Rewriting city growth -- for standard and experimental

[Sarlock]

Thinking more about the sign idea.  This, of course, is a pakset dependant graphic, so each pakset could decide what its abandoned building sign looks like: a for sale sign, a graphic depicting the unoccupied building, etc.  The question is whether this provides a meaningful enough clue to the player.

An area with poor public transportation is also not necessarily a candidate for abandoned buildings - many suburbs have poor (or no) public transport and rely on road travel for all transportation... yet still experience significant growth due to the services provided by the nearby city.  Buildings would become more tolerant of poor transportation options if other economic reasons are sufficient to still build there.

Detroit's inner city has hollowed out considerably over the past few decades but the metro area population has remained fairly static (thus the metro area is still growing while the inner city is declining): there is something to be said about a large city having enough of its own momentum to still grow (or at least resist decline) just by the virtue of it being large already.  At some point, commercial and industrial jobs (of which there are still plenty in the Detroit suburbs) become the most important driver of city growth.  Available transportation options merely influence where in the city this growth concentrates.

[jamespetts]

Ahh, but the success rates system is specifically not limited to public transport: if enough people can get where they want to go by walking or driving, that will suffice.

[Sarlock]

That's where Experimental becomes much more powerful in simulating a good growth distribution, given that it has the calculations in place for these transportation methods.

[jamespetts]

Indeed - the idea is that we can abandon the idea of artificially making cities more dense in the "middle" (and coming up with some definition of what the "middle" is), as the only reason that cities are more dense in what becomes their centre is precisely because it is easier to get from places to there and from there to places than other than in that area.

Edit: One challenge with having building growth being global as opposed to based on individual cities, however, is how we account for things such as electricity supply to cities and supply of nearby goods - we need to come up with a way of factoring in those things, I think.

[wlindley]

The latest Experimental in 'git' seems to have a tendency to grow linear cities -- especially at map edges but also in the centre of the map (see city, centre right).  Hoping this gets resolved as well.

[jamespetts]

Hmm - which branch are you using? I haven't noticed this.

[wlindley]

devel-new ... or should I be on another, for testing?

[jamespetts]

Hmm - the "passenger-generation" branch is derived from "devel-new" and does not show this. If you want to test with the current code, using 11.x is better: "devel-new", "stacked-stations" and "passenger-generation" are for testing new features, which are not yet complete (with the possible exception of "stacked-stations", but I have not tested that yet).

[waerth]

At least cities like that make it easy to transport pax LOL

[jamespetts]

The Economist has an interesting video on city decline and growth here.