Bay 12 Games Forum

Please login or register.

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

Author Topic: Why number of sites explodes in only 20/30 years....?  (Read 2053 times)

Nilsou

  • Bay Watcher
    • View Profile
Why number of sites explodes in only 20/30 years....?
« on: August 15, 2017, 10:56:21 pm »

Ok, I was trying a little experiment : Is dwarf fortress stable without the population/site cap during world gen... ?

I mean, it should, with a choose of good initial condition... you except that war between populations regulate each other. And you except that as pops have to make babies and dies (except for gobelin and elves but this could be compensated) that the explosion of population should be verrrryyyyy sloooowwwww.

Annnnnd surprise ! This is absolutly not the case. In only 20 years i had an explosion of sites on my maps... and not specifically gobelins ...

So, i really don't understand some points, here's my questions :

- What is the starting population of each civs in theirs first site at the beginning ? Is this moddable ?
- Is populations REALLY controlled by baby's born and death during worldgen ?
- How many dude is necessary to form a new site ? Is this number moddable ?
- How small site with only some dude can resist just near big gobelins pops stack ???? Is it a kind of hiden logic ?
- What define the ability to one race to reproduce ? And more precisely :
                        - Does the age play a role in the ability of reproduction (i imagine child does not reproduce so child age play roles, but what about old men ?)
                        - What define the probability for couple to form for each races ? [ORIENTATION] ?
                        - Is LITTERSIZE the only parameter that defined number of baby when baby borns ?
                        - What defined the time where female are pregnant (does the game have a pregnancy notion ??)

- What about food ? Does it facilitate population expansion in world gen to have the NO_EAT tag ? What does regulate the cunsumtion of food for creatures ? I have no information on this point, is this hardcoded, related to size ?
- Is it possible to make creatures eat other thing that vegetables and meat ? (like other materials, metal etc ... ?)


Thanks you ;)



 
« Last Edit: August 15, 2017, 11:33:21 pm by Nilsou »
Logged

Nilsou

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #1 on: August 15, 2017, 10:58:21 pm »

Erroneous post. Please ignore
« Last Edit: August 16, 2017, 01:31:18 pm by Nilsou »
Logged

KittyTac

  • Bay Watcher
  • Impending Catsplosion. [PREFSTRING:aloofness]
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #2 on: August 15, 2017, 11:13:36 pm »

(How do you delete message on this forum ?? :p)

Ask Toady or ThreeToe to remove it if you reaaaaaallllllly want to remove it. Otherwise it's best to not bother the developers.
Logged
Don't trust this toaster that much, it could be a villain in disguise.
Mostly phone-posting, sorry for any typos or autocorrect hijinks.

Nilsou

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #3 on: August 15, 2017, 11:34:11 pm »

No problem, it's just for my second error message though, my first question message is a real set of questions XD :p
Logged

Nilsou

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #4 on: August 16, 2017, 01:28:54 am »

Erroneous post. Please ignore
« Last Edit: August 16, 2017, 01:31:27 pm by Nilsou »
Logged

PatrikLundell

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #5 on: August 16, 2017, 04:03:08 am »

You can edit your messages using the Modify link. If I understand it correctly, your second post really wanted to say you didn't intend to make that post, rather than that you didn't want the original post, as it normally is interpreted as. If so, just edit it to say something like. "Erroneous post. Please ignore".

Assuming you actually want answers to the OP:
1. entity_defaults.txt has a number of parameters for each race indicating max site pop, max number of civs, and other such things. Look these up and check the wiki for what they control.
2. Probably not.
3. Don't know.
4. It's a result of the somewhat dubious world gen combat process. If I understand it correctly each side essentially form a line and the front character performs a duel to the death against the other side's front character. Repeat until one side is eliminated (or possibly some kind of yield condition is reached). It's far from ideal and is going to be replaced eventually (but "eventually" does not have to be this decade...).
5A. Being adult. There is no menopause for either dwarves or humans, so they're reproductive up to the end.
5B. The history logic is different from the fortress one (there's rampant incest, for instance). I don't know what the logic is, but it probably doesn't check most of the fortress mode restrictions.
5C. I believe there's some parameter controlling how frequent multiple child birth is.
5D. In fortress mode females get pregnant (can be seen with DFHack, but no other indications). I think the pregnancy period is fixed at 9 months throughout, but I'm not certain. I'm not aware of any parameter.
6. Don't know about the history phase. The raw for each creature defines what they eat (such as grazers, which require grazing in fortress mode). Grazers have a common parameter that determine how much they need to graze, but the base factor is size, with the parameter being a multiplier. Apart from grazers, only citizens need to eat (in fortress mode: I don't know about adventure mode), and I suspect everyone eats the same, but I don't know. Get yourself a number of large and small animal people citizens and observe them, but I suspect they'll all eat one meal at the same intervals.
7. Better asked in the modding forum. I have no idea.
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #6 on: August 16, 2017, 04:37:59 am »

You need to have patience, for the DF community isn't that big. Truly, game's fame far exceeds its grasp.

Also, large question. PatrikLundell posted before I did, so there's some overlap...

Now, as for your question, I don't know everything - I have barely modded atm - but to answer you...


Starting population for civ placement is 100 for each of the civs, expect for dwarves, who start with 200.

The minor sites (hillocks, small forest retreats, dark pits, hamlets) are usually around 100ish population, which I presume is controlled by
Code: [Select]
[MAX_SITE_POP_NUMBER:120]in entity.default. However, I've seen thousands - up to 4200 - in minor sites on rare occasions. Not certain what causes those.

The major sites are a little more complicated. First, some attributes shared by major sites:

 • Past civ placement phase of worldgen, you can't establish a major site in 7 tile range of another major site, even if the first major site is destroyed.

• Major sites get access to the resources of all subordinate sites oof their range. Their markets have better, greater variety of choices. Though I don't think goblins sell anything....

 • They're a target for immigration of animal people/proto-civilization joining. Large biomes give more animal people, and they can do so across the oceans. Though goblins only get their pets.

Ω - (dwarven) fortress. Seems to have cap of 200 dwarves in it, but libraries and joining animal men/protocivilizations can inflate your world's fortresses sites and pops .txt  waaaaay past that. Most common seem to be gorlaks and plump helmet men, of which I've seen thousands of each.

They also often die relatively quickly to FBs (within a century) in worldgen unless they get to build a hillock nearby, which will repopulate them after a rampage. After hillocks are built they can further build dwarven mountain halls Ω, which are entirely underground settlements.

Dwarves always start with a major site - very convenient for restraining human towns and elven retreats.

R - (goblin) dark fortress (R is closer to curses cabinet than pi symbol).
Usually gets a structure of 10k goblins and 2527 "pets" - beak dogs, trolls and ogres. This takes around 2-400 years and can even occur if the site is completely isolated - say, by being placed in the middle of evil area with 7 tiles of terrifying plains or ocean all around.

It is not just the immortal, sleepless, not thristy or hungry goblins that give these their power. Being led by a demon leader is a considerable boon, which allows them to usually survive and thrive even in megabeast-heavy worlds; unlike the other races.

Should the site be conquered - by, say, elves with giant elephants - the resultant site will still be able to have 10k civilized population, expect it might now be 10k elves.

Goblins always start with one, much like dwarves. Though they can be put in much wider variety of locations.

- Elven forest retreat.
The civilized population, while inconsistent, seems to range from 400 to 3000 for longer-term worlds. I'm not sure of the mechanics.

Elven forest retreats tend to have numerous animals available, though some get capped at 10 or 35 pop. Others rise with population much like the goblin pets, which means hundreds for .

+⁎# - human town.
Probabably most complicated site, the town is built in two phases - first, the keep, and then, the sewers.

Keeps typically have 500-1500 population and + or symbol.

Once the sewers are built, the population explodes into # or , which have 10k population cap much like R. This can happen from breeding, but if it took a bit the immigration is going to be main factor. If you have two isolated islands, one getting a dark fortress and other a town, then the time the sewers are estabilished is going to be only time dark fortress population drops by the thousands, resulting in partially-goblin race town. Elves and dwarves can immigrate too, but they won't be doing it in so great numbers.

With first keep, you also get the first temple in the world, which means the start of werebeasts and vampires - unless you set the numbers of those to 0 in advanced worldgen. Only towns get temples.

A vampire, or a cabal of vampires, being one of the few things that can keep the lid on town's population. If your town is struggling and goes to + from #, it's likely there's at least one vampire with up to tens of thousands of kills present.

New sites.

Honestly, if you have unsavage suitable biome nearby with no adjacent tower, vault or other such thing blocking it, expect them to build one if they can breed.

Unless, you set the site cap in advanced worldgen to some positive number. This has the effect of preventing the founding of new sites (including towers) once that number of sites is reached). Civilizations will still be placed if you, say, set it to 1 with 4 civs, but they'll never make towers or smaller offshoots in that case. There is also a max civ pop, but I'm not sure what effect that has.

As sites reduce savagery near them, I find it better to restrain site building with pre-set values. Here's some numbers:

30: Calm. Given 0, 10, 20 and 30, human civs in civ placement phase tend to be placed on 30 calm plains the most for some reason.

67: Not going to place a civ here, but when a site gets nearby it is eligible shortly.

73: Take a bit of time to break through, around 400-700 years though can be up to 1200 on rare occasion (you'll probably get them to conquer another site and build from there sooner if they have the range).

90: Next millennium is a likely possibility. Up to 1800 years, though you're more likely to see half that.

100: Take up to several millennia to break through, if this is all you have to work with from a dinky hamlet.

Bigger and multiple ones reduce faster. Also, elves don't reduce the savagery of joyous wilds, and can settle on there fine - use partial taiga instead if you need a large forest region for good biome.

o Death.

Your small sites may not just be attacked by goblins - they may be at war with others at the time, or even. themselves.

Need for food and death reduces site populations. Childs also do not reproduce, but old men/women do. [BONECARN] is a tag that limits creature to eat just meat, which adds a complication to Kobold Camp.

Can't say much on the effects of orientation. All usual rules, such as age gap, are not followed. You can have werebeasts who give birth to themselves and then get cursed, though usually the marriage waits until they're an adult.

Instead of littersize, you could use egg laying. Kobolds do manage to reproduce, so it could give you more granular control, though fortress mode would be a hassle to manage.

Yeah, the game has a pregnancy notion. Sentient beings are pregnant for 9 months, non-sentient 6. Birds lay fertile eggs for that duration instead, which take a season to hatch. I'm not sure pregnancy duration is tracked for histfigs, though - I've seen no miscarriage in legends, and can't find a data structure for it.
« Last Edit: August 16, 2017, 04:42:41 am by Fleeting Frames »
Logged

Nilsou

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #7 on: August 16, 2017, 02:23:56 pm »

Hi, and thank you very much for your answers !

Sorry for the erronous posts, i didn't mean to up the topic artificially it was just misclick while editing ;)

A LOT of very good informations shared by you two guys, and, if i resumed some point and understand some others, i can ask new questions :p :

- The mechanism of army invading settlement doesn't exist ? Am i right ? If i understand it's a kind of line between civilization with type of individual artificial combat ...
Does it mean that armies are only tracked after world gen ??? And, is the "neighbours" a hard criterion for that ? Is it really impossible to attack for an army that is not included in the neighbours list or is it a soft criterion which is only informative ? Does this criterion applied for this "combat line" mechanism ?

- Eating seems to impact the population in your answers, but one more question : is the production of food simulated by site and how ? Because, i guess, if the production is so easy (like in fortress mode XD) i imagine that it does not reduce a lot the population ...

- As i understand of your answers, all the subtle management of couples, like ORIENTATION and necessity of shared the same god, are not used during world gen. BUT, according to some point of the wiki, if we create civilization of homosexual or purely assexual dwarves, they all died quickly in world gen... if it's the cas does this mean that ORIENTATION is fully taken into account finally and that we could regulate with precision population with it ?

- The more critical point of your answer is the number required for the generation of a site. I have no clear answer about that into your both answers concerning funding of new site (i miss it maybe ^^), but clearly i can't assume that the maximum for a site is necessary to found a hilllock, for example. What is the minimum number for funding one of those, 7 dwarf ? :p ? Did the population really substracted/tracked from the mountainhome to the new hilllock ?

- Because it's this point the "problem". How, in 100 years, could exist 30 new human/dwarf/whatever hilllock without depopulating drastically the first town ??? I understand that elves/gobelins don't die so it could be logic, but in my tests human and dwarf spread quickly too... and it is kind of strange that these hilllock survives. Beside, how could these multiple little small town survive to megabeast and other thing like that. Did the megabeast/were beast/titan limitation of population/etc... to attack also applied to ALL the site during worldgen, regardless of civs and not only to yours in fortress mode ? If it's the case, could it be possible to regulate populations by lowering the criteria that trigger attack of titan/megabeast in worldgen ?


Logged

PatrikLundell

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #8 on: August 16, 2017, 03:35:41 pm »

The army engagement logic I mentioned is applicable to worldgen history, and I suspect it applies to non fortress battles in fortress and adventure mode as well. However, in an active world there are armies moving around (they were added early during 0.40.X: before that fortresses were attacked by armies out of thin air).
The pre embark neighbors list is currently lists at most one civ per race, and only those civs can attack (and trade with) you currently. In the next release others can attack you as well, and I think the range restriction is removed or at least relaxed.

- I've got no info on food production impact.

- I've never heard of a requirement to share the same god for couples, although doing it might increase the likelihood of a match slightly. I don't know how you'd regulate population with orientation flags, since that's a static raw setting (apart from ensuring non reproducing non aging creatures eventually die out).

- Fleeting Frames has some knowledge of the site founding mechanics, while I don't. The tracking of non histfig populations is currently buggy, however, so even if there is some accurate logic it might not apply fully due to buggy input data.

- Titans/megabeasts attack everyone, which is a reason a lot of sites get razed even without the help of goblins or humans. Again, you can't really "regulate" populations with monsters, as they attack when they feel like it and who they feel like, so they're random killing factors. Also, I doubt they respect attack triggers during history, since that's a logically odd feature created to give fortress players some breathing room.
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #9 on: August 16, 2017, 06:22:38 pm »

Quote
- The mechanism of army invading settlement doesn't exist ? Am i right ? If i understand it's a kind of line between civilization with type of individual artificial combat ...

Yeah. "Army" goes to site A, fights a series of duels, and then either is defeated or conquers or pillages the site.

Quote
And, is the "neighbours" a hard criterion for that ? Is it really impossible to attack for an army that is not included in the neighbours list or is it a soft criterion which is only informative ?
Well, kobolds are not listed there iirc, but can send you ambushes once greedy enough.

Quote
Does this criterion applied for this "combat line" mechanism ?

During worldgen? Mm. Distance and natural obstacles can near completely remove attacks, but on rare occasions army can manage to jump over an ocean. I always have land connections so it just looks like going beyond 30 tile range, but I know at least one case in DF general discussion where elves did just that.

Quote
As i understand of your answers, all the subtle management of couples, like ORIENTATION and necessity of shared the same god, are not used during world gen.
I dunno if they are, which is why I didn't answer that *shrug*

If fully homosexual fail, they might be, though you might need crazy ratios like 1 hetero for 500 homosexuals to make breeding speed reasonable. Can always test it, I suppose.

I've read that if you have some super-rare caste of a common species you can start as that caste in adventure mode even if none should exist, though.

Quote
The more critical point of your answer is the number required for the generation of a site. I have no clear answer about that into your both answers concerning funding of new site (i miss it maybe ^^)
Nothing to miss, it just wasn't there. As you've observed, if there's any limit it is not an undue burden.

Maybe related to site founding:

With human town and sewers, it's possible to not get a town, or sewers in town in suitable location on initial founding or keep construction, even with dozens of hamlets (blocked by major site) nearby. If that happens, it can still possibly happen later.

In my current world, iirc the keep was built in 917 and sewers in 1287. (Hamlet was founded in 0)

I dare speculate that it takes it takes a certain type of person to found things, maybe present in 0,4% of population - so if it didn't happen in initial 100 random citizens (67% chance with 0,4%) it might happen with next 100.

When the hamlets aren't blocked from becoming a town, it typically is built within first 3 founded hamlets, so it'd seems right. Maybe a little low.

If that mechanism is also present on making new hamlets, it'd make founding them costless but needing more hamlets to be founded so that there can be more such founder humans. In that case, you could limit making new hamlets by dropping the population cap on old ones from 120 to, like, 2.

Personally, I'll stay with using pre-set savagery, but a general solution would be better for personal application.

Quote
Beside, how could these multiple little small town survive to megabeast and other thing like that. Did the megabeast/were beast/titan limitation of population/etc... to attack also applied to ALL the site during worldgen, regardless of civs and not only to yours in fortress mode ? If it's the case, could it be possible to regulate populations by lowering the criteria that trigger attack of titan/megabeast in worldgen ?
No, I don't think they're limited by their raws for worldgen sites.

Multiple sites seem to share population relatively freely in worldgen (but not during post-worldgen). As such, if a dwarven fortress isn't wiped out before making hillocks it likely survives the FB attacks.

Though note that during worldgen, the megabeasts are limited by ocean and distance, while past that they're not and only attack your fort after worldgen.

Trying to use megabeasts to cull site populations tends to be almost all or nothing results, where either megabeast dies or non-goblin civs suffer and lose most of their sites.
« Last Edit: August 16, 2017, 06:26:07 pm by Fleeting Frames »
Logged

KittyTac

  • Bay Watcher
  • Impending Catsplosion. [PREFSTRING:aloofness]
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #10 on: August 16, 2017, 10:19:44 pm »


Trying to use megabeasts to cull site populations tends to be almost all or nothing results, where either megabeast dies or non-goblin civs suffer and lose most of their sites.

Except titans can't be killed in worldgen! But the problem with losing sites still stands.
Logged
Don't trust this toaster that much, it could be a villain in disguise.
Mostly phone-posting, sorry for any typos or autocorrect hijinks.

Nilsou

  • Bay Watcher
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #11 on: August 17, 2017, 05:10:12 am »

Ok, because i was REALLY suprised. I had managed to create a VERY difficult world, with lot of titan/megabeast/cursed/tower etc... And lot of gobs.

What happen and make me came here is that where set of site are limited it's OK, "normal civs" have lot of difficulties to survive, which was my objective.
Buuuuut, if you just let population and number of site without limitation i was REALLY surprised to see an explosion of site, even more with humans and dwarf. Who managed to survive with no problem but also make more town than gobelins ...

It was... hum ... surprising. So i guessed that in a way on an other the mechanism was not simple :p It seems, with your answers, that this is not the case ^^.

I think i will try to make experiment on auto limitation with some parameters like orientation. 
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #12 on: August 17, 2017, 07:10:25 am »

Except titans can't be killed in worldgen! But the problem with losing sites still stands.
A quick gen of the world I created for cowmuffin (outdated now due knowing better worldgen techniques, but still neat) reveals this:

||

Eyeless meaty dino, struck down by giant monarch butterfly in forest retreat. FBs can also die, so maybe your worlds don't have enough fighting for titans?

KittyTac

  • Bay Watcher
  • Impending Catsplosion. [PREFSTRING:aloofness]
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #13 on: August 17, 2017, 07:57:32 am »

Except titans can't be killed in worldgen! But the problem with losing sites still stands.
A quick gen of the world I created for cowmuffin (outdated now due knowing better worldgen techniques, but still neat) reveals this:

||

Eyeless meaty dino, struck down by giant monarch butterfly in forest retreat. FBs can also die, so maybe your worlds don't have enough fighting for titans?

Hmmm. Either I didn't look at the version shown on the wiki, or it was disinformation.
Logged
Don't trust this toaster that much, it could be a villain in disguise.
Mostly phone-posting, sorry for any typos or autocorrect hijinks.

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: Why number of sites explodes in only 20/30 years....?
« Reply #14 on: August 17, 2017, 08:02:44 am »

It's not listed on DF2014:Titan, so what page did you visit?
Pages: [1] 2