Cannonball

Owen Rubin (first Atari coin-op game programmer) - Cannonball game story: I have been asked about the very first game I worked on at Atari and how my first creation took place. That would be the coin-op version of Cannon Ball. The game was one of the very early microprocessor games created at Atari in early 1976. I liked playing the game Stunt Cycle, a hardware only game where you jumped a motorcycle over an ever increasing array of buses, and wanted something that easy with increasing difficulty. At that time, video games were almost completely created by four people: The hardware engineer, the software engineer, the mechanical designer (cabinet and controls) and a tech. There were some support people to the hardware engineer for building prototypes, and the software engineers had computer operators, but I did not even know about them at this time. The entire game, graphics and sound were created by the software engineer for the most part. I received a very basic motion object hardware (Black, White, and Grey objects) with a simple bitmap stamp playfield and told to create a game. I had not been at Atari long, so I had no idea HOW a game was built, just that there were a few people there doing it. I was amongst the first 5 programmers so we sort of made it up as we went along. Ok, so they set me up in a small lab next to my small office, and I got to work. First, I created the graphics for a small man running, flying through the air, and hitting the wall, including a "squish" animation, and a man walking off holding his back. This was all done in 8x8 motion objects, several put together to make a bigger object. So on graph paper, I created these objects. I was no artist, but it was easy at that resolution. I then did the same thing to create the cannon, the wall with a hole in it, a powder keg, and a smoke blast! These I hand programmed into a small PROM (this was before EEPROM), so you wanted to be sure you did it right. After I got the graphics installed, I sat down at a device called MicBug, a 6800 (NOT 68000) Motorola "development" system. This was a small computer board that had a cable that plugged into the processor socket of the game board with some RAM, You entered instructions at a model 33 teletype! With this device you could put in a breakpoint, alter memory and registers, run and stop the processor. Very basic. So, with a small card with 6800 instructions on it, I began writing AND HAND ASSEMBLING my code on paper, then I would type it into the model 33 teletype into MicBug, and save it on paper tape. Editing was done by creating small "patch" paper tapes with changes. I had a big box of changes, and would eventually punch out a full dump. After a number of months, my boss suggested a review of the game. No problem. We set it up. I had the game running, simple sounds, self test, and all coin routines. I would guess about 75% of the project. A man came out, stood next to the cannon and waited.

On the opposite side of the screen was a tall wall with a hole in it (that moved each round) Pulling on the handle caused the powder keg to add powder to the cannon. The more powder, the higher the shot. Releasing, the man jumped into the cannon. Push the fire button and the man flew out of the cannon towards the hole. Make it through and you scored. Miss, and you splat into the wall or floor and limped off. Simple but fun. Towards the end of the review, my boss asked if he could see my program code listings. Listings? What listings did he mean? Well, as it turns out, there was a way to create programs for games, I just never knew about it. Rather than hand assembling, one simply wrote the program on paper, handed it to a computer operator who then edited it into an early DEC PDP computer (like an 05 for example.). These operators would assemble the program and produce a paper tape output and a listing of the assembled program. Oh! Then the program remained on the PDP. Later changes were simply scrawled across the listing