Re: [computer-go] Changing the subject

["Antoine de Maricourt" <antoine.de-maricourt@xxxxxxx>]

I've been thinking about developping a go program for years (since 1990 ?).
At that time I wrote a free SGF viewer under X11, called xgoban (not 
maintained anymore).

3 years ago, I decided to start seriously to write a program (on spare 
time).
I focused on the design of a flexible framework / library for games (so far, 
mainly based on search), that would be game independent, highly modular (I 
wanted any type of search strategy to be quickly implemented and tested) and 
highly efficient (wrt speed).

So far, I only used it for Atari Go (as a testbed) and I reproduced already 
known results for small boards (5x5, 6x6 crosscut) more efficiently in the 
begining of 2004. This was enough for me to decide that the library's design 
was good, and I spend most of 2004 to rewrite the code in a more robust way 
(status raised to experimental and buggy to robust and stable).

I now would like to test more elaborated search strategies, as well as 
bigger board sizes (first with Atari Go, and then with regular Go).

I have absolutely no knowledge about Go. I only played once (with Martin 
Mueller, in a workshop in Cannes, around 1991, because he was kind enough to 
offer me a 9x9 game, and I was very quickly defeated).

I consider search to be one of the key points.

Antoine.

PS: will somebody summarize the information in this thread? Could be 
interesting.


_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/