On Sun, Dec 6, 2009 at 1:06 PM, 80n <80n...@gmail.com> wrote: > On Sun, Dec 6, 2009 at 5:47 PM, Tobias Knerr <o...@tobias-knerr.de> wrote: > >> Using the object history is just an approximation based on the >> assumption that mappers will usually keep an object if they are >> improving existing data, and create new objects if they add completely >> new information. It's not really possible for an automated process to do >> anything else. >> >> Actually even this doesn't work. If a way is split into two (using JOSM) > then the database does not record any information about the split and the > history is kept with only one of the ways. > > This means that the history and original attribution for half of all the > split ways is just not there. I don't think there's going to be a way of > deleting all the data for those people who don't accept the new license. >
There's not going to be a way of doing it perfectly. Consider the reason people get so paranoid about someone tracing copyrighted maps. If you accept that the data is copyrighted, then a single contribution has the potential to "taint" a large portion of the database, depending on how strictly you want to interpret what constitutes a "derivative work". Part of me suspects that this whole notion of removing contributions from people who don't agree is going to get dropped. At least for the contributors who don't respond one way or the other. It's just going to destroy too much of the database.
_______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk