Here are various factoids, presented in random order, discussed during dinner on May 22nd, 2013. The author apologizes for not remembering all the interesting stories, as I was actively engaged in the entertainment of the moment.
- Pfutz brought a children’s book centered around Joust which includes some history about the game development process. (ISBN-10: 0896862410 , ISBN-13: 9780896862418)
- It is easier to kill a “Tery” when the mouth is open, rather than closed. You have 3 pixels to aim for in an open mouth, only 1 in a closed. No players at the table were aware of this, after how many hours of play?!?!
- Python caused a stir by suggesting they use chocolate brown for the cabinet sides.
- Dual player mode was intended as a cut-throat battle between players to increase coin box profits. Employees at Williams would compete head to head over vending machine items. But when released to the public it turned into a method of collaboration to help the game play time extend.
- John was high school buddies with Crystal Castles developer Franz Lanzinger. Both were on the chess team. Thusly, Joust has a lot of flow and strategy based on past experiences with pinball and chess.
- Prior to Williams, John was a toy designer with an ability to build storyboards to aid the design process. Just what Williams needed to continue their future as a video game designer. Especially on the heals of their top two stars, DeMar and Jarvis leaving their inhouse facility to produce Stargate and Robotron via VidKidz. President Mike Stroll and Engineering leader Ken Fedesna found their future star in John.
- John wrote out many story ideas for Williams games, but it was the concept of Joust which resonated a bond of excitement between Bill and he.
- John used a drafting board and dry erase markers on the monitor to keep track of ledge design and bird movement during development. After Pfutz would change variables in the code, John would re-evaluate accordingly and adjust. The end result is the most perfect simulation of flying one can experience in a classic arcade game.
- The Joust logo in the attract mode was vectorized.
- Joust was indeed created at the N. Kedzie office. I don’t remember the address, but it was just a few blocks from the California Ave. main office. The main office was where pinball manufacturing was, marketing and upper management. Video manufacturing was at Gurney. Kedzie was where video engineering, art and pinball designers were. Ken Fedesna was the highest ranking person in the building. The building was shut down before Narc and after Star Rider, but I don’t know exactly when. It might have been when I left the company for a brief stint at Mylstar (where I first met Warren Q-Bert Davis). -John N, May 2013
- The “real” birth place of Joust was a building that WMS rented on N Kedzie Ave (I forget the address, it was near W Addison St). Eventually, we all moved to California Ave. As a matter of fact, Python got to rent an apartment that was across the alley of the Kedzie building. And all he had to do was to walk across the alley to get to work… – Pfutz, May 2013
- Due to the hardware and memory limitations, Bill had to use many ingenius workarounds and “pfutzian physics” to build the code for the game
- Although the word “Pfutztonian Physics” is not in the Dictionary (patterned after “Newtonian Physics”), this spelling is probably “more” better “pfutzian logic”. PS I hope that Newton is not turning over in his grave… -Pfutz, May 2013
- The “final” concept for jousting the other player based on vertical height was originally a typo in the developing code, which they discovered made the game “fun.” John also mentioned the original concept was for larger birds but it didn’t work out in the screen space available.
- There were several concepts/ideas before finding the “fun” one….
- I am fuzzy on the whole lance height thing. I thought the rule was always that the higher ostrich wins in a collision. I had a lot of other rules regarding dismounting that were ridiculous and we got rid of them. For sure the elegance of only dismounting by colliding with a lance higher than the other came about during development.
- As I remember the very first rule was to hit the player with the lance (not the armored bird) to dismount the player (hence a “realistic” Joust). This was one of the rules that I really liked.
- When we implemented this rule we found out the bird was alot larger than the player and my physic’s mostly missed hitting the small player and instead mostly hit the big bird. This made dismounting any enemy/player hard and frustrating.
- So for the sake of “fun” we had to throw out that rule and come up with another idea… Several ideas happened, do not collide with the birds head/neck (still not fun), then parts of birds body (still not fun), etc.-Pfutz, May 2013
The Legendary Artwork of Joust
- quoted from Python, June 2013…with the ever present technical assistance of Kotlarik, sketched,shaped,formed,defined,colored and animated everything that moves and lives on screen in the video game Joust.
- I had nothing to do with the font and the stone clouds on screen or the marquee art and control panel art on the
- I only did the side art stencil art-
- and the chocolate brown I had to have because of design integrity; the stone clouds were predominantly brown,
weren’t they? The Warrior Rider and the Ostrich were yellow,red and white -which on a black background would of defined the Color Combo called in design 101 and color psychology: The “Stay away”, “Warning”, or “I am Deadly” visual message !!! SO BROWN WAS THE BEST CHOICE IN THIS DESIGN SCENARIO.
- While on this subject, I must also tell you all that I painted on my very own, on my own time, at home.
- Note- Constantino “Connie” Mitchell painted the marquee artwork. Click HERE for a list of his prolific work in the industry
- Harry Williams was the game designer.
Harry (of course) created Williams, then sold it to Louis Nicastro
- Flight 2,000 was manufactured by Stern who owned URL (Universial Research Labortories),
where I programed Flight 2,000.
- At the time,Harry was living California, which was a long distant to Elk Grove Village IL ! – Pfutz, May 2013
The infamous “Tery bug” in the original Red Romset
- I could have sworn that it never automatically impaled itself until the last minute art change to the ptero.
- The Pterodactyl bug was not found in our testing phase, maybe because the Pterodactyl striked fear in the hearts of mortal men (However, I guess that we forgot to test with immortal men!).
- After we started to ship Joust, we heard the bug happening on one of
the east/west coasts.
- At first we did not believe it and tried to get more
details on how the game was really getting beat. That is, until the opposite
coast also reported the bug and the same details started pouring into
- Actually, the “quick” fix was that Pterodactyl aimed lower, at the legs of the players bird (after all, legs
are tastier!), preventing the Pterodactyl from impelling itself onto the players lance.- Pfutz, May 2013
- Lon is confident there is a difference in enemy dynamics of the Shadow Lord between the 2 romsets.
- John mentioned that Lon’s observation that the shadow lords tend to get stuck riding on the “Tery” in red roms due to the pterodactyl dynamics, would be logical per design.
- During dinner conversation, the topic of a row of pixels in the Pterodactyl came up, but on recounting Bill doesn’t recall this change in design. The author visually inspected the Tery’s between both rom sets and couldn’t identify a difference.
- I honestly do not remember the art change.. It might have happened.
- Again I do not remember any other game changes, just the Pterodactyl. I do remember we were in a rush to find/get a fix into the field (and alot of people had alot of suggestions). -Pfutz, May 2013
- We will leave this unsolved mystery to the future MAME developers in the crowd. An open-ended challenge to identify any coding changes in flight dynamics!
- UPDATE- Sean Riddle confirmed in the program code that the pterodactyl graphic is identical between the 2 romsets.
Rom versions and High Scores?!
- Fundamentally, I am trying to remember why there were 3 colors of ROM’s (Red, Yellow, Green). It could be that one was a prototype (typically, the prototype color is red, for “DANGER: do not ship”) . If so, there are usually 100 or less of them (unless sales had some anxious customers…). The prototype software would also probably explain Lonnie’s statement that the Red ROM’s are easier to play (we were not done tweaking the game…)
- And I do remember using the default high score to indicate ROM versions. It was an easy way to identify the ROM’s without opening the cabinet. I probably forgot to change the default high score when we did the Pterodactyl fix (we were in a hurry….). I think that I constructed the score by a version and something else that I do not remember
High score 107212 = V1.07 ?212?
High score 109102= V1.09 ?102? – Pfutz, May 2013
The CPU in a Joust machine
- ……This brings to mind an interesting story. I was handed a “faulty” CPU chip. I was told a game was built that would not display the players score during the game (it was blank). However, the game played perfectly. You won & lost men AOK.
- And your score would be put in the high score to date tables correctly (as best as they could determine with a blank score during game play).
- This problem was isolated to be the CPU chip. And I was asked to look into this.
- The blitter chip was supposed to stop the CPU when it started. My routine to display the score during the game was so optimized that I would start the blitter, expecting the CPU to stop, then write more information to the blitter chip on the very next instruction.
- However, for some reason this CPU did not immediately stop confusing the blitter chip when I wrote to it again. The CPU would stop if I added 1 nop instruction after starting the blitter.
- Out of the millions of silicon parts inside the CPU that have to all work together, I am still amazed that CPU’s can work. -Pfutz, May 2013
There are Roms, and then there are ROMs
- I am trying to remember the technology(s) used in the ROM’s. Please excuse me if I get something incorrect. But at the time there was fusible link PROM, or EPROM, or even mask ROM. I am not sure which was used in Joust.
- EPROM was guaranteed for a finite amount of years (I think 10 years), but usually lasted longer. I thought that EPROMs might fail by slowly changing bits from 0’s to 1’s (back to its erased condition).
- Fusible link PROMS, blew a small fuse from 0 (erased) to 1. And sometimes the fuses would regrow back (due to electrical charges when accessing the byte) from a 1 to a 0.
- Mask ROM had long lead times to manufacturer and were cheaper in large quantities, but lasted a long long time. However, predicting a game with large quantities months again of production was next to impossible, so I assume this was never used. -Pfutz, May 2013
- Joust Wave layout- http://seanriddle.com/jwaves.html
- Tery bug explained- http://seanriddle.com/ptero.html
- Joust Sprites explained- http://seanriddle.com/ripper.html
- Score actually rolls at 100 million, not 10 million- The score on the screen is 7 digits, but memory has room for 8. The ones-digit isn’t read from memory; a ‘0’ is always printed, regardless of the value. The 8th digit is incremented correctly when rolling over each 10,000,000 points until reaching 99,999,999; then memory rolls over to 10,000,000.
- Waves roll at 99- The wave # is kept as 2 BCD digits. I didn’t see anything odd when it rolled
over; it just went to 0.
Click HERE to return to the WMS Joust page