Grab is messing up OpenStreetMap data in Southeast Asia

Re:Grab is messing up OpenStreetMap data in Southeast Asia by londons_explore on Thursday December 20, 2018 @10:22AM (#18725613)

OSM has full edit history like wikipedia, so 'undoing the damage' doesn't need to be hard.

They could have a simple flag on edits which say if they're done from satellite data, and if so what age the data is. Any real-world surveys shouldn't be allowed to be overwritten by older satellite data without good reason (eg. spam, or the real world survey was done badly)

[ Parent | on HN ]

Re:Grab is messing up OpenStreetMap data in Southeast Asia by thomersch_ on Thursday December 20, 2018 @10:42AM (#18725809)

Objects are dependent on each other and new, legitimate edits can happen on top of damaging ones which we would not like to revert automatically.

Yes, there are tools in place which can revert edits, but there is still manual checking required. The annoying thing is that companies sometimes go ahead and "invest" money into edits which turn out to be garbage and then volunteers need to revert or fix them up.

[ Parent | on HN ]

Re:Grab is messing up OpenStreetMap data in Southeast Asia by ris on Thursday December 20, 2018 @12:01PM (#18726633)

There are also a couple of subtle details of the OSM editing model that could make it a lot easier for a knowledgeable vandal to make changes that are rather more painful to undo than they should be.

[ Parent | on HN ]

Re:Grab is messing up OpenStreetMap data in Southeast Asia by Piskvorrr on Friday December 21, 2018 @09:28AM (#18733471)

One word. Merging. The database is massively multiuser, plus the process is far worse for cartographic data than for code.

[ Parent | on HN ]

Re:Grab is messing up OpenStreetMap data in Southeast Asia by rmc on Friday December 21, 2018 @03:17AM (#18732054)

You can add a "source" tag to changesets. But if someone doesn't want to use that, they can just turn it off.

[ Parent | on HN ]