Author: Bas Hamstra
Date: 06:31:52 04/26/00
Go up one level in this thread
On April 26, 2000 at 03:59:17, Tom Kerrigan wrote: >On April 25, 2000 at 10:06:38, Bas Hamstra wrote: >>To make a 2150 program is a piece of cake. But to make it progress to 2350 is >>years of hard work. I think there is hardly any PC top program with less >>development history in it then at least 5 years. >I don't think writing a chess program is a piece of cake at all. Simply >understanding the rules of chess is not very easy. Implementing the rules in >a reasonably fast, accurate (!) way is difficult. Judging from students I >know, I think it would take 3 semesters of a CS degree program before someone >is equipped to tackle the problem. On top of that, you have to know about >alpha-beta. I have met a number of CS grads who remember that minmax/alpha >-beta is an algorithm but can't remember what it's used for, much less how to >implement it. And I'm sure that even the best AI classes would not cover stuff >like quiescence searches. Of course, now that massive online resources exist, >people can simply surf the web and immediately see exactly how to do this >stuff. It took me about a year to write a solid program (I started in ~1994). >Now I get all sorts of e-mail from people who download TSCP, read it >carefully, and write their own strong programs in a matter of days or weeks. >It's a little annoying, but I guess that's progress. After having thought about it I think you are probably right. If you have no experience, it is not a piece of cake. And if you are willing to figure out most of the details yourself, it will take you a lot of time. But then you will understand things that you will never understand otherwise. The question is where do you draw the line. Are you going to re-invent every little wheel? Well, no. On the other hand, for me there is absolutely no fun in copying chunks of code from other programs or even looking at other sources. So personally I draw the line at ideas from other people in "normal language". If someone explains to me how to do a hashtable, it's something else to me than copying code. Anyway, if you have gone through the process of figuring it all out once, to write a new chess program from scratch is a piece of cake. For a 2150 rating it would need little eval and complexity. Regards, Bas Hamstra.
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.