Board Thread:Suggestions forum/@comment-25129884-20160825161951/@comment-26172435-20160827225136

LOTRMod wrote: AlteOgre wrote: Conclusion on v29.2_25 after 100% generation of PF, incl. bordering areas and structures: no settlements at all, notable lag when generating new bordering (sub-)biomes. I stick to my gut feeling that doubling the grid size reduced the effective generation chance with a factor 4. Yes, doubling the grid size had that effect. Would you recommend reducing it a bit again? @Mevans, With the other, more significant improvements you've made, I guess it would be very interesting now to see how the original v29.1 generation settings would work out. On the other hand, from what I saw in the pictures of Obsidianman and Drago, of the earlier high-spawn versions, I'm actually afraid that would lead to the conclusion that will still be not acceptable. I feel the increased grid size doesn't only reduce the number of generation instances that are being checked for fitness, but also that this leads to an even stronger reduction for smaller biomes, that have relatively much 'collision chances' with unfit areas for settlement generation. The value of 24 for instance means that the grid over PF only has about 20-30 grid elements. Over half of them, maybe even almost all, collide with structures and other (sub-)biomes. So, in order to get an acceptable number (4-8) of settlements to generate in PF and Pelargir, I'd say reduce the grid size to 16 to get ca. 60-80 grid elements in these biomes. This may significantly increase the number of suitable grid elements and will make further tuning of the spawn chance (that float parameter you use) meaningful. A test with gridScale 16, gridRandomDisplace 2 and combined spawnchances of up to 1 (like: Vanilla Gondor 0.4F + PF Gondor 0.6F) may give an interesting result. Just my guts ... don't trust them. :)