I am reading this info from the FCC published list of PSAPs and individual PSAP websites. You can do it as well if you like.

I have not looked at importing these data. I add this information when editing because it seems important to have these kinds of anchor institutions mapped.

Thanks David, that issue is pretty annoying.

County Road XXX and State Highway YYY and such are often the name as well as the ref.

There is ambiguity by state I guess, sometimes it is Statename YYY or some other variation.

Feel free to revert.

Apologies.

Many of the entities I saw on the map were well lined up with the imagery I was using and I only moved things that were not.

Is there local warping to the extent of several meters or is there just patchy misregistraion of the OSM objects?

Thanks for finding this, that must have been a copy and paste error.

Walmart Supercenter buildings are distinct semantic entities that can be tied to a well defined geometry. The solar panels on the roof can be attributed to Walmart inc and estimates made of their output. The loading docks on the boundary of the polygon can be counted and assessed against the floor area and local road capacity. The subway inside the store is semantically inside the store instead of merely near the store.

I have been modelling the world in a manner that uses the expressive power of the medium to enable the best geographic inference.

I am sorry, I was not aware of a local norm mandating all point venues. If you would like to give me a bbox delineating the area of this convention I will avoid it in future.

I am not sure I understand.

The name of the road is Farm-to-Market Road 685 or similar, in the same way that there are roads named County Route 652, US Highway 22 and State Highway 13.

Thanks for spotting that, I must have confused "New Londons".

If you emit a lat-long in the parking lot of that church, your semantic activity should come out as religious from my code. Not residential.

If I were standing on a parking lot of a church, ready to walk to church, I would not feel like I lived there.

If you park your car there, the lat long I get from it as I powers up again should be semantically tagged as religious in nature. Not residential.

As I said, I am completely fine with not further tagging the landuse around single worship entities. But am pointing out the implications of the current scheme.

This is interesting but leads to tricky use cases. Looking at the logic of automated analysis:

First, any piece of land should in principle have some sort of landuse attached, even by default implicitly or explicitly stating unused. If the landuse for this small region is not religious, what is it?

Further, a car parked in that parking lot is engaged in religious activity. It is a private parking lot owned by the religious entity associated with the place of worship. The attributes for the parking lot, were tagging complete, might indicate it is not public and for the use of the facility attached. A land use polygon of course encodes that naturally avoid the entire issue of parsing a parking lot tagset.

Assume I have a billion accurate lat-longs and I want to find which of them are attributable to religious entities. I would want any falling into the church grounds, especially the parking, to come up with a religious tag.

Inside the US I can do this with land parcel data, but that is messy compared to just having broadly implemented landuse=religious tagging scheme.

I am not disputing your assertion or the practices in the wiki and will avoid edits contravening that. I am noting that this practice leads to inconsistencies when parsing the model of the world.

Thanks for the info. Is there a place where these things are documented so I can understand them in future?

Sorry, I was trying to separate out the home depot building from the other buildings and must have forgot to put them back in.

I found a bunch of situations where a building marked as a Home Depot store actually included several other stores. Also situations where the Home Depot was a point and needed it's building separated out from the larger instance.

Thanks for catching that.

Thanks, fixed that.

From consulting the source, jw.org

I also merging GNIS data already in OSM. At some point I will clean up the outdated or otherwise incorrect related GNIS.

I also use county land parcel govt opendata, for example

https://marioncounty.maps.arcgis.com/apps/webappviewer/index.html?id=b41e1f1b340a448682a2cc47fff41b31&find=072W20CA03000

This org has financialized a bunch of their real estate assets recently and knowing what they own vs what they use for their core mission is very interesting.

An additional twist is that all 151 of the PA locations have a litigation hold and subpoena on them for CSA documenting material at present. It is likely this legal action will imperil the finances of the entire org, leading to increased interest into their assets.

This is a data journalism submarine waiting to surface. I am getting things somewhat in order for whoever picks up the story down the line.

First hand knowledge, I stayed the night there. They are weird hippies with conspiracy theories about microwave ovens but it is a nice cheap guest house that is clean and functional.

A few pics here

https://imgur.com/gallery/nAD53qj

Thanks for catching that, clearly an error.

Sounds like a job for a fixing bot. I am not sure how you feel about it but is certainly violates the no abbreviations rule here osm.wiki/Editing_Standards_and_Conventions and quite apart from that offends my sensibilities immensely.

Seems like an error, will fix thanks.

Sorry about that, "roads" and "buildings" are the two dominant things I do and if I switch between the two I sometimes forget to switch the comment.

The next edit was the buildings on this location, mobile home parks are important concentrations of living units that tend to get neglected in a variety of spheres.

A bunch of my edits are driven by news events. A bunch are driven by an infrastructure analysis project I am tinkering with. Another set is driven by global completeness concerns for specific classes of entity. Another is driven by demographic underepresentation.

The worklists are all built by my analysis tools for myself.