Martin, On 10/14/2015 11:18 AM, Martin Koppenhoefer wrote: > frankly, if there was a halfway usable repository of open > addresses that could be merged with OSM for those who want it, and if > open addresses become available for regions where OSM already has > addresses, I'd not be opposed to dropping the addresses from OSM in > those regions. > > Really? I've always thought our user's ground truth would be trumping > data we'd import, i.e. we'd request from importers not to drop features > that are already there, but to conflate in the opposite way, drop from > the external data set the stuff that we already have, before importing > the rest. Did I read you right? Can you explain why you changed your mind?
You read me right and Michal did too. Yes, I always said that we would want to be able to import Open Data at the processing stage (i.e. into Nominatim etc.) instead of importing it into OSM, so Michal is right. This of course has the drawback that you can't edit the government data sets, and this is why Tom Lee would prefer to import the government data sets into OSM to make them accessible for editing. To which I replied that I would prefer to have this data in a non-OSM editable repository, rather than in OSM, because I feel that there is a disparity between the amount of address data and the number of mappers actually interested; I would prefer if those who want a crowd-sourced address data set would not burden OSM with that. Tom wrote that there's not enough manpower in openaddresses to actually edit the data, and I cautioned him against assuming that the OSM manpower would automatically be available for editing addresses. Now if there *was* a crowd-sourced address collection project, then I would not object to OSM deferring to that for addresses. If, say region X made their address data openly available, it would be possible to conflate that with the (supposedly better) stuff we already have in OSM, add the result to the crowd-sourced address collection project, and drop it from OSM. If the alternatives are to either add the gov't data to OSM or move existing OSM address data into a separate project, I'd clearly prefer the latter, although I recognize that there might be people who would like to keep "their" address data in OSM. It is something that would have to be discussed. It might be possible to piggyback the crowd-sourced address collection project onto OSM but I would really think that if crowd-sourced address collection is not viable as a project in its own right (because of lack of people willing to give away their spare time to improve it), then it will not be viable in OSM either - only that the situation would be less obvious. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" _______________________________________________ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
