Author: Robert Hyatt
Date: 07:36:53 07/23/01
Go up one level in this thread
On July 22, 2001 at 12:39:57, Roy Eassa wrote: >I made an appointment to work with Richard Greenblatt when I was undergrad at >MIT. It was about 1979, and we were going to resurrect that old program >(MacHack, as I recall) and improve it. I had to get up early on a Saturday >morning (death for a college student!) to make the appointment. Greenblatt, who >had a well-slept-in cot in his office and never made eye contact with me when we >set up the meeting, never showed up. That was the end of my chess programming >career. Greenblatt was a "character" to say the least. The main thing new about the "program" fischer played was that it was the first real effort to use special-purpose hardware to search a game tree. Belle (early versions circa 1977) had hardware move generation and evaluation, and makemove/unmakemove, but the search was software. Greenblatt and others developed a thing called "CHEOPS" (CHEss OPerationS if I recall correctly) that could search very quickly. It had no positional evaluation at all, and was used to accept or reject moves proposed by the normal MackHack chess engine by doing a significantly deeper search on them but only looking at material. It never played in any computer event so we all assumed it didn't work very well, in practice... IE no MackHack version ever played in any ACM event whatsoever...
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.