[Freeciv-Dev] Re: (PR#13845) Increasing the appeal of very large cities

["Jason Short" <jdorje@xxxxxxxxxxxxxxxxxxxxx>]

<URL: http://bugs.freeciv.org/Ticket/Display.html?id=13845 >

Benoit Hudson wrote:
> <URL: http://bugs.freeciv.org/Ticket/Display.html?id=13845 >
> 
>>In mom it was dependent on the terrain.  Each grassland allowed 1 point
>>of population, each hills and forest and ocean 1/2 point, each mountain
>>0 points (or something like that).  Something like this wouldn't be that
>>hard to implement (see my "death to smallpox" ticket).
> 
> Sounds reasonable.  How does that work when cities are built near each other?

mom had a pretty big citymindist, like 3 or 4.  Tiles covered by more 
than one city got only half-credit (I think they got half-credit even if 
covered by 3 cities).  The death-to-smallpox patch addressed this by 
simply dividing the bonus by the number of cities covering the tile. 
The main problem here is integer rounding.

>>>My formula can generate maxsize < size, which means either my formula
>>>is bad or we need to special-case that.
>>
>>It's bad because eventually all cities will end up at size N.
> 
> Why is that bad?  If you don't build aqueducts, you *should* top out
> at N=8 according to the default ruleset.

Hmm.  Right.  The problem though is the max population will vary based 
on the amount of food coming in.  It should instead depend on the 
maximum amount of food provided by all tiles, regardless of if they're 
actually worked.

-jason