[gdal-dev] Transform produces different results from 2.2.4 and 3.0.2 or 3.0.3

[Even Rouault]

On mercredi 29 janvier 2020 05:11:10 CET mike Null wrote:
> Even, Carsten, thanks.
> 
> Indeed I looked through what Even sent, and it worked. I did very much the
> same thing as the C# example you just posted.
> 
> However it got me thinking. Will I need to do this in the future? Or are
> there plans for coordinate transformation to use the new axis ordering
> system?

It is up to you to decide if within your application you want to stick with 
the not-so-well-defined 'traditional GIS order', or want to embrace natively 
the axis order of the CRS you manipulate.

GDAL internally uses a mixed position regarding this. All the raster side uses 
the OAMS_TRADITION_GIS_ORDER strategy. The vector side mostly defaults to it 
too, except a few places like the GML driver if you specify the
INVERT_AXIS_ORDER_IF_LAT_LONG=NO open option

Even

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com