Tord Romstad wrote:Thanks for confirming that it works, Dann!
Before people rush to download the binary, I should warn that this engine is not very interesting to general users. It is basically just a heavily lobotomized version of Glaurung, which is probably at least 200 Elo points weaker than its big brother. And unlike most programs, Viper will probably get progressively weaker with each new version. I will keep removing and simplifying code in order to make the program shorter and easier to read, with the goal of eventually ending up with a TSCP-like program with a parallel search. My hope is that this will help beginning programmers who wants to implement a parallel search themselves.
Right now, Viper is nowhere close to achieving this aim. The code is still far too big and complicated. In fact, the current version is probably not interesting for anyone except those who wants a chess engine with a name beginning with "V". This is, of course, the only reason why I released the program so early, and also the reason for my choice of name.
The two losses against Spike are expected, of course. Spike would probably score more than 90% against Viper.
Tord
Thanks for trying to help programmers to understand parallel search and I may look at it in the future but
I do not understand it.
The normal way to write a program is to start with something simple and add knowledge.
Why did you start with something complex and remove knowledge?
Did you start viper from glaurung and removed knowledge from Glaurung to make it simpler?
Another comment:
Personally I think that it may be interesting for me to follow the process of programming and not to see only the final product.
I wonder if there is some free source code for some thinking game(does not have to be chess) when the programmer did not release only one big source code but many source codes with the names
X1,X2,X3,X4,...X300 when for everyone of the source code there is some readme file that explain in english the task that the programmer decided to do in order to get Xi+1 from Xi
I think that something like that may increase my motivation to look at it.
I also think that it is better if something like that is done not for chess but for a game when there is no competition for it so the code is not going to be used for cloning(the programmer may define rules for a new thinking game for that purpose).
Uri
Posted: 23 May 2006, 13:58
