By the way, I'm not asking you to fix this - just to tell me (a) where you imported the building footprints from, and (b) where you got the address information that you attached to them. I've not been able to find the data sources unambiguously based on the changeset comments. I'm pretty confident that I can design a mechanical edit to undo most of the damage, but it would help to know the root cause (and also to know that you're not continuing to install corrupted addresses. I've found them as recently as changeset/88825072 - which is over a year old but I see that you haven't done all that much since then.

Plenty of examples in Little Falls, too!

http://overpass-turbo.eu/?q=LyoKVGhpcyBoYcSGYmVlbiBnxI1lcmF0ZWQgYnkgdGhlIG92xJJwxIlzLXR1cmJvIHdpemFyZC7EgsSdxJ9yaWdpbmFsIHNlxLBjaMSsxIk6CsOiwoDCnCJhZGRyOnN0csSMdCJ-xZZNYcS6fFByb3NwZWN0fE1vbnRnb23Eknl8UmFuZHxGaXLFkXxTxaLFp8WzxIPFtsWzRm_EqMScxbRpZsaEU2l4xoRDxIhybGVzfE5ld3TFp3xDxadjb8SxfFdpbHPGmEfFnsShfELEsMS3bsSQcnxKb2huxp_GkcWkxaZyZ8WxfE90xL_FqiQixYjCnQoqLwpbxoJ0OmrGo25dW3RpxazHiToyNV07Ci8vxI_ElMSdciDFk3N1bMa9CigKIMeqxYR5W8WLxY3Fj8WRxZNlxZUhxZcgIl0oe3tixKp4fX0px5nHqsiDCseax5xwxqt0x6HGkcekx6bHicSYb2R5x5k-x5nIkXNrZcS9cXQ7&c=ArCCAh6VEQ&R

Here's a shorter query that also shows the problem. It'll give a false positive on one-word street names like 'Broadway', but those are rare.

http://overpass-turbo.eu/?q=LyoKVGhpcyBoYcSGYmVlbiBnxI1lcmF0ZWQgYnkgdGhlIG92xJJwxIlzLXR1cmJvIHdpemFyZC7EgsSdxJ9yaWdpbmFsIHNlxLBjaMSsxIk6CsOiwoDCnCJhZGRyOnN0csSMdCJ-xZZNYcS6fFByb3NwZWN0fE1vbnRnb23Eknl8UmFuZHxGaXLFkXxTxaLFp8WzxIPFtsWzRm_EqMScxbRpZsaEU2l4xoRDxIhybGVzfE5ld3TFp3xDxadjb8SxfFdpbHPGmEfFnsShfELEsMS3bsSQcnxKb2huxp_GkcWkxaZyZ8WxfE90xL_FqiQixYjCnQoqLwpbxoJ0OmrGo25dW3RpxazHiToyNV07Ci8vxI_ElMSdciDFk3N1bMa9CigKIMeqxYR5W8WLxY3Fj8WRxZNlxZUhxZcgIl0oe3tixKp4fX0px5nHqsiDCseax5xwxqt0x6HGkcekx6bHicSYb2R5x5k-x5nIkXNrZcS9cXQ7&c=Aq-p8Igp4Q&R

Sure, here are a bunch. There are so many that it seems to be a systemic issue.

http://overpass-turbo.eu/?q=LyoKVGhpcyBoYcSGYmVlbiBnxI1lcmF0ZWQgYnkgdGhlIG92xJJwxIlzLXR1cmJvIHdpemFyZC7EgsSdxJ9yaWdpbmFsIHNlxLBjaMSsxIk6CsOiwoDCnCJhZGRyOnN0csSMdCJ-xZZNYcS6fFByb3NwZWN0fE1vbnRnb23Eknl8UmFuZHxGaXLFkXxTxaLFp8WzxIPFtsWzRm_EqMScxbRpZsaEU2l4xoRDxIhybGVzfE5ld3TFp3xDxadjb8SxfFdpbHPGmEfFnsShfELEsMS3bsSQcnxKb2huxp_GkcWkxaZyZ8WxfE90xL_FqiQixYjCnQoqLwpbxoJ0OmrGo25dW3RpxazHiToyNV07Ci8vxI_ElMSdciDFk3N1bMa9CigKIMeqxYR5W8WLxY3Fj8WRxZNlxZXFl17FmcWbxZ3Fn8WhxaPFpcWnxanFq8Wtxa_FscaAxbbFuMW6xpzFsnzFvsaexoHGg2jGhcaHyJDGicaLyJDGjcSwxpDGksaUxpbGmMaabsacxp7GoMaixqTGpmXGqMaqxLrGrcavxrHGs1fGtce-xrjGusa8xr5vx4BdKHt7YsSqeH19KceZx6rJgArHmseccMardMehxpHHpMemx4nEmG9keceZPseZyY5za2XEvXF0Ow&c=Aq9HDMUm_O&R

Tags from TIGER/Line 2021 that I missed removing after I copied/pasted them to the `nist:fips_code` and `gnis:feature_id` on the relations. They're gone now.

This changeset (and the others committed in the area) have a systemic problem with addresses. Address prefixes and suffixes are stripped from the street names, so that a building that should have the address "332 South Washington Street" shows up with the address "332 Washington".

This abbreviation is not tolerable: "1234 State Highway 5" shows up as just "1234 5" which is a nonsensical address; "Third Avenue" and "Third Street" are not distinguished. Communities often repeat housenumbers between streets differing only in the direction prefix (301 East Main Street and 301 West Main Street are distinct addresses).

Obviously, with the great data volume, this will need to be repaired with a mechanical edit. Since I have access to the NYS street and address mapping that you apparently used, I can do the point pairing and override any street names that are unchanged from your import, but I'd appreciate it if you could help me with identifying the scope of the work - how many of these huge changesets are there, and how much of the state do they cover? Does the problem affect all ways imported by `NYBuildings`? Does it extend beyond `NYBuildings` to other import work you may have done?

Most states don't have nearly the same amount of overloading of names: Village of Schoharie inside Town of Schoharie inside Schoharie County.

New York even overloads names at the same admin_level: City of Tonawanda is admin_level=7, same as Town of Tonawanda, which borders it; dittto City/Town of Plattsburgh. I don't have a good way to make the boundaries in those cases comprehensible without going to the longer names.

'"Town of" is not needed for New York'? Can we discuss this on Slack or one of the mailing lists please? It's extremely common that a name will be duplicated among counties, cities, towns, and villages. It may be redundant for the specific instances you've been editing, but I've been going toward using a bare 'NAME' on the place node (the `label` role in the relation, which I'm trying to have always present), and 'NAME County', 'City of NAME', 'Town of NAME', 'Village of NAME' on boundaries. Otherwise, how will you handle the boundary between the Town of Plattsburgh and the City of Plattsburgh, or show the Village of Nunda within the Town of Nunda?

Right, and the settlement agreement lists tax parcel numbers, so when in doubt I can examine the individual parcel records - at least in the counties where they're licensed permissively. The map you link to is at an awkwardly small scale to answer the detailed questions when conflating boundaries.

Hi, I'm in the process of adjusting the minor civil division boundaries if New York to align to NYSGIS. The ones from the 2008 import of Census Bureau data are often horribly misaligned - I've moved some by a quarter-mile or more.

Some of the boundaries that I've been editing have been conflated with land belonging to the Haudenosaunee nations. In particular, I'm writing this message to warn you that some of the boundaries of the Onyota’a:ká: nation are getting edited because they were glued to town lines that are being adjusted.

I'm also fixing some topological problems as I go.

Could I ask you to have a look and make sure that I haven't broken anything horribly? (JOSM's validator doesn't find anything to complain about, and by eyeball it appears to match the map from the state and BIA, but that's no guarantee that it's actually correct!

For what it's worth, at the time I'm writing this (version 4 of relation/13737250), I've not done anything to the boundary south of Oneida Lake and Rome - but I'm working acctively in the area, so I may have done more by the time you read this.

Ah... I see what happened here. 146 Front Street is the historic village hall, which has been sold and repurposed - but still has the name in the masonry on the building.

I'm updating minor civil divisions in New York, incliuding making sure that administrative centers are present, and I see that Village of Deposit gives 61 Front Street as the address of the village hall. Is the web site correct? (I mapped a village hall at the address given, and then discovered the conflict.)

I checked, and this is not a confusion with the Town of Deposit offices, which are at 3 Elm Street.

Did you perhaps merge a misplaced address point from the GNIS import?

I'm attaching this message as a comment to several changesets,
because I see that there were several cooks in the kitchen here.

I see that there's a polygon,
way/156508026,
that's gone through several rounds of editing, and which I don't
understand. I'm entering this message on several changesets that
have edited it in hopes that someone can explain to me what's
supposed to be going on here. Multiple users have been
adding/deleting `landuse=residential`, and adding/deleting
`place=town`.

What is the actual intent of this area? If it's io indicate a
residential landuse, then it is overbroad, because I can see that
it includes several schools, several religious facilities, a
couple of parks, a cemetery, a golf course or two, and what
appears to be a commercial strip. It would be possible to make
the area the outer way of a multipolygon and then cut out these
inner land uses as inner ways, or to break the residential area
up into smaller pieces.

If the intent is to mark the boundaries of the Town of DeWitt, as
Stupefacient's comment on changeset/9116121 would tend to
indicate, that's also not correct. This irregular round area is
about half in the City of Syracuse and half in the Town of
DeWitt. Th portion that is in Syracuse appears to be something
that's informally identified with the Meadowbrook neighbourhood,
while the portion that's in DeWitt township coincides almost, but
not quite, with the DeWitt census-designated place - which
I plan to upload as `boundary=census`.

I'm in the process of updating minor civil divisions in New York
from 2020 census data and 2021 New York State reference data,
and I hope that someone who's manipulated this area can give me
some insight into what should be done with it.

I'm attaching this message as a comment to several changesets,
because I see that there were several cooks in the kitchen here.

I see that there's a polygon,
way/156508026,
that's gone through several rounds of editing, and which I don't
understand. I'm entering this message on several changesets that
have edited it in hopes that someone can explain to me what's
supposed to be going on here. Multiple users have been
adding/deleting `landuse=residential`, and adding/deleting
`place=town`.

What is the actual intent of this area? If it's io indicate a
residential landuse, then it is overbroad, because I can see that
it includes several schools, several religious facilities, a
couple of parks, a cemetery, a golf course or two, and what
appears to be a commercial strip. It would be possible to make
the area the outer way of a multipolygon and then cut out these
inner land uses as inner ways, or to break the residential area
up into smaller pieces.

If the intent is to mark the boundaries of the Town of DeWitt, as
Stupefacient's comment on changeset/9116121 would tend to
indicate, that's also not correct. This irregular round area is
about half in the City of Syracuse and half in the Town of
DeWitt. Th portion that is in Syracuse appears to be something
that's informally identified with the Meadowbrook neighbourhood,
while the portion that's in DeWitt township coincides almost, but
not quite, with the DeWitt census-designated place - which
I plan to upload as `boundary=census`.

I'm in the process of updating minor civil divisions in New York
from 2020 census data and 2021 New York State reference data,
and I hope that someone who's manipulated this area can give me
some insight into what should be done with it.

I'm attaching this message as a comment to several changesets,
because I see that there were several cooks in the kitchen here.

I see that there's a polygon,
way/156508026,
that's gone through several rounds of editing, and which I don't
understand. I'm entering this message on several changesets that
have edited it in hopes that someone can explain to me what's
supposed to be going on here. Multiple users have been
adding/deleting `landuse=residential`, and adding/deleting
`place=town`.

What is the actual intent of this area? If it's io indicate a
residential landuse, then it is overbroad, because I can see that
it includes several schools, several religious facilities, a
couple of parks, a cemetery, a golf course or two, and what
appears to be a commercial strip. It would be possible to make
the area the outer way of a multipolygon and then cut out these
inner land uses as inner ways, or to break the residential area
up into smaller pieces.

If the intent is to mark the boundaries of the Town of DeWitt, as
Stupefacient's comment on changeset/9116121 would tend to
indicate, that's also not correct. This irregular round area is
about half in the City of Syracuse and half in the Town of
DeWitt. Th portion that is in Syracuse appears to be something
that's informally identified with the Meadowbrook neighbourhood,
while the portion that's in DeWitt township coincides almost, but
not quite, with the DeWitt census-designated place - which
I plan to upload as `boundary=census`.

I'm in the process of updating minor civil divisions in New York
from 2020 census data and 2021 New York State reference data,
and I hope that someone who's manipulated this area can give me
some insight into what should be done with it.

No idea what you saw - I don't see it on current Bing, or Mapbox Satellite, or NYS Orthos Online..Am I missing something? Should we keep this?

Oh, and I'm a little puzzled that so much of your mapping of political borders shows Bing Aerial as a data source - political boundaries are seldom visible and almost never verifiable from imagery. I'm guessing it's just that you were combining these with edits to objects of other types?

Hey, ECRock,

I've been going through New York administrative boundaries trying to get them reconciled with the fairly authoritative data set at http://gis.ny.gov/civil-boundaries/.

I'm starting to move in on your turf - there are a lot ot townships that you mapped originally that appear to have differences with the NYSGIS data set.

We seem to have different approaches to conflation - you seem to favor conflating a town line with any road or waterway that appears to follow it - I've even seen a few conflated with driveways and pipelines (although it may have been a different user that did that). I strongly prefer conflating those features only where I know that one defines the other. Otherwise, there are too many cases that we'll get wrong. An area like osm.org/#map=16/42.5576/-79.0906 might look a mess, but it's correct - the river moved and the political boundary didn't! I find the practice of zealous conflation particularly dangerous when the ways in question are tiger:reviewed=no, since users are likely to review the way, drag its geometry around, and damage the boundary.

Finally, I'm looking to revise names like 'Newfield Town' to 'Town of Newfield' - no local says the first form, nor do the signs carry it.

I'm hoping to get this done without firing the first shot in an edit war, so I hope you're OK with most of this! If you spot a problem, be sure to let me know, so that I can revise or revert.

Hi, Steve.

I'm taking the liberty of removing the 'Town of Ellery' way in the middle of Chautauqua Lake, since it never was useful (admin boundaries have to be closed), and I'm in the process of importing all the missing townships from NYSGIS data. Expect to see Town of Ellery with a complete boundary in OSM shortly.

It's a temporary tag to keep track of associations between an incremental upload and an external database. I included the username so that people like you would know whom to ask. Plan is to remove the tag once the upload no longer needs to refer back to the ways - that'll happen once townships start bing completely bounded. Expect to see some of these tags start to disappear in a few days.