Interview: Alexander G. M. Smith (2025-07-28)
Alexander G. M. Smith speaks with Alexander Rojas in this phone call interview for Sega Retro.
by Alexander Rojas with Alexander G. M. Smith
Sega Retro · July 28, 2025
This Sega Retro interview with Alexander G. M. Smith was conducted by Alexander Rojas via phone call.
AR:
Hi, Mr. Smith, this is Alex. How are you today?
AGMS:
Hi, Alex. It's Alex here.
AR:
[laughs]
AGMS:
Nice warm day here, 90 degrees Fahrenheit. It's sunny.
AR:
Yeah, us too. It's pretty hot out here in Vegas.
AGMS:
Well, it's always hot there.
AR:
Yeah, exactly. I wanted to say thank you so much for taking my call today. It's just very nice of you to give me your time like that.
AGMS:
Yeah, I don't know if I can tell you that much, so it might not take that long, but we'll see.
AR:
[laughs] Yeah, and whatever you can remember, no worries if you can't remember- just to document what you can and get it written down. Again, I do appreciate all that.
AGMS:
Okay! Oh, by the way, you sent over the David Javelosa interview. Interestingly, there's an Artech reference there to how Artech got started with Sega. Artech is the games company I was working for here in Canada.
AR:
Yeah, I was going to actually… we can go ahead and start with that. How did you first get started with Artech?
AGMS:
How did I get started? Well… I have habit of seeing people doing interesting stuff and sort of following them around. I saw a guy who's at the local Amiga computer club meetings, and he's doing interesting stuff. David Buck doing ray tracing, DKB trace, at the time. He was a student at Carleton University. "It sounded like you're doing cool stuff there at Carleton." OK, so I went there and got a degree in computer science.
AGMS:
Then I met another guy called Steve Tibbett, who's in Ottawa. He does a lot of Amiga stuff, like VirusX is a big thing. He's an amazing programmer. He was working at Artech Studios doing computer games, sort of Amiga versions of stuff. I said, "Hey, maybe I apply there." So in 1991, December, I got a job there.
AR:
'91 December, gotcha.
AGMS:
Yeah, and it's kind of neat this. I worked for a week, then I had two weeks of vacation.
AR:
Oh, right off the bat?
AGMS:
Yep. So in '91, the work really started. I was doing mostly C coding then, PC games initially. I was looking at a simulator for the B-1 Lancer bomber. Just B1-B, I think. Because they'd been doing a lot of other games for Three-Sixty [Three-Sixty Pacific], which is a computer games company. Megafortress was the big one. They had some sort of Tom Clancy tie-ins back then, before Tom Clancy went bigger. I actually went out the air show and saw the B1-B and made up some sort of simulators for various… taking damage. You hit the fuel tank, you lose speed, this sort of thing. They had other games like that, based on DOS-based games.
AGMS:
I started work on December 16, 1991. There was a week of training, learning the company's code library, and playing around with VGA routines. More serious programming on Megafortress II and not just reading up on the B1 Lancer bomber docs started in January 1992. Though I got pulled into the Theatre of War project with frequent side jobs in Patriot. I guess some other company got the Tom Clancy rights. In that summer of 1992 I started learning C++, rather than plain C. In October 1992, the first Sega mentions (for Light Harp), calling up Cross Products UK with a question about their emulator.
AGMS:
There's a few DOS games. That one didn't actually get finished. Then we went to Theater of War, which was kind of a predecessor of real-time strategy games. Chess, but with extensions, like nuclear bombs. It was actually multiplayer over the telephone lines with CompuServe and other places. We had the user interface wrong. It could have been a big hit if you could control groups of units. We thought it was more like chess, you could control one unit at a time. The computer could actually issue commands so quickly, so it always beat you. That was one of the first Super VGA games, too. It was at 640 by 400 or something.
AR:
Was it challenging working on that for the first time, SVGA?
AGMS:
Not that much. There's some notes there about actually rewriting drivers and getting it to speed up like a factor of three or something. It wasn't too hard. We had a little graphics library for all sort of in-house library of code and things.
AGMS:
Let's see, then we got to do Patriot, which unfortunately got shipped.
AR:
[laughs]
AGMS:
That was the coffee table book, which turned into a game, based on the Gulf War. Descriptions of all the units in the war, pictures and things, and no actual gameplay, unless you're an actual general. Because there's like three directors of the game going through it, that's a bad sign. Some of them had a lot of military experience, so it ended up basically giving orders to your troops and watching them run around. It's not that interactive.
AGMS:
Those are all at a fairly small place. The company was about… 1993. We're around Bank Street in Ottawa, near the Lansdowne Park Stadium, sort of across the street from Almos. We had about a maximum of 17 people at the time.
AR:
Wow, so pretty small.
AGMS:
Yeah. We had the last one- around 1993, when they sort of left the place, we were doing the Sprint Cars for me on the Sega [Genesis], and Larry and company (Larry Donais) was the programmer for Crystal's Pony Tale, which was coming out then, because we had a connection there with Sega [of America], so we got some work from them. There's also a prototype for the Air Drums game for the Light Harp slash Activator.
AR:
Oh what was that? That prototype?
AGMS:
Well, I actually have- the guy who made it, I think he's from France. It's a ring on the ground with LED beams that you can interrupt with your hands and feet. Because it's an octagon, there's eight beams. I think he actually had 32 beams on his device. He said, "It'll play music and do lots of stuff." Sega took it and cost-reduced it to be too slow and too low-resolution to actually be useful. They mapped the eight directions into joystick directions and heights and levels into buttons. You can actually play it as a joystick. Again, it lagged a lot.
AGMS:
We were doing the Air Drums game. We had the artists drawing a nice animation of the graphics, and it was literally, what do you call it, disembodied drumsticks. It was fun to kick the bass drum there and all that stuff, but it didn't actually go anywhere, other than a CES demo or something.
AR:
Right, I think I remember seeing one screenshot of that. I think we have an article on that, so I'll expand that with whatever you can tell me.
AGMS:
Well, I can't say too much, because it was mostly the artists were doing that. I wasn't actually directly involved other than trying it out once in a while. Like I said, it was disappointing, because the actual hardware was cost-reduced to not 32 levels, only eight directions instead of 32, and only, I think, two levels in the final hardware, and a big lag as well. It was quite not useful for games, because you have a giant lag.
AR:
Yeah, that's unfortunate, because there's a lot of potential there, but Sega of America just has to bring everything down to a certain cost-price.
AGMS:
Yeah, I guess they were aiming for a target price, and then they lost the functionality along the way.
AR:
Yeah. [laughs] Exactly.
AGMS:
Anyways, so shortly after that, we started doing Sprint Cars; prototype for a 3D car-driving game around an oval track. Carl Mey is the director for that one.
AR:
Right, Carl.
AGMS:
He's a real sprint car fan, so he sent us videotapes of actual sprint cars, mud and so on. He also gave us a lot of clues on how to actually do 3D on the Sega Genesis, which has no bitmap mode, it has just character map graphics, as you know.
AGMS:
You can actually set up your font, [your] character map cells to be almost a bitmap, except the columns are all sort of… not in sequential order, so they go back and forth, and so on. It's really awkward. The trick for doing graphics on that was to actually write code, if you want to draw something on it, to actually take a texture, your 64 by 64 pixel or whatever, maybe smaller, and then draw it on the screen using this funny mapping of bits.
AGMS:
You actually write a copy of the code for each possible height you want to actually draw it. You're going to scale a thing from 1 to 32 bits tall. You have 32 different subroutines, each one hand-coded, just a bunch of move instructions in a row [laughs]. Skipping over pixels or doubling pixels or whatever, as needed. You have a program to write the scaling code, and then you just compile that. Because you have two megabytes of ROM, so you can stick in lots of routines there.
AGMS:
I wrote a lot about that. There's a bit of a story and stuff on my blog about that, some video clips.
AR:
Right, about the project blog. That's super helpful, by the way. We really appreciate you doing that. We've learned so much from that.
AGMS:
Yeah, it was fun game to write, with a few other people helping out, and the artists taking advantage of all the existing character map graphics tools we had for all the other games, Air Drums and the Crystal's Pony Tail. We developed a whole pipeline for that, mostly using Deluxe Paint for the animation, I think.
AR:
Oh, OK. Can I ask a little more about Air Drums? You kind of told me how it came about. You said that did reach a playable phase, and it was shown at least one CES, right?
AGMS:
Well, demo phase, not much of a fun thing. You could kick some of the drums.
AR:
A demo, but playable, quote unquote.
AGMS:
Yeah.
AR:
Do you remember why… I'm assuming that was Sega of America that canceled that. Did they ever give you a reason why they didn't follow through with that?
AGMS:
No, I didn't hear anything about that. No, it's… suddenly we were working on another project. They didn't actually mention that.
AR:
I don't think we knew that Artech had done that. We knew that it existed, and we had maybe like one screenshot and a little bit of information floating around about it. We had not been able to tie it to Artech.
AGMS:
Let's see. I think Phil LaFrance and Cory Humes were the artists on that. They're really good now. They're actually at some of the animation companies in Ottawa here, and moved up to management, actually, for Phil LaFrance. He was a really amazing animator.
AR:
Could you give me those two names again?
AGMS:
Phil LaFrance, Quebecer, and Cory Humes, C-O-R-Y space H-U-M-E-S. Phil LaFrance is probably the one who actually knows the most of it.
AR:
You also did work on that as well, right?
AGMS:
Not really. I tried out a bit. I talked to him once in a while, but that was mostly someone else doing that project.
AR:
Gotcha.
AGMS:
You can ask them if they remember doing that game.
AGMS:
They asked to do the other game. They're the artists. The Corel stuff? I'm not sure of the timing. There was a whole bunch of character games with a giant bunny rabbit and other things. Was it… No, you have to ask them. They came out of the Algonquin College here in Ottawa, which has an animation program, or had at least, and we sort of got some of the best people out of that and they came up to work here. Then after a decade or so, they'd move on to somewhere else, start their own business or whatever.
AGMS:
They also, later on, they had a computer game programming program at Algonquin too. I think Tony Santamaria, one of the Artech programmers slash managers, became a professor over there. You always get the best pick of the crop.
AR:
Do you recall how Sega of America first reached out to Artech in the first place? Was that before you had arrived?
AGMS:
That was before my time. Was it the Sega Music Development Kit [Sega Music Development System] or whatever it was?
AR:
Yes, yeah.
AGMS:
That was 1989 or something. Because Artech had been around since the early 80s, 82 or at least… They started off doing… there were a bunch of people working on the NABU computer. You've heard of that before?
AR:
No.
AGMS:
The NABU PC, sort of an Ottawa- it's a Z80 processor, Texas Instruments 9918A video chip, 64K of RAM, which was a lot at the time, and a 6 megabit cable modem at the time everyone else had 1.2 kilobit modems. It was kind of… basically you download all this stuff off the cable TV network using this modem. It sort of a…. Round robin, it had a cycle of stuff to repeat everything every minute or so. It's actually been- I'm doing some work now on a retro version of that, because someone released a whole warehouse of those on eBay, because when the NABU was shut down, they warehouse all that stuff, and someone said, "Hey okay, we can get rid of these, it's taking up space."
AGMS:
Suddenly there's a whole bunch of people interested in that. Some of the people recovered the software that was on the server side, and some development programming libraries. I'm actually using that to do some Z80 programming, because it's kind of fun.
AGMS:
Anyways, so NABU had been doing a lot of games for this computer. Simple stuff, because it's a simple video processor. They hired a bunch of college, high school students mostly, and various other people. They've been doing the games, and as contractors, Rick Banks and some other people had an idea. "Well, why don't we form our own companies to make games too, because we know what it's like now." So they started up, they actually sort of subcontracted out NABU games, and started developing games for the Colecovision. Coleco was surprised that they actually made BC's Quest for Tires for them.
AR:
Oh, that one's stellar. What a great example.
AGMS:
Yeah. Coleco didn't realize that people could just break their encryption or whatever it was. It was actually really simple. I think they [NABU] had to retro-license it or something. They went on to make versions for the 64 [Commodore 64] and so on. I've seen some of the graph paper they used to actually draw the graphics on it, and sort of tile little pixel patterns in it, and they had to encrypt it into hex and put it in.
AGMS:
They did that. They spun off from NABU into Artech Digital Entertainments, is the official name. Let's see. The 1980s had the C64, other consoles at the time, and they did various games for that. There's a bunch of early PC ones they did as well. We had plaques up on the wall. Then, let's see, I joined in 1991, and then they were moving more into PC and console games. I've been doing all sorts of odd things on and off.
AGMS:
There's a US special project… there's a museum ship simulator, so you can try docking a giant ship, cargo ship at a dock, in one of the museums in Canada. There was the failed air traffic control simulator. The local air traffic control school wanted to replace their slide projector with… y'know we'd done Megafortress and other games with flying around, its got sort of early 3D with polygons. We wanted to upgrade the slide projector with something, and we said "Okay." So we made a bid, and we had to send in our resume and prove we're all Canadian, and do a whole bunch of paperwork for that… and got turned down. CAE in Montreal got the job for several million dollars more than what would actually charge them. I guess it's politics.
AR:
What was early Artech like? In terms of the environment, you said it was pretty small.
AGMS:
Mostly one big room, pretty low ceiling. In front you had…. Part of the room continues on, a good nice window on the street. There's also a closed-off room beside it where the two bosses had their big desks, very big desks. We sort of had the meetings there. The main room was fairly dark, there's a kitchen at the back, and people would stick… was it pieces of fabric, or actually I think, blinds between the sections there. You have your desk there, and you have sort of a fabric thing hanging from the ceiling, to mark off where the territory ends.
AR:
[laughs]
AGMS:
And back and forth. There were floppy disk, no networks. For Patriot and other ones, we actually had RCS as the version control system. We have one PC with the RCS main thing on it, you check out your file and lock it, and put it on a floppy disk, and walk it over, and come back and try and merge it again. It was sort of a sneakernet with the central server. Again, using PCs is, "Oh hey, Windows came out, and OS 2 came out." It's kind of neat stuff. My first PC was a 386 of some sort.
AR:
Oh yeah, that was actually going to be another one of my questions. Was that your first PC in general, or your first PC at Artech?
AGMS:
First PC at Artech there, yeah. The guy, Steve Tibbet I mentioned, had an Amiga 1000 or something on his, and later on, he had an Amiga 4000, so it's kind of cool. Mostly PC and DOS.
AR:
Right, I saw you did a lot of development on the Amiga.
AGMS:
Yeah, he did, yeah. I didn't get to use it too much. Other than for artwork stuff, we use Art Department Pro on the Amiga for rendering and merging stuff. You can set up batch jobs to compress files and make animations, and we had it as part of our pipeline. We also have Imagine as a 3D editor there, so you can actually, if you want to occasionally stick in a 3D graphic, you have it render it into in 2D, and you take it over to whatever game you needed it for. I think one or two of the people actually knew how to use it.
AR:
So, to hop to Outlaw Racing here. You said that originally that was, and you'd also said this in the project log, that it started out as an experiment to just kind of see how the Genesis could handle 3D, and also there was a stock car racing angle, but the Sega VR implementation, the VR headset implementation, that came later in the project, correct?
AGMS:
I think there was that in mind, if we wanted to do a 3D game. Yeah, we started off with just drawing panels, just some example code from Sega, and then wrote our own code, and code to write the code. For the first few months we were doing test images, we were trying to make wall panels, and then have them scroll by at different speeds. Then we had to do, "Okay, stick in some cars there," so you have to draw sprites which are semi-transparent, so you can see around behind them, and add some obstacles there, and you go, "Oh, it's going kind of slow." Okay, so you can't draw the ground, so we actually have to- the whole ground is transparent with occasional small objects to make you think there's ground there, but it's not.
AR:
Yeah, that was a creative workaround.
AGMS:
Then after that you have some sort of game logic, you have cars moving around, you have virtual lanes for the AIs to drive in, and the whole physics and collision stuff, and of course music and sound effects. Who was the musician then? …Hmm, not sure who was back then, I think it was Mark Mitchell? Or he was more… 1994 afterwards.
AR:
It might actually be on the [Outlaw Racing] page.
AGMS:
I'll check some of the games around that time to see who the musician was. Of course, the funny thing about the company is all the directors are musicians. We have Rick Banks, who's an academic musician, studied in Europe, he's also the programming manager at the company, one of the partners. Paul Butler, who's a… not sure what kind of music he does, country or something? He's actually a bit of a name… he's the man in the suit, he goes out and gets the business, does all the traveling and jets and back and forth. Those are the two owners, initial owners.
AGMS:
Later on, I guess mid 90s, Dick Cooper from the Cooper Brothers became another director, sort of partner, and he's more in charge of the video stuff, and scripts and directing people in the studio, voice actors, that sort of stuff. He's actually gone back to music, has sort of a second career with the Cooper Brothers, doing tours and things. He was getting up on the age there, but he's had a great time doing that.
AGMS:
Some of the company directors could have done their early music.
AR:
Interesting! Sorry, just to hop back to the question that we had just discussed. Speaking of the VR headset, because that's kind of an area of Sega history that we know very little about… I believe in your project log, you said something of the effect of Sega of America never actually gave you one of their VR headsets to test the VR capabilities, is that right?
AGMS:
Yeah, we never saw one, I never saw one.
AR:
Did you have to make your own implementation?
AGMS:
No. We did have to make the game so the screen can move up and down, as if you're looking up and down, so there's actually a lot of control, you can tilt the camera up or down.
AR:
So all done in software.
AGMS:
Most of it was scrolling, because there's not time to actually do much camera moves, there's no real transformation matrices you can change, because it's all sort of… you don't have the CPU power to change transformations that fast. It's sort of kind of cool if we're looking straight ahead. It's not as flexible as a full 3D system. It did have to do the up and down and looking around, so you could actually move the camera sideways, or up or down, while still driving forwards; in a 3D world driving forwards, but you're looking sideways.
AGMS:
I'm not sure we had any demos of that, but that's part of the game, there's a lot of controls you could do, you could look sideways. In the VR headset looking sideways, up or down.
AR:
Near the end of that project, it also looks like, or at least from the project log, it looks like you were kind of investigating the feasibility of bringing that to the 32X (or the Mars), or was that in conjunction with Motocross, or did you want to bring them both?
AGMS:
No, that was just talk most of the time, no actual hard looking at it, just people coming, "Hey, maybe we should do this on Mars, whatever, what would it take?"
AR:
Right, it's the next system.
AGMS:
Just a couple of meetings basically.
AR:
You had said that you DID look into the 32X a little bit for Motocross. Was that also just talk?
AGMS:
Oh, 32X? There was a little looking into it previously, then actually we decided to take on the project. Chris Chan, the lead programmer actually went to Japan for lectures and lessons, and brought back a couple of prototypes to look at, or test our code on.
AR:
Right, that was one of the [32X development systems] that got broken in the mail, right?
AGMS:
Yeah, well actually he took it with him on the airplane. Or was it by mail, separately? I think it was on the airplane. [laughs] Yeah, so the guts got shaken up too much.
AR:
Well that's unfortunate. Hey, you made due with what you had.
AGMS:
So we had one working one, which, y'know… Yeah, and it's like Chris Chan and, what's his name, Michael? Was it Mike? Yeah. Michael Stevens, yeah. We're actually looking at the actual hardware there, and it's two processors interfacing with the Sega Genesis. They were concentrating that stuff. I was doing all the game logic, and physics, and I was prototyping all that stuff on a PC version. It could compile the code for the PC, or the 32X, or whatever.
AR:
Right, move it over.
AGMS:
So that would get going for the emulator would, yeah.
AR:
Speaking of the SH-2 that you mentioned. I actually I have two questions from one of our fellow researchers, Katsushimi. The first one was that there were rumors that very early 32X development hardware only contained one SH-2 chip. Did the development hardware that you had have one or two SH-2s?
AGMS:
Oh, it had two, yeah. We were actually doing tests there, like I mentioned there. They found it was slower.
AR:
Right. Just to follow up on that, so with the 32X, at least on the Sega of America side, there wasn't a whole lot of faith in that. It was just kind of like a stopgap thing. No one was really too passionate about the project. When you were working on it, did you have much faith in the 32X, or did you just kind of see it as kind of like a pointless venture?
AGMS:
Well, it seemed interesting at the time. Of course, Sega sending prototypes to a second-rate company like ours was a sign that they were not super interested in it. One of the top-rate studios wasn't actually working on it that much. We did all this experimental work for them, basically. Their test lab.
AR:
Right, yeah. You did a lot of hardware, a lot of interesting implementation of things.
AGMS:
Yeah, so we're trying it out, seeing if it works. Didn't expect this to be- it wasn't a top-level game. It was supposed to be just to show the hardware off and what it could do. The main thing with the hardware was actually just the video display. True-color graphics, no longer palleted stuff. Actual bitmap with 15-bit colors. Red, green, and blue, no limit in the palette. That was what people noticed the most. The performance improvement wasn't that much better. Maybe they could have actually released the 32X which only had just the video graphics card, basically, for the Sega Genesis. That could be kind of interesting.
AR:
Was the increased color usage… did that tie into the decision to use early JPEG implementation in Motocross?
AGMS:
Yeah, because we wanted to stick in some pictures and screenshots and things. I said, "Okay, well, rather than trying to do tiles and all this other stuff, we just take and do JPEGs." We had JPEG code. One of the students, I think it was a co-op student, Sebastian, sort of adapted that to run on the… I think it was the SH-2 maybe, and just decompress your bitmap into a full screen. That's mostly what it used for, just the menu screen and stuff like that.
AR:
Speaking of Motocross Championship, you said that a lot of the, or at least… you had taken some of the code from Motocross and rewritten it, reworked it, and then used that for Corel Moto Extreme, correct?
AGMS:
Yeah, some of it. Well, I didn't actually… the next programming team was working on something else. I was doing Corel Chess then. They pretty much took it and re-adapted it for the Corel thing and added silly stuff like the lunar gravity level, which is kind of neat. The graphics are a giant Buckminster full-type dome, and you're hopping over the hills on the moon.
AR:
I saw a screenshot of that. That actually looks pretty fun.
AGMS:
Yeah, so, I'll just say about that one. Can't really say too much about it. We had a deal with Corel to do several games. There's Corel Chess, there's the one of the cartoon characters, I forget what it's called. Oh, and Corel Super Putt.
AR:
Oh, was that Corel Wild Cards? With the cartoon characters? Wild Cards?
AGMS:
Wild Cards, yeah, that's the one, with the characters. That's when the animators went to town, had a blast doing all those characters.
AR:
Right. Speaking of Corel, what was your favorite project through all the Corel work that you did. What did you enjoy the most?
AGMS:
Well, I didn't do that many.
AR:
[laughs]
AGMS:
Of course, Corel Chess, I was working on it. [laughs] It was actually going pretty smoothly, too. I was making it nice and general so it could actually… it was basically supposed to be Mac and PC, then Windows came out, Windows 95. We missed the deadline and the Mac version fell behind, because we had to do more Windows stuff. I wrote using the Windows Graphics Device Interface, no direct access, just general, regular Windows calls for most things.
AGMS:
Like, the start menu is a dialog box with- if you don't know about Windows, you can actually specify the graphics. There's bitmaps for the buttons in the background, all that, so it looks nice, but is usable like an ordinary dialog box. Keystrokes work and all this other stuff. Didn't have an installer, it was just an EXE, you'd run it, and it would play chess and it could play across the network.
AGMS:
We had several different chess sets and 3D graphics. Each piece is rendered from several different angles and the game engine would actually play the appropriate pictures, make it look like it's moving or whatever, or the camera is changing.
AGMS:
We had Corel also given us a, or we bought really cheap, a Silicon Graphics Crimson, which is a coffee table sized red box, it's a SGI computer. Needed to have a special 30 amp power plug put in for it. It was so noisy, we'd actually stuck it in the air conditioning room in the building, and had the cable through the wall for the monitor.
AGMS:
You had like boards and boards of stuff to do this. We're using, I think Softimage from Montreal as the rendering software. Our Daniel Tosti, Italian graphics artist who's originally in TV in Italy and came to do graphics there, came to Canada, and worked for us for a few years doing games. He designed and made those nice, elegant sets and rendered them all, and he had the boards and backgrounds from different angles, and the game engine would stick them together and make it look like a 3D world which you could rotate with the scroll bars and pick up pieces and move them around and have them animate as they walk over and fight and so on.
AGMS:
That was pretty cool. Actually, it still works, because Windows GDI still works on news computer still looks nice.
AR:
Oh, awesome, yeah, I know a lot of other Windows stuff doesn't work anymore.
AGMS:
We had to actually use a third-party chess engine. Was it from Harvard or someone, somewhere?
AR:
You know, just to hop around a little bit here. I did want to ask a little bit about Crystal's Pony Tale. You said you were kind of involved in that. What was your involvement in that?
AGMS:
Looking at my logs… [it was] mostly burning EPROMs.
AR:
Oh, gotcha.
AGMS:
Actually, we moved to the new building in 1993, November. It was a large open space, and one of the 2D artists is also a carpenter. He built us sort of wooden partitions around and you have like an area of two or three desks, or have a bigger square area with maybe six desks in it. There's six desk areas, one of us doing Motocross. Next to us is the Crystal's Pony Tale with Larry Donais, so we hear their music all the time, occasionally have to ask them questions about things.
AGMS:
What do I remember about that one? Let's see, I remember complaining about one of the bugs they got from Sega saying, "If you bump into this thing 256 times, you can go through it." Why is that a bug? It doesn't really matter. No one's going to do that 256 times.
AR:
[laughs]
AGMS:
I also see I was helping out with physics with that one. They wanted to jump to a different level at a certain height and make it look natural. I was like, "Okay," I had taken physics at university, "Here's the equation for velocity and time. You can just figure it out." Using some of my math background to help them out occasionally. I didn't have much to do with it.
AR:
Speaking of… just to kind of cover everything that you've worked on with or through Sega. There's Outlaw, there's Motocross, we've got Pony Tail, and then we've got Air Drums. Are there any other Sega-Artech projects that…
AGMS:
Let's see, well, the 1989 music system [Sega Music Development System].
AR:
Oh, and the music system, I apologize.
AGMS:
Which I think was made by Paul Butler, or at least he mentioned something about it, for the directors who were doing that. Let's see…
AR:
This can be released or unreleased.
AGMS:
Some 3DO stuff.
AR:
Oh, M2, correct?
AGMS:
No, we didn't get the M2. We had a 3DO development system, nice big Mac Quadra and the emulation hardware. We were playing… Actually, Steve Tibbot was playing around with that, trying various things, and I was interested. Cause the actual 3DO operating system is very similar to AmigaDOS. The same people wrote that. That's pretty cool.
AGMS:
We did some internal demos on it, but we ended up trading the system to Gray Matter in Toronto, and they sent us back a Nintendo development system so we could do a version of Tetris for it, I think.
AR:
Oh, interesting.
AGMS:
Something like Tetris.
AR:
Yeah, there were a lot of people who were very hopeful about the system [3DO].
AGMS:
Well, it almost took off, but Sony came around and had something better. Of course, everything has changed so fast with technology, and you had to keep up. Sony has been keeping up with the PlayStation, and so has Microsoft. They've slowed down a bit now. Back then, it was changing so fast.
AGMS:
32X was there for how many months? They got superseded by the Saturn?
AR:
Yeah, not a whole lot.
AGMS:
Oh, by the way, the only other Sega thing I had was… I sort of quit in 1999, December 1999, just before Y2K, and the going-away present from the company was the Sega Dreamcast, which was pretty cool.
AR:
Oh, wow. That was pretty much soon as it came out.
AGMS:
Yep. Pretty much, yeah. I'm playing it because Shenmue is probably my favorite one. Then I came back [to Artech] in 2003, doing other stuff.
AR:
What were you doing for Artech at that time, when you came back?
AGMS:
2003? I had some projects, but they yanked me off that after a week, and then I was working on DVDs. What can you do with a DVD player? How can you program it? There's a small computer in there running the menu system. It has 32 bytes of memory and 4GB of instruction space.
AGMS:
The first thing we did was the various demos and trying things like, can you use the alternate video tracks to do alternate storylines, like here's a train, do you turn left or right here, and what happens? Various button pushing tricks and stuff. It's very limited, but you can do some things. The first one we did was actually Time Troopers, the trivia game, which had various historical questions.
AGMS:
They actually spent a lot of money and got John Cleese to be the actor there. This one weekend, they took down a whole pile of costumes, maybe 50 or so different costume changes. He'd be playing a Viking or a robot or something else and asking questions and doing things. We had various minigames using these DVD tricks we found to actually answer the things. Not just all button choice, sometimes it would be timing based or other things.
AGMS:
Using that 32 bytes of RAM, we tried to keep track of score and keep track of the last five questions you'd seen, because that's as much memory as you could spare. You sort of identify questions by four-bit number. You have more than four bits, but you have the last four bits, and you sort of save the last four bits for the last few questions. I'm not sure if there was more or not. I tried to avoid repeating things too often.
AGMS:
That was a really hectic one, partially because we were developing the system. We had a kind of database and a compiler, which would compile DVD projects out of this database of questions. They also had of course artwork for the questions in different languages. Eventually, we had tools we would actually use. We would have the question itself, run Adobe After Effects to generate stills or video for the questions with the text from the database in different languages. Also use SoX to mix sound samples from the different actors recording it and different things.
AGMS:
It got quite automated there. We could actually have a huge DVD with all these complex paths through it. Maybe string together templates of questions or other things. Eventually, we got to use that in Monopoly Tropical Tycoon, was this sort of highlight. We had a whole board game for Hasbro with that.
AGMS:
Yeah, that was a chaotic project. At the end, some of the projects the artists were working on were cancelled. Suddenly we had 20 artists working on questions. The end result was kind of chaotic. Everything looked different because we were all rushing. We hadn't learned how to do it before. It was all a mixture of styles and things.
AR:
Right, because I don't think there were a whole lot of DVD games at the time. You were kind of a pioneer in that.
AGMS:
Yeah, the only big other one was Scene It? The Movie Game, which is ideal for DVDs because you had movie questions. They had all the questions pre-made. You just had to get the rights to all the movie clips. We actually had to use John Cleese and the artists to make all the questions and things.
AR:
You even got a patent for that, right?
AGMS:
Yep. So, the whole DVD trivia game implementation, like I say, with the history of things, so we get to keep track of questions we've seen and score. I found out that some Chinese video players are pretty crummy, software-wise. The implemented integer is a signed integer, but it's supposed to be unsigned. The high bit of the number would be garbage at times. We actually had only 15 bits of value in each of those. It shows that, well, 16 registers. Supposedly 16, but only 15 bits actually work.
AR:
You know, speaking of integers. I remember, I want to say it was in your project log. You mentioned that the 32X dev system that you had actually had an error with short integers? Is that right?
AGMS:
Yeah, that's right. The GCC compiler we're using had not implemented shorts correctly for 16-bit integers. Occasionally they're getting mangled and they're put in the wrong place in memory.
AR:
Was that development system particularly difficult to work with, or was it kind of like, once you figured it out, it was OK?
AGMS:
It was OK. It's only had to compile stuff on it, using the GCC to get it running on the Mars side of things. I did all the development in regular IDE on the PC first. Most of the bug finding [in Motocross Championship] was found there. Just, "OK, compile it all now for the final one," then we find out what works and doesn't work. Usually all the bugs that are game logic bugs and missing artwork and all that, we find on the PC version first. We didn't really care that much about that side.
AGMS:
Of course, Chris and Michael had more trouble trying to get things working, because they developed the communications between the two processors and all this other stuff. It's kind of hard to get a handle on.
AR:
So on Outlaw, you said you tested the game on custom hardware, which contained a DSP chip for assistance. Could you tell us about that?
AGMS:
Yeah. I think Michael Stevens was doing that code for that. The DSP chip, he was trying to do the bitmap stretching algorithm on that. It was pointedly slower. Our fastest technique of just having a sequence of move instructions in ROM for each different possible choice was faster.
AR:
Actually, I know I'm kind of hopping around quite a little bit, but I wanted to… Outside of Artech, I wanted to ask about your early background in computers. How did you first know that computers, or working in computers, could be something that you wanted to do for a career, like very early?
AGMS:
Hm. I was always building things and taking things apart as a little kid. We even had some school teachers who encouraged that. Yeah, "Take apart a vacuum cleaner or a radio and see what's inside it." So I was building little things. We had, let's see, RadioShack logic kits, like a 4-bit counter with lights and so on. You'd just sort of play around and solder it together and try various things with it.
AGMS:
For computers, besides these toy ones, the first one we had, I guess, was a friend's Commodore PR-100 programmable calculator. We were doing Lunar Lander-type games on that. That was fun. Eventually I got a TI-58 calculator, I could do my own.
AGMS:
Then in the summer of May 1977, this was when Star Wars came out. Our neighbor had got on a course that had an extra KIM-1 computer to play with. I got to play with that. That's really the one I got really involved with programming. I said, "Hey, it's cool. You can see stuff happening." Also building some electronics for that. Made myself little power supplies and a memory board for it.
AGMS:
Got a used stockbroker's terminal, which is a 1960s Bunker Ramo. It has a giant keyboard on it. Which is arranged sequentially A to Z [laughs], not touch type, and vertical columns, so it's very tall. It had a small CRT inside it. I couldn't use a CRT for anything much. I tried, but it's all vacuum tubes and strange signal levels. I could hook up the keyboard to the KIM-1 through a parallel port.
AGMS:
Because the device has a circuit card inside full of diodes where it codes the keys into whatever Bunker Ramo keycode numbers are. They're still binaries you can read them in. I was using that for various things. After that, I gave up on the CRT. It had a diode matrix. You could actually put up binary codes. I could read those in the KIM-1. Later on, I added my own little display with two rows of eight LED digits, sort of orange. It was a giant size, like they're about more than half-inch tall.
AGMS:
At the time, I'd seen a lot of games, handheld ones using Football and other ones, we have a little LED display. Eight digits and two rows gives you eight wide by six tall pixels. We have a 48-pixel screen, ooh. [laughs] So I did some games on that, chasing after each other and running around. That was kind of fun.
AGMS:
There was actually sort of intensive coding there for a whole year. I was doing stuff, and ended up doing my own floating point math routine and a little calculator. It's kind of cool doing a square root and seeing square root two show up. Then my dad sort of caught on to it and got us the Apple II, or II Plus, actually. It was more basic programming and some actual commercial work for family tree software. So we had some assembler code.
AGMS:
Later on, there was a third-party company called Quinsept doing family tree software. I did a package for them for printing out family trees. Using all this graph theory stuff we've learned at university, you can take the data of whose parents and shuffle things around and insert new people there and then print it out nicely. That was kind of cool. I don't think anyone's done fancy printing graphs as complex as that since then.
AR:
Did you like working on that with them [Quinsept]?
AGMS:
It was OK, except it was a bit awkward. The two-letter variable names and not much in the way of comments. You'd actually code your program first in parts.
AR:
Did you like doing coding on the Apple II?
AGMS:
Yeah. In BASIC, it was kind of annoying. Lack of REMs and so on. Assembler is a lot better. I used the Lisa assembler. It's actually like you have disk files. Later on we had a disk drive. You have large programs and it would assemble it, which is kind of what I was doing in BASIC but automatic.
AGMS:
I had fun writing very small games. The biggest one was a clone of the Tron Light Cycle game. Sound effects and little things running around. Let's see. After that, I actually did a Pac-Man [clone] on it in BASIC at university on my friend's Apple II there. That was when they made Apple II clones all over the place. We had one. That was kind of cool.
AGMS:
Let's see. Then in 1985 we got an Atari ST for a couple of years. Did a few things there. The big… In 87 we got an Amiga 2000, which is the West German model, which is even better because I had seen it at a friend's place. He was playing Marble Madness. It looked pretty cool. It has multitasking and all this fancy operating system stuff.
AGMS:
Usually when I acquire a new computer I actually get the manual to it. For the Commodore 64, other ones… I'd be excited by the manual, read about the SID chip and everything. "Oh, this was neat." Amiga had a whole stack of manuals and had a real operating system there. That was fun. I did a lot of Amiga stuff, various utilities, nothing big in the gaming world.
AGMS:
I did a video player [on the Amiga] where it would stream video into memory and swap the video registers to actually switch to the next frame while it was loading the next one using the hard disk DMA. You'd get about 15 frames a second in full screen there, which was pretty nice.
AR:
Wow. I'm sure that was a lot of the time.
AGMS:
Yeah. I also had some audio recording and playback stuff. The other cool thing about the Amiga was if you want to play back 20 sound files, you could do AGMSPlaySound, AGSMStart, and it started up sound, whatever. It would actually open all the files before it starts playing and then as it's going just reads from each one to play it.
AGMS:
Just a point, other operating systems like BeOS or Linux, you only have a thousand file handles. You want to open 500 files, what happens, y'know? They can't do the trick of opening everything in advance.
AGMS:
Tons of Amiga stuff. Then in the 2000s I got a PC so I could run BeOS. Dual processor, Pentium 3s. That was another decade, lots of fun there, and various BeOS utilities and things. Doing some radio automation software on the commercial side for radio station automation, time downloads from FTP servers, a touch tone sound recognizer, various utilities for radio station automation… That was kind of cool. It's mostly Linux on a newer PC in 2013, 14. Of course, mostly these days it's a Z80.
AGMS:
Doing the NABU… [and] Nth Pong Wars, I call it. Sorta working on that game right now.
AR:
Oh, right. I remember seeing that on your website.
AGMS:
Yeah.
AR:
I want to say we discussed this, but you were programming at home and you had made your own games. I apologize if you said this before, but did you work in video games before Artech or was Artech your entry into the game industry?
AGMS:
No no. Before that I was working at Bell Northern Research Telephone Company doing telephone network simulations for the statistics department. You'd have the… this was in Pascal. You'd have all the data from the cross-Canada telephone activity for a few days on giant piles of magnetic tape. You'd have a simulation of the network. You'd run all the traffic through the network, see what the results were. Try failing a connection, see what the results were. By trying it with the improved failure routing, so it actually routed around failures, you could actually handle these amounts of traffic and delay the building of the second cross-Canada fiber optic cable by a year or two.
AGMS:
That's Bell Canada. Big company, parent company of this sort of… it's interesting to know that they could delay their giant expensive cross-country cable while investing in some better switching equipment. That was sort of stuff we were doing then.
AR:
Did Artech reach out to you or did you reach out to Artech? I apologize, I think you might have answered that too.
AGMS:
It's cause I met Steve Tibbett at an Amiga user group meeting here and he's talking about the stuff he's doing at Artech and was part of the demo there. I said, "Oh, okay, can I work there too?" So he said, "Oh, okay, send in your resume. So sure."
AGMS:
That's because I'd finished my master's about that time. 1988 to 91, I guess. Again, computer science, doing computer graphics. My thing [thesis] was a distributed multi-user puppeteering system using networking and Macintoshes and Smalltalk.
AR:
Oh, that was Handyman, right?
AGMS:
Yeah. That was pretty fun actually, using Dave Buck's ray tracer to render the final version of that on multiple Unix workstations because it took a lot of CPU time there. Then after that I said, "Okay, what should I do next?" Then I talked to Steve Tibbett. "Okay, sounds like a fun place to work at, so try that."
AGMS:
One drawback is when you work at a games company, they pay about two-thirds what you get out working elsewhere, but it's more fun.
AR:
Right. Yeah.
AGMS:
[laughs] That's a small company [Artech] too. It's not like working at EA, where you're just working on one little thing. Here you're sort of jack of all trades trying all these new technologies coming out or setting up the office network or… actually there was a big debate there. "Should we use Novell or LANtastic or what's this TCP stuff that's coming out?" So we had several generations of the network before we ended up with Gigabit Ethernet. We had the coaxial, cable (LANtastic). All of our systems joined together sequentially and there's a terminator there. We eventually had a PC server in the middle with, I think, four network cards in it and wires going off to different parts of the office, each of the different rooms in the office. Quite often you find that the artists had disconnected the network and rewired it so they could play, I guess it would be Doom at the time, over the network. You'd come back in the morning and "Okay, why is it not working? Oh OK… Put them back."
AR:
You know, looking back at your time on the game industry, was it an interesting experience, was it rewarding? How do you feel about working in the game industry?
AGMS:
Exciting. Lots of fun. We've got to try all these things do all this stuff and sort of… actually you can talk to people about it, what you did. I could talk to people about telephone network simulations, but most people won't appreciate it as much as if I talk about "I worked on Outlaw Sprint Cars" or whatever.
AGMS:
I got to use some of this stuff… Oddly enough, the first year physics came in useful quite a lot. There's an unreleased game called Cloudships, or well, it had sort of… Cloudship Traders, other [names], where you're flying a floating ship in the air. It's sort of a… land with mountains coming up occasionally that you can dock at. I used the wind resistance equations from first year physics to actually get nice motion there. You have the wind, like you're on a boat with waves lapping against you and it's pushing you towards the shore and you have to avoid it.
AGMS:
That was a nice game with physics. It had detachable parts, so your ship has masts and engines and things dangling off it so someone could shoot off one of your engines and you start going sideways. Of course, because you lost weight, you start going up. In the same way, whenever you fire your cannon, you get a nice recoil, the whole thing swings in the space and the cannonball goes over and hits the other guy and he swings off. That was fun. At this point, no one actually took that up.
AR:
What was the name of that game again?
AGMS:
Well, we never actually had it finalized. It had several names. It was Cloudship Traders or Skyships, and some nice art designs by the artists, sort of fantasy world stuff. They did do Cloudship Traders on the iPad, which never went anywhere much. It was released and the company shut down in 2012. They sort of finished the game and shut down the company at the same time.
AR:
What was the original intended platform for that?
AGMS:
Well, it was being made on a PC and using a joystick. Probably a console game would have worked [too]. I shopped it around, we had demos, but no one wanted it.
AR:
Did you ever shop that to Sega or was this past the time? I'm sorry, I didn't get the….
AGMS:
That was past the time, I guess. In the 2000s or some sort.
AR:
So you've been maintaining your personal website since at least the 90s. That's actually really impressive. What first kicked that off?
AGMS:
96, yeah! First of all, just web pages on the various ISPs. Mostly Freenet, which was National Capital Freenet, sort of a volunteer-run ISP here in town in Ottawa. They're still going. They're reselling DSL lines and so on, but they have space for hosting websites. Then I got my new PC in 2013, 2014 and decided, "Oh, okay, well, it actually has some active stuff there running on it. Not just static pages. Okay, I can do that here as well." But I kept all the content, so I had all of my 1996 review of James and the Giant Peach, other movies and things and notes and all the software. When I was doing software, I uploaded it to my website so people could find it there and download it. Now I'm actually doing a bit more web server-side stuff. I have the Fringe Festival Visitor Schedule Optimizer for making schedules and trying to maximize the number of shows you see.
AGMS:
I have sort of an aborted attempt at doing a social media website with forgiveness and forgetfulness.
AR:
Right, the less evil social ranking system.
AGMS:
Yeah, that one. That's sort of unfinished. Actually, I had to partially shut down because the spammers were trying to sign up to it and just fill people's email boxes up with sign up requests, which is kind of silly.
AR:
That's kind of the internet. I'm sorry to hear that, though.
AGMS:
Yeah, I have to have a better system. Maybe have to vet the sign-up requests in advance before they actually send out an email saying you want to sign up. I was thinking of doing physical postcards as a sign-up mechanism. "Here, send me a postcard."
AR:
Yeah, that'd be a lot easier. [laughs] Could you tell us what your passion for theater and the Fringe Festival, specifically?
AGMS:
Oh, it's like popcorn. See one show and you're going to have to see another one. Even if it's a bad show you want to see another one. It's fun. I also get to see stories. It ties in a bit with my philosophy about the meaning of life being trying to make the universe a more interesting story.
AR:
Oh, I like that.
AGMS:
Yeah, it actually works as a whole ethics system, you could say. If you want to optimize the size of the story of the universe, you have a sort of reference audience and you can measure the size using laws of compression to see if they're interested in stuff or not. If you make everything interesting, you have some rules to follow. It's obvious that World War III would be bad because there'd be no more stories.
AGMS:
Same for exploring space, having diversity there. Elon Musk wants to get off planet. It means more stories. Hitler is bad because less stories or nothing's happening because everyone's being oppressed. This whole idea of taking stories as a worthiness of actions actually works for the whole philosophy. It's really hard to use, though, because you have to predict the future. If you're in a lifeboat, it's women and children first, easy to use philosophy. This one, you have to actually know the background of everyone and predict what their future lives will be. Will they be interesting in the future? It's kind of hard to use.
AR:
I like your outlook on it, though.
AGMS:
[laughs] Anyways, you end up using several philosophies in a layer. First of all, you have the built-in reactions and instincts and you have reptile reactions and you have human-type philosophies. If you have time, you can figure out the optimal approach to doing something that's the best way of making things interesting. That's kind of cool that you can use information theory like lossiness to actually decide what is interesting, measure interestingness with a reference audience.
AR:
I like that. The last question I wanted to ask, and I appreciate the patience: Of all the work you've accomplished through your career, just everything you've done, what are you the most proud of?
AGMS:
Probably Monopoly Tropical Tycoon. The whole DVD compiling system, database lines… that's sort of the ultimate version of it. The artists, including Dick Cooper, had time to rewrite that as a two-person…. Well, before you had a news anchor telling things happening, and [he] switched it to two news actors, the dumb guy and a smart blonde, to have a whole lot of character interaction there. You roll the dice, go to the DVD, it tells you stuff, and actually it keeps track of things happening using this little DVD menu scheme. It's pretty cool. It's all backend. You can even play the DVD online on a web server. You can go to the emulator with the DVD opcodes running on a server and you can test things out that way. That was pretty nice. I got a patent out of it, too.
AGMS:
Monopoly's pretty cool, too. The whole rules engine for Monopoly was nice because it's the only one that actually implements the full rules of Monopoly as they are on the paper rules. It was just kind of cool.
AR:
It is nice to have a digital version.
AGMS:
Yeah. There's the whole Monopoly Star Wars thing in 1997. We actually got paid overtime that summer.
AR:
Is there anything that you would like to contribute or anything that I didn't ask you that you'd like to mention?
AGMS:
I pretty much exhausted all the Sega aspects there. You might want to talk to some of the guys if you wanted to find out about Pony's Crystal Tale [Crystal's Pony Tale], but that's about all I know.
AR:
All right! Well, thank you so much again. Again, it's a pleasure speaking with you. I learned a heck of a lot here. Just hearing your outlook on things, both back in the day and now, was super enlightening.
AGMS:
Yeah. I would say it's fun to work for a small company. You get to do everything.
AR:
I bet.
AGMS:
Yeah. Okay. Thanks for the interest. Hopefully, some of that will be useful to you.
AR:
Absolutely. Thank you so much again, Alex. I really appreciate everything today.
AGMS:
Okay. Thanks. Pleasant talking to you. Okay, bye.
AR:
Bye-bye.
Transcript sourced from https://segaretro.org/Interview:_Alexander_G._M._Smith_(2025-07-28)_by_Alexander_Rojas and migrated into devquoted with linked people, tags, source metadata, and media where available.