- William Lachance arranged a host of electrons thusly: - > > Having the uuids around also make it easier to talk about > > differences/errors between OSM and geobase data. Someone can look > > at a road in OSM and easily find the original GeoBase road (using > > your favourite gis tool) and compare them. Without the uuid somehow > > connected to the OSM road you'd be guessing at which road in the > > geobase data it comes from. > > I'm not sure if I understand, can't you just compare roads based on > geometry? A UUID is basically just a pointer to a piece of data with > particular properties. Why can't you look up the same thing based on > the properties themselves? In the case of the road, can't you just > compare based on name, nodes, etc.? Is it just a tools problem or > something more?
Well, as a developer with a bunch of experience manipulating XML data, and no experience writing routines that deal with spatial data, it would sure be easier for me to tackle a problem where there are common identifiers to identify the same data so that they can easily be identified. It lowers the barrier of entry, I suppose, to those with more common knowledge (I expect that people with knowledge of spatial algorithms & data structures is small compared to the developer pool in general). Of course, roadmatcher (or similar) could be used to do the identification, the the output could be processed to provide a mapping file between OSM IDs and NIDs, which could be used to do something similar to having the NIDs in the database. Perhaps the OSM ways could be kept together, and a machine readable tag could be added that describes the breaks that the NRN data would induce, and what the NIDs would be. I can't say I'm much of a fan of this idea, but it could work. *shrug* Tossing out ideas here -- no solutions for the issue. Austin. -- Build a man fire, he'll be warm for a day. Set a man on fire, he'll be warm for the rest of his life. _______________________________________________ Talk-ca mailing list Talk-ca@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-ca