[Proj] Convert a planetographic latitude to a planetocentric latitude

[Martin Desruisseaux]

Le 18/08/2018 à 22:29, Kristian Evers a écrit :

> Good catch. I wonder why this was changed to the current
> implementation. The differing ellipsoid parameters for CRS1 and CRS2
> can be regarded as datum information (albeit rudimentary datums). 
>
I see the same as Kristian; "ellipsoid shift" is incomplete compared to
"datum shift", but still seem less surprising to me than doing nothing.

Anyway, a library can chose to do nothing, or to approximate an
ellipsoid shift with "Abridged Molodensky" method on the basis that
there is no point trying to be too accurate in the absence of more
complete datum shift information, or chose to do a full transformation
going through geocentric coordinates. All those choices are okay. But a
suggest that coordinate transformation libraries should provide some
mechanism telling to the user which operations the library choose to
apply for a given pair of CRS. The WKT 2 (ISO 19162) expression of
CoordinateOperation can be one way.

    Martin