So wait, trees do not respawn?

[loftar]

In hindsight, I think I vastly overestimated the strain that natural forest regrowth would put on the server; redoing my calculations seems to indicate that it oughtn't be much of a problem at all. However, I empathically agree with Rift that cutting down an area should have severe effects -- I wouldn't want a situation where trees are completely ubiquitous. If anything, I'd implement the exact opposite of Coriander's suggestion, so that trees regrow faster the less civilized an area is. I'll have to keep pondering the issue.

[sabinati]

If anything, I'd implement the exact opposite of Coriander's suggestion, so that trees regrow faster the less civilized an area is

i agree wholeheartedly.

[Antioch]

can't you just make a Cumulative distribution function with the ammount of trees on the Y-axis with 0 as minimum and the ammount of forest tiles x a specific ammount as maximum? This way clear cutting an area will severly reduce tree growth and it is fairly realistic

so tree growth would look like this:

edit: (removed pic)


with X time and Y ammount of trees in the forest

[loftar]

I'd probably use the current "map decay" loop, which picks out a point at random every now and then, and spawn a tree there only if 1) it's a forest tile, 2) there are other trees within a certain radius and 3) there is nothing else within another, smaller radius. That way, forests would only grow back of there are trees left, it would only grow at its own edges, and it would reach a certain maximum density. What I like the least about that manner of implementation is, really, that I'd have to "fine-tune" the target tree density so as to match as approximately as possible what the mapgen produces, which seems kinda ugly.

[Peter]

Ugly smugly. Just have the worldgen run 5,000,000,000 cycles of that treeplanting loop given several seed trees.

[deathc4]

Good lord my firs are growing slow. Its agonizing. I NEED ME SOME WOOD STAT!