Topic: Advanced World Gen !Science! (Read 22113 times)

JohnieRWilkins · « **Reply #15 on:** June 26, 2012, 12:33:57 am »

I just confirmed your findings. Here are my results:

http://imgur.com/a/4SyzJ

SAFry · « **Reply #16 on:** June 26, 2012, 01:22:31 am »

@JohnieRWilkins interval, yes thanks for the correct mathspeak, been a while since my formal education. Interesting the similarity between interval 1 and 15 but also interesting how different they are in that the 16-17 world has tundra where all the interval 1 worlds seem to be quite temperate. *edit* I meant all the interval 15 worlds are quite temperate

@knutor anecdotally site cap and SECRET_NUMBER: are supposed to influence necro towers but this remains to be proved. I would have to break out the legends viewer to answer your question properly. I have some more basic questions to answer but hopefully I'll get on to it.

A couple of observations though, at the moment my world gen is ending at 10 years as I'm just looking at geography but I'm fairly certain that undead is not one of the civilisations placed during the 'placing civilisation' phase. There don't seem to be any necro towers at year 10 on the DF mode embark map. Finally, in the past from looking at legends viewer Kobolds do use the 'placing civilisation' phase and quota however they never seem to spread from their initial site. In fact their history often peters out after 80 years and I was wondering if they were even reproducing at all?

More !science! is definitely required but it will have to wait for after work

JohnieRWilkins · « **Reply #17 on:** June 26, 2012, 02:49:40 am »

The necromancers must raise a large horde of zombie followers (they recruit once every 10??? years from old battlefields) in order to build a necro tower. You'll have to gen for longer.

Eric Blank · « **Reply #18 on:** June 26, 2012, 03:55:06 am »

Quote from: knutor on June 25, 2012, 11:41:57 pm

What I would like the most out of Adv Worldgen SCIENCE! is if someone could define the range of possible entities that occupy the sites in the [SITE_CAP:1040]. Just who and how many civilizations can settle in these 1040 sites? I'm positive the two newest occupiers Kobolds and Undead have their own place at the dinner table with the legacy civilizations(human, goblin, elf and dwarf). But do they use SITE_CAP, to determine their location, or some other mysterious mystery.

But just what does the civilization dinner table look like? Must [SITE_CAP] > Caves + Evil regions, to provide for the full variety of Civilization options? It would appear so, if we are to expect Kobolds and Undead ARE actual civilizations, and not just figments of our imagination.

The fact that the default map SITE_CAP hasn't been altered since these two new civilizations have been entered makes me believe, that their placement is not by mere chance alone, but a constraint of CAVE mins and Evil region mins. Thanks, Knutor

What do you mean kobolds didn't exist until recently? They've been here for longer than I have been playing (3-4 years.) and were just having toruble avoiding starving to death early in worldgen without modding in mid- to late-31.xx versions. Are you talking about the legacy version? Does it not include them? I've never actually downloaded it, and can't right now. (because my computer hates downloading things, apparently.)

I know evil regions don't factor into the site cap, becuase those are biomes and considered prior to the accumulation of entities. I can say kobolds have only ever occupied one site per civilization; the cave they start in. I also understand from Toady's devlogs and such that bandits and criminal gangs in cities, as well as necromancers, count as a sort of pseudo-entity that follows different rules from normal civilizations.
I think caves are directly factored into the site cap, though, as I had a lot of trouble previously where I would raise the cave counts to 1600 total but forget to raise the site cap well above that, so it would only place the starting sites for other civs like humans and wouldn't generate further sites.

So, kobolds are a true civilization that is just restricted to their home cave, probably because further caves can't be created after initial placement and all the other caves become occupied by various beasties that the kobolds can't evict the way the game works now. That's still a guess, though. I'd suggest researching how entities and pseudo-entities, and sites are generated would make for valuable science in this. It always confused the hell out of me how the game chose to place sites. The world seeds don't seem to affect it much, especially for necromancer towers. It will come up with a new number and distribution of towers each time, it would seem.

Also, SAFry, did you consider you compiling your research into the two initial posts, or writing a world generation guide based on it? So that it can become useful to any player that doesn't really read into science threads but wants to generate a world to their liking, of course.

vidboi · « **Reply #19 on:** June 26, 2012, 05:22:31 am »

Quote from: darkrider2 on June 25, 2012, 05:28:49 pm

I think the world genning differently is due to the way procedural generation works.

The seed remains the same but the way the seed gets manipulated to create the world is altered by the range of available drainage/rainfall levels. And thus a different world.

It's actually due to the rejection parameters. The decisions the seed generates are not changed by the settings, but how they are implemented are. For you average world a result of 0.5 (from and random number between 0 and 1) would give a drainage of 50. With the settings changed to drainage: 0-1 it'd give a drainage of 0.5. As a result the seed produces the same result for every region but is implemented differently giving badlands instead of deserts for example, whereas maintaining the same world. On other occasions the changes will result in the world being rejected for some reason and thus producing a completely different world.

SAFry · « **Reply #20 on:** June 26, 2012, 11:06:15 am »

@JohnieRWilkins I'd be happy to look into it once I've covered the basic geography research. You're welcome to post any of your own research here or if you do your own thread please drop a link so we can follow it. *edit* sorry that was more aimed @Knutor who wanted more on necros but thanks, I didn't know that about necros recruiting every 10 years, in fact I must admit I know very little about the pseudo-entity types generation that Eric describes.

Quote from: Eric Blank on June 26, 2012, 03:55:06 am

Also, SAFry, did you consider you compiling your research into the two initial posts, or writing a world generation guide based on it? So that it can become useful to any player that doesn't really read into science threads but wants to generate a world to their liking, of course.

It would be good to write a guide, it's like the information I gave in the first section is all pretty much available in the wiki but the wiki doesn't really explain things in an accessible way. You have to read the whole page and comprehend it before you know where to start, it's more of a reference guide for the individual parameters. Only thing is where would I put this guide? Forum posts get lost quick, would it be appropriate to put it in the wiki? Somewhere else?

@Vidboi, with respect it's not the rejection parameters or minimum counts as I call them in my original post. (I may go with rejection parameters in future as the phase is more intuitive) I disabled all the rejection parameters in my 'blank' word recipe and you can see in my screen shots in the top left of the screen there are no rejections. I was surprised that the seed is implemented differently as I believed the same as you, that the world would be the same but drier or colder or whatever. Turns out no.

All this interval stuff is little more then a curiosity though, I can't see any practical application in researching it further.

Tonight I want to see which of the parameters do and don't trigger a different implementation of the seed then if I have time start working on meshes.

Thanks all for your interest though!

knutor · « **Reply #21 on:** June 26, 2012, 11:45:50 am »

Here is a some WG Science, Garath and I worked out from another thread. See Beast parameter below, and notes. I'd take a pocket region world parameter set, something we all have and can recognize from,

C:\Program Files\LazyNewbPack\DF_34_11_win\data\init\world_gen.txt

Code: [Select]

[WORLD_GEN]
	[TITLE:POCKET REGION]
	[DIM:17:17]
	[EMBARK_POINTS:1274]
	[END_YEAR:1050]
#Beast end year, the year when titans and megabeasts begin their reign of terror.  
#Beast refers exclusively to the whole number of surviving titans and megabeasts left. 
#With the exception of 4 and under titans and megabeasts stopping at 30yr.
#Reign of terror starts at 30yr and stops when a rate of 80% is slain.
	[BEAST_END_YEAR:30:80] 
	[REVEAL_ALL_HISTORY:1]
	[CULL_HISTORICAL_FIGURES:0]
	[ELEVATION:1:400:52:52]
	[RAINFALL:0:100:26:26]
	[TEMPERATURE:25:75:26:26]
	[DRAINAGE:0:100:26:26]
	[VOLCANISM:0:100:26:26]
	[SAVAGERY:0:100:26:26]
	[ELEVATION_FREQUENCY:1:1:1:1:1:1]
	[RAIN_FREQUENCY:1:1:1:1:1:1]
	[DRAINAGE_FREQUENCY:1:1:1:1:1:1]
	[TEMPERATURE_FREQUENCY:1:1:1:1:1:1]
	[SAVAGERY_FREQUENCY:1:1:1:1:1:1]
	[VOLCANISM_FREQUENCY:1:1:1:1:1:1]
	[MINERAL_SCARCITY:2500]
	[MEGABEAST_CAP:1]
	[SEMIMEGABEAST_CAP:2]
	[TITAN_NUMBER:1]
	[TITAN_ATTACK_TRIGGER:80:0:100000]
	[DEMON_NUMBER:20]
	[NIGHT_TROLL_NUMBER:10]
	[BOGEYMAN_NUMBER:10]
	[VAMPIRE_NUMBER:10]
	[WEREBEAST_NUMBER:10]
	[SECRET_NUMBER:20]
	[REGIONAL_INTERACTION_NUMBER:20]
	[DISTURBANCE_INTERACTION_NUMBER:20]
	[EVIL_CLOUD_NUMBER:10]
	[EVIL_RAIN_NUMBER:10]
	[GOOD_SQ_COUNTS:1:0:0]
	[EVIL_SQ_COUNTS:1:0:0]
	[PEAK_NUMBER_MIN:1]
	[PARTIAL_OCEAN_EDGE_MIN:1]
	[COMPLETE_OCEAN_EDGE_MIN:0]
	[VOLCANO_MIN:1]
	[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:250]
	[RIVER_MINS:1:1]
	[PERIODICALLY_ERODE_EXTREMES:1]
	[OROGRAPHIC_PRECIPITATION:1]
	[SUBREGION_MAX:2750]
	[CAVERN_LAYER_COUNT:3]
	[CAVERN_LAYER_OPENNESS_MIN:0]
	[CAVERN_LAYER_OPENNESS_MAX:100]
	[CAVERN_LAYER_PASSAGE_DENSITY_MIN:0]
	[CAVERN_LAYER_PASSAGE_DENSITY_MAX:100]
	[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:1]
	[NON_MOUNTAIN_CAVE_MIN:2]
	[ALL_CAVES_VISIBLE:0]
	[SHOW_EMBARK_TUNNEL:2]
	[TOTAL_CIV_NUMBER:5]
	[TOTAL_CIV_POPULATION:15000]
	[SITE_CAP:18]
	[PLAYABLE_CIVILIZATION_REQUIRED:1]
	[ELEVATION_RANGES:36:72:36]
	[RAIN_RANGES:0:0:0]
	[DRAINAGE_RANGES:0:0:0]
	[SAVAGERY_RANGES:0:0:0]
	[VOLCANISM_RANGES:0:0:0]

and add comments to it.

Not sure how to add comments myself, but someone here might. In some files it has to start with a # or be placed within some kind of bracket. Guess one way to make sure, save it with notes in and if it bugs out, its not right note prefix. HEH! I'm really bad at Shark Dentistry. *blush*

Then what we'll have at the end of the day is an outline, or blueprint, or finished product. That would be more beneficial than the Wiki, which like we both can agree upon is lacking.

Incidently, I'd find your original SCIENCE simpler to read if instead of using.. 0-100, 0-0, 100-100 shorthand, that it be written using the actual commands straight from the proverbial horses ~~patoot~~mouth. [DRAINAGE:0:100:26:26] That maybe nitpicky, and I appology ahead of time for being nitpicky, but it would go along way into helping me grasp the science going on. Maybe others can chirp up in disagreement or agreement. Or tell me to shut my pie hole.

Like that the reader can compare it directly to what is going on their own files and worldgens. Sincerely, Knutor

knutor · « **Reply #22 on:** June 26, 2012, 11:48:34 am »

edit: Double post, my bad. Clicked SAVE button too many times during lag. I'm a casualty of a 504 gateway error. Please delete me. Yeah, SAfry. I cannot remove it either. I'll add something. Since I took up the space. I reported this double post to a Moderator, already tho. This might poof.

[MEGABEAST_CAP:18]
# Called Megabeast Cave in DF, but called Cap in worldgen.txt
[SEMIMEGABEAST_CAP:37]
# Called Smimegabeast Cave in DF, but called Cap in worldgen.txt

Appears to be a soft Cap, as both of these have mating offspring potential. Unsure if Megabeasts even use Caves anymore, with Kobold changes. They do use Lairs, which in the current version is only Caves.

SAFry · « **Reply #23 on:** June 26, 2012, 12:13:56 pm »

Interesting Knuter, I like the idea of your protocol of using the actual parameters from the .txt file. Like [DRAINAGE:0:100:200:200] meaning minimum 0, max 100, X variance 200, Y variance 200. I guess I do it differently as I load the world gen recipe into the worldgen.txt file then edit it in DF where each of the parameters are on a separate line and it doesn't look like that but I will try to give the parameters in future.

The idea of an annotated worldgen file is good too but really only useful if you edit the parameters in the .txt file and not in DF like myself. If we do end up writing a guide I think it would be great to include the annotated file even if only for reference. I think you're right, you just put a # in front of each line of text.

*edit* can't delete your post BTW, use it for your next bit of research if it bothers you

SAFry · « **Reply #24 on:** June 26, 2012, 12:44:04 pm »

Also Knuter, I'm not sure about how correct the information in that file you found is, [BEAST_END_YEAR:200:80] means at the year 200 worldgen starts checking the number of megabeasts killed (although this may actually refer to titans) and if this reaches 80% then the world gen finishes. At least that was my understanding of it, I had a recipe once that kept finishing at 200 years, I increased the [TITAN_NUMBER:9] to 360 and the world recipe was able to continue beyond 200 years.

Also see http://dwarffortresswiki.org/index.php/DF2012:Calendar#Ages

orius · « **Reply #25 on:** June 26, 2012, 01:49:49 pm »

Interesting info.

knutor · « **Reply #26 on:** June 26, 2012, 02:04:28 pm »

Refers exclusively to Roc, Hydra and Wyrm.

Nothing else, does that beast kill rate, count when determining premature history stops. We determined that a couple weeks ago with Garath, by putting 0s in and rolling out histories, against 0s in for other megafauna. No history rolls out to a premature end, with 0 for Megabeast Cave/Cap. Unless of course we squelch the kill rate, with -1.

Anotherwords, Titans and Semimegabeasts are not counted towards premature history stops. I thought they were, too. Titans, minotaurs, seemed mythical enough to me. And after seeing a noob slay a 1343yr old dragon and die to a minotur in Legends made me sit back and ponder, boss battles.

SAFry · « **Reply #27 on:** June 26, 2012, 03:02:19 pm »

OK well I used a pocket region with the following parameters:

Spoiler (click to show/hide)

So it has [BEAST_END_YEAR:30:80], after 30 years start checking right?

I set [TITAN_NUMBER:10] and it got to 422 years, note that the end year is 1050 so something cause it to finish prematurely.

Now using the same exact seed I change [TITAN_NUMBER:X] a few times and get the following results:

Titans,End Year
1,30
2,30
3,225
5,286
10,422

I think this suggests a correlation between [BEAST_END_YEAR:30:80] and [TITAN_NUMBER:X].

As I had this pocket world test all ready to go I then put [TITAN_NUMBER:1] back to 1 and set [MEGABEAST_CAP:1] to [MEGABEAST_CAP:10] and it got to 833.

Put [MEGABEAST_CAP:10] back to [MEGABEAST_CAP:1] then set [SEMIMEGABEAST_CAP:2] to [SEMIMEGABEAST_CAP:10], got to 30 years. Set it to [SEMIMEGABEAST_CAP:20], got to 30 years, set it to [SEMIMEGABEAST_CAP:200] got to 30 years.

Well that's interesting, the wiki says [MEGABEAST_CAP:X] and [SEMIMEGABEAST_CAP:X] refers only to the number of caves available not to the number of monsters but this seems to be wrong. It also looks like [BEAST_END_YEAR:X:X] checks for the number of titans and megabeasts but not semimegabeast.

So like you said, you can set titans to 0 and it won't necessary cause a history stop but try setting titans and megabeasts to 0 at once and I bet it will stop as soon as it hits whatever you set X to in [BEAST_END_YEAR:X:Y]

knutor · « **Reply #28 on:** June 26, 2012, 03:04:12 pm »

This topic interests me, hope you don't mind me chirping in again.

Quote from: SAFry on June 25, 2012, 12:00:22 pm

Also one might assume that a 2x2 mesh size is like a small brush and a 16x16 is a large brush but is that true?

That's how I imagine it, too. I've been toying with volcanism and mesh sizes in DF in order to get larger ore deposits to create a reason for using mine carts. But haven't been able to prove anything, because I haven't yet installed DFHack. I think I need that to uncover the subterranean fog.

I'm guessing Toady uses a Cubic Spline formula to draw our worlds.

One correlation, I've found visually pleasing during WG is to hold the variances to the same numeric figure as the mesh.

Lets say 8. A 8x8 mesh for elevation, with 800x 800y for variance.
Lets say 2. A 2x2 mesh for drainage, with 200x 200y for variance.

I try trim subregion total as small as possible, while still getting ample Site Caps for all the possible Civs.. I can easily find an embark region this way, without having to do a bunch of F1F2F3 clicking, which is what is required with 5000 subregions. This also speeds up the embark filter region feature, as less subregions exist to bog the process down.

The three variables, I use large meshes on are Volcanism, Elevation, and Savagery. The others I tend to make small meshes, or no mesh. I'm not sure if this trend is popular or not, but it has been making me pleasing worlds for awhile now.

One thing I did read tho, is changing the mesh does nothing if frequencies are left at 1,1,1,1,1. If I leaving frequencies 1,1,1,1,1, I don't bother with adding a mesh. Knutor

knutor · « **Reply #29 on:** June 26, 2012, 03:22:32 pm »

Cool test, that explains it. The more megabeast caves you have the longer it'll roll.

Look how long it rolled out with 10 megabeasts/1titan. 800ish years.
Then look at how long it rolled out with 10 titans/1megabeast 400ish years.

Titans don't stop the history, they only factor into Age changes. Go one step further, and in that region, with 1 megabeast, open Legends and look at when it passed away. d. is Death. It'll be 400ish. As soon as the rate of deaths for megabeast, and only megabeast, is met, history prematurely ends. No other beasts factor into that. So in theory if you crank megabeasts up to 10000 and they start mating, game will never leave Age of Myth, and never stop, until the end of days, 1050. Don't crank it up to 10000, game years will slow to a crawl and if your PC is as weak as mine, it'll crash DF, trying to find enough Civs for them all to Rampage.

Yup, Wiki is out of date, when 34.xx first came out they took off the worldgen page. *shrug* Someone put it backup, its got some nice content on there, just some info leaves me spinning and bobbing, I guess.

News:

Author Topic: Advanced World Gen !Science! (Read 22113 times)

JohnieRWilkins

Re: Advanced World Gen !Science!

SAFry

Re: Advanced World Gen !Science!

JohnieRWilkins

Re: Advanced World Gen !Science!

Eric Blank

Re: Advanced World Gen !Science!

vidboi

Re: Advanced World Gen !Science!

SAFry

Re: Advanced World Gen !Science!

knutor

Re: Advanced World Gen !Science!

knutor

Re: Advanced World Gen !Science!

SAFry

Re: Advanced World Gen !Science!

SAFry

Re: Advanced World Gen !Science!

orius

Re: Advanced World Gen !Science!

knutor

Re: Advanced World Gen !Science!

SAFry

Re: Advanced World Gen !Science!

knutor

Re: Advanced World Gen !Science!

knutor

Re: Advanced World Gen !Science!