Open mikelmaron opened 9 years ago
at the moment results are "hardprojected" to the http://spatialreference.org/ref/epsg/3785/
The more common name for that string is EPSG:3857.
It's also worth noting that it only attempts the string if the SRS https://github.com/hotosm/hot-exports/blob/master/custom_data_export/cde.cpp#L72 fails.
As a projection for transferring data in, 3857 is reasonable provided you aren't working outside +/- 85 degrees latitude, particularly for OSM data when most work in creating that data will have been done in that projection.
The C++ utility does calculate way_area, which is particularly meaningless in WGS84, being a non-conformal projection.
If we change the hardcoded projection to 4326 not only does that somewhat address this issue, but it will likely fix the bugs assosciated with issue #75 as well.
at the moment results are "hardprojected" to the http://spatialreference.org/ref/epsg/3785/ projection at ... https://github.com/hotosm/hot-exports/blob/master/custom_data_export/cde.cpp#L75 I don't know what were the reasons to do the reprojection the in the first place, or why is epsg:4326 a better choice at the moment...
This change requires modifications to the CDE C++ utility, backend export processors and RubyOnRails frontend applcation