Who wrote the patch ?
It is Larry DeMar who created the patch using the listing of the diagonal explosion that I commented while travelling by train fromQuebec citytoChicago. I explained verbally the problem, then offered to Larry this colorful manuscript which was about 6 pages.
Instead of disabling the protections like I did for my own tests, which would require modifying 6 EPROMs, Larry preferred to use his algorithm to re-compute a new checksum that satisfy simultaneously each of the 6 internal protection. The patch for the diagonal explosion is located in only one EPROM, the second EPROM patch is needed to complete the updated checksum.
You may deserve to know: if any of the 6 checksum thread find a discrepancy, either because an illegal modification have been made or because of any hardware problem, the retaliation consist in corrupting a random part of the RAM. The result would be random reset or any other strange behavior. The would-be thief trying to counterfeit a Williams video game would spend a large amount of time trying to understand why any change even in only copyright text would crash the computer. Obviously, they gave up, so Williams could make more money than other company thanks to the genious of Eugene Jarvis and another Williams engineer that Eugene considered as more brilliant than himself (maybe Mark Ritchie?, I forgot the name…)
Larry sent me a copy of the 2 EPROM by mail, he few weeks after my visit.
Eugene was willing to give me one of the 6809 based development computer that was used to design Robotron or Joust or … However, as I could not carry this heavy machine by train, I did not get this museum article.
Larry gave me one of the 3 PCB that was used as model for the special chip. This 40 pin chip which act as fast X/Y DMA, to copy an object from EPROM to video RAM is the actual integration of approximately 40 standard TTL chips. It uses synchronous 4 bits counters, 2 of them cascaded to form 8 bit (for size in X and Y) or 4 of them cascaded for 16 bit counts. The schematic would look very similar to the section of Robotron board which generates the horizontal and vertical trace for the PCB. The DRAM chip is refreshed by reading each pixel to display on the CRT. The vertical lines are actually incrementing by 1 count in actual RAM address and the horizontal in incrementing by 512 in order to read a distinct RAM page during the display of each pixels; The Williams video were quite advanced compared to all others, they were the only one to use DRAM for more than 10 years.
Larry printed the source code of Robotron:2084 V5. For that, he used his own 6809 computer connected to a dot matrix printer with 3 heads. It took less than 1 hour to print this. Meanwhile, Larry showed us his prototype of a revolutionary pinball machine that he was developing in his apartment. High Speed is the first pinball machine which tries to detect bad contacts and change dynamically the rules to allow good players to complete a scene even when a specific target would not to be touched by the balls.
Naturally, Larry had a Robotron, and a few other machines, in his place. I have pictures ofEugene sweating on Robotron like any good player in a typical arcade.
Click HERE for the next page in the story
Click HERE for more technical information about the shot-in-the-corner patch
Return to the main page for Christian Gingras HERE