An Interview with Atari 2600 developer and Imagic Co-Founder Rob Fulop

Paleotronic had the great fortune to chat with Atari 2600 developer and Imagic co-founder Rob Fulop about life as a rockstar videogame developer in the early 1980s, and what came after.

Born in 1959, Rob Fulop graduated from college with an electrical engineering degree in 1979, after which he went to work for Atari. During his time there he ported Missile Command to the Atari 2600 (which became one of its most popular games). He later wrote Demon Attack, another of the 2600’s biggest hits, while at Imagic, the company he co-founded.

We’ve read in other interviews that you were first exposed to computers in high school. Where was that? What did your class use them for? What did your classmates think of them? Did you learn BASIC programming or some other language? Did you design your first game on them?

The first computer I ever saw was a teletype machine linked via modem to a local museum. It was 1974, at Skyline High School in Oakland, CA, in the makeshift “computer lab” – basically the storage room of our math class. One could write simple programs using the BASIC language – I was instantly hooked and joined the small group of geeks who signed up each day for 15 minute blocks of time on the machine.

The very first program I wrote flipped a coin 100 times and tabulated the results. After that, I was only interested in writing programs that “played” with the user somehow.   The second program I wrote played the game of Takeaway – a simple pile game similar to NIM. I programmed the game so that the computer could “cheat” and thus never lose. The third program I wrote was an attempt to attract girls into the computer lab – a “Boy Analyzer” – the user could enter the first name of any boy – and the computer would write out what sort of boyfriend he would be. Of course the program was rigged to give high ratings to the names of me and my friends, and thumbs down ratings to the names of guys we didn’t like.   Unfortunately this was before we figured out that labeling a guy as “bad news” was basically giving him an A+ rating – but the program did succeed at filling up the computer lab with girls.

Was that experience what led you to pursue a computer science degree at university? Which university did you go to? What was it like taking computer science in the late 1970s? Did you have a university mainframe? Which model? What did you do with it?

Yes, the experience in my 11th grade math class led directly to my choice of a major “EECS” as well as school “University of CA, Berkeley”.    At that time, we wrote our programs in FORTRAN, had each line of code put onto a punched card – then we would submit our “decks” of cards to the basement of the Engineering Building, and pick up the results of our output a few hours later.   It wasn’t until my junior year that I got to sit in front of a screen and learned about the UNIX environment.

When you got that first summer job with Atari, designing sound on the Superman pinball game, what was it like working with other technically-oriented people? Were you excited to be there? How did you design and generate the sounds? What hardware was used to do that?

Atari was very challenging in that I had never programmed the 6502 microprocessor, and the environment was quite unforgiving.   We would hand write our code on paper, turn it into the data processing window, where typists would type the code into the machine and generate a paper tape.   

We then took the paper tape to our development stations and fed it through the reader – then we could see if the program worked.   Usually it did not, after which we had the option of hand “patching” the assembly code to try to work out the glitch – or if it was to much data to hand patch, we would hand rewrite the errant section of code, and turn it in again for another try.    

The system had the advantage though of teaching us to really THINK through one’s code before just hacking it in there.    But it was a slow process, nonetheless.    The sounds themselves played on a simple sound chip – I don’t recall the specific one but it was used in virtually ALL the pinball machines of the time.

After you graduated, you returned to Atari and they put you to work in the consumer division, initially creating an Atari 800 version of Space Invaders. Could you please outline a bit about the technical processes you used? Was everything written from scratch in Assembler or did you use another programming language? Did you have any help or were you on your own? How did you feel about the game when you were finished?

We were given a bare bones hardware manual for the TIA chip set – the hardware that is inside the Atari 2600.   I was told to make a game.   I spent the first six weeks visiting the Atari coin-op demo room every day, where I could play all the Atari games – as well as playing all of the existing 2600 games.    I finally decided that I possibly could get a version of “Night Driver” to work on the 2600 chipset.    Everything was written in 6502 Assembly.   I received a bare bones lesson on how to write the display system from Larry Kaplan, a senior programmer at the time.    

When the game was almost ready, we had a big meeting with “Marketing” where the executives came over and saw the game and seemed to like it – Ray Kassar, the CEO asked me “where did you get the idea for this?”, which blew me away given Night Driver was one of their most popular Coin Operated games at the time!    Apparently he had never seen it.     I was happy to be done with it – it shipped with an ugly bug that nobody seems to be bothered by but me.

Moving from the comparatively luxurious Atari 800 to the Atari 2600 must have been tough, from a programming perspective. Could you go a little into the process of programming an Atari 2600 game at that time? What did you write the code on? How did you execute it on the 2600? Did you write algorithms on paper first? You developed a sprite generator for the 2600 on the Atari 800, could you talk a little about how that worked?

After Night Driver for the 2600, I moved over to the Atari 800 to do a version of “Space Invaders” – in retrospect it is remarkable that it was just up to me – nobody told me to port Space Invaders, even though Atari had purchased the license.   Also, I decided on my own that I would not “copy” the original game – as I was far “too creative” to simply copy an existing game like I did with Night Driver.    So I changed Space Invaders all around- I added a big dorky spaceship on the left of the screen – I did my own character designs – the top two rows make an “R” and an “F” when they march – I made up my own scoring system.   The finished game looked and played totally different than the actual coin op Space Invaders by Taito.   Marketing came to look at it and said “it’s ok, but why didn’t you just make Space Invaders, I mean, that’s what people will expect when they buy it”.   And in that instant I felt like a total moron.   They ended up shipping the game anyway.     

The technical process of game development became a lot easier at this point because programmers could each have their own TERMINAL to write their own code – thus short cutting the need to turn one’s code into the dreaded paper tape queue.   This meant one could turn their next version of the game around MUCH faster, sometimes several times per hour as opposed to several times per day.   So the games got a lot better since we could try more things.  The sprite generator wasn’t developed until I was at Imagic – at the time of Space Invaders I did all the graphics myself on graph paper, and then converted each line of dots into corresponding hex data which I would hand enter into the program. 

How did you end up with the task of translating Missile Command to the 2600? What were some of the personal and technical challenges you faced doing that? Why did you hide your initials in it? Was that a result of the anonymous corporate culture Warners had implemented at Atari? Did that culture hinder your ability to develop games? How did your co-workers get along with your corporate overlords?

Missile Command was my favorite coin op game at the time – Brad Stewart and I were at lunch one day talking about if it would be possible to convert Asteroids and Missile Command to the 2600.   He had an idea for how he could get the display engine to display a blocky asteroids playfield – it involved “flickering” frames so that essentially you use the hardware “twice” and every frame it goes back and forth displaying one set of graphics, or another.   I thought the idea may also work to allow the 2600 to generate the many graphics needed to do MIssile Command – we both left lunch that day determined to see who could be the first to have a working display up on their system.   He won – but I was close behind, about a week later.  The working displays inspired us to push forward and get the game logic working – both Asteroids and Missile Command were completed about the same time.   At the time, the lessons of my prior “less than original” Space Invaders game were burning in my soul – so I was determined that Missile Command would be as EXACT a replica of the coin op version as I could possibly get – down to the very last sound effect – EVERYTHING would look and feel like the original Missile Command.   And i think that’s exactly why it has remained as popular as it has.

The initials were just simple Easter Eggs – at the time it was just “fashion” for us to hide some sort of message inside the game – mostly for our amusement.   The challenge was to do the Easter Eggs so that they didn’t consume too much memory space, which at 4K, was at a total premium.    Missile Command was at about 4,500 bytes when it was done – I had to “squeeze” over 10% of the code down and still try to get it to do the same thing – crunching, or squeezing code was the bane of a game programmer’s life – it was dreadful work since nothing looks any different, everything must remain the same, it just takes less memory.   The last 100 or so bytes are the worst, we would do all sorts of ugly things in the code to save a byte here and there.

I never had a problem with anybody at Corporate Atari.    I just thought it so weird that they knew absolutely nothing about video games.  Zero.    Like one there was a brainstorming session where all the marketing people came over, and we went to lunch at a pizza place where they had a bunch of games, and we walked thru and I overhead one of the women say “Wow, this is so fun, we should visit arcades sometimes!”    Stuff like that, I just couldn’t figure out how they could have jobs there and not be really into games.   But there were almost proud of the fact that they knew nothing about the games, had never played them, etc.

Once your version of Missile Command took off, you understandably wanted to stop working like a studio musician and live the life of a rock star instead. Some of your other co-workers seem to have had similar desires. Did you collectively approach Atari first about renegotiating your arrangement, or did you simply decide one day to jump ship and form your own company? Did you give notice, or one day did you all just not turn up? Did Atari try to talk you out of it?

A few months after Missile Command shipped, we received Xmas bonuses, and I thought maybe I’d get like $5,000 or something – and was thinking I would probably purchase a better used than my grandma’s old Dodge Dart.    Instead I received a coupon for a free turkey at Safeway – I remember thinking “boy these guys are dumb – I’m 23 – I’d be thrilled with $5,000 and happily do another three games here”.   Instead I decided at that moment to pack up my Dart and leave.   My boss, Dennis Koble was planning to leave as well, along with William Grubb (marketing) and Mark Bradley (Sales).   So I asked to join the team and they said yes right away.    We all gave notice with very little ceremony – it was only after we left that Atari panicked and thru crazy bonuses at the remaining programmers.

What were the early days at Imagic like? Any interesting anecdotes about startup culture, and the comparison with Atari’s culture? What was it like competing with Activision?

Startups are fun – it’s like Us against The World.  My primary motivation was to make Atari cry when they saw my next game.   That’s it.   Just break down in tears.    I was motivated to do the very best I possibly could.    Activision was our main competitor – and they had the A list talent – David Crane, Bob Whitehead, Al Miller, Larry Kaplan.   These guys were the Belles of the Atari Ball when I was there.   The guys to beat.   But at first, I was singularly focused on making Atari regret how they mistreated me with the Turkey Bonus.   We didn’t think we could even touch Activision.

How did you come up with the idea behind Demon Attack? What was your creative process? Was it your first wholly-original Atari 2600 project? How long did it take to develop? Did you run into any roadblocks? How did you test it? Who came up with the marketing? What was your “90% rule”? How did you feel once Demon Attack was released? How did Atari’s lawsuit over Demon Attack affect Imagic?

Demon Attack was derived from the popular game “Galaga” where hoards of attacking flying thingies “peeled off” and came right at the player – it had a particular panic-inducing moment where the player realizes that they are going to be facing off directly with an alien mind – and it was this panic-inducing moment that I wanted to capture in a game.   I also wanted to make a game that featured MANY levels of play and have as many different looking enemies as I could fit.    Actually, Demon Attack was my second original game, the first original was at Atari after Missile Command, but I never finished it and the prototype unfortunately is long gone.   I did Demon Attack in about 8 months, and it was a very clean vision – just one play pattern that repeated over and over.   

I worked on the swooping motion for a long time, never was happy with it,  then moved on to finish the game, then I went BACK to the primary motion and tweaked the hell out of the motion routines to end up with the organic motion that gives the game it’s distinctive feel.   I had to fight to keep the game in the lab for the last month, my mistake in finishing everything before polishing the motion – marketing was eager to ship it – I insisted on holding it back.     It got quite heated.   

Finally I resorted to removing the backup copy of the source code from the company backup disks that were kept offsite, so that they had no choice but to wait for me to give them a final releasable version.  It was not pretty.   But the game did great – we were awarded VIDEOGAME OF THE YEAR by Billboard Magazine – which we used in all of our marketing.    It was my first lesson in the life lesson that as long as a game is GOOD, despite all the screaming, nobody is going to remember that it was LATE!    It’s always at the end of a project that a GOOD GAME becomes GREAT – but it takes a dedication to polish and a willingness to hold steady despite the shrieking of Sales and Marketing.    Many a games marketing manager at the time would say something like “The game itself doesn’t matter.   I could put dog pooh in a box and sell it”.    And sadly, they typically did just that.   Thus the big game crash of 1982.

Shortly after Demon Attack hit the market you arguably did become a video-game rock star – how did that affect your personal life? Did that success create any friction between you and other game designers? Were there any rivalries? Was there a lot of pressure placed on you to one-up yourself? Where did you go from there?

My name was on the back of Demon Attack, plus I received royalties.   There is nothing like handing six figure checks to an insecure 25 year old to change his worldview.   I upgraded my car, house, girlfriend, pot dealer, everything.     There was definitely a noticeable change in my long term friendships – many of my long time pals were still living the post-collegiate bum lifestyle and now their friend Rob hit the “big time” and doing so lit a fire under their butt – they moved out of mom’s basement – got themselves into law school, etc.     I never felt professional rivalries at Imagic .   I remember going to a CES party and talking to a woman who told me “I heard that Rob Fulop may be here later” … I thought that quite amusing and think maybe that moment was the peak of my rock stardom – but for the most part I never took any of it seriously since basically I was the same geeky guy who just happened to like computers and games.   I knew at work that my next game wasn’t going to be as strong as Demon Attack  – I wasn’t hungry in the same way – wasn’t as into “Making them cry” – Cosmic Ark was based on an unusual graphic glitch that I had stumbled into that filled the screen with stars – I was eager to use to to stump Activision – I’m still not sure why it worked.

Did you see any signs of the video-game crash before it happened? How did you feel once it became apparent the bottom had fallen out of the market? Why do you think it happened? After the video-game crash set in, how did Imagic try to cope? When was the decision reached to effectively pull the plug? How did that affect you personally? How were you able to move on?

The video-game crash caught all of us by surprise.   In retrospect, all the signs were there – at the 1982 June CES, many retailers complained about inventory that wasn’t moving – but our optimistic sales team was pretty much looking the other way, attributing the problems to poor quality product flooding the shelves.   There was no discussion that Imagic games would be affected, since we had a reputation of top quality games.     

It was a very difficult thing to deal with, and took me a long time to fully wrap my head around – particularly how CLOSE we had come to going public (literally Imagic was on track to offer our stock to the public market a few days after Atari let out the bombshell that they had the worst quarter ever).    I left the company soon after, there was an effort to keep things going, but the spirit of the company was gone.    Personally I went from very much believing the New York Times article that listed my net worth after the public offering at $8MM – to calculating I had about a two year cash window in the bank before I would need another job.    

Still not fully believing the crash was real, I found some freelance development work doing a few games for Parker Brothers, but soon after they bailed out of the market.   I started an independent game based on Robots called Actionauts, had a working prototype before realizing that the game had better potential on the Commodore 64, so I got a development deal from Simon Schuster to publis≠h the game.    Six months later, Simon Schuster also bailed out of the software business leaving me to market the C64 version of Actionauts myself.  I ended up giving the game away as “Freeware” and sign people up to a Commodore based bulletin board for updates – Commodore was excited about this idea (of using electronic distribution to release a software title) so they ran an article about the game, and PUBLISHED the bulletin board phone number so people could connect directly with me and download the game.    I upgraded my “server” and waited patiently for the article to be published – but no calls came – not ONE.    When I actually received my hard copy of the magazine, I found to my horror that Commodore hand published the WRONG PHONE NUMBER of my bulletin board – thousands of calls had been placed to the wrong number!    I was devastated.    

I ended up doing an online casino game for Q-Link, the predecessor to America On-Line.     Rabbit Jack’s Casino would become the first online casino, and the game earned royalties for years, ported to the IBM and Apple platforms.   It was a difficult three-four year transition.   Soon afterwards, I hooked up with Nolan Bushnell’s company Axlon and started working on the NEMO Interactive Movie project with led to Night Trap.

Do you have any other creative hobbies, such as playing a musical instrument? Do you feel that they helped your ability to design games? Would you recommend budding game designers take up other creative pursuits as well?

I’ve been an amateur jazz and ragtime pianist for many years – my musical background definitely helped with the sound design in my work – as well as the inherent sense of craftsmanship that comes along with basic musicianship.   I’ve also been an aspiring writer on and off at various stages of my life.     I imagine any modern game designer needs to immerse themselves in at LEAST two  other creative mediums – writing – music – graphics – animation – and basic design sense.   Games and stories have essentially intertwingled at this point – there is no point where the story stops, and the game begins – as far as I’m concerned, the game design role is AUTHORSHIP, plain and simple.  If it’s not on the page – it won’t be on the stage – so it all stems from the design documents.   And who creates the design documents?   The game designer.    

New games written for classic consoles are all the rage these days. if someone wanted to create a new Atari 2600 (or Atari 8-bit) game, what would be some of your top tips? Do you have any advice about programming or game design in general? What about on the business side?

Obviously one needs a strong VERTICALLY moving game playfield for a game to be 2600-worthy.    I think several mobile games meet this criteria – particularly a game like Doodle Jump  which in my opinion would make a beautiful 2600 game.  Anything where the action flows from the top of the screen down to the bottom, and not too many game objects are needed for any particular horizontal plane.   That’s why Doodle Jump would work well.   

My basic theory of game design is there is very little reason to invent new games – the good ones already have been invented and we don’t really need any more.  The trick is to know where to look for the good games that already work, that have proven the test of time, that have been played and replayed countless times by people.    The best game I ever made was DOGZ and CATZ – modeled after puppy dogs and kitty cats.    People like puppy dogs, it’s not surprise they also like the digital versions.    Rabbit Jack’s Casino simply took four basic casino games and made them work online.  An adventure game is simply the game of TREASURE HUNT, always fun.   Most multiplayer shooters are simply versions of an age old playground game called TAG that we all played as kids.   Etc.    There is little reason to try to dream up a brand new game.   Sports offer tons of good game design ideas.   As do card games.   As do countless good board games.  We really don’t need new play patterns, kids are just as happy playing the same play patterns that they have always enjoyed.   

Your story would make a great movie! Has anyone ever approached you about that?

No, but thank you – that’s very funny!   Like who would the hero and villain be?   Or would I be both?    How would the movie end?   Would people cry or laugh ?   I’ve toyed with doing a book, I probably have enough content for a book, just not convinced enough people would want to read it.

Please consider subscribing to Paleotronic. Subscribers get a full digital PDF edition and an optional glossy full-colour printed copy, and all content is exclusive to subscribers until the following issue is released. Plus, by subscribing you support our efforts to spread understanding of the history and fundamentals of electronics! Thank you.

Be the first to comment

Leave a Reply