|This is the talk page for discussing improvements to Commons:Geocoding.|
|Archives: 1, 2, 3, 4|
Viewing Geocommons on GoogleMaps won't work after February 2015
I was viewing some geocoded images on google maps through the View this and other nearby images … feature of the location template, when a note popped up that some things might need to be changed in order to have this work after February. It had a link to this page, which links to this page. Looks like KML overlays won't be possible anymore after February – at least not in the way they are now. Just thought I'd give you a heads up – pinging some people who might be interested: @Para, Kolossos, Jarekt, Dschwen: --El Grafo (talk) 13:39, 24 November 2014 (UTC)
Any guidelines for precision?
Are there recommended practices for when to geocode and how to indicate uncertainty? The section on Precision merely says "If we are aiming to a precision of about 7 m..." (emphasis added). What should we be aiming for? For instance, if a photo was taken in somewhere a public park 500 m wide, would it be appropriate to geocode the center of the park? What about a city or county? (not that Google maps usually centers and drops a pin in the geographic center/centroid of the place). How much error is too much for geocoding? I think there should be explicit (even if loose) guidelines for when not to geocode (e.g. "where the camera location cannot be estimated to within approximately 50 m, the location should be stated in the description but not geocoded"). Animalparty (talk) 22:25, 17 December 2014 (UTC)
- There are many cases where a precision much bigger than 50 m cannot be obtained, but where geocoding is still useful, i.e. pictures in the middle of deep jungles or in high mountains (Himalayas, etc.). There, even a precison of 1 km would be useful. Reagrds, Yann (talk) 22:36, 17 December 2014 (UTC)
Yes. By coincidence I encountered this question yesterday for Category:John Ericsson monument (Manhattan). This object is in an urban park which is densely photographed and geotagged, so 50m or the center of the park would be entirely unacceptable. Web searching was able to confirm a location within 20m, and I'll probably visit the place next month and locate the best pictures to the usual 5m precision of Google Earth aerials of this island. For a picture in the middle of the nearby Category:Long Island Sound, 500m might be tolerable, depending on the object and circumstances. Jim.henderson (talk) 16:41, 19 December 2014 (UTC)