Astro Blaster

As anyone who's been following the MAME WIP page at Davesclassics.com knows, Mike Balfour has been quite busy lately, moving the scope of the MAME project further back in time, adding drivers for the video games of the seventies. So far, his credits have included Sprint 2, Space Odyssey, Astro Blaster, Monster Bash, and several others. For MAME .31, he has already added Atari Basketball, Sprint 1, Night Driver, Boot Hill and Dominos, among other things. I have conducted a brief e-mail interview with Mike concerning his efforts to bring back these lost classics:

First of all, how and when did you find out about arcade emulation? What got you into emu programming?

I've actually been interested in emulation programming for many years. The first emulator I used was a TRS-80 emulator for the PC about 7 yearsago. The whole concept fascinated me, so I eventually wrote an Apple II emulator for the Macintosh. Of course, STM got released before Ifinished and was much better than mine so I gave up. I also started writingZelda (my all-time favorite game) for the Mac from the ground up (clean-room reverse engineering), but discovered Marat's Gameboy emulator in '95. Once I saw that consoles could be emulated, rewriting Zelda seemed pointlessand emulating it seemed more interesting. At this point, I still didn'teven dream it was possible to emulate arcade games (a mythical untouchableGrail to me) until Sparcade. Sparcade was the final push that got me into emulation programming. I started digging around for knowledge, but was having trouble finding *any* information on the net. It wasn't untilMAME had been around for a while that I finally started getting enoughknowledge to join in productively. I discovered that going to a local arcaderepair shop and asking for old schematics can do wonders. :)

Sprint 2

Why did you choose to start concentrating on the older games?

I never really intended to. The games I choose to work on really hasbeen driven by information available to me. I had just finished Space Odyssey when Al Kossow began his push to archive all the Bronze Age game information. Never being one to pass up schematics, I read through the Sprint2 Manual and was surprised at how straight-forward and basic the design was. I figured I could knock out the game in a couple of days,so I went ahead and added it (it actually took a couple of weeks). Ironically, it wasn't until after I had it partially emulated that I recognizedwhat the game was and remembered it. After I finished Sprint2, I realizedmost of the other old Atari games had a similar design, so I figured I mayas well work on them too. :) Now, it's become sort of a quest to get asmany of the Bronze Age games done as I can, because I'd like to see themhave a fair representation in MAME and in people's nostalgia. Plus, it's funfor me to find more games that I didn't remember until I get the chance toplay them again. :)

Super Breakout

-Is it necessarily any harder to emulate something that was made in 1975 than it is to emulate something that was written in the early 80s, where most of what has been emulated in MAME shows up?

Not really. In fact, I'd say it's probably easier to get it to atleast 75% emulated, because the hardware is generally so basic. The last 25%can be tricky because they used non-standard controllers (there wasn't a standard yet for them to follow), or because they do hardware collision detection, or because some graphics are generated by logic circuitsinstead of graphics ROMs (Night Driver's rectangle generator is a good exampleof this). It usually takes me a few hours to get one of these games up toa title screen, but another week to make it playable. In contrast, Space Odyssey took me about a month. Of course, I'm biased because Atari has some of the best-documented schematics in these early games. I didn'teven try to do Targ and Spectar from Exidy because of their sickeningly bad design. I have tremendous respect for Dan Boris for tackling those. Of course, later Exidy games continued on the same miserable hardwaredesign concepts, so I guess it's really more a question of company than time period.

-Some of the earliest video games (Pong, Computer Space, Etc.) do not use either ROMs or a CPU, and are composed entirely of TTL logic. For this reason, it seems impossible that these can be emulated. It has been mentioned in channel #mamegames that there has been some dialogue among the MAME team about finding a way to actually do this. Do you have any thoughts on this?

Some of these are on my list of "games to emulate someday". It's by no means impossible to emulate these games. It's just a question ofdrawing up a state machine and writing code that duplicates the state machine. Conceptually, though, it's moving away from emulation and closer to aport of the game, particularly because you would be able to run them without needing any sort of ROMs or external files. Once you're emulatinghardware at a high enough level, it's hard to distinguish it as "emulation". Al Kossow posted an excellent description of this at one point to comp.emulators.misc. It's easier to recognize it as emulation if you emulate each TTL chip and their interactions, but you can accomplishthe same effect by rewriting the logic as carefully-timed C code. I guessit's just a question of what your definition of emulation really is.

What was your favorite arcade game way back when all these games were out in the arcades?

Heh, it's a tough call, I've pumped coins into so many and excelled at none. It's probably a toss-up between Xevious, Elevator Action,Pacland, Arkanoid, Moon Patrol, Bump 'N' Jump, Tron, and Discs of Tron, but Ihate to neglect the classics - Asteroids, Space Invaders, Donkey Kong, and Pac Man.

Thanks for your time.

Return to Warrior's EMU Rumor page