few weeks later
on a prototype SNES. I told them
that this is as good as it’s going to get unless they let us design some
hardware to make the SNES better at 3D. Amazingly,
even though I had never done any hardware before, they said YES, and gave me a
million bucks to make it happen. I
hired a team of expert chip designers from Cambridge – some of whom had worked
for Sinclair Research in past lives, and the Argonaut 3D game/tech engineers
worked together to make the first ever 3D graphics accelerator and one of the
first ever RISC Microprocessors. And
since the SNES hardware wasn’t documented very well, we reverse engineered the
machine and built our chip to work directly with it (either inside the SNES, as
was originally intended – or, outside the SNES – on a cartridge).
The Super FX chip was originally codenamed (by us) the “MARIO chip”.
Mathematical, Argonaut, Rotation & I/O chip.
It was Nintendo who renamed it the Super FX chip.
Nintendo and Argonaut had a joint venture company called A/N Software
Inc, based in Nintendo of America’s HQ, to commercialize the chip.
It was used in StarFox, StuntRace FX, StarFox 2 (which never shipped),
I’ve always had this theory that Videogame Technology is extremely competitive technology when measured against non-Videogame Technology. For example, videogame programmers are probably The Best Programmers. Period. 3D algorithms invented for videogames end up being used for other uses because they’re so good – like the movie industry. Similarly, microprocessor and graphics acceleration technology that we literally invented at Argonaut for Nintendo was world class 3D and microprocessor technology in the “computer” business at the time. Back then, in the early 90’s, the Super FX chip was the world’s best selling RISC microprocessor-- outselling the ARM and MIPS chips by millions in those years. It was only when the PlayStation came out in the mid 90’s that we got beaten on RISC sales. And also, then ARM chips were adopted by the GSM phone market, and sold in their gazillions inside every phone. But we were innovators and pioneers. The entire 3D acceleration market that NVidia and ATI now dominate, Argonaut was there first and we’ve got the patents to prove it. Nowadays, Nintendo does. But they’ve got my name on it. Do an EDGAR search for Jeremy San if you want to see them.
After we built the Super FX chip for Nintendo, we went on to design a chip for Philips for a videogame machine that never came out (codenamed GreenPiece aka CD-I 2). We also did one for Apple – also a videogame machine that never came out (codenamed VeggieMagic) – and one for Hasbro – for a Virtual Reality game machine that never came out (codenamed MatriArc). Eventually we got a bit fed up and tired of using our technology to design 3D and microprocessor chips for other people – with amazing horsepower in little inexpensive chips that might never come out - so we designed our own chip – ARC – Argonaut Risc Core – and spun off a company that sold its design non-exclusively to many many companies. This company still exists today, but is no longer connected with the videogame industry. Yet again, proof that videogame technology is the best technology, period. http://www.arc.com
now own PKR.COM, an online Poker Game site. How did that come about and tell me
about some of the technology behind the site?
I've always enjoyed playing poker; both in real life and also online. I’ve played online for years. My favourite sites were PartyPoker and Paradise Poker. But I’ve played on many others as well. It always used to amaze me how popular these sites were; where thousands of people at any one time would play poker against each other – for money – and yet the software was a bit dull, flat and lifeless. I thought we could do a lot better - yet again, by using Videogame Technology.
I had started to think about this while at Argonaut – but alas, after 22 years in the business – Argonaut had grown to such a size (300 people) that each month its salary bill was $2 million! If a few game contracts got deferred and you don’t keep signing new business then you hit a brick wall. We tried to downsize, but it wasn’t fast enough and the company died. So, after getting over the initial shock of losing “my baby”, I made a few calls and started a new company – initially called Crunchy Frog Ltd. We eventually renamed it PKR Technologies Ltd – and hired 10 of the best game development people from Argonaut for the new company. The goal is to use our skills and combine that with Game Technology to make the best online poker anyone’s ever done. My theory was that it’s a lot easier for game development people to learn about the Online Gambling business, than it is for Online Gambling companies to learn how to make good Game developments. I was right. To this day, PKR.com has no competition and is the only one (of thousands of poker sites) with a 3D realistic online poker game. Literally we built everything from scratch – we built the servers – the payment processing systems – the customer services – and of course, the game client application, and we keep on going. Our team hasn’t stopped, even though the game released almost a year ago, we’ve been improving it each day and we keep issuing updates every month or two. PKR’s game engine downloads new updates automatically so every player gets to play the latest version at any moment. In fact, our new release that goes live later this month is a little bit of “Steam…Lite” in that it downloads the game updates while the player is sleeping so it’s ready for them to play when they wake up so they don’t have to wait around while it updates. We’re very proud of the game we’ve built and how much more enjoyable and immersive it is than any other online poker game. Not only that, but the web site is stickier too-- with a poker-centric social-networking site in there. Our PKR players love what we’ve done… and in less than a year, we’ve signed up almost a million poker players to the site. And the signup rate is accelerating!
been in this industry for 25 years. Looking back, what would you say is your
personal favourite game from Argonaut’s library?
What about your favourite classic game from other developer’s?
my favourite games were the ones I made myself.
StarGlider and StarGlider-2
and a little bit of AfterBurner. And
also key milestone games, like StarFox – because they were innovative, not
just in development and execution but also in business relationship.
For StarFox, we taught Nintendo how to make 3D games.. and we worked so
closely with them that we had staff from Argonaut
Argonaut ran into financial difficulties, and eventually closed its doors, you
were able to take control of two subsidiaries: Morpheme and Just Add Monsters.
Morpheme makes mobile games, and Just Add Monsters released “Kung Fu
Chaos” before being rebranded as Ninja Theory.
You’ve now got “Heavenly Sword”, one of the most anticipated PS3
titles, just about ready for release. Tell us about it.
I didn’t take control, as such. I provided funds and help to the management teams of those studios to buy themselves out of their awful predicament. Both Morpheme and Ninja Theory will tell you I haven’t set foot in their offices since that MBO (Management Buy Out). All I did was help keep the wolf from the door – to save them from dying at the same time as Argonaut died. Unfortunately, Morpheme since found it rough going and sold out to Eidos not long ago. The sale was more or less at cost price, so no one got rich. At least the studio continues and I wish them all the best. As for Ninja Theory – I have faith that they’ll do well. They’re a good bunch and have been working for 2 years while at Argonaut, and post Argo, for a further 3 years on an amazing game that finally ships next month. I can’t wait.
has your role been in the development of Heavenly Sword?
My only role in the development of Heavenly Sword was of recognizing, and identifying talent – Nina, Tameem and Mike – when they were just three people.. and of backing them with significant funds, while at Argonaut – and also, after Argo died – of continuing to back them personally – until such time as they were able to close their own publishing deal – which I always knew they would – albeit it took a long time. Such an ambitious game was very risky for any publisher to take on. I provide a sounding board and non-executive advice to the company, but the Ninjas run themselves. They work closely with Sony and are completely independent. As I said, I haven’t even visited their offices, although we email and speak quite regularly. I should change that soon… I’ll go visit them this summer I think! Maybe to celebrate the game’s release, after such a long journey.
you compare “Heavenly Sword” to “Starglider” (your first commercially
successful game) how would you sum up the changes in the technology? The
industry as a whole?
Well, StarGlider was written in my bedroom and cost the publisher less than $20,000. Heavenly Sword was developed by a massive team of over 100 highly talented people – over quite a few years – and probably cost the best part of $20 million! But that’s progress! Seriously.. there’s only really superficial comparisons. StarGlider was one of the first 3D games – ever – and as such had pride of place and little competition. It was also a relatively simple, yet enjoyable game. Heavenly Sword is extremely rich - and complex – in the tech that makes it look so beautiful, and in the way that it’s controlled, played and enjoyed. And it has had an army of talented people working on it for long periods of time – more like the way a movie is made. In fact, many of the “creatives” on the project are exactly that – movie people using movie techniques. And the $20 million is “on the screen” as they like to say. When you play the game, you appreciate the work that’s gone into it. It’s one of the most beautiful games I’ve ever seen. I think this will probably be true for many people who play it, come September or the demo that’s on the PlayStation Store.
no doubt heard that “Heavenly Sword” is often called “Goddess of War”,
comparing it to the successful “God of War”. Do you think this is a fair
comparison or are you struggling to separate the identity of this new franchise?
Heavenly Sword was started before God of War came out. You can see the early demo videos for proof that the Heavenly Sword game that’s shipping now – looks identical (or better) than the original demos of 4-5 years ago. It’s probably not a coincidence that two teams of people thought they could do a fighting game as good as any fighter – against multiple opponents - and combined with other gaming aspects to make a deeper, more fulfilling and enjoyable game.
Given that they’re both excellent fighting adventure games – then yes – they’re similar in genre. There’s another similarity of sorts that isn’t as well known. Both games were built – at least in part - by EX-Argonauts. Not many… but there’s a few ex-Argonauts in senior positions on the GoW team – and quite a few ex-Argonauts at Ninja Theory. I sometimes think that Argonaut was one of the prime training grounds in the world for game development – and the alumni of Argonaut went on to produce some amazing Games… for someone else!
So there you have it. While you might not know who Jez and all those Argonauter’s are – you’ve probably heard of or even used their technology. It just goes to show you that the genius within is often the strong, silent type. I think we’ll find Jez’s belief in the Heavenly Sword team was well founded – and that’s just a little bit more than a Ninja Theory.
Bolton has been collecting game consoles for over 20 years and has dozens of
systems and over 5,000 games. He even owns an original copy of Star Glider. Do
you? Tell us about it [here] if you do.
All articles ©2000 - 2006 The Armchair Empire.
All game and anime imagery is the property of their respective owners.