Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Misterstone]

I have noticed something weird that occurs from time to time in world gen.

A certain combination of 200-300 elevation, 300-400 elevation, perhaps volcanism factors, who knows what else, and a high erosion + rivers count causes super high peaks to appear along where erosion is presumably cutting channels into the ground.  I am assuming that it is a bug, that perhaps erosion goes so deep that it resets at the top value for elevation or something?

Anyway, run these worldgen params and you should see what I am talking about.  Solid triangle peaks will appear out of nowhere during the rivers/erosion process.  It is actually kind of neato, but if its a bug it needs to be fixed I guess.

Code: [Select]

[WORLD_GEN]
[TITLE:NEW]
[DIM:129:257]
[END_YEAR:1050]
[BEAST_END_YEAR:300:96]
[REVEAL_ALL_HISTORY:1]
[CULL_HISTORICAL_FIGURES:1]
[ELEVATION:1:400:401:800]
[RAINFALL:0:100:200:400]
[TEMPERATURE:25:75:200:400]
[DRAINAGE:0:100:200:400]
[VOLCANISM:10:100:1600:1600]
[SAVAGERY:45:100:200:400]
[ELEVATION_FREQUENCY:4:3000:1:1:1500:150]
[RAIN_FREQUENCY:6:1:1:1:1:1]
[DRAINAGE_FREQUENCY:6:1:1:1:1:1]
[TEMPERATURE_FREQUENCY:1:1:1:1:1:1]
[SAVAGERY_FREQUENCY:1:1:1:1:1:1]
[VOLCANISM_FREQUENCY:6:20:20:20:20:160]
[GOOD_SQ_COUNTS:20:150:200]
[EVIL_SQ_COUNTS:1500:3000:6000]
[PEAK_NUMBER_MIN:25]
[OCEAN_EDGE_MIN:1]
[VOLCANO_MIN:75]
[REGION_COUNTS:SWAMP:0:0:0]
[REGION_COUNTS:DESERT:0:0:0]
[REGION_COUNTS:FOREST:0:0:0]
[REGION_COUNTS:MOUNTAINS:0:0:0]
[REGION_COUNTS:OCEAN:0:0:0]
[REGION_COUNTS:GLACIER:0:0:0]
[REGION_COUNTS:TUNDRA:0:0:0]
[REGION_COUNTS:GRASSLAND:0:0:0]
[REGION_COUNTS:HILLS:0:0:0]
[EROSION_CYCLE_COUNT:1000]
[RIVER_MINS:400:400]
[PERIODICALLY_ERODE_EXTREMES:0]
[OROGRAPHIC_PRECIPITATION:1]
[SUBREGION_MAX:3500]
[CAVE_MIN_SIZE:5]
[CAVE_MAX_SIZE:15]
[MOUNTAIN_CAVE_MIN:100]
[NON_MOUNTAIN_CAVE_MIN:50]
[ALL_CAVES_VISIBLE:0]
[SHOW_EMBARK_RIVER:2]
[SHOW_EMBARK_POOL:2]
[SHOW_EMBARK_M_POOL:2]
[SHOW_EMBARK_M_PIPE:2]
[SHOW_EMBARK_CHASM:2]
[SHOW_EMBARK_PIT:2]
[SHOW_EMBARK_OTHER:2]
[SHOW_EMBARK_TUNNEL:2]
[TOTAL_CIV_NUMBER:25]
[TOTAL_CIV_POPULATION:20000]
[PLAYABLE_CIVILIZATION_REQUIRED:1]
[ELEVATION_RANGES:0:0:0]
[RAIN_RANGES:0:0:0]
[DRAINAGE_RANGES:0:0:0]
[SAVAGERY_RANGES:0:0:0]
[VOLCANISM_RANGES:0:0:0]
Pardon me if it has already been reported!

[Havlock]

Well, I can confirm that this worldgen config does indeed do what Misterstone says it does. I've gen-ed up about a dozen worlds using these parameters, and every one has huge mountains spring up from lowlands that should be eroding away, such as near the coastline or rivers. While this is a bug, I suggest finding how much these parameters can be changed before the monster cliffs go away, and sharing that info, 'cause I know there are players that would love such terrain.

[Draco18s]

Wow, that is weird.  I have no idea what could be causing it.  Fiddled with a few things and they showed up (interestingly just changing "periodicaly erode extreme cliffs: yes" does some pretty funny stuff).

[Misterstone]

I am not sure exactly how to tweak it, but I am pretty sure that a certain balance of low mountains (200-300) to high mountains causes this to happen when lots of erosion takes place.  Tweaking those elevation params is the only way I know to turn it on and off at will.

[Draco18s]

Changing the 0-20 portion of the mesh from 4 to 1 reduced their probability (either they'll show up in droves or not at all).

Got one gen that had them with a mesh of 1:1000:1:1:1:1, didn't record the seed and couldn't get it to happen again.

[Keldor]

This actually makes sense, since when you put in the elevation ranges or 1000:1:1:1000 or such, you're telling the world gen to bias the heights so that extra high and extra low locations are far far more common than mid elevations.  As a result, the peak top ends up biased into very high, while the base ends up as very low.  I actually like this, since it makes it possible to have tiles with ocean coasts AND HFS.  Or volcanic islands with HFS...

[Draco18s]

This actually makes sense, since when you put in the elevation ranges or 1000:1:1:1000 or such, you're telling the world gen to bias the heights so that extra high and extra low locations are far far more common than mid elevations.  As a result, the peak top ends up biased into very high, while the base ends up as very low.  I actually like this, since it makes it possible to have tiles with ocean coasts AND HFS.  Or volcanic islands with HFS...

The thing is, it's not 1000:1:1:1000, the original is more like 1000:1:1000:10, but you get large numbers in that last range (I opened up the heightmap, both raw and as a 3D image and those blocks max out the peak hight in a square, like this:

........
........
...#....
........

Where a . is normal land (a middle grey with some variations) and the # is white, pure white (the edges a little fuzzy due to the fractal algorithm).

[Toady One]

There was a mistake in the part where the erosion sim invisible rivers got confused about very flat areas and placed the null value (which happened to be the peak height).  I'm not sure what it was about the parameters that made such flat areas (mostly near the oceans), but the peak-forming bug has been fixed for next time.