Print

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

[grenedle]

Increasing site cap has DRASTICALLY increased civ survival rates (even without increasing cave amounts). Practically all of them survive and get to an average of 2000-5000 individuals each. By totals, Dwarves or Humans are first/second, then Elves or Goblins are third/fourth. I generated a world up to 250 years and the trend still holds.

I also upped the Rain and Savagery Frequency and the Good/Evil square counts. I haven't really noticed a difference with the former two (although I was already happy with them pre-change) but I really noticed the latter change. I think I might also have more Good than Evil squares when I generate the world I will play.

I noticed that Goblins always seem to have fewer civs than the other sentients. Is this normal, or are they not having enough places to establish civs? I'm worried about running out of goblins. I feel like this is unlikely unless I play in a world for a while, or make an adventurer to deliberately hunt them down. Then again, I've never played long enough to depopulate a civ. If so, how would I go about upping goblin numbers?

Relatedly, when I have 40 civs, the worldgen makes 2-3 goblin civs and 9-10 civs of the other four sentients. I assume that this sort of distribution is common Is it possible to directly affect the distribution of civilizations?

I don't really mind too much about not having a lot of Towers. When I said that I wanted a similar world to Salkryn's request, I hadn't realized he wanted Towers. Although the discussion is useful for when I may want more Towers. As a test, I upped Secrets to 100 and kept all other parameters the same (except for upping uninvited guest numbers as I usually do). I only got 5 towers, versus the 2-3 Towers with the 14 secrets normally in Medium worlds.

[Asin]

Hey. What conditions are needed to generate savannas? I desire to make a world with a bunch to test something.

[Fleeting Frames]

@grenedle: Strangled, then.

One other effect more sites has is that often you will not see a hamlet become a town, or greater one, until more sites are built (this can be hastened or delayed by manipulating savagery, and entirely prevented with (dark) fortresses within 7 tiles of them). It's a bit weird, but it seems like it is rng if the initial hamlet population has capability for township, and if not they need to be purged and recycled through natural death before there's another chance.

For goblins, are you upping your demon number? Goblin dark fortress are founded in -1 by demons striking a spire (more like a circular tower) when making a pact with human or dwarven mean gods. (and not all demons qualify, oddly)
The screenshot I posted on last page had 19 to 81 goblin-elven distribution - I probably didn't have enough demons there, either (didn't check, but convenient given how animal people almost never join goblin dark forts).

Given their relative domination in worldgen, limiting them is nice, and on the other hand for a source of constraints on major sites or visitors to buff those elsewhere they're more easy to place than dwarves, which is also nice.

Other way to affect distribution would be modifying the raws, which has the advantage of allowing mocking with demon number to mess with history or generally more interesting hfs; default is
[MAX_STARTING_CIV_NUMBER:100]

While I'm not precisely sure what you want, other than all civs and multiple starting sites, not wanting towers makes it a lot easier, as it allows one to separate most megabeasts from civs by way of ocean barriers.

@Asin: Answered in Q&A forum.

[Max™]

Oh yeah, that's another case I forgot about for editing stuff mid-worldgen: setting the values low for the first stretch of time can end up pretty easily getting some "barely scraping by" civs which you can then give breathing room with a raised cap.

Started this one out with a low cap, it hit year 103 in a few seconds:

Spoiler (click to show/hide)

Upped the sites/pop cap until about year 300 then I dropped the pop cap to -20000 to see if it would encourage a culling, it didn't really but it did slow the growth/bit by bit chip away at the number of histfigs left at 400:

Spoiler (click to show/hide)

This one I set the site cap up and let it run for a bit before dropping the pop cap to -10000 and leaving the site cap at 725 I think so it ran to 625 pretty nicely:

Spoiler (click to show/hide)

Both were results of the same worldgen:

Code: [Select]

Created in DF v0.43.05.

[WORLD_GEN]
[TITLE:SLIGHTLYMODDED]
[SEED:A4Yk8IUau0SaMcqk6oqK]
[HISTORY_SEED:M4s8WG2qQ84Q2KsoO8A4]
[NAME_SEED:ySM6m2YigAoo4OaIQOug]
[CREATURE_SEED:MgoAmSowQkiAuuuSmEwE]
[DIM:129:129]
[EMBARK_POINTS:1504]
[END_YEAR:625]
[BEAST_END_YEAR:750:-1]
[REVEAL_ALL_HISTORY:1]
[CULL_HISTORICAL_FIGURES:0]
[ELEVATION:1:400:808:808]
[RAINFALL:0:100:202:202]
[TEMPERATURE:-10:30:202:101]
[DRAINAGE:61:100:202:202]
[VOLCANISM:75:100:202:202]
[SAVAGERY:0:100:303:404]
[ELEVATION_FREQUENCY:5:66:0:11:0:444]
[RAIN_FREQUENCY:3:1:0:0:2:3]
[DRAINAGE_FREQUENCY:3:1:0:0:2:3]
[TEMPERATURE_FREQUENCY:2:3:2:1:1:1]
[SAVAGERY_FREQUENCY:4:1:2:3:4:5]
[VOLCANISM_FREQUENCY:3:0:0:0:2:3]
[POLE:NORTH_AND_SOUTH]
[MINERAL_SCARCITY:1500]
[MEGABEAST_CAP:8]
[SEMIMEGABEAST_CAP:3]
[TITAN_NUMBER:0]
[TITAN_ATTACK_TRIGGER:80:0:100000]
[DEMON_NUMBER:28]
[NIGHT_TROLL_NUMBER:0]
[BOGEYMAN_NUMBER:0]
[VAMPIRE_NUMBER:0]
[WEREBEAST_NUMBER:0]
[SECRET_NUMBER:0]
[REGIONAL_INTERACTION_NUMBER:28]
[DISTURBANCE_INTERACTION_NUMBER:28]
[EVIL_CLOUD_NUMBER:14]
[EVIL_RAIN_NUMBER:14]
[GENERATE_DIVINE_MATERIALS:1]
[GOOD_SQ_COUNTS:125:75:0]
[EVIL_SQ_COUNTS:125:75:0]
[PEAK_NUMBER_MIN:15]
[PARTIAL_OCEAN_EDGE_MIN:0]
[COMPLETE_OCEAN_EDGE_MIN:4]
[VOLCANO_MIN:200]
[REGION_COUNTS:SWAMP:0:0:0]
[REGION_COUNTS:DESERT:0:0:0]
[REGION_COUNTS:FOREST:0:0:0]
[REGION_COUNTS:MOUNTAINS:6500:1:1]
[REGION_COUNTS:OCEAN:125:0:0]
[REGION_COUNTS:GLACIER:16:0:0]
[REGION_COUNTS:TUNDRA:0:0:0]
[REGION_COUNTS:GRASSLAND:0:0:0]
[REGION_COUNTS:HILLS:0:0:0]
[EROSION_CYCLE_COUNT:500]
[RIVER_MINS:1:650]
[PERIODICALLY_ERODE_EXTREMES:1]
[OROGRAPHIC_PRECIPITATION:1]
[SUBREGION_MAX:5000]
[CAVERN_LAYER_COUNT:1]
[CAVERN_LAYER_OPENNESS_MIN:90]
[CAVERN_LAYER_OPENNESS_MAX:100]
[CAVERN_LAYER_PASSAGE_DENSITY_MIN:0]
[CAVERN_LAYER_PASSAGE_DENSITY_MAX:10]
[CAVERN_LAYER_WATER_MIN:0]
[CAVERN_LAYER_WATER_MAX:100]
[HAVE_BOTTOM_LAYER_1:1]
[HAVE_BOTTOM_LAYER_2:1]
[LEVELS_ABOVE_GROUND:15]
[LEVELS_ABOVE_LAYER_1:5]
[LEVELS_ABOVE_LAYER_2:1]
[LEVELS_ABOVE_LAYER_3:1]
[LEVELS_ABOVE_LAYER_4:1]
[LEVELS_ABOVE_LAYER_5:2]
[LEVELS_AT_BOTTOM:1]
[CAVE_MIN_SIZE:5]
[CAVE_MAX_SIZE:25]
[MOUNTAIN_CAVE_MIN:0]
[NON_MOUNTAIN_CAVE_MIN:0]
[ALL_CAVES_VISIBLE:0]
[SHOW_EMBARK_TUNNEL:2]
[TOTAL_CIV_NUMBER:24]
[TOTAL_CIV_POPULATION:-20000]
[SITE_CAP:1250]
[PLAYABLE_CIVILIZATION_REQUIRED:1]
[ELEVATION_RANGES:500:0:6000]
[RAIN_RANGES:0:250:1000]
[DRAINAGE_RANGES:0:0:1000]
[SAVAGERY_RANGES:264:0:4160]
[VOLCANISM_RANGES:0:0:528]


[Fleeting Frames]

Honestly, I'm not sure what exact effect pop cap has; #730 had it set at 13k, but maxed out both a major town and three (2 initially) dark forts.

But I just realized a potentially most useful application of site cap: Perhaps, a source of nomadic necromancers. See, they can still create the zombie groups they lead, just not build towers when site cap is too low. Issue is that most necros will go straight to tower or simply stay in site till next war (that said, this can still happen).

Also, was rethinking of human hamlets and townbuilding (as well as humans having support biome:any wetland), and did a quick 10k test with 4 civs on marshland dotted with plains and hills. (Same-ish result if dotted with 40 civs.)
I think this indicates...*town with nearby but not directly adjacent hamlet in 29*...Something? I'm not certain, beyond it being unwilling to use isolated spots for towns normally, being even unwilling to go to them when sinister/serene is in the way (making that jump without good/evil large regions, though). *Confused*

On that note, Max™, any reason you didn't remove forts from dwarves when modding them to make cities?

[Max™]

I added a somewhat altered version of the vanilla fort making dwarf entity and a different one for the cities because I was experimenting with getting isolated chunks of map with mountains between them and then having them linked later with tunnels and liked having the deep/surface dorf civ mix.

Oh another bonus of worldgenedit: 10k is by no means the limit, I ran a world to year 1,111,111 for a test.

[Fleeting Frames]

Ah. Though I suppose there's low chance those two go to war.

And true; I don't think I've seen a town being founded past 1800 years of the founding of initial hamlet, and construct sewers past 2200 years iirc; so 10k was sufficient for my purposes where 1k wasn't.

[Max™]

I think the deep/surface dorfs only end up in territory disputes occasionally, but my entities are all sorts of modded to where I've got various nobles with squads you can join like a lord hearthperson thing for quests and various name changes of all types, plus the ethics are very different from vanilla, like elves taking slaves to encourage warfare more often.

[feelotraveller]

I noticed that Goblins always seem to have fewer civs than the other sentients. Is this normal, or are they not having enough places to establish civs? I'm worried about running out of goblins. I feel like this is unlikely unless I play in a world for a while, or make an adventurer to deliberately hunt them down. Then again, I've never played long enough to depopulate a civ. If so, how would I go about upping goblin numbers?

Yep, increase demon number.  The game by default tries to place equal numbers of civilisations by race at beginning.  Though it is not unusual for goblins to dominate later in history...  If you are thinking of adventuring increasing night trolls and bogeymen would be good too; and vampires and werebeasts for general fun and good times.  Have a look at the defaults for other medium worlds for a rough guide.  Only nearest site of civ attacks you, and it is quite possible for even a couple of sieges to deplete their numbers massively (and forts tend not to run long enough for them to recover significantly).  Better to start another fort elsewhere in the world for a fresh supply of goblins if they are not - or no longer - nearby in significant numbers.

The rain/savage frequency adjustment effects the 'mesh' size basically making the grid smaller (making the map bigger made the grid bigger), and the points of the grid get distibuted according to the following 5 numbers in 20% blocks.  Then the random variations fill in the in between bits but in Quilt the variations dominate.  Trust me, it's a small touch but it does make it a little bit better...

[PatrikLundell]

I've made a script that causes world gen to pause at a number of placed during world gen, allowing manipulative users to tweak their worlds at specific times recognized by DF: placed_caves,  placed_good_evil, placed_megabeasts, placed_other_beasts, made_cave_pops, made_cave_civs, placed_civs, and finished_prehistory. It currently takes no parameters, but stops all all of those places (including when the world is rejected half way through). There is also no way provided to stop it once started: it has to detect the finished_prehistory state, after which it finishes. It is NOT a plugin, so it will have to be run manually for every world gen. I have no knowledge of how to make plugins, but it should be fairly easy to add parameter to only stop at the points selected.
The script acts by requesting a callback for every DF frame, so it will slow things down a little, but the lengthy parts tend to be the calculations in between the frames.

Spoiler (click to show/hide)

Code: [Select]

function worldgen_breakout_box ()
  local state = df.global.world.worldgen_status.state
  local placed_caves = false
  local active = not df.global.world.worldgen_status.placed_caves
  local placed_good_evil = false
  local placed_megabeasts = false
  local placed_other_beasts = false
  local made_cave_pops = false
  local made_cave_civs = false
  local placed_civs = false
  local finished_prehistory = false
 
  local num_rejects = df.global.world.worldgen_status.num_rejects

  function callback ()
if state > df.global.world.worldgen_status.state or
  num_rejects ~= df.global.world.worldgen_status.num_rejects then
  num_rejects = df.global.world.worldgen_status.num_rejects
  active = true
  state = df.global.world.worldgen_status.state
  placed_caves = false
  placed_good_evil = false
  placed_megabeasts = false
  placed_other_beasts = false
  made_cave_pops = false
  made_cave_civs = false
  placed_civs = false
  finished_prehistory = false
  dfhack.println ("Starting new run")
end

if num_rejects ~= df.global.world.worldgen_status.num_rejects then
  dfhack.println ("Reject detected", df.global.world.worldgen_status.state, df.global.world.worldgen_status.placed_caves)
  num_rejects = df.global.world.worldgen_status.num_rejects
end

if active and df.global.world.worldgen_status.state == 9 then
  if not placed_caves and df.global.world.worldgen_status.placed_caves then
    placed_caves = true
    dfhack.println ("Placed Caves")  
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
  end
 
  if not placed_good_evil and df.global.world.worldgen_status.placed_good_evil then
    dfhack.println ("Placed good/evil")
placed_good_evil = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
  end
 
  if not placed_megabeasts and df.global.world.worldgen_status.placed_megabeasts then
    dfhack.println ("Placed megabeasts")
placed_megabeasts = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
      end
 
  if not placed_other_beasts and df.global.world.worldgen_status.placed_other_beasts then
    dfhack.println ("Placed other beasts")
placed_other_beasts = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
  end
 
  if not made_cave_pops and df.global.world.worldgen_status.made_cave_pops then
    dfhack.println ("Made cave pops")
made_cave_pops = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
  end
 
  if not made_cave_civs and df.global.world.worldgen_status.made_cave_civs then
    dfhack.println ("Made cave civs")
    made_cave_civs = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
  end

      if not placed_civs and df.global.world.worldgen_status.placed_civs then
    dfhack.println ("Placed civs")  
    placed_civs = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
  end

      if not finished_prehistory and df.global.world.worldgen_status.finished_prehistory then
    dfhack.println ("Finished prehistory")    
    finished_prehistory = true
local screen = dfhack.gui.getCurViewscreen ()
screen.worldgen_paused = 1
dfhack.println ("The breakout box is now done")
  else
      dfhack.timeout (1, 'frames', callback)
  end
else
      dfhack.timeout (1, 'frames', callback)
end
  end
   
  local started = dfhack.timeout (1, 'frames', callback)
  dfhack.println ("The breackout box is now started.")
end

worldgen_breakout_box ()

[Fleeting Frames]

Huh. Can't think of an use for 1,2,4, but others can possibly be useful to modify in some form sometime. Though I'm sad to say that I currently don't have a reason to use it.

[PatrikLundell]

I didn't put too much thought into the breakout locations, but rather provided those that DF contained internally. Personally I intend to use it (or a superior replacement by someone else) the next time I make a world for playing, which I hope will be around September (my hope/guess for when DFHack catches up with the next DF version). As a minimum I intend to mess with plants and animals without having to make sure I hit <ESC> at the right time.

[EPM]

I've been experimenting to try and deliberately cause bizarre/glitchy map generation. To, thus far, no success, but at least educational failures.

1: Generating a world, then cutting out its feature files from the save data and transplanting them into an entirely different world.
This just led to DF crashing when attempting to load the region. My first attempt was to generate a large region world, and then paste the feature files into a pocket island world. Crash on load. I then limited the feature files to the numbers the pocket world originally had. Crash on load. Generating two different pocket-size worlds also led to crash-on-load behavior. Notably, deleting the feature files entirely seems to cause no problems subsequently.

Future Science: Try generating worlds with identical seeds but minor tweaks made to mesh settings or civs, see if "feature transplants" are made more possible with that sort of high similarity. Beyond that, there may be some kind of header or ID in the feature files that invalidates them outside their correct worldgen.

2: Generating a world with a gigantic HFS bottom layer and minimum layer sizes, including only 1 (max openness) cavern layer. Maximum elevation mesh variance and volcanism was attempted to be included, though the elevation variance had to be scaled down due to endless rejections. Embarking on a region tile with a mountain peak at the lowest possible map was done in hopes of the HFS layer breaching an upper one, but the magma sea seemed to behave itself.

Anyone know ways to force weird/glitch worldgens? My next idea is to try deliberately duplicating various raws.

[PatrikLundell]

You can cause weird things with my regionmanipulator tool, but that's pre embark, not world gen. You can also mess around with plants (and probably creatures) to get surface ones to appear in caverns on embark (I haven't checked whether new ones will spawn, or the darkness prevents that), but I've cause custard apples to grow in the first cavern (as shown by "reveal all").

If you're going to transplant things from one world to another, you need to connect the pieces. It sounds that what you've done is akin to doing a "heart transplant" by cutting out the old heart, dump the new one in the hole, and then stitch up the wound. No re-connection of veins, arteries, or nerves (and, given the world size attempt, no check for tissue type compatibility or blood groups either).
I don't really see the point in trying to move things from one world to another when it's probably a lot easier to modify the destination world structures to be what you want. Rather than trying to transplant the complete set of features, it's probably more fruitful to add/modify the existing features so that the ones you care about (your future embark(s), typically) look the way you want.

[Fleeting Frames]

For 1, I've noticed that sometimes Pre-set value world's can have certain stubborness. Changing an elevation of a tile near river by 1 might result in nothing, the river course changing subtly, or different river map. In first case, you have different parameters result in exact same worlds.

For 2, well there's earlier in this thread mentioned no hfs/magma resulting in caverns open to surface + empty volcano behaviour (unless the volcano is in NW corner).

There's also the supertall cavern behaviour that can be forced to happen with painted strips of 1 elevation ocean - 100 elevation land - 400 elevation mountain all one after another, and then embarking in the 100 elevation land.

Areas with deadly cold or deadly hot can be weird as well, especially adjacent to normal areas - serving as kind of superevil weather where organics disappear, creatures go missing (and can't walk on in case of too hot). I recall vjek managed a reanimating waterless embark where dwarves died if they got caught in normal rain back in DF2014, I expect history would change but geography would stay the same at least in 43.03. There's also vaporizing waterfall floating around somewhere, and colder instead downfall creates a natural water cistern - that's then going to cause spring floods if the colder area is merely "temperate" cold.

Air biomes are neat in how you can have other creatures, undead, reanimation, evil rain, cold areas with tropical areas on the side/down/up. You're always going to get the one in 1 region tile NW of your embark region tile, and possibly other directions as well.

What else, uh...Well, the people in camps you embark on might be hostile, might ignore you or might even petition to join you fort. Last could be interesting RP-wise, though I'm not sure I'd call it weird.

PS: you likely want to remove minimum counts to avoid rejections - they're there so that you'd generate a "standard" world, after all.