Collating and conflating is one thing, but we really need to encourage custom data apps like https://www.zap-map.com/ to use OSM as an active database which they feed back to. This will only happen when private companies realise that long term value is not in the data itself (because other people can collect it too) but in the extra deep knowledge gained in curating it, and from services on top of the raw data.
On Tue, 21 Jul 2020, 23:12 Nick, <[email protected]> wrote: > Could the data be included in https://osm.mathmos.net/survey/ ? > On 21/07/2020 22:42, Colin Smale wrote: > > On 2020-07-21 22:54, Mark Goodge wrote: > > It's the errors which are more of a problem, because it's generally better > not to map something than to map it wrongly. > > This is a difficult point. Data is never 100% complete, and frequently not > 100% accurate. At what point it becomes better not to have the thing in OSM > at all, is rather subjective. > > If the location was only accurate to ±50m, would it still be good enough? > If the operator was not tagged, would it still be good enough? > > Is an "imperfect" object in OSM more likely to get corrected than a > missing object is to get added? Should I not add a missing object because I > cannot be sure of the "operator" for example? Talking about the charging > points data set, how can one detect what is an error? > > I would say, get the data out there, and let the world feed back any > inaccuracies to the source for inclusion in the next version. > > > _______________________________________________ > Talk-GB mailing > [email protected]https://lists.openstreetmap.org/listinfo/talk-gb > > _______________________________________________ > Talk-GB mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/talk-gb >
_______________________________________________ Talk-GB mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk-gb
