On 8 August 2012 17:03, Matthew Landauer <matt...@openaustralia.org> wrote:
> As far as I understand the LGA boundaries totally fit within what OSM > should provide, doesn't it? > > Well, there is a school of thought that says that data that is fully external to OSM, and is updated on a fairly regular basis by a third parties there is limited value in performing an import into OSM. The data in OSM quickly gets out of date, and updating it once it has been modified and integrated with other OSM data can be hard - near intractable. Ultimately, it is easy to import data into OSM. Much harder to maintain it The problem being OSM doesn't have a good method right now of supporting layers of separated data. Projects like Common Map are looking at this issue (lots of separate data sets with a common schema). It is yet to be seen if the solution lies there. That said, I'm personally happy to move forward with importing the boundary data. Mainly because it is important for the map, and we have no other way of supporting it. But I think we need to look carefully for any lessons we learned from the ABS_2006 suburb import before we do. We never did figure out whether we were importing actual boundaries that could be corrected or whether the import was what it was and shouldn't change. We never agreed on whether they should be moved to align with coastline and other features when it looked like they were incorrect. We never had a process to update the information when new external data became available. We had many ugly maps with coastline and boundaries overlaid and crossing. We did have the benefit of using the boundary data where it aligned with some natural features, although I get the feeling with better Bing and AGRI data the value of that is diminishing. Ian.
_______________________________________________ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au