Author: Christophe Theron
Date: 23:26:36 05/18/02
Go up one level in this thread
On May 18, 2002 at 15:35:15, Fernando Villegas wrote:
>Probably there are some pals here that not even know what I am talking about
>with the smashing, mighty word "Colossus". Perhaps they will believe I wanted
>to go to an archeological site and I pushed the wrong address in my favorite
>list. But for, let us say, more "experienced" fans -meaning above 35 years old-
>Colossus is a respectable although obsolete name, a venerable piece in the
>growing colection of the chess programs museum. It was, in the 80's, one of the
>most succesful of the several that were aimed to the mass market, to begin with.
>At least it was succesful enough to let Martin Bryan, his father, to buy a
>expensive new brand car, so he say. That was not a minor accomplishment. After
>all the 80's were the initial, heroic but pennyless era of chess programming and
>besides that many many candidates to glory were in the shells asking for
>attention from a public than, in those times, was even lesser in number than it
>is by now. Well, at least they were prepared to pay sums that for today
>standards seems almost ridiculous. Anyway, most of those products were scarcely
>worthy of his glorious names; we had too much pieces of junk christened as
>"Master", "Great Master", "Champions" and "Experts" with codes of about 50 to
>100 pitiful Kb that scarcely played at 1300 elo, tiny books counted in hundreds
>of moves in the best cases, without the most elemental ending knowledge and
>prone to be cheated with 3 plys combinations.
>It was not so with Colossus. It played a respectable chess and what's more, it
>delivered the very first program -version X for PC- with an algorythm capable of
>some learning in the opening. It did so with a basic device: each time after a
>move his search showed a dramatic fall of the score, it marked that move as a
>criminal to be kept for ever out of circulation . Besides that it had options
>for selective of full width search and a book better provided of lines. With
>this and that, my Colossus IV, one that runned in a Atari console, I remember
>well, was generally capable of beating 60 or 70% of the time my old beloved
>dedicated unit, Chess Champion Challenger. This last one had a 1770 USA elo,
>then I guess Colossus was worthy of about something more than 1800 or so. Not
>bad at all for a engine that examined no more than 100 or 150 moves per second.
>A meditative sould could wonder to himswelf how much could Martin have got if he
>had his program running in the currents processors and with the added experience
>of some 20 years of extra work. But then Martin Bryan left chess programming for
>good. He got bore wit it. Besides he falled in love with checkers and,
>commercially speaking, discovered in it a new gold mine almost empty of
>explorers, so he begun to do some programming on that and in due time he
>produced one of the strongest, if not the strongest commercial checker software,
>called, of course, Colossus. It was second only to legendary Chinook.
>
>But what happened with chess? Did Martin left completely the field? What he
>thnks looking at the new stars and his creations? I made myself those questions
>and I decided to contact the man. He was kind enough to answer and you willm
>concur with me that some of his statements are somewhat contradictory as if in
>Martin are two opposed forces fighting inside him: one, dicided to forget all
>about chess and even to pose as absolute out of it, the other one toying with
>the idea of a eventual comeback.
>
>Some friends here suggested more questions than those you will see here, but a
>problem with my email service made impossible to send in time a second bath of
>inquiries, so I ask benevolence for all that is not in the following note.
>
>Qustions and answers:
>
>
>a) Have you, in these last years, given a look at the chess programming scene?
>Do you know the current trend of it? What do you think of it?
>
>Actually not much really. I do keep a mental note when new program versions come
>out, but I couldn't even tell you who the latest computer world champion was!
>
>
>
>b) Have you given a thought to the idea of a comeback? As you said to me in an
>emai about your current task, at least to give another shot could be a matter of
>fun and good experience
>
>I think a 'comeback' would be the wrong word :-) Any chess program I release
>would be a spare time effort, either for fun or experience. I doubt that I would
>ever get back into full time computer chess programming.
>
>
>
>c) Nevertheless I remember that in our previous intercourse you said you felt
>capable of tweaking anything in your hands to make of it something better. Have
>you tried with your own Colossus?
>
>Unfortunately I've not really found any time to try new/improved algorithms. I
>still do believe however that there is nothing stopping anybody keen enough to
>enhance things if the desire is there.
>
>
>
>d) What is going on with Colossus checkers? It is by now a finished adventure?
>
>I haven't touched the commercial version in about 7 years. However I do have a
>new experimental version. This would eventually be included as one of the game
>engines in the new generic interface I'm working on. I also have a Reversi
>(Othello) engine done.
>
>
>
>e) If ever you venture again in chess programming, in what conditions you would
>do it? Independently, as in the past? As part of a existing company? Which
>techniques you would use? What kind of product you would try to deliver to make
>a difference with the current crop?
>
>I would almost certainly do it independantly and wholly for fun. I work as a
>contractor for a large chemical company to earn my living so I'm not trying to
>make a goldmine out of the games stuff. I kinda feel it's more relaxed that way
>too and I can just enjoy the games stuff for what it is. I don't know that there
>would be anything new in Colossus that would distinguish it from the rest, but
>over the years I have received so many emails from chess enthusiasts who are
>happy just to experience the different playing styles of the programs and asking
>me to come back with a new version that I guess I will.
>
>
>
>f)Anything else about this field I have not asked but you would think fit to
>say, please do.
>
>As I say I don't really follow the computer chess scene much (or even the human
>scene!). In fact I don't even know if Kasparov is still World Champion?! I
>vaguely recall there was some sort of split a few years ago? That's how out of
>touch I am :-) The computer Draughts scene has been interesting over the last
>few years as several new GM strength programs have appeared and Jonathon
>Schaeffer (Chinook program) has now released the 8-piece endgame databases to
>the public which will only improve things further. They're even having a
>computer tourney in Las Vegas soon to decide a challenger to Chinook for its
>world crown. Say hi to all the old fans of Colossus and White Knight for me!
>
>
>
>Cheers,
>Martin
>
>And cheers, Fernando....
Thanks for posting the interview Fernando. Too bad it is so short, but it
deserves its place in the CCC museum!
Well done!
Christophe
This page took 0 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.