[Talk-us] US-wide highway reclassifications by Fluffy89502

[Minh Nguyen]

Vào lúc 18:16 2021-10-16, Bradley White đã viết:
> I don't think a wholesale revert is necessary given that it sounds
> like from other users that some of Fluffy's changesets have been
> legitimately useful. Changing road classifications tends to ruffle
> feathers of local mappers so I think most of these less-helpful
> changesets will be reverted naturally anyways (though it is certainly
> frustrating to have to waste time cleaning this mess up). I certainly
> would be in favor of a block until they agree to better changeset
> comments and participation in discussions over road classifications if
> that's what the majority of their edits are going to be.

A block could be an opportunity to ask Fluffy89502 to engage with the 
various ongoing discussions about road classification, but I agree that 
it shouldn't be accompanied by a wholesale revert. It wouldn't be worth 
the effort at this point; this isn't a few changesets concentrated in a 
specific area. A revert probably wouldn't have the same effect on a 
reclassification specialist as it would on, say, someone who goes around 
drawing parks.

It isn't as though Fluffy89502 is randomly vandalizing the map and 
sometimes getting it right by coincidence. They're consistently focused 
on downgrading highway=trunk to primary based on physical 
characteristics, which is certainly one school of thought. But at one 
point, they stated personal criteria that were oddly specific. [1] 
Lately, however, I've noticed them focusing instead on downgrading short 
trunk segments that had been trunk only for their physical 
characteristics, and questionably so. [2] So I wonder if they've taken 
notice of our feedback or the reclassification proposal under 
discussion. [3] Many of their edits also demonstrate a clear preference 
for ending highway=motorway at the first at-grade crossing and no 
sooner. [4] This is another topic that periodically comes up both here 
and on Slack, for which there doesn't seem to be overwhelming consensus yet.

By the way, this slow-motion, distributed edit war between motorway and 
trunk and primary is larger than any single user. For example, 
Fluffy89502 has effectively reverted some changesets by maper_t, who has 
a similar history of trunk fiddling, unhelpful changeset comments, and 
lack of engagement -- but globally. We can play whack-a-mole with these 
users, but they're simply filling a vacuum in our documentation. Our 
energy is best spent developing clearer, widely accepted guidelines 
about what qualifies as highway=trunk *and* where that classification 
begins and ends. If we can get there, then I don't think even the lack 
of communication would necessarily pose a day-to-day problem. In fact, 
these mappers could become an asset to the project through their 
diligence, if not their camaraderie.

[1] https://www.openstreetmap.org/changeset/89839376
[2] https://www.openstreetmap.org/changeset/112226209 had been 
erroneously tagged trunk based on a lack of driveways (going up a 
mountainside or down a cliff) and a median barrier (installed as a 
safety measure on this windy stretch).
[3] https://wiki.openstreetmap.org/wiki/United_States/Highway_classification
[4] https://www.openstreetmap.org/changeset/112225863

-- 
minh at nguyen.cincinnati.oh.us