On Wed, Nov 20, 2013 at 1:35 AM, Even Rouault <[email protected]> wrote: > > Le mercredi 20 novembre 2013 03:29:25, Trent Piepho a écrit : > > That file had both StdParallel1 and ScaleAtNatOrigin defined. > > > > I've uploaded two new files. This one has only StdParallel1 defined with > > no ScaleAtNatOrigin. > > https://drive.google.com/file/d/0BybuTedE9CLxRnNlNUl2Zm5JRzg/edit?usp=shari > > ng > > > > This file has the lat_ts stored in NatOriginLat and no ScaleAtNatOrigin > > defined. > > https://drive.google.com/file/d/0BybuTedE9CLxMzltYlBFWU51RzA/edit?usp=shari > > ng > > > > I think the latter might have better compatibility. Yet it seems to > > clearly be the incorrect way to do it. > > If the way of encoding/decoding is changed, it would be best to do it the > "right" way and eventually other vendors will catch up with this > interpretation.
Right now the way GDAL handles Mercator_2SP doesn't work with anything, including GDAL. So clearly it should be changed to something else. Of the two options, one (StdParallel1) would appear to be the correct way and the other (NatOriginLat) might have better compatibility with existing software. It would be nice to see the results of testing the file using NatOriginLat, to see if it does indeed work correctly with other software. GDAL could be patched to read both options as Mercator_2SP. It's not correct, but given that there appears to be a lot of software that places the latitude of true scale in NatOriginLat and that using an origin other than the equator for Mercator is not something commonly done, it's going to be the right guess almost all the time. _______________________________________________ gdal-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/gdal-dev
