[PROJ] Rads/Degs

[Paul Ramsey]

On Wed, Feb 13, 2019 at 10:22 AM Even Rouault
<even.rouault at spatialys.com> wrote:
>
> On mercredi 13 février 2019 10:04:59 CET Paul Ramsey wrote:
> > > On Feb 13, 2019, at 12:06 AM, Even Rouault <even.rouault at spatialys.com>
> > > wrote:>
> > > On mardi 12 février 2019 11:16:56 CET Paul Ramsey wrote:
> > >> Is there some clean way to tell proj_create_crs_to_crs() that I would
> > >> like my input angular units and output angular units to be degrees? As
> > >> it stands I'm just using proj_angular_input and proj_angular_output to
> > >> test and apply the change pre/post transform.
> > >
> > > Hum, with which PROJ version ?
> > >
> > > I guess with PROJ 5.x, the angular unit was radians
> > >
> > > With PROJ master, proj_create_crs_to_crs() object will expect (resp
> > > return)
> > > the angular unit of the source (resp target) CRS, which will generally be
> > > degrees (except a few ancient French SRS that take grads)
> > >
> > > The function at
> > > https://github.com/OSGeo/proj.4/blob/master/src/apps/cs2cs.cpp#L212
> > > can be used to get the angular unit of a CRS.
> >
> > Hokay, that makes the renovation a little more complex, good to know! :)
>
> I'd advise to "skip" PROJ 5.x, that is handle it just like the PROJ 4.x series
> using the good old proj_api.h, but target more PROJ master / 6.0 (which is
> within a few hours before seeing a beta/RC I think)

Yep, that's where I've basically ended up in this journey. The proj5
API isn't rich enough to convert to entirely, and there's no benefit
to it. Having magical "best transform" and associated vertical datum
stuff in Proj6 makes the pain worthwhile.

P