Link to [MetaCRS] Common SQLite-based dictionaries for SRS
A heads up with regard to a discussion taking place on some lists, started on the low-volume MetaCRS list, and reposted elsewhere by Even Roualt: http://lists.maptools.org/pipermail/proj/2015-August/007196.html It strikes me that the response by Huw James is very well founded, and that the data model for a future common dictionary for coordinate reference systems should permit the optional use of historical CRS. This is an enduring problem with EPSG codes, which can change with each release, so making reproducible research harder (especially as the EPSG CSV files in PROJ.4 and GDAL, and versions in the GeoTiff software, do not report their versions). I would invite ideas now, to assist in guiding MetaCRS towards a robust and time-aware dictionary structure with self declaring versioning. There is a further question of the proposed use of SQLite as the backend, as R does not have built-in DB drivers, but has an RSQLite package on which future CRS handling might have to depend Roger
Roger Bivand Department of Economics, Norwegian School of Economics, Helleveien 30, N-5045 Bergen, Norway. voice: +47 55 95 93 55; fax +47 55 95 91 00 e-mail: Roger.Bivand at nhh.no