[OSGeo-Discuss] Detect and define the coordinate system of gis data with no projection information automatically

[Luí­s de Sousa]

Dear all, thank you for the many contributions to this discussion. Some background:

If you have not done so, please watch the address by Calvin Metcalf on Proj [0]. From his reflections and the discussion that ensued it became apparent to me that presently there is no CRS/projection registry in the public domain under maintenance. Take the EPSG registry for instance: not exactly open and does not provide essential information, particular on projections. I myself have struggled recently to find information in the public domain on recent projections such as Eckart's or Snyder's. Digging into scientific publications is pretty much the only way to go (something not every one can do). The original message by Manfred was just a catalyst to this idea of an OSGeo sponsored/sanctioned registry.

My idea is to use GitHub and its pull request mechanism as way to force human verification of contributions. Contrary to other registries, this would be a completely transparent process. The end result would be a structure of MD files hooked up with ReadTheDocs. It just happens the Proj4 documentation that Even linked to [1] is already a step in this direction.

I agree that any new initiative should not overlap with ongoing projects - in particular since Proj4 is an OSGeo project. Therefore the route action right now would be to contribute to the Proj4 documentation, as Even suggests.

There is though a major issue to solve, the Proj4 handles. They are ambiguous in some cases (as Calvin reports) and do not cover all projections out there. So, there is some hard rock to break.

Regards,

Luís


[0] http://ftp5.gwdg.de/pub/misc/openstreetmap/FOSS4G-2016/foss4g-2016-1163-nobody_cares_about_your_datum_or_the_kleinstaaterei_of_spatial_reference_systems-hd.webm

[1] http://proj4.org/projections/index.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/discuss/attachments/20160902/b2b3e4f4/attachment-0002.html>