> A way to do this would be to map any fence lines marking the boundaries.
That's a possibility. Though I have no way to be sure where the fences extend to in the front half of the properties. Probably not worth it now that I think of it. > Have you checked it for accuracy? The DCDBLite data for Qld had ok > accuracy in some areas, but in others it has lots and lots of mistakes > and out of date information and incorrect positioning. So far the data seems pretty accurate. It's the IDENTICAL data that is on the city's website, which lines up perfectly to the high resolution (<1m) aerial imagery they have of the entire city. When I brought it over into OSM as large "landuse blobs", it straddles the existing OSM streets quite nicely. > We ended up using the DCDB data here to map other features like > railway lines and waterways, rather than the property boundaries > themselves. The railway lands and park boundaries are another great use of this data. Especially given how out of date the satellite imagery of my city is. :) > Depending on the quality of the data it may be a very bad thing to do, > this is typically referred to as implopping and with everything else > everyone wants to map the bad data is never touched or updated and > left to rot. > > If this is mainly for your own use, you may be better converting it to > KML or your own pgsql DB and making a transparent layer, rather than > putting it directly into the OSM DB. Yeah, the more I think of it, the less I want to use this to mark property boundaries. I think I'll stick with just deriving the landuse areas out of it and be done with that. It's still a great source of information in that respect. Thanks, Tyler _______________________________________________ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
