[PROJ] Motion: promote PROJ 6.3.1RC1 to final release

[Bas Couwenberg]

On 2020-02-12 13:07, Bas Couwenberg wrote:
> On 2020-02-12 12:47, Lesparre, Jochem wrote:
>> And why aren't at least the
>> 2008 grids used instead? It looks like cs2cs EPSG:4937 +to EPSG:7415
>> uses no grid at all, but a 7 parameter transformation, introducing
>> errors up to 25 cm horizontally and large errors vertically.
> 
> If I understand correctly the grids will only be used by cs2cs if they
> are locally available, so you need to have the grids installed in
> PROJ_LIB already.
> 
> The first option suggested by projinfo 6.3.1 used the 2018 grids, but
> this may not have been the solution used by cs2cs.

Just tested with PROJ_DEBUG=5 and it uses the 2018 grids if available, 
but the margin is still exceeded.

  Test:   01 Texel
  Exec:   cs2cs EPSG:4937 +to EPSG:7415 -f %.4f
  Input:  53.160753042 4.824761912 42.8614
  STDERR: pj_open_lib(proj.db): call fopen(/usr/share/proj/proj.db) - 
succeeded
  pj_open_lib(rdtrans2018.gsb): call 
fopen(/usr/share/proj/rdtrans2018.gsb) - succeeded
  pj_open_lib(nlgeo2018.gtx): call fopen(/usr/share/proj/nlgeo2018.gtx) - 
succeeded
  [...]
  Output: 117380.1232     575040.3400 0.9794
  Expect: 117380.1200     575040.3400 1.0000
  x coordinate exceeds 0.001 meters: 0.00320000000647269
  NAP exceeds 0.001 meters: 0.0206
  Test FAILED: From ETRS89 to RD/NAP - 01 Texel

These are the expected values as documented in the 2008 documentation:

  
https://salsa.debian.org/debian-gis-team/proj-rdnap/raw/upstream/2008/Use%20of%20RDTRANS2008%20and%20NAPTRANS2008.pdf

Kind Regards,

Bas