Yes, if it can be reverted cleanly. Otherwise the refs and relation would need to be removed manually. SR 42 is already tagged in old_ref=* where it belongs. (Though it should also be mapped in https://www.openhistoricalmap.org/ where it *really* belongs.)

There’s no such thing as an “Old State Route” designation that would justify network=US:CA:Old on this or the other changesets.

Here’s a tip for preventing this situation in JOSM: https://community.openstreetmap.org/t/us-89-wyoming-relation-truncated-apparently-accidentally/146147/2

Interesting, they used to advertise themselves as half grocery store, half flooring store. Liquidator of random things describes it better than anything else though! I guess I’ll have to find another favorite example of a store that’s impossible to map as a single POI in OSM.

Grandview isn’t just a flooring store. When you combined the POI with the building, you deleted a few other copies of the Grandview POI that we had to keep around to represent the grocery store and shed store. Are you sure they’ve stopped selling anything other than flooring?

Both dictionaries are under copyright. But they were published in the U.S., so they’re only eligible for very limited copyright protection that would cover a creative layout decision or the exact wording of a nontrivial definition. In particular, lemmata cannot be copyrighted. [1] Both are translation dictionaries, in which the definitions are as brief and utilitarian as possible, so no copyright interest is possible in the portion that has been used.

[1] For a more in-depth legal analysis with the usual caveats, see for example https://meta.wikimedia.org/wiki/Wikilegal/Lexicographical_Data

https://community.openstreetmap.org/t/aligning-state-borders-to-pavement-quality-transitions/145948

boundary=administrative is an administrative boundary, not a drawing of where Caltrans and ODOT left their mark. DOTs usually run over the boundary for practical reasons.

changeset/186667493 undoes the boundary changes.

You didn’t align the road with the boundary, you aligned the boundary with the welcome signs. This is taking the welcome signs far too literally. Caltrans often puts the signs slightly before or after the state line as circumstances allow. [1] These are informational traffic signs rather than legal boundary markers. They can satisfy the requirement that a boundary be verifiable but they don’t define reality.

[1] https://commons.wikimedia.org/wiki/File:California-Inyo-Nevada_border_signs.jpg

More explanation, particularly about the unlikely jog in the Ashland–Iron county line:

https://community.openstreetmap.org/t/county-lines-in-the-great-lakes/128172/8

changeset/186152590 restores the normal name and branded features per survey.

Hi, thanks for adding the missing bit of cycleway. Are the YMCA’s parking lots really named “Parking Lot” followed by the secondary intercardinal directions? way/1537841823 It’s OK to leave a parking lot unnamed if it doesn’t go by a common name in practice.

By the way, if this street is highway=service, some routers will avoid using it as a cut-through. For example, Valhalla is passing the exit and taking the next exit, then doing a U-turn back onto Highway 85 in order to get to Almaden Expressway. Routers often try to avoid cutting through service roads because they aren’t intended as part of the public road network.

During its lifetime, Almaden Plaza Way has been through seemingly every possible highway classification except highway=raceway, so anything is possible. 😉

In isolation, the eastbound lanes do look like a candidate for highway=motorway_link. But we would need to either make the westbound lanes a link way that leads nowhere, or else the two sides of the street would have wildly different classifications. Most mappers avoid classifying a street differently on either side. There are occasional exceptions, typically because both sides are considered different streets.

A similar thing happens routinely in Texas. Freeway exits there typically empty onto a frontage road. If the ultimate destination of that exit is highway=primary, a short portion of the frontage road gets upgraded to match. Here’s an example: way/599989943

This situation is so rare here that I can understand how highway=primary would seem like a mistake.

I’ve reverted this changeset because Almaden Plaza Way is a connector for an Interstate exit onto an arterial road. Even if not for the ramp for I-280, service=driveway is for residential driveways and the like, not parking lot access roads.

changeset/185985290

We do use superrelations for some things like routes, but not for boundaries. Anyways, here’s the forum discussion on whether we should even have MSA boundaries in OSM:

https://community.openstreetmap.org/t/chicago-metropolitan-area-relation/131055/9

The handful of classification changes appear to be incidental to the route updates in the changeset. More discussion in https://community.openstreetmap.org/t/wholesale-missouri-route-edits-by-ne2/145510

https://community.openstreetmap.org/t/chicago-metropolitan-area-relation/131055/9

The sign is mounted on a streetlamp pole, actually a rigid metal sign now that I look more closely.