Talk:United States/Boundaries

Capitalization

I reverted edits such as 1488696 in order to apply title case consistently to headings on the page and fix incoming links. I don't understand why it's so important to capitalize the words "territory" or "commonwealth" when they refer to the concept in general as opposed to a specific entity. Notice, for example, how w:Territories of the United States lowercases the word many times, for example: "The United States currently has sixteen territories." This article is not a legal document or piece of legislation. Ultimately, the question of whether to capitalize such words is hardly consequential, but that's why I think we should err on the side of writing for laypeople. – Minh Nguyễn (talk, contribs) 08:05, 7 July 2017 (UTC)

New Orleans wards

I removed this text:

In cities which contain both wards and neighborhoods (as does New Orleans), it may make sense to tag the wards admin_level=9 and the neighborhoods admin_level=10, to preserve this hierarchy. New Orleans might chime in and suggest that wards are administrative and neighborhoods are not; there is little or no consensus here.

It's true that New Orleans has both wards and neighborhoods. In fact, there are both the less formal, amorphous kind of neighborhood that a lot of cities have, as well as the formally defined "neighborhoods" primarily used by the City Planning Commission. There are significant differences between the two sets of neighborhoods. As far as I can tell, the former shouldn't be mapped with boundary relations, and the latter are no more administrative boundaries than the NOPD precincts or New Orleans City Council districts – none of which line up, by the way.

Moreover, none of these things are mapped in OSM anyways. The article's stated purpose is to describe the database as it is. As a major contributor to OSM's coverage of New Orleans, let me go out on a limb and say the local community isn't going to map wards and neighborhoods as described above. But if anyone feels strongly that they should be mapped that way, there are other venues such as United States admin level, the talk-us mailing list, and OpenStreetMap itself. Put down the wiki and start mapping!

 – Minh Nguyễn (talk, contribs) 08:26, 7 July 2017 (UTC)

Minh, to say "put down the wiki and start mapping" is disingenuous at best, and mean-spirited and callous at worst. Why would you suggest to another volunteer to "put down the wiki" when you yourself picked up the mantle of wiki-writing with this very article in response to another article with largely the same content? Ah, with your contradictory stated purposes to "not discuss novel tagging proposals...(but rather) ignore them in order to focus on how to tag new features as they come up." (See below). — Stevea (talk) 20:18, 16 July 2017 (UTC)

Introductory text

I find it contradictory that the introductory text states:

"this article does not discuss novel tagging proposals...(it) ignores them in order to focus on how to tag new features as they come up."

Huh? Perhaps clarification is needed to differentiate between "novel tagging proposals" and "how to tag new features," as these seem either exactly the same or quite similar.

And, if (as stated above) it is true that this article "describe(s) the database as it is," how/why is it that its stated purpose in the intro text is "how to tag new features?" I find this highly confusing.

— Stevea (talk) 20:18, 16 July 2017 (UTC)

Please delete Regions COG from Connecticut 5 and CDP from Connecticut 10

See [1] for reasoning. Stevea (talk) 02:37, 3 May 2020 (UTC)

Unique entities often known as Consolidated (was Consolidated County-Cities CCCs)

I'd like to get a consensus on this and then update the United States admin level page and this United States/Boundaries page to include tagging instructions for CCCs.

NOTE: This discussion has been made into a forum topic for increased visibility at the OSM Community Forum.

It may be that these 39 (40, 41...) things here are "unique" in USA and OSM simply agrees to tag these as we document here. "There are no rules" (for CCCs) in United States admin level because generalizable rules cannot be written for CCCs, they are unique and as we specify below. Stevea (talk) 22:41, 29 November 2024 (UTC) An update to this, as I have stayed out of the community forum discussion (so far) and Loren uses this table to document what is: a fluid, moving towards "all CCCs are tagged as two" but we're not quite there yet with wide understanding, agreement to some things and then consensus. My preference is to say "there are a few dozen 'double-things' one tagged 6, one 8." We're most of the way there, the way to get there (call it reality-based tagging) is to agree "a set of a few dozen things which we call CCCs are tagged with a double-method that we agree to, with no exceptions." The wiki remains / becomes more strongly definitively prescriptive, and the map (for a moment at least, hopefully longer) stays updated as things are or chanage in the real world. (There are CCCs, a few dozen of them, and we tag them with superimposed 6 and 8 polygons). That's crisply-defined and crisply-tagged (as so). I'm listening to downstream user concerns and I see no (or few) contradictions, so continue down the crisp path described here. With more discussion, I believe we'll achieve this consensus. And we will have "rules" for what CCCs are (co-terminous or mostly-so polygons, one tagged 6, one tagged 8), as we do now, notwithstanding my "they may be unique..." statement I made at the beginning of this (moot?). We'll get this tightened up; excellent leadership, Loren. Stevea (talk) 21:05, 1 December 2024 (UTC)

The lengthier guidance on US admin levels states "Currently there is no convention for single-county CCCs to be 'tagged twice' with two coterminous relations tagged boundary=administrative: one with admin_level=6 representing the county and another with admin_level=8 representing the city. (On these, a tag of border_type=county;city prevails)." I was short-sighted about this regarding San Francisco (only, it turns out). Stevea (talk) 22:43, 29 November 2024 (UTC)

Below is Wikipedia's list of the 44 consolidated city-county entities (I've added the five boroughs and associated counties in New York city to the original 39 initially identified, for a total of 44 CCCs). Of those eight are Alaska's "City and Borough of XXX" entries, of which only two have separate county and city entries.

Setting the Alaska entries aside, 32 of the 36 remaining CCCs have an entry for both the county and the city in OSM, with the county tagged admin_level=6 and the city tagged admin_level=8 just as any other county and city would. (I had initially reverted the delete of San Francisco County and Nantucket County and added Denver County and Broomfield County and also cleaned up the tags for the counties and cities to bring that total to 36 of 36, but have since undone those pending the resolution of the community discussion).

Additionally, 15 of those are CCCs without coextensive boundaries and would need two entries regardless (one for county and one for city), leaving 17 CCCs with coextensive boundaries (aside from Alaksa). There are four that I cannot determine whether they have coextensive boundaries or not.

Although for CCCs the county and city are consolidated, they still remain distinct entities and I believe OSM should have an entry for each, just as Wikidata does.

I would "second the motion," which essentially is (properly) duplicating the singular relation and making it two relations, each (properly and distinctly) tagged, for eight relations total, six in Alaska, two in Colorado. And we clean up any fuzzy business, like harmonizing (these in Alaska). Stevea (talk) 22:49, 29 November 2024 (UTC)

Adding Washington, District of Columbia (WDC) as a discussion topic (a "bubble being blown") in the "bubble-gum discussion space" of the Community Forum topic. A "lightning proposal" is whether WDC collapses to not one Single or two Separate such consolidated boundaries, but, in fact, zero of them: 2-4-8-9-10 -> 2-4-9-10. Stevea (talk) 00:14, 13 December 2024 (UTC) In fact, this was done. Stevea (talk) 05:10, 18 December 2024 (UTC)

Status	State	County	County QID	County relation	City	City QID	City relation	Nature	Coextensive?	Emerged	Comment
done	Alaska	former Greater Anchorage Area Borough	Q49297981	Deleted	Municipality of Anchorage	Q39450	2605259	Single		Merged 1975	Regardless of the "string of its name," Alaska state law and its consolidating charter of 1975 enshrines into Anchorage both division (or borough, both county-equivalents) and municipal (incorporated city) levels of government. Quoting Wikipedia, "Though its legal name is the Municipality of Anchorage, it is considered a consolidated city-borough under state law." The consensus reached with map data tags the single remaining relation with boundary=administrative + admin_level=6 + border_type=borough;city.
pending	Alaska	Haines Borough	Q512981	2605277	Haines	Q79546	110769	Dual		Merged	Not sure the OSM relations are actually referring to the same area
done	Alaska	former Greater Juneau Borough	Q56404707		City and Borough of Juneau	Q29445	2605281	Single		Merged 1970
pending	Alaska	Petersburg Borough	Q25408755	2605308	Petersburg	Q80027	3496169	Dual		Merged
done	Alaska	Former Greater Sitka Borough	Q13046085		City and Borough of Sitka	Q79804	2605309	Single		Merged 1971
done	Alaska	No former borough			Municipality and Borough of Skagway	Q615975	2605310	Single		Created 2007
done	Alaska	No former Borough			City and Borough of Wrangell	Q43983	2605235	Single		Created 2008
done	Alaska	No former borough			City and Borough of Yakutat	Q487681	2605317	Single	Deleted	Created 1992
done	California	San Francisco County	Q13188841	Deleted	San Francisco + place=city	Q62	111968	Single	Yes	Creation, now legally dissolved, remains as "consolidated nomenclature"	Consensus reached with map data. City-level deleted as a demonstrably historical vestige / legally dissolved, name-only "nomenclature" case (City and County of...). The single remaining relation is tagged boundary=administrative + admin_level=6 + border_type=county;city + place=city by consensus.
done	Colorado	Broomfield County	Q16088503	Deleted	Broomfield + place=city	Q492819	1411321	Single	Yes	Creation (as "consolidated nomenclature")	Deleted per community discussion; Believed to be the same as Denver, so consensus reached with map data. The single remaining relation is tagged boundary=administrative + admin_level=6 + border_type=county;city + place=city by consensus.
done	Colorado	Denver County	Q15906757	Deleted	Denver + place=city	Q16554	1411339	Single	Yes	Creation (as "consolidated nomenclature")	Deleted per community discussion; Noting that this consolidation is "nomenclature" only, consensus reached with map data. The single remaining relation is tagged boundary=administrative + admin_level=6 + border_type=county;city + place=city by consensus.
pending	Florida	Duval County	Q493605	1210719	Jacksonville	Q16568	119008	Dual	No	Merged
pending	Georgia	Bibb County	Q488171	1027990	Macon	Q219656	119694	Dual	Yes	Merged
pending	Georgia	Chattahoochee County	Q486394	1059842	Cusseta	Q2153623	119458	Dual	No	Merged
pending	Georgia	Clarke County	Q112061	1020709	Athens	Q203263	119353	Dual	No	Merged
pending	Georgia	Echols County	Q493029	1073778	Statenville	Q3116378	153836880 (node)	Dual	?	Merged
pending	Georgia	Muscogee County	Q492048	1073925	Columbus	Q239870	1850813	Dual	Yes	Merged
pending	Georgia	Quitman County	Q498684	1073928	Georgetown	Q2855909	13272053	Dual	No	Merged
pending	Georgia	Richmond County	Q498319	420422	Augusta	Q181962	11078175	Dual	Yes	Merged
pending	Georgia	Webster County	Q491514	1074324	Preston	Q951376	13656346	Dual	No	Merged
done	Hawaii	Honolulu County	Q487704	3861844	Honolulu: place=city (only)	Q18094	119231	Single	No	Creation (by traditional geographic ahupuaʻa)	"Honolulu is not an actual municipality." Consensus reached with map data: Honolulu County is tagged boundary=administrative + admin_level=6 + border_type=county while Honolulu (the city, properly the ahupua'a, a geographic area between volcanic ridges) is tagged boundary=place + place=city + no border_type=* tag.
OK for now, "wart"	Indiana	Marion County	Q506230	1812962	Indianapolis	Q6346	18438530	Dual	No	Merged into Unigov	Discussion here. It is clear that Marion County ≠ Indianapolis. While currently, OSM sees no need to further denote or focus on complex internal gearworks of CCCs, a small "wart" remains: Indianapolis’ territory legally and geographically overlaps that of the towns included (but not excluded) with consolidation which are at the same level of 8 due to taxing authority complexities.
pending	Kansas	Greeley County	Q377131	1070390	Tribune	Q1917357	130296	Dual	No	Merged
pending	Kansas	Wyandotte County	Q375652	1070421	Kansas City	Q486472	129568	Dual	No	Merged	Offered as an example here as possibly offering solution strategies to Indianapolis.
pending	Kentucky	Fayette County	Q491931	130537	Lexington	Q49241	11078268	Dual	Yes	Merged
pending	Kentucky	Jefferson County	Q500776	1804288	Louisville	Q43668	1804307	Dual	No	Merged
pending	Louisiana	East Baton Rouge Parish	Q491949	1836609	Baton Rouge	Q28218	132206	Dual	No	Merged
pending	Louisiana	Lafayette Parish	Q506892	1837810	Lafayette	Q128891	131949	Dual	No	Merged
pending	Louisiana	Orleans Parish	Q486231	1836428	New Orleans	Q34404	131885	Dual	Yes	Merged
pending	Louisiana	Terrebonne Parish	Q387555	1836620	Houma	Q388503	131586	Dual	No	Merged
done	Massachusetts	Nantucket County	Q2991355	2387088	Nantucket	Q18372627	Deleted	Single	Yes	Creation	The single remaining relation is tagged admin_level=6 + border_type=town;county by consensus. Also, there remains significant conflation in Wikidata between the Town and the island that needs to be cleaned up.
pending	Montana	Deer Lodge County	Q490774	142396	Anaconda	Q483539	150938709 (Node)	Dual	?	Merged
pending	Montana	Silver Bow County	Q484513	1740649	Butte	Q467664	6840978	Dual	Yes	Merged
pending	New Mexico	Los Alamos County	Q484296	1344294	Los Alamos	Q318637	171160	Dual	No	Merged
pending	New York	Bronx County	Q855974	2552450	The Bronx	Q18426	9691916	Dual	Yes	Merged	border_type=borough since further consolidated as New York City with admin_level=5
pending	New York	Kings County	Q11980692	369518	Brooklyn	Q18419	9691750	Dual	Yes	Merged	border_type=borough since further consolidated as New York City with admin_level=5
pending	New York	New York County	Q500416	2552485	Manhattan	Q11299	8398124	Dual	Yes	Merged	border_type=borough since further consolidated as New York City with admin_level=5
pending	New York	Queens County	Q5142559	369519	Queens	Q18424	9691819	Dual	Yes	Merged	border_type=borough since further consolidated as New York City with admin_level=5
pending	New York	Richmond County	Q11997784	962876	Staten Island	Q18432	9691948	Dual	Yes	Merged	border_type=borough since further consolidated as New York City with admin_level=5
pending	North Carolina	Camden County	Q504294	1242002	Camden	Q2934722	158270930 (Node)	Dual	?	Merged
pending	Pennsylvania	Philadelphia County	Q496900	417845	Philadelphia	Q1345	188022	Dual	Yes	Merged
pending	Tennessee	Davidson County	Q1177705	1847619	Nashville	Q23197	197472	Dual	No	Merged
pending	Tennessee	Moore County	Q494716	1848192	Lynchburg	Q1023048	196141	Dual	No	Merged
pending	Tennessee	Trousdale County	Q501939	1848203	Hartsville	Q3458917	153568121 (node)	Dual	?	Merged
done	District of Columbia	No county here, though Census Bureau calls WDC county equivalent	Q3551781 (not County, DC)	162069 (not County, DC)	Washington: place=city (only)	Q61	5396194	Neither	Yes, Washington (as place=city) with District	Washington practically, though not necessarily legally†, dissolved as administrative city	"Washington" is now (December, 2024) tagged boundary=place + place=city + no border_type=* tag, not boundary=administrative + admin_level=*. Consensus reached with map data, though see this† Wikipedia article.

-- Loren Maxwell (talk) 18:45, 28 November 2024 (UTC)

I agree that such tagging is inconsistent (and made that observation in the wiki text). One approach (which you just began in Georgia) is to "keep track," like in a big table. There are >3000 counties in USA, and getting consistency across tagging "standards" is both partial and improving. We partly do so in our map, we partly do so in our wiki. Thank you for your continuing contributions and improvements to the map. If this is "part-way to consistency among such tagging," we can say it is a success. (Wider understanding of "how we tag" and "how we mean to tag" is important!) Wider consistency? "Mixed success!" Growing this "organically" (from the ground up at a national scale) is doable, but a fair bit of effort. Still, I wish to be encouraging. Sharing knowledge of how things "should" or "might" be (in an idealized-tagging scheme), well, that's forward movement! Maybe you could "spark a fire" (or just did) to get all of Georgia "done" as you see fit. Stevea (talk) 17:45, 28 November 2024 (UTC)

Thanks, Stevea -- I was actually in the middle of building the table to track the CCCs but accidentally hit "Save changes" instead of "Show preview", so that's why I only have Georgia listed at the moment. I'll be adding the others, along with my initial thoughts on how the tagging should appear, although the tagging will probably just be the current county and city tagging without necessarily noting they are consolidated. -- Loren Maxwell (talk) 17:56, 28 November 2024 (UTC)

The short answer is, "it's complicated". The fact that it's difficult to query is a consequence of this complexity. It can't really be avoided. I operate a data consumer that consumes town and city boundaries, and this complexity is something I have to deal with on an ongoing basis. It's a big topic and deserves a forum thread for wide visibility, not just a wiki talk page discussion. --ZeLonewolf (talk) 20:20, 28 November 2024 (UTC)

ZeLonewolf is correct, that "it is complicated." And a wider (community discussion topic which should link here) is a good idea, giving correctly different exposure to this. I also seriously applaud this first draft at a table, with its brief explanations; nice. Stevea (talk) 22:27, 29 November 2024 (UTC) So, Loren, prompt to start a topic pointing to this table. You could make a case, you could "stand aside" (for the inevitable discussion that will follow) and simply listen. Brian is right when he says "complicated." Let's see if we can tease it all out.

Thanks, Stevea -- I've fixed (at least according to my sense of what should be done) all the counties except those in Alaska. I'm new to OSM -- is the forum you guys are referring to the US community forum? -- Loren Maxwell (talk) 00:31, 30 November 2024 (UTC)

Yes, that's the forum. If you haven't, join, look around (that forum, it might take a bit of orientation, you might feel quickly "at home") and start a new topic (in "United States") asking people to look at the discussion here. Raise Brian's concerns that there may be complications for downstream use cases. I think we're on the right track to "make all 39 consistent" (and we've got 'em all done correctly except in Alaska). But we do want "more eyes" on this and for Brian and others to be able to chime in and raise concerns more broadly. The community discussion forum should get us wide views, which can then be focused here. If I had to generalize, I'd say "discussion in the forum, finalities in the wiki(s)." After good discussion in the forum, that might mean a touch in this (here) wiki-space, then we tighten up the (two) wikis with the consensus achieved from the forum discussion. Stevea (talk) 01:47, 30 November 2024 (UTC)