[Proj] Regression 5.1.0 -> 5.2.0 ?

[Even Rouault]

Søren,

I think the issue is that the _TC32_TD32 definition in the DK file has a 
Helmert transform that lacks the convention used for rotation terms, which is 
required since PROJ 5.2. So likely adding convention=coordinate_frame would 
fix the issue (since that was what was implemented before PROJ 5.2), but I'll 
Kristian or Thomas check that this is appropriate.

Even

> Hi
> 
> In 5.1.0 the following command worked as expected :
> 
> $ cct +init=DK:S34S
> 
> But in 5.2.0 this happennes :
> 
> $ cct +init=DK:S34S
> cct: Bad transformation arguments - (missing required arguments)
>     'cct -h' for help
> 
> 
> Do I need to use some other initstring  in 5.2.0 ?


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