[Talk-us] Correct source for population=* tags on US metropolitan cities

[stevea]

On Jan 30, 2021, at 7:25 PM, Clifford Snow <clifford at snowandsnow.us> wrote:
> You may find this hard to believe, but I fully agree with you

Clifford, there is a great much we agree about!  Of course I believe you and it isn't hard at all.

> I was subtly trying to point out the futility of using data from these areas. The government can change them at will. And apparently it looks like they are going to adjust them again. I don't mind updating a city's boundary but this game they play with statistical areas is nuts.

Yup, let 'em play nuts, OSM doesn't want to chase after numbers of blobs that are always changing.  Federal government numbers for such things can be queried of the federal government.  I'm careful as I say this as sometimes there really are times and places where OSM wants a bit of overlap here.  Cadastral-level, no.  Protected areas, I (and others) think yes.  Sometimes these are governments who administer these data, sometimes non-profits or similar.  The openness of such data varies state-to-state (fully open happens, too, as in California regarding geo/GIS data), while at the federal level, the data are public domain / fully open.

> Sure we could add in all the census areas so someone could add up the number but how does that help OSM.

What I mean is that there seems to be a hunger for a datum that is "a greater city / conurbation's population including suburbs and exurbs."  What roughly colloquially is sometimes called "greater metropolitan area population."  Part of the reason is that there is "Philadelphia" and (it's precise, only-in-the-city-limits citizenry) and there is "Philadelphia (greater metro)" that are sometimes "both known as" Philadelphia.  Or you might want to see a map with "only" population or "greater" population, both with the name-label of "Philadelphia."  To do that, I'd think we might want to create a new key called something like conurbation_population or greater_metro_population (these terms vary and we're in the US here while keys in a UK English might be preferred, yet I don't know this phase/word).  That key might get added to the city polygon (or node, if there are entire cities which remain nodes only, possible and maybe even probable).  Then, it's up to a renderer (checkbox) which one might be displayed or used to determine a renderer font size.

Maybe this goes to the tagging list, as such things are a worldwide phenomenon of cities.  There might already be similar tags and/or strategies built into renderers to display capital cities, "first cities" (the largest in a country or state, for example...) and allow some renderer choices / checkboxes in toggling how features interpret / display such multiple levels of data.  I have read a fair bit of wiki, but the practices of how less-revealed tagging for capital cities or cities of a certain size or agglomeration might have some rendering handshakes of its own that I'm not aware of.  This would likely blend with / live in that realm somewhere.

I think that would help OSM, I think that would help many better understand that renderers decide (or don't) whether and how to display certain data.  It's not always a simple, neat, clean, easy-to-describe pipeline, it might be more complex.  With the right kinds of low gear instruction, it could even serve as a tutorial of how the OSM data-render pipeline / toolchains work.

SteveA