Just to give a bit of a summary of the review problem here, there have been 1164 changesets in the last 20-odd days roughly overlapping the UK:
changesets=> SELECT count(*) FROM osm_changeset c, (SELECT ST_SetSRID(ST_MakeEnvelope(-11.0,48.0,2.5,61.1),4326) AS geom) s WHERE ST_CoveredBy(c.geom, s.geom) and created_at > '2025-03-03 00:00:00' and tags -> 'comment' LIKE '%mpr.lt/c/%';
count
-------
1164
(1 row)
of those, some are straightforward (“added buildings”). Some are questionable (e.g. would normally require local knowledge). The changeset comments of many are entirely opaque (this appears to be a perfectly valid edit but the challenge has prompted the user to use a meaningless description).
On that screen there is a small white icon (“a flag”) that you can click on
On the next screen you have to enter an email address and some text. It says it will create a github issue.
After doing so the white symbol to the right of the text at the https://maproulette.org/browse/challenges/50062 page changes from white to red and the text “This challenge has been reported” appears.
Clicking on that goes to the github issue. It was opened by the repository owner and contains a link to the logged in OSM user.
Hey guys - I’m just wondering how specific you want people to be when fixing invalid ways. I mean, if a geometry is invalid due to being self intersecting and I’m just fixing it so it doesn’t self intersect what would you like as a comment?
I thought that adding a comment that just says that I fixed an invalid geometry would be enough. I could add a comment that says fixed a self intersecting geometry… would that be better? I’m referring to MapRoulette.
I’m just looking for instruction on what kind of comment you are looking for.
Thanks!
I think it would help to say that you were fixing (multi)polygons which previously had an invalid geometry. For example “Fixed an invalid way” doesn’t say what was wrong previously (geometry? tags? entirely correct mapping that no longer exists in the real world?).
For examples of the comments hat I tend to use for this sort of thing see https://www.openstreetmap.org/user/SomeoneElse2/history?after=164025427 - although some of those could do with a bit more description too; for example on that list: Filled in gap in Anglesey National Landscape. See https://www.openstreetmap.org/changeset/164127603 implies that everyone knows that Anglesey National Landscape is supposed to be a polygon, and Removed spur that had been accidentally added to NCN425_R 4247567 implies that everyone knows that NCN425 is a cycle route.
In the case where you’re working down a maproulette list, you know what the maproulette instructions said to do, but people looking at the changeset later won’t.
Before I make the change to the challenge I want to run this by you. The challenge always tells the mapper what the issue is
issues description examples:
“w123456 geometry has duplicate nodes”
“w54321 geometry is an area that is open”
“w76543 geometry is invalid”
I’m going to change the challenge to set the checkin comment to
"Improved geometry to fix: {issues description}
So they will look like this:
“Improved geometry to fix: w123456 geometry has duplicate nodes”
“Improved geometry to fix: w54321 geometry is an area that is open”
“Improved geometry to fix: w76543 geometry is invalid”
And then ask mappers to add any specific flavor that they think is necessary.
I am helping the map by reporting bad challenges. I should not be stopped doing so by a checkbox asking me to 1) lie, or 2) do more work. I don’t want to contact anyone, I don’t want to lie, if a challenge is clearly inappropriate, users should have the ability to stop it, at once.
Instructions on the challenge I’m talking about:
Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads Satu Mare Unnamed Roads
In this challenge they are not always multipolygons: for instance the duplicate node issue can be found on a linestring as well. But I will try to identify that in the future on challenges that are specific to multipolygons.
@SomeoneElse2 Please see the improvements that I have made to the challenge. Here is an example of a change I just made.
I added the “Removed a spike.” sentence. Everything else is the default commit text for the revised challenge.
If this is satisfactory please help to unflag the challenge.
This checkbox is probably related to the traditional expectation that someone spotting a problem should attempt to contact the person responsible before escalating to the DWG or calling them out in Slack or Discord’s #questionable-edits channel. I don’t know whether or not it’s appropriate in this case; it depends if the challenge’s author has an adequate opportunity to explain themselves.
In my case I am typically contacting author with text that I will use to flag challenge (if it is clearly broken) and proceed to immediately flag the challenge.
Another quality contribution from Maproulette users: Changeset: 164466661 | OpenStreetMap changes a recently-built signalized crossing to an unmarked crossing
I started suffering a batch of these kind of changesets around here today.
some are good, but a some others aren’t and are doing changes without reading the wiki or local project wiki.
and not from novel users this time, as most of them are experienced collaborators with several years and thousands of changesets in OSM…
@muralito This challenge is done, but if you want to work together to ensure that challenges targeting your community can be better designed I’m here to help.
Also feel free to flag challenges if you come across ones that generate questionable edits, I check those personally every other week.
Thanks. According to the changeset comments, the challenge had influence in that several users were doing the same kind of task, but the source of the problem seems to be iD is misguiding some users with some suggestion to change tags. That really surprises me because I would expect that experienced users were more resilient to this kind of errors, but any user, including experienced ones, would be confident that if iD, being the “official” editor, is doing a suggestion there is a problem with the data and it’s ok to “fix” it…
I started suffering a batch of these kind of changesets around here today.