Richard A. Bartle: September, 1993 interview

September, 1993 interview

MICHAEL: When did you first start to write MUD?
RICHARD: Well, it wasn't ME who started it originally! It was a chap called Roy Trubshaw, who's still a director of MUSE and very much a part of the company. Roy wrote the first version in 1978, followed by a second version shortly after. I was involved with both of these projects, but didn't do any programming. I only did THAT when Roy began the third version, which he got about 25% finished before his time at Essex University (which is where we were at the time) was up. I wrote the remainder of the program, and turned it into a game (prior to that, there was precious little to "do" in it). I wrote MUD2 pretty well alone, in 1985. Roy and I give ourselves mutual credit for MUD.
MICHAEL: What made you start in the first place?
RICHARD: I've always been interested in games - board games originally, but computer games too, now. The fun for me with MUD isn't in playing, though, it's in designing and developing the world. I'd probably work on MUD in my spare time even if there were no players whatsoever!
MICHAEL: What computer did it all start on?
RICHARD: A DECsystem-10 mainframe. This was a 36-bit machine which had an addressable memory of a whopping 256K. OK, so in bytes that would be equivalent to about 1M, but still, it's paltry compared to today's machines. Besides, we were only allowed to access 60K of that 256K at once. The first two versions of MUD were written in machine code for this very reason, but the third version was in BCPL (a beautiful language, of which C is the best-known derivative). So I implemented a paging system whereby text was swapped out to disc, thus keeping within the memory-usage constraints. What a way to carry on! Incidentally, MUD1 is still run in the USA on CompuServe, on account of how they must be the only place left in the world to run DECsystem-10s still (except for in museums).