On Tue, 2020-07-21 at 17:30 +0100, Mark Goodge wrote: > But it's not reliable enough for an unfiltered bulk import; there are > duplicate entries, incorrect coordinates and incorrect or missing > addresses.
We could turn this into an advantage of Open Data and OSM. Validation of other datasets makes them more accurate and in turn useful to everyone, including the original publisher. > What it would be useful for, though, is adding the technical data to > a chargepoint that has been mapped by on-the-ground observation. So > long as we can match an entry in the NCR to a mapped chargepoint, > the metadata (eg, connector types, owner/operator/network) can be > updated automatically. +1 But I think we should also add the other data (after other validation, like de duplication and null island), possibly like the naptan import with, naptan:verified, or a more generic import_verified. That could easily be fitted into on the ground verification tools like StreetComplete. Or perhaps a more visible option is to make notes for the ones that aren't matched to existing data allowing local mappers to check and add them manually. On Tue, 2020-07-21 at 18:44 +0200, Kai Michael Poppe - OSM wrote: > Again, this sounds like a perfect job for MapRoulette where the NCR > goes into a GeoJSON and people get to work on that data with what's > in OSM already - the part that requires effort is to correct what's > bogus ;) I think of MapRoulette as being more "arm chair mapping", I think you'd get people just adding it without any validation. Cj _______________________________________________ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
