Natives (aka. Minor Civs, Barbarians, etc.)

[LDiCesare]

Doesn't the economy model have population density as well?

Yes population matters for the econ, but is it worth running econ for a pop of 1000 in a square? And having a civ going along with it?

Do you know of a thread that details what data (and what size range of data) you need to store each Civ?

No. It depends a lot of what features we have. There's data that depend only on the squares (economy), or the amount of units. Then there's the number of provinces, which is likely to be higher when we have many civs than when we have few. Provinces don't hold much info.
Each civ has a world view (map knowledge) with explored/unexplored squares, seen units, ownership of the square, economic value of the square when last seen, and, in case of terraforming or whatnot, generally duplication of all possibly changing terrain data for every square.
The civ then has plans and a military ai, plus datai to handle its internal policies and government info, and various other stuff which are of fixed size.
The biggest points are the military and per square cost, the squares including squares not owned by the civ.
Then you have inter civ relations and diplomatic data, which is not yet coded. And you also have all tech levels.
It's pretty hard to give any accurate figure of what a civ costs, but mostly it costs by the fact it sees things around it, so its holding territory hardly affects its cost or size. Units and the amount of fog of war that's been lifted are what metter in terms of processing power/memory consumption..

[alms66]

Originally posted by yellowdaddy
If we set a working maximum number of civs at 50, and if we include the possibility for some of these Civs to exist as HGUs ("stateless" civs), then surely it can work?

IMO, your number of civs is rather low. I should note that I generally hinge my definition of 'civ' in these types of games as being equivilant to 'empire' or 'nation', if you use some other definition, you're sure to come up with different numbers. I would shoot for having a modern-day number of 200 civs (approximately the same as the number of real-world) nations. The number of 'civs' in that light, has only lessened over time, so it would not be inappropriate to consider that there were at least twice as many civs in the ancient world, or a minimum of 400 civs. If you want to include 'stateless' civs in that, you can probably safely double it again.

For EIT, Native populations was one way to reduce the number of civs, by filling the voids of the world with the danger of spawning a new, hostile civ - keeping the early civs from expanding too rapidly. Not to mention the fact that expansionist civs generally ran into native populations wherever they settled, in nearly all time periods, and then proceded to absorb them, so the idea of native populations is rather appropriate.

[Mark_Everson]

Originally posted by yellowdaddy

Doesn't the economy model have population density as well?

Yes, well actually it just has total population in the square. The only breakpoint right now for processing power is zero population. As the econ model exists a 50-person square takes the same comp resources as a 5,000,000 population square.

Do you know of a thread that details what data (and what size range of data) you need to store each Civ?

No. But it is the clock cycles used for AI and econ that is the limiting factor, rather than data per se. Once we get diplomacy and diplomatic AI fully functioning I anticipate that that will also be a clock cycle hog.

It sounds like a "HGU" Civ wouldn't take up too much processing power

It certainly could be done in a cycle-lite way. How much coding it would take to make them work with the rest of the system, I don't know. We obviously need to compare how fun HGUs are for the work involved and the resources used vs some sort of minor civ approach. It may be, FE if civs without central control end up using much less processing power than centralized ones, that "natives" can just be full-fledged civs with little centralization. Like the Celts that gave the Romans so much trouble, especially early on.

[yellowdaddy]

Thanks for such helpful replies everyone.

LdC: Yes population matters for the econ, but is it worth running econ for a pop of 1000 in a square? And having a civ going along with it?

LdC: ...mostly it costs by the fact it sees things around it...

M_E: Yes, well actually it just has total population in the square. The only breakpoint right now for processing power is zero population. As the econ model exists a 50-person square takes the same comp resources as a 5,000,000 population square.

M_E: ...civs without central control end up using much less processing power than centralized ones, that "natives" can just be full-fledged civs with little centralization. Like the Celts that gave the Romans so much trouble, especially early on.

I should point out, I envisage two categories of acentralised "civs" - the HGU (Hunter-Gatherer Unit) and SFU (Subsistence Farmer Unit).
The idea is that you have HGU->SFU->CIV.


So the size or density of population makes litte difference with the economy model, because you have to have to allow a similar range of things for each civ or population to "see"?

Could you create more simple econ model that applies only to HGU/natives? - Subsistence Economy Model? and Hunter-Gatherer Economy Model? - I mean, what do you need to model in that?
Do you really need sectors for this kind of model? (unless you count witch-doctors as service sector, smithy as a mfg sector) They don't need sites of arable land for HG Economy, though they do for Subsistence Farming Economy; they don't build permanent farms or invest long-term; they don't build permanent buildings of any kind (ok, wooden semi-permanent shacks for the Subistent Farmers), they don't have provinces, they take only from the tile they live in; they don't really need productive capacity, because of the lack of any market for goods.
They fit in that empty box just above "Early Civilisations" on you table on the Econ Model page.

Therefore, could the solution lie in the division between real (settled) civs and primitive units (HGUs and SFUs) that don't permanently inhabit or "own" a tile?
i.e. the economic model only has to function for squares that contain a settled, civilised population; meaning that all other squares have a population of 0, but can contain HGUs and SFUs, who equate to having a populated square?

Do you catch my drift?

Imagining a Dawn Scenario, the first turn, starting with 20 civs and 30 HGUs. You have only 20 populated tiles (populated in terms of the various models: econ, soci, riot), the rest are completely empty, except for these HGU-Civs (limited to say 20,000 max per unit before they have to split into two or evolve into SFUs).

Could this solve the problem? You can have most of the squares of the game "populated", but populated by single-unit HGU/SFU-Civs, who don't require the same detailed modelling as a settled CIV.
Does this not actually simulate human history in a satisfactory way, whilst utilising memory efficiently?

As for fun factor, well, lots of irrational, unpredictable HGUs and SFUs could add to the challenge of the game, by spoiling the plans of settled civs. Couldn't you perhaps use this HGU model to model (as alluded to early) guerillas, bandits, pirates etc... ?

For EIT, Native populations was one way to reduce the number of civs, by filling the voids of the world with the danger of spawning a new, hostile civ - keeping the early civs from expanding too rapidly. Not to mention the fact that expansionist civs generally ran into native populations wherever they settled, in nearly all time periods, and then proceded to absorb them, so the idea of native populations is rather appropriate.

Maybe you could actually have a lot more "civs" in the game, by turning most of them into trogs? I want to see loads too (my number of 396), but I want to have a realistic view of what the hardware can do.