Friday, December 22, 2006

Interview with Don Daglow re Early Game Programming

Student Viraj Tipnis conducted this email interview with Don Daglow of Stormfront Studios for his university gaming presentation in USC Gamepipe's CSCI 180, Survey of Digital Games & Technologies.

1. Is there any reason why you chose to write Baseball as your first game.You might have been a fan ,but were there any other technical reasons why you chose that particular game?


Apart from traditional games like chess and Monopoly, the first game that fascinated me was called All Star Baseball, which I started playing at age 11 in 1963. It was a board game simulation of American baseball that ignored the role of the pitcher but did a great job of modeling the performance of hitters. Kind of like a soccer sim that accurately modeled the skills and performance of every forward in the league, but had only one "average" fullback who played both sides of the field for every team.

I ended up inventing a way to modify the game to simulate both offense and defense, but did not have all the statistics needed to implement the design accurately until I went to college, when my passion turned instead to computer games when I programmed Baseball in 1971. So only a few college friends ever saw my "improved" version of All Star Baseball as a board game.

The minute I saw the computer terminal I realized I could build a far more accurate baseball game than All Star Baseball, so I immediately set to work on it.

I also designed a comic-inspired "Peanuts" dice-and-board game when I was about 12, which was a too-simple pursuer-pursued dice-rolling game in which Snoopy and Linus alternately stole Linus' blanket from each other. No one would ever say that it predicted my later success in game design!

2.I haven't used a PLATO or DECUS terminal before. It seems similar to todays filesharing networks.Was there any concept of copyright or intellectual property at that time?

I'm asking this because games were distributed 'freely'.

Actually, there was no such thing as a DECUS terminal: the games for PDP-10 computers (and I imagine other systems later, but that was "after my time") were shared between universities on "DEC Tapes." The school loaded the game on its computer's hard drive (precious real estate, hence the concerns re size) and THEN students could access them after logging into a terminal. So we thought of the terminal as a PDP-10 terminal, not as representing DECUS.
We had no thoughts of IP's because the idea that anyone would ever pay us to create games was so foreign. It was an underground activity one did for fun, not something one ever dreamed of being paid to do. When I applied for a job at Mattel in 1980 they thought I was lying about having written games for 9 years because they didn't know that era of game design existed -- fortunately, they figured out that I was telling the truth so in the end it helped this English Major get a job as a "Senior Applications Programmer."

3. You were one of the pioneers when it came to early computer games. Do you feel that if gaming was considered to be different type of field in computer science, we would have had better games today.?I mean was writing games and spending hours playing them not looked upon too favorably at colleges as a career prospect?

It was looked on as a waste of time UNLESS you were doing it to interest more people in getting involved with computers, in which case it was considered socially redeeming. Of course, the underlying flaw in this theory was the assumption that the minute someone got involved with the computer they would immediately regard games as pase and start doing "serious computing." They were half right: people who started wiuth games did use the computer for serious purposes, but they never stopped wanting to play games.

This led to all sorts of (now funny) contradictions. There were 10 of us in our computer science study group, of whom I was the one concentrating on games. When they wrote an article in the Pomona newsletter about us almost all of the examples cited were my work -- because the really inventive work everyone else was doing was "too hard to explain" or "didn't sound interesting to non-math-types." This was very unfair to the other students -- I heard that one of our group later led the project that computerized the Library of Congress.

Yet if we were caught writing games we'd be kicked off the system, despite the "good PR." Which was unfair to us!

4.Some of the major arcade companies during that time, had PLATO connections and had free access to these games. Companies were 'inspired' after looking at some of the coolest games and turned them to arcade versions.

Students who wrote them as free ware never got credited for their work...I have a few examples..

What are your comments on this?

This probably happened after I was already at Intellivision, but it doesn't surprise me. When Creative Computing Magazine started publishing code for microcomputer games in about 1980, everyone who had written anything during the 1970s probably saw their game code listed with their name removed from the source. IMHO that was actually the greatest impetus to the no-credit no-royalty use of a generation of work.

5.Do you think that game programming in colleges needs some type of boost,like having some kind of game sharing network like PLATO between colleges?. I feel that really good games were shared and developed by students at lot more in the 70s than now. Please correct me if I am wrong in thinking this.

I can't comment on the comparison because I don't know what happens in colleges now.

I can say that what games were picked up by DECUS (or shared less formally by computer center managers) was very flukey and unpredictable, and we had no influence over what other schools took or didn't take. I first found out Star Trek had become "the #2 Star Trek at most schools" (behind the MIT one that had quadrant printouts) when I started getting fan letters at my dorm. I didn't even know DECUS had picked it up.

Baseball was a "core hog" because it did so much calculation, especially in full season mode, which took 50 minutes per league on a lightly loaded system. So it was nowhere near as welcome as the 32K version of Star Trek, which as a turn-based game "rested light" on the need to be swapped into core for processing.

You may have been briefed on this, but the PDP-10 was a time-sharing copmputer with only one CPU. Each active user's program was "swapped" into core for a fraction of a second to process a few instructions, then swapped out for the next program. In theory it was supposed to be so fast that everything you tried to do endured just a short delay. Picture 55 users (including a hospital business office) electronically sharing the processor inside your cell phone to do all their work, and you get the idea why games that only had to be swapped into core once in a while fared better than those that wanted to do intense calculations.

Of course, sim designers like to hog core... when I wrote Utopia at Intellivision the programmers all kidded me about being the first guy ever to write five pages of assembly code that displayed absolutely nothing on the screen of a video game!

By the time I wrote Dungeon in 1975, which turned out to be arguably my most influential work, its size and its "calculate what every monster in this world is doing every turn" CPU load was a major liability to its distribution. The anecdotal stories I heard were that it was shared more often as an underground activity "off the record" than through above-the-table swaps, though I've had a variety of people tell me they played it on a PDP-10 in the mid and late 70's so I know it gained at least a little momentum.

6.You obviously received a lot of fan mail when you wrote Star Trek. Was there anything like students across Universities waiting for Don Daglow to come up with his next game? Did you share your source code with any body at that time?

No, I'm afraid that even though we got fan letters from students we had no influence on whether anything would ever be shared beyond our immediate friends, so I never thought of trying to provide something new to those same people. Some letters would have game improvement suggestions, though. Maybe someone somewhere was waiting for my next game, but we always thought of just doing things for the people at our school. We definitely were not looked on as people doing something significant on campus.

Since the only language used for most of the distributed games was BASIC, your source code was always right there for anyone to see and could not be hidden -- thus its later publication over and over in books and magazines. This was part of its appeal as something to draw more people into programming. The only other "lengua franca" available on all systems at the time was Fortran, which was a TERRIBLE game language, though I did a simple port of Baseball as an exercise. I did a version of Ecala in SNOBOL but only the BASIC version ever got shared. ELIZA was in compiled LISP, but our system did not support LISP as a language and we had no access to it. Assembly, which became the default game language family when microcomputers were invented, was used almost exclusively for operating system work on the PDP-10 because its instruction set was limited so doing anything with it was cumbersome. Of course, compared to the 6502 microprocessor that came later it was pretty cool!

7. Did arcade companies approach you with a lot of $$$ to sell them your ideas? Or did you approach any companies to sell your code?

No -- because the 70s work was all text-based there were no arcade versions. After all our work was published anonymously whatever benefit we could have had was lost. My one benefit was that both Baseball and Dungeon were so long and complex they were never published for the "type it in yourself" market and no one ever apparently figured out how to abridge them. By way of contrast, there must be at least 10 versions of Star Trek that are cut down in one way or another to fit on 2 or 3 pages and be "typeable." Almost all of them eliminate the work Jonathan Osser did to add the z-axis to the game's math.

And I wasn't smart enough to do what the Zork guys did and start a company when the Apple ][ (<-- dated symbol use!) came along. Once that market emerged I was already running the Intellivision game design team and not thinking about comercializing my older work -- I thought we were going to end up being one of two or three dominant long term design groups in the world and couldn't believe my luck at leading the Intellivision team. History served me a great dose of humility when the Crash of 1983 came along!

No comments: