> Our pathfinding algorithms are the state of the art; we cannot improve them in
> general,

I disagree.

Considering speed:
Reusing of old information like Bradley and I have proposed might
drastically increase the speed.

Considering effectiveness:
I proposed a heuristic to sometimes produce shorter paths
for "three-point"-pathfinding. We probably can do better.


> although we might tweak them for 1 wide passages.

Yes. We also could implement game features that prevent many 1 wide passages.

Maybe Leo's pavements.  Or how about allowing buildings
to be moved? Let's say globs have to bring in x times "building size" many
"wood" and the building can shift one field, if there are no obstacles.

-- 
Kai Antweiler


_______________________________________________
glob2-devel mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/glob2-devel

Reply via email to