Print

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

[ArmokGoB]

That shouldn't be too hard now that the problems have been solved. I'll try to remember that next time I have an opportunity to update.

Thanks!

[Fieari]

Would it be possible for the system to allow "broad strokes" designs for a world?  By that I mean that you would paint an area designated as "mountains", and it won't just stick a uniform cliff there, but simply weight the normal random fractal algorithm to be much more likely to stick mountains there, so that it still looks natural.  Likewise for hills, plains, coastlines, and water.  The end result is that you could stick in a sketched out general map, and the algorithm would "rough it up" and make it look naturally generated.

Then instead of requiring a detail heightmap, you could take the map from, say, one of the 8/16bit era Final Fantasies and generate a realistic looking world that ALSO conforms (in a general manner) to the old time map.  Or you could take a map from your D&D game that you drew with pencil and see it realized in game, and not have it look completely like crap with brightlines and whatnot.

I feel like it would work similar to seam carving, and so shouldn't be too hard, but then again, I know that things that sound easy tend not to be when you get down to brass tacks.  I'm just hoping.

[Vattic]

You can sort of do what you are asking Fieari at least for height-maps. I asked about it here and got an answer in the following post with some instructions and further down the result. The maps you input can be fairly primitive with just blobs for different heights.

[cephalo]

Yeah, what Vattic said. You can put some crude windows paint drawing in there and smooth it out automatically. Then add in some noise and you have what you are asking for.

[DizzyCrash]

Is this still active? I cant find any current 2015 downloads for this, but it seems like there are new options.

[Greiger]

Still active as far as I know, and it's not symbols or DFhack dependent like many other utilities so it's mostly version independent.  As long as Toady doesn't make any significant changes to the mapgen code¹ this utility should work just fine into the foreseeable future.  Though as far as I am aware it is completely up to date with the current DF version.

¹ And even last time Toady DID make a change to the map code this utility still worked fine before it was updated, DF just set the missing data to default and it worked hunky dory.

[cephalo]

Is this still active? I cant find any current 2015 downloads for this, but it seems like there are new options.

Yeah just release a new version a couple of months ago.

[thepaan]

First, cephalo, thanks for all your hard work in making this.

I'm putting together a matrix to show the resulting biome(s) for each combination of settings for rainfall, drainage, and temperature. However, preliminary testing shows that setting the slider for temperature % below scorching to 0 and noise control weight to 0 does not result in much or any scorching temperatures. The effect of this is that even with the temperature set to maximum, there are very little tropical biomes present in the world.

Is this a bug, intended behavior, or am I doing something wrong?

[cephalo]

First, cephalo, thanks for all your hard work in making this.

I'm putting together a matrix to show the resulting biome(s) for each combination of settings for rainfall, drainage, and temperature. However, preliminary testing shows that setting the slider for temperature % below scorching to 0 and noise control weight to 0 does not result in much or any scorching temperatures. The effect of this is that even with the temperature set to maximum, there are very little tropical biomes present in the world.

Is this a bug, intended behavior, or am I doing something wrong?

Make sure you set the pole settings to 'None' if you are using PWDF temperatures, otherwise you will get a combination of DF's temperature map and your temperature map.

Also, DF messes with the temperature a bit according to altitude and possibly also drainage. You should get a great many 'scorching' biomes, but there will be some temperate ones mixed in. The name 'tropical' is used for wetter areas like forests or swamps. 

[thepaan]

I just did another test world with % below forest set to 0, % below scorching to 0, and % below dune/swamp to 100. I also ensured that all noise weight is 0 and the pole setting is none. The map has very few scorching biomes; most are warm or hot. Are you sure the temperature has nothing to do with being considered tropical? This map seems to correlate well, having very few tropical biomes, all of which are either hot or more often scorching (I am unable to locate a warm tropical biome). Also, my testing indicates the lack of scorching biomes is the same regardless of drainage. I haven't finished testing lower rainfall settings. do you think that could affect the temperature?

[cephalo]

I had done some testing with DF 40.16 which I had on hand, and I got at least half scorching all over the map, with the rest being warm and hot and a few temperate. Maybe something has changed. I'll have a look at it when I get a chance.

I think 'tropical' is hot or above, but I don't remember.

[thepaan]

I used dfhack to get some region tile info and it looks like the upper end of the temperature range looks like so:

• ??-84 is hot (temperate)
• 85-89 is hot (tropical)
• >=90 is scorching (tropical)

Edit: It looks like these values are only true for worlds without a pole. With one or more poles, DF seems to allow more variance with some temps above 90 remaining temperate and some temps below 85 also being tropical.

Edit2: Another quirk I found involving the poles is that without them there will be no Mangrove Swamps or Tropical Dry Broadleaf Forests. Mangroves seem to be randomly placed anywhere a Tropical Saltwater Swamp could occur, or at least I can't find any parameter that differentiates the two. The same is true for Tropical Dry Broadleaf Forests and Tropical Moist Broadleaf Forests.

[thepaan]

I've finished the matrix I was working on. It shows a breakdown of what sort of biomes you can expect with different rainfall and drainage settings. This is more detailed than the descriptions given in the user guide.

PWDF biome distribution by rainfall and drainage:
http://dwarffortresswiki.org/index.php/File:PWDF_Biome_Distribution.png

I was originally going to use some qualitative criteria such as "many", "some", and "few", but the results of some sections were too complicated. I wrote this script instead and ran the same test world through each combination of drainage and rainfall to get quantitative results.

DFHack script to check every world region tile:
http://dffd.bay12games.com/file.php?id=10725

I did not differentiate between freshwater, brackish, and saltwater for lakes, swamps, and marshes. Neither did I account for frozen biomes; all land biomes which are frozen become tundra except those where drainage is 75 or above (everything 0% on the drainage tab).

The temperature threshold for freezing is -5 and below, unless you have one or more poles. Taiga replaces Conifer Forest (and maybe Broadleaf Forest too) between -4 and 9 inclusive, and anything 85 or above is Tropical. As I stated, the temperature rules become flexible when one or more poles are used. More weirdness when using poles includes apparently random substitution of Tropical Dry Broadleaf Forest for Tropical Moist Broadleaf Forest, and Mangrove Swamp for Tropical Saltwater Swamp.

Finally, cephalo, would it be possible to get additional sliders on the rainfall tab to allow for more granularity with regards to the grassland biomes? Currently you can get a mix of Grassland, Savanna, and Shrubland, but not really a map covered in any one of those. I think 5 sliders instead of 3 would be perfect.

Biome distribution by DF rainfall and drainage values:
http://dwarffortresswiki.org/index.php/File:DF_Biome_Distribution.png

[cephalo]

Wow, what a complicated mess! Yeah, there's a lot of little exceptions and weird rules in there. I gave up on understanding it in too much detail.

One thing does concern me, can you confirm that the pole settings can allow/prevent certain biomes from occurring? What if you use PW's poles? There may be some under the hood percentage finding inside DF, and when you present a map with all one temperature, this percentage finding might fail.

Instead of using all one setting per map, you might try banding without noise. When I did my research, I used a linear gradient from 0 to 100 in one direction for rainfall, and then another gradient for drainage in the other direction. That gave me boxed regions where I could examine the interplay between rainfall and drainage. Some of the other maps might influence this also, but I didn't research it. My temperature research focused on what was 'hot', 'cold', 'freezing' etc. rather than biome.

The drainage map doesn't allow gradients in PWDF, but you can fake it using a map that does, and then copying and pasting the resulting text output into another text file and doing a find/replace operation to change the name of the row tags.

[thepaan]

Thanks for the pointers. I finished running through a bunch of tests and discovered some things.

First, Mangrove Swamps appear to be Tropical Saltwater Swamps where drainage is 9 or less. The reason they don't show up when a map is generated through PWDF with poles set to none is that all tropical region tiles generated are inland. They are either freshwater swamps or not swamps at all. This behavior does not change between maps where the entire world is set to scorching and maps where a pole is simulated by PWDF alone. Using either the all-scorching world or PWDF-simulated poles world, and adding either a north, south, or north and south DF-generated poles, produces Mangrove Swamps.

Second, I continue to be baffled by the Tropical Moist/Dry Broadleaf Forest. It seems that the pole presence does indeed influence this biome. One map, using the gradients you suggested and the world temp map set to 200, had Dry and Moist variants adjacent to each other with north, south, or north and south poles. The same map had only Moist variants with no poles. Here are some of the values I pulled from one of the mixed maps:

Region  x:223  y:205    Tropical Moist Broadleaf Forest
elevation      = 130
rainfall       = 100
drainage       = 89
temperature    = 66
salinity       = 0
is land?       = yes

Region  x:220  y:206    Tropical Moist Broadleaf Forest
elevation      = 122
rainfall       = 84
drainage       = 88
temperature    = 82
salinity       = 0
is land?       = yes

Region  x:228  y:208    Tropical Moist Broadleaf Forest
elevation      = 118
rainfall       = 84
drainage       = 91
temperature    = 82
salinity       = 0
is land?       = yes

Region  x:227  y:211    Tropical Moist Broadleaf Forest
elevation      = 126
rainfall       = 100
drainage       = 91
temperature    = 66
salinity       = 0
is land?       = yes

Region  x:223  y:204    Tropical Dry Broadleaf Forest
elevation      = 145
rainfall       = 100
drainage       = 89
temperature    = 66
salinity       = 0
is land?       = yes

Region  x:222  y:204    Tropical Dry Broadleaf Forest
elevation      = 147
rainfall       = 100
drainage       = 89
temperature    = 66
salinity       = 0
is land?       = yes

Region  x:224  y:205    Tropical Dry Broadleaf Forest
elevation      = 121
rainfall       = 94
drainage       = 90
temperature    = 72
salinity       = 0
is land?       = yes

Region  x:225  y:205    Tropical Dry Broadleaf Forest
elevation      = 110
rainfall       = 94
drainage       = 90
temperature    = 72
salinity       = 0
is land?       = yes


The final thing I learned is regarding the temperature. With one or more poles set, biome distributions get very weird. For example, I found a Mangrove Swamp with a temperature of 27.