This is a question I’m considering a lot … filtered through the brief rushes of reading the amazing crisismappers list, diving into OSM on the wiki and IRC channel … Are we doing everything we possibly can in serving the responders? Can we coordinate our mapping work better? And once aid starts flowing and the immediate response turns to long term recovery and reconstruction, how will our process and community change, when more and more data to synchronize will be coming from the ground in Haiti? How do we operate better in the next disaster? Big questions, and glad that many folks are already compiling ideas. Please add ideas and needs there.
One immediate point that has been raised in private discussion, and now publicly, are the separate efforts of OSM and Google MapMaker in Haiti, separated by incompatible licenses. As Sean Wohltman points out, a crisis like this is hardly the time for discussions on licenses and community models, and whatever can be done for the best benefit to the crisis response should be done. All I can simply say with restraint is that I disagree with the assumption of Google’s position that the OSM license prevents community use of this data … rather it only prevents Google participating in that community, by their own choice. That shouldn’t stop OSM from considering doing everything possible.
There is likely much to be gained by everyone working in the commons. Muki Hakley has performed an analysis of OSM and MapMaker coverage in Haiti, and it shows complementary coverage. The analysis makes no assumption of correctness, time frame, size of community/number of sources, and only considers geometries not names and tags, but still shows that each have built up in different regions. The problem of how to potentially use these two together needs quite a lot of work.
Sean Gorman suggests a time and geography limited CC-0 license on geodata, in order to move things forward. I’m not taking a personal view on this possibility, actually undecided, though I would at least suggest attribution would be a courtesy, and impose no considered constraint. Instead, I would suggest the OSM community consider this matter, and if there’s a general consensus it would be valuable, perhaps by determined vote, then build a quick process of getting time and geography limited dual-licensing approval from the couple hundred active Haiti contributors so far. There would also need to be a way to get such approval from new Haiti editors.
Important I think to consider is the time and effort involved in discussing this, and the process of dual licensing, vs the potential benefit. This question shouldn’t distract us from the very real effort of mapping and producing amazing data products.
What do you think?