Comment by danpat | Hacker News Reader

Comment by danpat | Hacker News Reader

Parent Full thread

I work for Mapbox. We have some teams dedicated to improving OSM. A lot of our product stack is built on the OSM data set, so we have a vested interest in improving it wherever we can. We do a whole bunch of analysis to try to identify areas with errors and shortcomings.

We built and use this tool:

    http://osmlab.github.io/to-fix/

to queue up stuff that needs fixing, and we have people churning through it all the time.

There are some really cool things in mapping tech. For example, Strava has made available a whole dump of point data that can be used to look for un-labelled trails and roads (see the "Strava" layer in to-fix).

Sorry if my comment ignored the contributions many companies make to OSM - but I think the article expresses concern that OSM's potential as something that can be contributed to within a local community is threatened by people from first world cities mapping swathes of area from satellite images. I disagree with that idea of course - I think the nature of contributing to OSM already favours in the field volunteers.

In my experience, local ground-level contributors are often reverted by well-meaning remote mappers who rely on outdated satellite imagery. In areas without a strong community of local mappers, most of the contributions will be coming in from users who've never been anywhere near your actual location.

That comment made the problem much more clear to me -- more clear than the entire article

We do too. The to-fix tool that we use tends to focus on things we can identify algorithmically, like broken links, isolated lines, mis-spellings, etc. We encourage all our staff to get out and map locally, and some have been highly involved in local GIS communities for years.

Reply on news.ycombinator.com