Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1] 2 3 ... 32

Author Topic: DF v0.47.01+ Worldgen Cookbook Thread  (Read 106609 times)

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
DF v0.47.01+ Worldgen Cookbook Thread
« on: February 06, 2020, 01:41:37 pm »

Here's the guidelines & goals for this thread:

Post your interesting worldgen parameter sets for version 0.47.01+, along with a description of the 'ideal' embark you've found. Pictures of sites or directions to them (so others can find them) are especially helpful, but at the very least, try to provide others with the parameters for your new Fortress Paradise, roughly where your embark was and what you got out of it (features, resources). One man's Paradise is another man's Hidden Fun Stuff, so try to point out if you found any obvious problems (no underground plants, no iron known to playable civ, etc).

Got an embark situation you'd like to see? Ask for it, and perhaps a world that someone else has discarded would work better for you! Found an interesting embark with some unusual feature (Adamantine space elevators, tiny island off shore in the ocean, chasms open from surface to caverns)? Post it, and let others have some fun with what you've discovered!

A screenshot showing the embark location is ideal, if you are posting with a specific embark in mind, rather than just the entire world.
Please use spoiler and code tags where appropriate, to keep your posts a reasonable size.

To get things started, here's my first 0.47.01 parameter set, which features..
One living dwarven civ.
One enemy goblin civ.
Single cavern starts at 130, ends at 126.
Steel friendly (flux, iron, wood)!
Limestone from 132 to 128.
Magnetite, Limonite, Platinum on Z132 in NW quadrant.
Native gold also present (on 123 at least).
Slow filling fresh water from aquifer on 134 SE of center (or under any embark-layer red sand).
Magma Sea from 125 to 121.
Candy present. 
Sinister badlands + desert, adjacent to a Goblin Fortress. 
Dead surface trees, but no undead animals or re-animation, as far as I can tell.
Spoiler: 0.47.01 Desert1 (click to show/hide)
If someone could verify the world generates identically for them, and the embark is in the same location with the same features, that would be ideal.

EDIT:  Here is an HTML form you can save (as an .html file) and open locally (in your browser) for use as a template or starting point for making specific embark feature requests.  There's no logic/validation in this form, it just selects everything with a checkbox and then highlights the results so you can copy/paste that into a new post.  If you think something should be added to this form, PM me.
Also, the HTML form is hosted here, on github pages, and should be usable by anyone.
« Last Edit: August 11, 2020, 10:46:15 am by vjek »
Logged

PatrikLundell

  • Bay Watcher
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #1 on: February 06, 2020, 04:07:29 pm »

Looks the same as far as I can see (win 64).

Note that the surface vegetation death percentage is 100%, so there won't be any new trees, shrubs (for seeds), or grass (for grazing). I don't think death percentage affects caverns, though.

As suspected, there shouldn't be any reanimation, assuming the DFHack mapping of the field was correct.
Logged

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #2 on: February 06, 2020, 04:10:55 pm »

In this particular embark, breaching the cavern is a viable option for grazing grass, trees, etc, yep.
Thanks for the verification, PatrikLundell.

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #3 on: February 06, 2020, 08:50:54 pm »

Checked, looks identical: https://i.imgur.com/UN5CAGT.png

With 44.02, I and PL got different results with same params & no rejections, so this is good. Though there's less stuff here; comparing the previous test parameter set might be not a bad idea (though not necessarily good):

Spoiler: Params (click to show/hide)

Spoiler: Expected result (click to show/hide)

Mind, these are old params without the new addition to wg params. If these match, checking a world with rejections might be prudent:


As for the bad stuff, I already confirmed in release thread that previous worldgens that used to not get rejections can now get some, and get different geography (not just evilness, as was the case with .44).

(All tests done on Lubuntu.)
« Last Edit: February 06, 2020, 08:55:44 pm by Fleeting Frames »
Logged

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #4 on: February 06, 2020, 09:09:20 pm »

That's interesting.  When I just tested the ABCDPOCKET parameters, it's not identical to yours, yet, it generates consistently the same without rejections, on Win64:

It's like... the geography is identical, but the evilness and civ placement is different?  Some of the names are the same, but not the Age or figs/dead/events.

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #5 on: February 06, 2020, 09:17:39 pm »

Thought: Do the results match if one sets evilness and goodness to 0? test

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #6 on: February 06, 2020, 09:24:51 pm »

Unfortunately, nope.

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #7 on: February 08, 2020, 01:24:41 am »

Here's one for 0.47.02 that I've been using for testing, and it seems pretty good.
Spoiler: 0.47.02 4civs3 (click to show/hide)
4civ3 - Embark level: Z125
A few surface trees, plants, and shrubs
Z119, SW quadrant: chalk, hematite, coal.  Steel friendly.
Lignite as well on Z117
Sand from Z124 - 121
Chalk from Z120 - 115
Cavern from Z109 - 105
Magma Sea Z102 - 97
Light aquifer Z122 - 121, (despite the no aquifer claim at embark/from the site finder), trivially walled off. Good fresh water source.
Candy spires x 2
Elves, Humans and Dwarves will come to trade.  Many different visitors will come if you have a tavern/inn.
Civilized World Population

   78 Dwarves
   405 Humans
   192 Elves

   Total: 675

EDIT: same world/embark, but better populations / more civs:
Spoiler: 4civs4 (click to show/hide)
Civilized World Population

   298 Dwarves
   769 Humans
   1640 Elves

   Total: 2707

Similar world, larger populations, with a steel friendly embark:
Spoiler: 4civs5 (click to show/hide)
Z121 limonite, magnetite, platinum (SE quadrant for sure, at least)
Z112 - 108 marble layers for flux
Z116 - 115 (and more) Native Gold
Z106 - 102 Magma Sea
Z116 - 112 Cavern
Candy
Trees on the surface
NE quadrant has light aquifer on Z122 at least, for fresh water.
Civilized World Population

   898 Dwarves
   503 Humans
   1337 Elves
   4020 Goblins

   Total: 6758

Ok, I've never seen this before, maybe others have..
In this 0.47.02 world, as time progresses, all the civs are placed, the goblins start expanding, their evil starts to spread, then... they're conquered.  The evil geography recedes, and their dark fortress is surrounded by monastaries, in the region tiles where the evil geography receded from.
Spoiler: D_CONQ_G_1 (click to show/hide)
The dark fortress is now:
4: Ozudstran, "Ruthlessbiles", dark fortress
   Owner: The Rags of Hoping, dwarves
   Parent Civ: The Flag of Lashing, dwarves
   administrator: Rimtar Figureconvents, goblin
   108 dwarves
   589 goblins
--
Civilized World Population

   971 Dwarves
   878 Humans
   1504 Elves
   1484 Goblins

   Total: 4837
--

I've gone to the monasteries in Adventure mode and completed quests for their NPCs. (retrieved a buckler artifact from an underground dwarf fortress in the northwest)
Earlier today, I saw a similar worldgen where the goblins were eliminated from the world.  Honestly, I've never seen that before utilizing this world layout. 
Invariably, in the past, they dominated everyone/everything, if they existed at all.  Now that their existence is actually variable?  That's pretty amazing, to me.  :o
« Last Edit: February 08, 2020, 11:03:23 pm by vjek »
Logged

DrCyano

  • Bay Watcher
  • For !!SCIENCE!!
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #8 on: February 08, 2020, 11:56:36 pm »

Here's some world-gen parameters I cooked-up to focus your world on certain challenges (copy-paste params into data/init/world-gen.txt and then use Design new world with advanced parameters to use these world-gen formulas):

The Misty Isles
This world is designed to create several full of temperate-climate islands with varied amounts of evil and savagery, allowing you to choose the level of danger for each embark.

For an extra fun challenge, visit your embark locations in Adventure Mode before embarking in Fortress mode, building bridge forts to connect the islands.

Example map:
Spoiler (click to show/hide)

params:
Spoiler (click to show/hide)



The Infernal Desert
Dare you try to colonize the Infernal Desert? You may have once scoffed at trees and water, but here you will miss them dearly.

Good for fortress players looking for a less typical experience or story tellers wanting to tell a "wasteland survival" fortress story.

Example map:
Spoiler (click to show/hide)

params:
Spoiler (click to show/hide)


The Savage Jungle
The story: you lead a team of Dwarven explorers who seek to cut their way from the coast into the heart of the jungle, hoping to pry treasure from the ruins and remains who had come before you, only to be consumed by the ever-hungry jungle.

Example map:
Spoiler (click to show/hide)

params:
Spoiler (click to show/hide)


The Frozen (evil) Volcano
This is hard mode! In this evil frozen land, can you claim the three volcanoes for all of Dwarf kind? Or will you just feed the undead monsters who prowl this Armok-foresaken wasteland?

Example map:
Spoiler (click to show/hide)

params:
Spoiler (click to show/hide)
Logged


A mad scientist necromancer, a peasant with a dream, a secret message inscribed on a gem, a giant bridge.
Come witness the Saga of the Puzzling Sea!
A documented Dwarf Fortress v0.47.xx game combining Fort Mode and Adventure Mode.

delphonso

  • Bay Watcher
  • menaces with spikes of pine
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #9 on: February 09, 2020, 11:49:25 pm »

So I'm trying very hard to get a pocket-sized world with a dead civilization. All the civs I get are dying, but not dead. Is there a quick way to check this, or a way to encourage it in world gen parameters?

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #10 on: February 10, 2020, 01:31:39 am »

This civ shows as dead or dying, and there are no dwarves alive in the world.
Spoiler (click to show/hide)
From what I recall (I don't recall which version), the 'c'iv screen has to be empty (immediately after embark) in order for a civ to be considered truly dead, I think?  That's not the case here, so.. it might still be considered dying.
« Last Edit: February 10, 2020, 11:50:08 am by vjek »
Logged

delphonso

  • Bay Watcher
  • menaces with spikes of pine
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #11 on: February 10, 2020, 03:04:33 am »

I've heard that 'c' thing before, but have never confirmed it for myself. Every world I've generated has been dying according to that - and 3 trials (getting to autumn) has shown that the civ was dying and not dead. Meaning that hasn't disproven or proven it yet.

PatrikLundell

  • Bay Watcher
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #12 on: February 10, 2020, 03:20:57 am »

The empty civ screen civ sub screen is the only known way to establish that DF considers a civ to be dead. Due to bugged accounting DF thinks there are still pop in most civs that haven't had a member for 1000 years, doesn't have any site, etc.
Thus, most civs that should be dead aren't, according to DF's decision process.
Note that this was valid in 0.44.12 and earlier, but nothing I've heard indicates it should have changed.

Also note that DF culls dead civs from the civ selection screen pre embark, which means you can only embark as a truly dead civ if ALL playable civs are truly dead. As a result of that, you should change the entity_defaults.txt entry for dorfs to only generate a single civ, and then you might get a dead one among the first 20 or so that appear to be dead.
The method I used when relying on pure vanilla methods (which is what's available for 0.47.X currently) was to generate a world, see that the civ looked dead, do a default embark at a random location (i.e. just get it over with), check the civ-civ screen, and kill DF. If the civ-civ screen showed the civ was truly dead, I'd then embark properly, otherwise it was time to generate a new world.
Note that you have to ensure auto save on embark is disabled for the method to work.

The world gen encouragement I use is to have many megabeasts and/or titans to get a good chance of the civ being killed off quickly and early, before it establishes a second site. This makes it tricky, though, as those same monsters also tend to wipe out everyone else, leading to a balancing act between dorf destruction and total destruction.

Finally, note that truly dead civs are completely incapable of getting any units sent out to either reach a destination or return. Presumably dead civs don't have to move armies, and so don't move your fortress' armies...
Adding to that, pocket worlds are lousy at providing your fortress with visitors (I've had a total of 1 in a dozen or so attempts before giving up, but others have reported it actually does happen). Thus, don't be surprised if it might turn out that all your visitors are the new nefarious ones...
Logged

delphonso

  • Bay Watcher
  • menaces with spikes of pine
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #13 on: February 10, 2020, 03:56:27 am »

Thanks PatrikLundell - I had just gone back looking at old posts of yours on this topic. I managed to get 3 dead civilizations (according to c>c). I generated pocket worlds, low sites, high beasts. And would test them out whenever DF crashed in world generation (about 1/40 generations). Mostly this was for speed, but ensured there was only 1 dwarven civ per generation, and increased the likelihood of the original site immediately dying.

I got about a 10% hit rate, I think.

I think there's something up with forts (not fortresses) - aside from the infinite fort wars. A couple of these that looked like they should at least be dying wouldn't even give the dead/dying warning on embarking. This only happened on worlds with forts/monasteries - even if they weren't dwarven controlled.

I can upload those dead civ saves, if anyone wants them.

vjek

  • Bay Watcher
  • If it didn't work, change the world so it does.
    • View Profile
Re: DF v0.47.01+ Worldgen Cookbook Thread
« Reply #14 on: February 10, 2020, 12:43:34 pm »

This 0.47.02 worldgen has an empty 'c'iv screen, and no living dwarves.
I remember some details about this from previous versions now, after reviewing some old posts.  If you have enough semi/mega/titans in the world, you can examine the world_sites_and_pops.txt and note that there is a fortress, but no living dwarves, and just animals there.  I call it the "2 dogs, 2 cats" indicator, because that's ~always the first two members of the dead civ fort. 
That, from what I saw this morning, increases the chance of the civ showing up dead.  But if you just have goblins+demons, it has no bearing, doesn't matter, and is not an indicator.
I only mention this as a bit of a potential time saver.
No outpost liaison sent in the first year, no migrants after the first two waves.  Definitely a dead civ.

Another empty civ screen example for 0.47.02; this one with different results in world_sites_and_pops.txt. 
This world still has 2 living dwarves, but the fortress site has no entity entries.  No forgotten beast, no megabeast, no animals, nothing:
Sites

1: Idashoddom, "Moistcloister", fortress
2: Ospurar, "Fountainmaligned", dark fortress

...
Also, this one has just a single goblin civ, and 4 megabeasts. (dragon, roc, hydra, bronze colossus).  The entire world has just 12 sites, total.
Spoiler: empty_civ2 (click to show/hide)

Third example, forgotten beast resident in the single fortress, 27 dwarf prisoners (of 27 total dwarves in the world), dwarf fort (with zero dwarves in it), dwarf camp conquered by goblins(?), empty 'c'iv screen on embark:
Spoiler: empty_civ3 (click to show/hide)

Fourth example of an empty 'c'iv screen at embark, for a dead dwarven civ; 26 elves, no goblins, while 2 megabeasts were generated, neither apparently survived to Year 200, as there are no lairs for them.
Spoiler: empty_civ4 (click to show/hide)
On the plus side, once embarked, you're at war with the remaining elves.  8)

Ok, this one is pretty awesome.  Dead civ, no sentients. Dragon in the ruins of the fort.  I checked it out in Adventure mode, sure enough..  Dragon just hanging out, near the fort entrance.  :o
World has 7 sites total, and 5 of them are just normal lairs.  If you ever wanted to fight a Dragon in adventure mode, as the only hero in the world, here's your chance.
Spoiler: Dead Civ, One Dragon (click to show/hide)
« Last Edit: February 10, 2020, 08:30:52 pm by vjek »
Logged
Pages: [1] 2 3 ... 32