Author: Angrim
Date: 03:23:01 02/14/04
Go up one level in this thread
On February 13, 2004 at 09:41:32, Janosch Zwerensky wrote: > >>I took a look at GO a few years ago, my conclusion was that once everyone >>agrees on what the rules of GO are, it will be an interesting project. > >I don't think the fact that there are several rulesets in use each of which >slightly differs from the others makes the project any less interesting (also >note that for the vast majority of positions possible, the same moves will be >strong or weak no matter which ruleset is used). It made it less interesting to me. Before I start writing an engine, I want to know the exact rules to the game. This may be less of an issue now than it was back then, several of my friends from FICS are now working on GO programs, and I am considering just useing whatever rules they decide on. >>Some people say funny things. A century ago computers were the stuff of >>science fiction. > >Were there actually science fiction novels a hundred years ago that featured >computers? I don't actually know, but the point was that if someone wrote about computers a hundred years ago, it would have been speculative fiction. Thus I found it funny that someone would be making assertions about what computers will be able to do given another hundred years. >>(...) >>>First, I think it's quite a bit easier to create an engine that plays Go legally >>>than it is to create one that plays chess legally. Even doing the GUI yourself >>>isn't too hard, since it's mainly just black & white stones on a grid. That >>>should encourage more people to try. >> >>There are multiple rulesets for GO, not only national rule sets, but even >>between different tournaments. Tromp-Taylor helps with this, but last I >>checked that ruleset wasn't universally accepted. > >As long as you don't get a superko situation or something similarly exotic on >the board, all the rulesets give the same set of legal moves. If I write an engine that uses a tree search, and searches a million or so positions per turn, then an "exotic" problem that happens one time in a thousand is a problem a thousand times per turn. Also, there is more to the rules than move legality, you also need to know when the game is over, and be able to figure out the score at that point. While in practice the Japanese and Chinese rules produce scores that are nearly identical, that isn't instantly obvious from reading them, and they add another layer of confusion to someone who is just getting started. >>(...) >>So to make money at GO, I would almost certainly need to learn at least >>one of Chinese, Japanese, or Korean? > >Compared to the problem of writing a dan-level program, any internationalization >issues will be trivial. That wasn't exactly an answer, makeing the program work in other languages might not be hard, but I would still have to talk to the people who run the company that sells the program. >Regards, >Janosch Regards, Angrim
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.