25
Mar

Losing Good People

In my work for the OSMF Data Working Group, I often have to mediate in conflicts between mappers. I’m not a trained psychologist, and sometimes it feels like it would be good to have one to deal with these kinds of situations. The process often leads to unsatisfactory results. Let me sketch a typical interaction.

Mapper 1 (to DWG):
Hey Data Working Group, this guy Mapper 2 is reverting almost all edits on his home turf and doesn’t respond to questions. We’ve discussed him in the forum and others have the same problem with him.
DWG (to Mapper 2):
Hey there Mapper 2, is it true that you’re reverting all edits in your area?
Mapper 2 (to DWG):
Yeah but they’re all idiots who don’t know the area and local practice. I told them they’re doing it wrong but they won’t listen.
DWG (to Mapper 2):
But what’s wrong about these houses that you have reverted here?
Mapper 2 (to DWG):
Ah, they’re breaking so much of my mapping, I can’t be bothered to look at the details.
DWG (to Mapper 2):
Please stop these wholesale reverts, and engage in a discussion with the community. Maybe they’re not the idiots you think they are. Here’s a forum thread that you could participate in.

(time passes)

Mapper 1 (to DWG):
Hey DWG, Mapper 2 is still reverting stuff without talking to us.
DWG (public blocking message to Mapper 2):
Dear Mapper 2, please stop wholesale reverts of edits in your area, and engage in discussion with the community if you have issues.
Mapper 2 (to DWG):
This is outrageous! You’re publicly criticising me! You’re taking sides! You don’t understand anything! I’ve had it with OSM, I’m leaving, good bye.

Of course we use a lot more words in reality, and try to talk them out of leaving in a huff, but we have indeed lost a couple of people that way. The sad thing is that these are often good and diligent mappers with excellent first-hand knowledge of their local area; they watch everything that goes on in OSM and they are really concerned about the quality of the data in their region. These are the kinds of people that can really boost our data quality – if only they could be made to play well with the rest of the community!

These people are protective of their work – too protective. And they believe that they are right. When they leave, they are frustrated; they fear that without their continued watch OSM is condemned to mediocrity.

There is probably a grain of truth in that. It is quite possible that in the immediate aftermath of the power mapper leaving, data quality in the region suffers. But new mappers – more mappers! – will join, and instead of a rigidly curated and over-protected landscape in which they daren’t touch anything, they will find a living project that invites them to participate. Maybe there is mediocrity in that, but it is sustainable – in stark contrast to data curated by one person alone according to their own rules.

Making everyone see that is one of the continuing challenges faced by the Data Working Group. We don’t want to lose good people; we want them to understand that there’s more to improving OpenStreetMap than mastering tags and relations. OpenStreetMap is a social project too, and we will have to continue to remind people of that.

There's 2 Comments So Far

  • Tordanik
    March 25th, 2013 at 09:59

    Some aspects of the problem doesn’t even have to do with psychology, but with our tools. Say, there is a new, inexperienced mapper making a lot of edits in your area. You want to maintain high data quality. What do you do?

    You could just ignore the edit. But that means that any mistakes will remain unnoticed. Because everyone makes mistakes, especially in the beginning – and especially if no one points them out to them -, and because new people (and possibly vandals) will always show up, this means that our data quality will be bad-to-mediocre forever.

    So when you have witnessed other mappers accidentally deleting data, breaking the routing, ignoring basic tagging standards and so on for the umpteenth time, you might be tempted to just revert any changeset that looks even slightly suspicious. But clearly that’s not the right answer either. We’d like to have more than one mapper per town after all.

    To me, it seems that the proper response would be to look over others’ edits (particularly the first few edits of a new mapper) to identify potential problems. If you find any, you can discuss them with the mapper, and if you need to correct them yourself, you can do so more surgically instead of reverting everything. However, that option is not available for any but the smallest edits because we have no usable presentation of the differences in a changeset.

    There are many other facets of the problem, of course, not all of which can be tackled with technology. But this one can.

  • aseerel4c26
    March 25th, 2013 at 22:59

    @Tordanik: I agree – knowing what happens inside changesets is awfully hard. Especially if there are big changesets with no (good) comment.
    Recently that pain was a bit (not totally) relieved:
    * https://wiki.openstreetmap.org/wiki/Quality_assurance#WhoDidIt
    * https://wiki.openstreetmap.org/wiki/OSM_History_Viewer – that should be available on osm.org – directly. I noticed it sometimes misses new created nodes, but am not sure on this one.

    More tools like those are needed.