Broad undiscussed New England place name reclassification

ezekielf · 2024-04-24T13:27:33.958Z

edops:

There are no efforts by anyone to come up with a “consensus” all talk and no proposals.

The OpenStreetMap community certainly does love endless discussions . Criticism well taken. However this is simply the reality of a volunteer driven, self organizing, open data project. Discussion must be held to establish consensus. I understand this is frustrating. I’ve been frustrated by it myself, but unless you plan to somehow get yourself appointed dictator of OSM, there is no other way. Discussion is effort and it has been under way for a number of months. See the latest topic for proposals: Framework for aligning New England place nodes to census categories

ElliottPlack · 2024-04-24T16:21:04.170Z

edops:

I also find it amusing that you are from Maryland, yet you are trying to dictate the status of small towns in Maine

I am here as moderator and community member, not dictating the status of any towns. The community here, especially other locals in the New England area, can come up with a solution that works.

Ideally the meaning of city, town, etc. would be uniform across all of OSM so data users know what they are getting when using the data.

ZeLonewolf · 2024-04-25T01:43:10.436Z

Zeke Farwell:

this kind of lone wolf behavior

Heyyyyyy now!

stevea · 2024-04-25T07:25:08.388Z

Over the last ten, twelve, fourteen years or so, many OSM contributors in the USA have hammered out methods and guidelines we choose to assign values of both admin_level=* (2, 4, 6, 8 as nation, state, county, city/town is well-established, but the devil is in the details and there really, REALLY is more to it than that) and place= (city, town, village, hamlet). Especially in New England, this is difficult, partly for historical reasons, partly for linguistic, political and “that’s how people call things around here” reasons. (Which are often very hard-fought and colloquial, but may or may not have much basis / grounding in legal reality). It may be that “lots of people think that ought to be a town” (or village or hamlet…) but if one person can point to a statute or city charter or state constitutional article that states black-letter law, that can be (and has been) accepted as definitive in a particular case. We could use more of those, although they are hard to come by and don’t always break ties; ambiguity often seems to remain. Still, we get closer and closer, and that is the ideal, so let’s keep “getting closer” in that vein.

I, we here don’t expect everybody who contributes to the betterment of this to have absorbed the entire history of it, as watching sausage get made isn’t ever as tasty as the final product. But we’re well along the way of at least “roughing out” what we mean by [city, town, village, hamlet]. Yet we still have a ways to go, especially in New England, where there are both quirkinesses and idiosyncrasies in about three states, where the other 47 look at the results with “huh?” (Yet, all fifty states must be seen as unique, even as we share many similarities, too)…

As with any “argument” in OSM discussion channels (not as in a rancorous cat-fight of opinions, but of a civil dialog with points, counterpoints, evidence and good ideas and suggestions bolstered by reason), poo-pooing the process of consensus while indicating hostility and impatience with the process will doom your efforts faster than anything. Please offer reasons, evidence, law, history, fact-based data and you will have many listening ears here, and that is the beginning of consensus. Yes, it can be difficult to achieve, and for something like this will have “lumpiness,” regional differences that are simply inevitable. But, that’s the way these things have, do and will get solved in OSM.

Look for commonalities, “show your work,” and don’t assume that “what you know must be correct because I live here and therefore must know best.” Maybe. But, maybe not. OSM is most certainly a learning experience for absolutely everybody. Know-it-alls are not really welcome. Experts, who can support their views, hey, great. Know-it alls? Sorry, no.

This is hard, yes. But it is doable, too. Let’s give ourselves the time and space to tease it all apart. We’re going to take those anyway, so let’s just say we are and get that understood by any and all who want to contribute.

stevea · 2024-04-25T07:53:28.833Z

Uh-oh, double-post by me.

While it isn’t done, but it is evidence of much improvement over our decades, what if we took a state-by-state approach for [city, town, village, hamlet]? You’d be correct to accuse me of “loving a big table of the fifty states with row-by-row results…” but this approach HAS worked to achieve SOME consensus w.r.t. highway classifications, admin_level, public_lands classifications and more. I’m just spitballing here, but it could both work and be an explanation for sharp differences that might be seen (in renderers at particular zoom levels) across state boundaries.

Each state could be population-based (with numbers / thresholds different from here vs. there), “must have a grocery, post office and hardware store to be called a village,” some combination, or whatever.

Edit: This does toss out the window Elliott’s “ideal” that a town is a town is a town (everywhere), but that does seem like an almost unobtainable ideal. Still, I keep an open mind.

miela404 · 2024-04-25T14:51:16.657Z

In New England, the city tag should be limited to Micropolitan and Metropolitan areas. This includes large towns like Brunswick, and small cities like Waterville and Sanford. It does not, however, include towns that are called cities, such as Rockland or Presque Isle.

This is a fair compromise for the two sides, and has a third party definition to pull from.

www2.census.gov

NECTA_WallMap_Mar2020.pdf

ZeLonewolf · 2024-04-25T14:52:33.331Z

I think the issue there are places like South Burlington, which I think folks feel should NOT be a place=city.

miela404 · 2024-04-25T15:35:29.056Z

This is not a problem. Because Burlington is the more prominent city, it gets the label. The same way that, currently, South Portland doesn’t get the label because Portland is more prominent, and Auburn doesn’t get the label because Lewiston is more prominent.

ZeLonewolf · 2024-04-25T16:28:59.268Z

miela404:

This is not a problem. Because Burlington is the more prominent city, it gets the label. [… more examples …]

By that logic, let’s just tag all places place=city because label collision algorithms exist.

place= values are not just used for map rendering. There’s value in getting this right and not over- or under-classifying places for all sorts of use cases that you or I haven’t thought of.

The collection of places we tag place=city needs to at least be internally consistent to have data value. If a rule we create results in an outcome where the consensus view is that it’s wrong, then we came up with the wrong rule

ezekielf · 2024-04-25T17:12:29.449Z

miela404:

Because Burlington is the more prominent city, it gets the label. The same way that, currently, South Portland doesn’t get the label because Portland is more prominent, and Auburn doesn’t get the label because Lewiston is more prominent.

I agree in these cases but there are others where multiple cities in close proximity are all highly prominent/significant. Perhaps not in New England, but Minneapolis-St Paul, MN and San Francisco-Oakland, CA come to mind. Whatever guidelines we come up with should result in a single most prominent place=city for many/most urban areas while still allowing for multiple place=city nodes in some dual/tri/(quad?) city areas.

uk4sm · 2024-04-26T00:24:12.877Z

This is the only actual proposal anyone has posted on this discussion. Well done, @miela404, I’m all for it. If New Englanders refuse to go by population, this is the best thing for city classification. I would personally take it a step further, and make it metros only, but then rural states would only have one city (which isn’t a bad thing, but some people might not like it for their rural state).

Unless anyone comes up with something that is more logical than US Census classifications, this should be adopted, at least for New England.

clay_c · 2024-04-26T01:58:31.010Z

Thank you for deleting your more inflammatory comments, @uk4sm. Please take care to engage with the community in a more constructive manner going forward.

What you are referring to is not a proposal, but just an idea. A consensus process takes time. There is an ongoing discussion about classifying New England place names here—I encourage you to contribute to it.

uk4sm · 2024-04-26T03:05:13.855Z

Oh I didn’t delete them because they were inflammatory (they weren’t) I deleted them because people generally can’t take criticism, and when they do receive it, they turn on whoever criticised them. I’m more interested in coming to a solution, rather than being ignored for my critical comments, as @edops has been for theirs.

ezekielf · 2024-04-26T03:24:22.935Z

uk4sm:

I’m more interested in coming to a solution, rather than being ignored for my critical comments, as @edops has been for theirs.

Are you two and @miela404 all friends? Your communication styles are all quite similar…

uk4sm · 2024-04-26T03:34:39.152Z

Ha ha I do most of my edits in and around London actually, so no I don’t know who they are. My interest in New England comes from my family that lives in Maine, (and the fact that it’s named after England). so It’s a place I visit often.

Not sure what you mean by “communication style” but I actually disagree with @edops on quite a few of their points, particularly regarding places like Brunswick. I think they should look into the US census Micropolitan and Metropolitan areas for New England, instead of going off their “local knowledge”

ezekielf · 2024-04-26T16:04:54.403Z

uk4sm:

I do most of my edits in and around London actually

I noticed that: Your OSM Heat Map @ neis-one.org. Looks like you’ve done a lot of good work on the London Zoo. I’ve been wondering why a relatively new user who’s mapping interest has been limited to very specific London locations would all of a sudden taken a very serious interest in New England place classification and Maine specifically. You joined the forum 16 hours ago after a four month hiatus from your London mapping. Why the sudden involvement in this topic? How did you become aware of the discussion?

uk4sm:

My interest in New England comes from my family that lives in Maine, (and the fact that it’s named after England). so It’s a place I visit often.

And yet you haven’t done any mapping there. Curious.

uk4sm · 2024-04-29T18:13:16.607Z

How about “none of your business” for an answer? Don’t like that? Well I don’t care. Such nosy questions.

clay_c · 2024-07-30T21:46:46.300Z

@edops has deleted their account, but now a new account by the name of edemes appears to be making broad changes to place classification in New England.

Portsmouth, NH
Bennington, VT
Waterbury, CT
Bristol, CT
Rochester, NH
Claremont, NH

I can think of two possibilities here: either New England is a mysterious abyss that new users worldwide are fixated on classifying, or @edemes and the various inactive accounts in the above discussion are all just one very opinionated person trying to endear themself to the community with sockpuppets and edit wars.

ZeLonewolf · 2024-07-31T13:11:23.086Z

I added a changeset comment on their most recent changeset asking them not to make any further changes to place classifications before discussing it here.

ZeLonewolf · 2024-07-31T13:31:22.091Z

I reverted 56 changes to place nodes in this changeset.