Hi Frederik, I never thought of it being used in the context of the overpass API. But, yes I think it work. You could supply 1 to N OSM entities to the overpass API (perhaps just an OSM file), and it would return a "best" match and a score to data in the OSM database. It could be an interesting way to deploy this kind of functionality to everybody that needs it.
Thanks Jason. On Sun, Jan 13, 2013 at 12:20 PM, Frederik Ramm <frede...@remote.org> wrote: > Hi, > > > On 12.01.2013 19:05, Jason Remillard wrote: >> >> I'm not sure if this has been discussed or not, if yes, a pointer to >> the old thread would be much appreciated! > > > It hasn't been discussed in this form but it has been touched on in several > discussions revolving around "permanent IDs". See > http://wiki.openstreetmap.org/wiki/Overpass_API/Permanent_ID for a (very > basic) implementation. > > In contrast to the approach you sketched, such "permanent ID" schemes would > usually find the best match for *one* input object (which could conceivably > be described by an input geometry too). If I understand you correctly, you > would be trying to match a set of, say, 1000 houses from an external dataset > to a selected region from OSM with the aim of having as little average error > as possible - which would be slighly different from doing one-by-one > searching. > > Bye > Frederik > > -- > Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" > > _______________________________________________ > Talk-us mailing list > Talk-us@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-us _______________________________________________ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
