Reviewing recent changes by Flap Slimy Outward

There’s a thread in #local-california on OSM US Slack about recent changes that @Flap_Slimy_Outward made to about 4000 ways in and around US 101. I’ve asked DWG for help to revert that changeset so we can figure out what changes, if any, should be kept. But while reviewing recent history on OSMCha, I’ve noticed several other recent changesets that might be problematic.

changeset/184114344 is a large scale reclassification of highways around Las Vegas, NV.

There’s also a sequence of changesets from yesterday that modified road names and classifications. It’s not clear to me how much of this was productive, but there are already some comments questioning the work in the changesets.

It would be helpful to have reviews of these changesets from local mappers, particularly in Nevada. Were these changesets productive, or are there issues that need to be addressed?

2 Likes

Sorry. I may have grabbed the wrong link from OSM US Slack. This is the thread about the changes around US 101.

I’ve left the other link in place because it’s also about recent changes that should be reviewed.

Also for reference: Mountain West Large Road Route/Road Name changes - Communities / United States - OpenStreetMap Community Forum

@Flap_Slimy_Outward FYI

This user is known to frequently make controversial edits: 30 “reviewed-bad” changesets on OSMCha, a long history of changeset comments, over 80,000 reverted changes, and many discussions about their problematic edits across Discord, Slack, and the Forums.

1 Like

Many of these changes are in Las Vegas. I adjusted lane counts, added sidewalks, and cleaned up TIGER tags.

If it wasn’t obvious by my Slack status, let’s just say that I am one of these “local mappers […] in Nevada.”

“I conducted an internal review of my edits and found no concerns.”

1 Like

Hi Flap,

First, I’ve appreciated your engagement in Slack and willingness to discuss the changes you’re making, even as you are frequently brought into a thread of critical comments, so thank you.

Knowing you are willing to engage constructively after the fact, and are as passionate about a complete map as the many others in the community, I’m left wondering how we can get to a point where you conduct proactive engagement before making the big changes. Would you be willing to post here or in Slack to share a plan before making large scale changes to highways? Do you feel like you have a place where you know you could engage the right people?

5 Likes

What specifically was wrong with this edits?

Probably something was given that you created this thread and asked DWG for reverts, but you do not mention any problems.

Large scale road reclassification is typically associated with bad edits, but then it should be easy to mention what was wrong or dubious.

Here’s one ongoing changeset discussion that elaborates on some of the issues. I’m trying to find one of the others.

Edit: Here’s another on the same stretch of road. There are many more, but these ones are recent: Changeset: 180790829 | OpenStreetMap

In the case of assigning names to highways in states other than my home state, I’m guessing the best places would be the “local [state name]” channels on Slack. Unfortunately, I have not had an easy time finding them.

I’m pretty sure the edits that will be reverted are the ones where I attempted to add names to US 101 in California.

I wasn’t going to try to review a changeset that touched 5000+ elements around Las Vegas without some local knowledge or input from people who are familiar with the area. And, as you may have gathered, @Flap_Slimy_Outward has a history of large scale edits that weren’t discussed beforehand and didn’t align with the community consensus. That’s why I posted here – to get some help with deciding where there might be issues with the changesets in Nevada.

As FSO mentioned, I have asked for the changeset in California to be reverted and I think there’s consensus that’s the right thing to do. The changeset modified a lot of road names, and in that case it’s going to be easier to start over with a clearer plan of what should be changed.

2 Likes

I noticed there were significant changes where highways were reclassified. But there are so many changes that it’s hard to review them. And I don’t generally work on highway classification in urban areas, so it’s not easy for me to say what may or may not have been appropriate.

I am one such local, if it wasn’t obvious by now. The “mass edit” you linked was actually two changesets in one. For the first part, I restricted highway=motorway (and highway=motorway_link) to Interstate Highways. The second half basically reverts the first half while cleaning up duplicate/useless tags (e.g., one leftover highway:motorway tag and the tag motorway_link=yes on a highway=motorway_link).

(See my entire paragraph above)

Since I’m a local, I know how the major roads are used.

I wish we could trust you to do the right thing with these changesets, but there are so many cases where that hasn’t been true that we really need someone else to look things over.

And, to begin with, it really isn’t a good idea to make large scale changes without discussing the changes first. I know you’ve heard that before.

4 Likes

Here’s a brief summary of the tag changes in changeset/184114344, with a count of the number of times a tag was changed, - for deleted tags, + for added tags, and ~ for modified tags.

EDIT: See updated analysis below.

I’m sure fixing a typo in a destination tag was fine. But removing 1062 noref=yes tags seems counterproductive. And what was going on with all the changes from highway=motorway to trunk or primary?

@Flap_Slimy_Outward can you explain why these changes were necessary? Are you sure that other mappers in the Las Vegas area would agree with these changes?

The California changeset’s problems are related to some points that have been made in other threads over the last several months. I wasn’t inclined to call anyone out on the forum over it. Now that it’s a scene, I’ll also try to explain the issue with some historical background that out-of-staters may not understand. None of this has any bearing on the changes in Nevada, for what it’s worth.

One road, multiple numbers

The State of California owns many miles of state highways (which we model as highway=* ways). The legislature has organized the State Highway System into a numbered series of state routes (which we model as route=road relations). These routes are enumerated in the Streets and Highways Code (yes, the same legal code that gave rise to the infamous service=driveway2). Section 401 describes Route 101:

Route 101 is from:

(a) Route 5 near Seventh Street in Los Angeles to Route 1, Funston approach, and, subject to Section 72.1, the approach to the Golden Gate Bridge in the Presidio of San Francisco via Santa Barbara, San Luis Obispo, and Salinas.

(b) A point in Marin County opposite San Francisco to the Oregon state line via Crescent City.

These routes used to be known as Legislative Routes because they completely differed from the Sign Routes posted by the Department of Highways (now Caltrans). The Sign Routes distinguished between State Routes and U.S. Routes. In 1965, the Legislative Route Numbers were updated to match the Sign Routes, without distinguishing between these two systems. (Interstates would come later.) At that time, Legislative Route 2 became Route 101 to match the U.S. Route 101 signs.

One road, multiple names

Since before the highway numbering system was introduced, various portions of these highways have been officially named by the legislature, Caltrans, local authorities, and private interests in an uncoordinated fashion. Back when Legislative Route 2 still existed, the legislature named it “El Camino Real”, since it follows a famous historical path from the Spanish colonial era. The name comes from Spanish, meaning “The Royal Road” or “The King’s Highway”, but English speakers only use the Spanish name without translation.

Much of the original Legislative Route 2 is still part of U.S. Route 101, but some parts have been bypassed, especially in the San Francisco Bay Area. The name “El Camino Real” stuck to much of the original highway, parts of which are now State Route 82. You can still see this name on street name signs.

Meanwhile, the new highways got different names. If the “Bayshore Freeway” were a surface street, the name would appear regularly at intersections and in addresses. But since it’s a freeway, the name is only rarely posted. Instead, it appears on maps and lives on mostly by word of mouth.

Over the years, the legislature has continued to apply memorial highway names, such as “James Lick Memorial Freeway” and “Colonel William R. ‘Bill’ Lucius Highway”. A memorial highway name gets two guide signs, one on either end of the designation. The designations often overlap, so that a given stretch of U.S. Route 101 might have as many as four different memorial highway names. No one ever uses a memorial highway name for anything. No one.

Regardless of all I’ve written above, most people just refer to the highway by the number of the route that it carries. Depending on the Caltrans district and other factors, signs may say “101 Freeway”, “Highway 101”, “U.S. 101”, or very occasionally “Route 101”. Colloquially, people say “the 101”, “Highway 101”, “U.S. 101”, or just “101”, depending on their dialect. So does Caltrans in their publications, very inconsistently.

As you’re probably aware, we have had a lot of debates over the years about whether the highways that carry numbered routes can be named after those route numbers. I think there’s some merit to pasting a route name on individual ways. But of all the possible number-derived names, “Route” is the least likely to be used on a U.S. Route.

One changeset, multiple errors

Changeset 183,896,352 tries to ensure that every way along U.S. Route 101 has a name=* and official_name=* tag by:

  • Adding official_name=Route 101 to each way. This name exists almost exclusively in the Streets and Highways Code.
  • Adding official_name=* based on names described in this article section on the English Wikipedia.
  • Applying “El Camino Real” to sections of present-day U.S. Route 101 that did not exist back when that name was designated on Legislative Route 2.
  • Making that official_name=* if there wasn’t already a name=*, regardless of the facts on the ground.
  • Changing old_name=* where one memorial highway name overlaps another.
  • Changing alt_name:en=El Camino Real name:en=The King’s Highway even though no one ever calls it that.

@Flap_Slimy_Outward notes that they got very confused trying to model the Wikipedia article in OSM. I don’t blame them. As an encyclopedia rather than a GIS dataset, Wikipedia sacrificed precision in the interest of readability. For one thing, they never note when a stretch of the highway has no name, practically speaking, because that would be so boring to read over and over again. Another problem is that the English Wikipedia’s roadgeek community has moved on to edit the AARoads Wiki instead, so many road articles are now being maintained by less knowledgeable editors who might not get these nuances right. I think we all take the shortcut of relying on an encyclopedia from time to time, but a change this dramatic would benefit from digging a little deeper for the tertiary source’s own sources.

The closest thing we have to an authoritative source for state highway names is Named Freeways, Highways, Structures, and Other Appurtenances in California, an annual listing published by Caltrans. Last year’s 2024 edition is the latest I can find. The introduction covers what I’ve explained and more. It pulls together multiple sources of names, from legislative acts to names that have survived from the pre-highway era. I highly recommend preferring this document over Wikipedia as a source for name=* and official_name=* on state highways.

(Ignore the “All rights reserved” on the cover page. That’s a great example of a government agency unaware that they have no right to any copyright.)

10 Likes

A few remarks with my DWG hat on:

  1. If changes are questionable and if there are too many to individually review, it is not unusual to revert the lot and ask the uploader to split things up in pieces that are easier to review. If someone overwhelms the community with their sheer amount of problematic edits (in between many good edits) it is all right for the community to call a stop to the whole endeavour. The user has received three blocks already; if they continue to do stuff that upsets the community then we’ll start giving them longer blocks (starting with a day and then doubling each time they do another undiscussed mass change).
  2. I have reverted changeset 183896352 entirely.
  3. The DWG has a separate complaint about changeset 180474880 (adding “Grand
    Army of the Republic Highway” as the name to all of US 6), and 179924057 (“Dinosaur Diamond Prehistoric Highway”), which we haven’t acted upon yet; I’ll refer the complainant to this thread.
7 Likes

Might be related to his thread: [RFC] Feature Proposal – Motorway classifications

1 Like

This was done on motorway_links. Since Vespucci’s “check rule” for either name=* or ref=* also includes motorway_links and trunk_links, I added noref=yes to tell Vespucci that motorway links usually don’t need a ref=* tag.

Like I said, this changeset was actually two changesets in one. I’ll just quote myself from earlier:

For my “motorway classifications” proposal, I wanted to illustrate how highways would be classified according to this new scheme. I quickly changed them back to prevent disruption of the database (I’m currently working on this in the sandbox). The second half of my changeset changed them back to highway=motorway (and highway=motorway_link), so, overall, there weren’t that many changes.

So there was a second changeset that reverted some or all of changeset/184114344? Which changeset was that?