Ideas for combating vandalism about openstreetmap-website HOT 11 CLOSED

Yes but community.openstreetmap.org is a better place to discuss with the community - here you're only going to find a small number of developers who are just going to get annoyed by a never-ending thread with no actionable conclusions.

I also largely disagree with the idea that something more needs to be done, or rather I think I have already identified a way to prevent the type of action used in the recent attacks from causing significant disruption and have taken steps to action that by opening osm2pgsql-dev/osm2pgsql#2185.

from openstreetmap-website.

The editing app can be easily faked, as it has happened before. You could pretend to be iD and still do large scale changes.

The dormant accounts idea might work from a technical pov, probably needs a bit more thought.

Sorry, we don’t have review queues or anything like that. It would be a huge change to the architecture and overall mapping process.

from openstreetmap-website.

Also see https://community.openstreetmap.org/t/the-osm-standard-tile-layer-looks-wrong-white-lines-abusive-comments-etc/111583/193 for why a "review queue" is a nice idea but a huge task with significant problems.

from openstreetmap-website.

No. Even in the browser editor iD this is a mouse click away and freely changeable:

Not quite - users can type here, but it won't actually accept their change. Those fields are locked to their original values.

(The original point is correct though that it's trivially easy to fork iD, or any other OSM editor, and have it call itself something else, making the idea of combating vandalism by looking at the created_by tag a non-starter.)

from openstreetmap-website.

Segregate suspicious edits for review, or commit on a delay. Specifically "large bounding box from newish account" and "golf-related from a new account" seem to be pretty obviousmarkers.

from openstreetmap-website.

None of this is something that is best discussed here - discuss it with DWG and come back when you have concrete requests.

Note that "chessiegp9" was a relatively minor even and didn't actually make a significant number of edits so it's unlikely rate limits would have much effect there.

from openstreetmap-website.

The editing app can be easily faked, as it has happened before. You could pretend to be iD and still do large scale changes.

The changeset tags for the app maybe, but wouldn't the 'Client ID' for the app be harder? That would require changing and building from source wouldn't it? It would still be possible to fake this, but it would add an additional barrier that they would need to cross for it.

from openstreetmap-website.

discuss it with DWG and come back when you have concrete requests.

Actually, we'd probably just say "discuss it with the community first so that people are happy about the way that contributing to OSM has changed". There was significant pushback against rate limiting from organisations that rely on putting lots of newbies in front of computers and doing some remote editing; we'd need to ensure that if we're going to change the way that people can contribute that people are (a) aware and (b) happy about that change.

from openstreetmap-website.

discuss it with DWG and come back when you have concrete requests.

[...] "discuss it with the community first so that people are happy about the way that contributing to OSM has changed". [...]

Yeah, I opened this issue so a discussion could take place here. Its obvious that something needs to be done, but nothing will get better if we just shut down the conversation and kick the problem down the road to different groups.

from openstreetmap-website.

The editing app can be easily faked, as it has happened before. You could pretend to be iD and still do large scale changes.

The changeset tags for the app maybe, but wouldn't the 'Client ID' for the app be harder? That would require changing and building from source wouldn't it?

No. Even in the browser editor iD this is a mouse click away and freely changeable:

from openstreetmap-website.

The editing app can be easily faked, as it has happened before. You could pretend to be iD and still do large scale changes.

The changeset tags for the app maybe, but wouldn't the 'Client ID' for the app be harder? That would require changing and building from source wouldn't it?

No. Even in the browser editor iD this is a mouse click away and freely changeable:

Thats not at all what I am talking about though, I'm talking about the Client ID that the editor uses to identify itself to the OSM API. While this is still something that can be changed, it would require rebuilding the application from source which is adding a higher barrier to entry for these vandals. Sure it won't stop a persistent vandal, but it would make them need to work a bit harder for it and I'd imagine some would probably just give up if its too much work to be worth it for them.

https://github.com/openstreetmap/iD/blob/develop/config/id.js#L14

from openstreetmap-website.