It’s technically the same one. Notice that I said, “first/second half of one changeset,” not “two changesets.” I did this by leaving the changeset open (this is something you can do in JOSM), restoring highway=motorway (and highway=motorway_link), and closing the changeset. One thing I learned is that JOSM will say something like, “…to the currently open changeset 184114344,” for example.
I think if you’re going to make large scale changes and immediately roll them back, you should not do that with the public OSM data set.
That said, I guess the question is whether that changeset actually did roll everything back or not.
3 Likes
I just verified with Vespucci that the highway=* classifications are back to normal.
Also, as previously stated, I’m doing this on the sandbox.
I’m not sure what you think the sandbox is, but that’s live OSM data you’re modifying.
Given the understanding that changeset/184114344 contains both modifications and the reversion of the modifications, here’s an updated summary of the net effect on tags after the changeset:
781 - noref = 'yes'
760 - motorway_link = 'trunk_link'
55 + motorway_link = 'trunk_link'
28 + motorway_link = 'primary_link'
21 + motorway = 'trunk'
15 ~ motorway_link: 'trunk_link' -> 'primary_link'
12 + hov:minimum = '2'
10 - motorway_link = 'yes'
6 - motorway = 'trunk'
2 ~ destination: 'Russelll Road;Frank Sinatra Drive' -> 'Russell Road;Frank Sinatra Drive'
2 ~ destination:ref: 'US 95' -> 'I 11 North;US 95 North'
1 ~ turn:lanes: 'none;slight_right' -> 'through;slight_right'
1 - traffic_sign:direction = 'forward'
1 + service = 'emergency_access'
1 + motorway = 'primary'
1 - layer = '1'
1 + lanes = '1'
1 + junction:ref = '9'
1 - highway:motorway = 'primary'
1 ~ highway: 'motorway_link' -> 'service'
1 + direction = '130'
1 - destination:street = 'Falcon Ridge Parkway;West Mesquite Boulevard'
1 ~ destination:ref: 'US 95 North' -> 'I 11 North;US 95 North'
1 ~ destination:lanes: 'Russel Road;Frank Sinatra Drive|none|none' -> 'Russell Road;Frank Sinatra Drive|none|none'
1 + destination = 'Falcon Ridge Parkway;West Mesquite Boulevard'
… and most of those changes are benign.
Removing 781 instances of noref=yes isn’t ideal, but it looks like those were mostly tags you added earlier. And the main value is to indicate to mappers that a ref tag is not missing.
I don’t know if anyone is looking at the motorway_link or motorway tags other than you, so those changes are not likely meaningful. And if the destination*= updates used local signage they would be good.
All told, there’s a lot less there than a changeset that touches 5000+ elements might suggest.
But if there’s a lesson here, it’s that tweaking public OSM data for a short term experiment is not a great idea. There are other ways to do that without making changes public.
4 Likes
I guess there’s another lesson. 5000+ element changesets that haven’t been previously discussed are going to attract some attention. It would be better to discuss them first.
And FWIW, I was able to get a summary of the other changesets around Las Vegas over the last day or so. They were mostly TIGER cleanup, which is good. And some other changes that seem reasonable. And there were some name changes for roads that look questionable but I think most of those were already noted in separate changeset comments.
3 Likes
None of discussed edits were done in sandbox, discussed edits were done on osm.org live data that is NOT a sandbox. Please do not treat it as a sandbox.
We in fact have sandbox. You can create account at https://master.apis.dev.openstreetmap.org/ for testing editing without affecting the real database.
Note that it has no dedicated map rendering or overpass server and data is different than one on live osm.org
If you want overpass, rendering etc you will need to provide it on its own. If you want to test edit on data matching real database you will need to add it first.
5 Likes
I already did.
I am writing regarding a series of widespread tagging edits made by user @Flap_Slimy_Outward across two states, which have introduced incorrect highway names and broken local data consumption (such as in Mapillary and DoorDash).
I attempted to resolve this directly with the user last week, but they either avoided my questions or only partially answered them. So I was asked to bring it here for additional insight from the community.
1. Incorrect Naming of US 6 (Changeset 180474880)
In this changeset, the user replaced or added the name=* tag as “Grand Army of the Republic Highway” across the entirety of US 6. Another mapper, @ZLima12, reached out to them and later reverted the edit. When I asked the user for their source and whether they had consulted local communities, they stated they had not because the name “applies to US 6 nationwide.” They cited the Wiki as their source. While the Wiki confirms U.S. Route 6 is designated as the “Grand Army of the Republic Highway,” this is a route designation, not the primary street name used for local navigation.
2. Incorrect Naming of I-70 and US 6/50 (Changeset 180334382)
The user changed the name=* tag of US 6 & US 50 from “Grand Army of the Republic Highway” to “Dinosaur Diamond Prehistoric Highway,” and applied this name to major sections of US 6 and I-70 in the Grand Junction area.
-
The Issue: “Dinosaur Diamond Prehistoric Highway” is a National Scenic Byway route * designation, not a local street name. Furthermore, according to official definitions and the Wiki, this specific scenic byway designation does not even extend into Colorado on these segments; it covers the Utah side of the relation. (I previously * created the correct Colorado relation for this scenic byway to prevent this exact issue).
-
I asked the user to revert these changes, but they misunderstood the request, and the incorrect names remain live.
Request for Assistance
Due to the large scale of these changes across multiple states, it is incredibly difficult for me or other local mappers to manually audit and verify which specific edits were valid and which need a full reversion. Because the user has been uncooperative in providing their data sources or acknowledging the distinction between name=* and route relations, I would appreciate the assistance in reviewing these changes and potentially executing a clean reversion where necessary.
Thank you for your time and help.
Oh, you meant removing the name “Dinosaur Diamond Prehistoric Highway” from US 6, US 50, and I-70?
I guess I can do that.
If you’re having issues with reverting changesets, contacting DWG would be a good idea. It also helps keep them in the loop on patterns of behavior.
And for completeness, there was another recent discussion on OSM US Slack about the use of Google Maps as a source. @Flap_Slimy_Outward it might be good if you could follow up on that thread.
1 Like
I did contact them regarding this. They suggested I bring it here to get more feedback from the community to see if the community agreed that these edits should be reverted before they rolled them back.
1 Like
I already did it: https://www.openstreetmap.org/changeset/184596068
Yeah, I’ll go check that out 
1 Like
@Flap_Slimy_Outward Can you please NOT change any name or any ref or any classification of any highway in the US or elsewhere until you have proved yourself more capable of “reading the room” when it comes to working out what is appropriate and what not.
People have been incredibly polite with you - perhaps more than you deserve, but there comes a time when we all have to recognise the cost of the rework that you are causing the community.
If you like to make one of these sorts of changes, please bounce it off this forum (or the DWG) first, and don’t proceed without unamimous approval.
– Andy (from the DWG, though not directly involved with recent FSO tickets).
11 Likes
This edit simply undoes an earlier edit I made, based on this post from a little earlier in the thread.
Also, as I’ve stated before, I’ll be more careful when editing highways (especially outside my local city) from now on.
1 Like
Don’t just be more careful, talk about any proposed changes before you make them!
2 Likes
I will do this more, starting with this strange set of roads in Chicago.
and do not make even 100% fine edits if anyone at all is opposed in any way to them (" don’t proceed without unamimous approval.")
1 Like