Re: [gnugo-devel] A speedup

[Paul Pogonyshev]

nando wrote:
> > I think we should look for exact timing results, however.
>
> [...]
>
> Ok, I'm gonna try it, in defend2 and defend3 as well I think. And I'll
> check with fastlib to see if it makes any difference.

in case fastlib gives worse results (which i doubt since it should work
most of the time), you can toss in a slight optimization here:


> +  for (k = 0; k < liberties; k++) {
> +    gained = accuratelib(libs[k], color, MAX_LIBERTIES, NULL);
> +    if (gained > 4                      ^^^^^^^^^^^^^^^

replace with

+  for (k = 0; k < liberties; k++) {
+    gained = accuratelib(libs[k], color, 5, NULL);
+    if (gained > 4                      ^^^

Paul