Re: Woody Woodpecker | The AI 50% Forest Nothing 50% BF
Posted: Mon Jan 04, 2021 12:11 pm
Thanks! This is great, and it seems to function and play in a team game; it's especially good in Feudal to early Imperial Age just because it's harder for a human to micromanage tree cutting.
I'm running a game and everything is going smoothly. (I actually really enjoy watching the AI play forest nothing; it's surprisingly relaxing for me.) I think debug-chat is okay once in a while; however, in Imperial age (but before trading is possible), it seems to repeatedly spam "an ally has a market". It doesn't seem to affect its performance in terms of the other things it's doing, but it is spamming it a lot. Looking at the code, it seems it's only intended to be triggered once or occasionally, not repeatedly -- I'm not sure why the condition "(not (up-point-contains ally-market-point-x c: market))" is repeatedly satisfied after it already set the ally-market-point-x on an allied market.
See screenshot: https://i.imgur.com/QrDiwQH.png
After the path is cleared, it takes a while but most of the AIs eventually figure out that they can train trade carts. However, it's slow because it will only start trade if the specific player indicated by "this-any-ally" (from "(players-building-type-count any-ally market > 0)") is reachable; otherwise it will wait until a path is cleared to that specific player, even if other allies can be reachable. I don't know if there's a way to have it cycle through all the allies. I suspect the "any-ally" search always just returns the same ally if all 3 allies have markets.
I'm running a game and everything is going smoothly. (I actually really enjoy watching the AI play forest nothing; it's surprisingly relaxing for me.) I think debug-chat is okay once in a while; however, in Imperial age (but before trading is possible), it seems to repeatedly spam "an ally has a market". It doesn't seem to affect its performance in terms of the other things it's doing, but it is spamming it a lot. Looking at the code, it seems it's only intended to be triggered once or occasionally, not repeatedly -- I'm not sure why the condition "(not (up-point-contains ally-market-point-x c: market))" is repeatedly satisfied after it already set the ally-market-point-x on an allied market.
See screenshot: https://i.imgur.com/QrDiwQH.png
After the path is cleared, it takes a while but most of the AIs eventually figure out that they can train trade carts. However, it's slow because it will only start trade if the specific player indicated by "this-any-ally" (from "(players-building-type-count any-ally market > 0)") is reachable; otherwise it will wait until a path is cleared to that specific player, even if other allies can be reachable. I don't know if there's a way to have it cycle through all the allies. I suspect the "any-ally" search always just returns the same ally if all 3 allies have markets.
