Re: [gnugo-devel] do some work for gnugo

[Arend Bayer]

Hi Kevin,

we are always happy about more contributors, great!

I think the most productive thing to do is to browse through
http://www.public32.com/regress/, pick some failures that annoy you
most, and try to understand what is going wrong. It's definitely a good
way to learn GNU Go's weaknesses.

The TODO list is worth looking at but has mostly fairly general items.

A pretty self-contained project that comes to my mind would be to
rewrite the persistent caches in such a way that the algorithms don't
take O(cache size) for lookup etc. The code for this is all in
persistent.c.
If they were O(1), we could increase the persistent reading cache size
for a nice speedup.

Arend