OK, a bit of a rambling post that I split off from the last post when looking over cities, and giving them more thought...
Toady's got pics of the city up!
And by Armok, they're ENORMOUS! I'm looking forward to doing some exploration.
Those are max-sized cities, so I'm looking forward to them.
I hope they will have more than z-level.
Here's the problem I have - cities seem to have really funky models for how their population goes up and down right now. They seem to need the caps that are placed over their population to prevent rampant growth at times, because there is no downward pressure on their population growth unless it actually gets to starvation.
I'll ask my question first, then let the rest of the post get taken up by all the random musings I have on the subject.
What models are there governing population growth and decline? Where do you want it to go - a simple and streamlined abstract model of growth and decline based upon a tally of upward population growth forces and downward pressures on populations, or a very gritty and detailed model where every wave of disease is tracked, so that there will be history events of the great city fire of 231?The problem with that is that starvation doesn't seem to be modelled very well at all - people just keep eating normally right up until there's only food for 30% of them left, and so 70% of the population just sits there and waits to die.
Realistically, you'd have hoarding before the crisis point, and inflated food prices so that only the really wealthy or those who saw the crisis coming furthest off and hoarded early would have a chance to really stockpile some food to weather the storm. People are going to want to protect themselves and their families, after all.
After that, people who are faced with starving and their families starving are going to want to protect their families, after all, so they're going to start rioting or trying to kill and steal food from those who are hoarding it for themselves.
I doubt a city would actually see 70% of its population just plain starve - you'd rather see 80% of its population die in riots and an orgy of crime, and the other 20% probably not be able to put the city back together again, except in a fractured state, and most of the survivors would probably scatter as refugees.
In a real city of the Ancient or Medieval world, there were serious downward forces on population thanks to the utterly rampant diseases that flourished in crowded cities with close human contact for disease transfer, with water sources tainted by sewage, with no proper sewer system, especially in the most crowded, poorest parts of town that received no city planning, and as such, the larger and more crowded and less-maintained a city was, the more ravaged by disease it would become.
The Total War series models this, especially in its earlier, more abstract and "clean" games by just plain having a -0.5% population growth rate modifier due to "Squalor" per every arbitrary amount of population in a city. Public sanitation projects (among others) give a +0.5% or +1% population growth benefit to help counteract this, although at the highest levels of population, it's pretty hopeless to try to stave off the problems of overcrowding. As population grows geometrically quickly, but the time it takes to build more complex sanitation systems becomes even more lengthy and cumbersome, and there are only a finite number of things you can do, population levels just naturally plateau on their own as disease and overcrowding's downward pressure on population hits equallibrium with the upward pressure of birthrate.
Now, that's a fairly abstract system (and not even Total War stuck with it, which is a bit of a shame, honestly), and I know that DF likes to go for the nitty-gritty realism...
I honestly wouldn't be surprised if there were procedurally generated diseases with procedurally generated mortality rates in DF. Each one coming through some different disease vector.
Infant mortality rates can be modeled abstractly or with greater detail, as well - areas with more crowding, more pollution, less sanitation, and poorer food quality and availability just plain have lower "fertility rates" in an abstract model, versus having the children being born then dying off in greater numbers than adults due to weaker immune systems and tolerances for the lack of food.
(Oh, and now I just thought about how childhood diseases and malnutrition also leads to poor development of the mind and body - leading to permanent loss of intelligence and smaller body size and disease immunity. That would mean tracking some sort of average population malnutrition factors for different population centers in worldgen... I honestly like abstract models better than the really gritty ones most of the time, but being able to tell that someone is from a city rather than a village because they're shorter would be kind of cool...)
Especially since it's possible to split off each district of a town as a separate area (where the best parts of town have better sanitation than the worst parts).
DF could simply have a simple birthrate bonus or penalty system, or it could actually model waves of diseases killing off individuals, food riots, infant mortality rates, violent crime rates, etc. and do these by individual town district (so the docks are a nasty part of town, but the clothier's district is fairly safe).
As I said, though, I honestly like Rome: Total War's simpler model (although that was when I had to actually control it, personally), but given how much gritty realism DF crams in versus how difficult it would be to code and store many of these things in memory, I honestly think Toady could go either way on how much detail he wants, here, so I'd say it's worth asking which way his winds are blowing.