Re: [gnugo-devel] Re: Re: zergling

[Arend Bayer]

> There are many experiential definitions in his post. And he must write many
> many codes to implement it.
That is true for any Go program. There are currently 46720 lines of code in
GNU Go's engine. Have a look at the manual, where most parts are described.

>          C) at each point ask question - can enemy place a living stone
> here?
>             [Go4++]
> 
> I prefer c, because it is just the meaning of territory. And A and B is
> experiential.
But then you need a VERY fast life and death reader. This is not easy
to program.

> make some current useful module better, and look for a good architecture.
> And I donst think regression test is
> suitable if you focus on pattern-based search. Perhaps it is better to
> create a test suit using some
> dead/live puzzle of different difficule level. I have thousands of puzzles,
> if u need, I can send it to you.
This is exactly what the regression tests are doing. There currently ~500
life and death puzzles in the regression test suites. Look e.g. at the file
regression/owl.tst, you can also look at the regression failure html-views
(link on the web page).

However, I agree it would be very useful to have more such tests. If you can
send us such test suites, that would be extremely useful -- not only for
GNU Go, all go programs can use these files if they implement the
protocol.

> I run gungo --life at 9*9 board. but It is running about ten minutes but no
> move is generated. :(
The --life code is experimental and has, as I understand, been abandoned
because of its slowness.

> If computer go's local search is weak then people, then it is no chance to
> win people in global scope.
You are very welcome to try a different approach. GNU Go is free software.

Arend