On 11/25/2017 05:49 PM, Edzer Pebesma wrote:
> On the system I work on (Ubuntu 17.04 with ubuntugis-unstable GDAL),
> haveGEOS() seems unreliable (giving false, where it is true).
I'm sorry, that was a wrong diagnosis of a self-caused problem.
--
Edzer Pebesma
Institute for Geoinformatics
Heise
On 11/25/2017 02:47 PM, Even Rouault wrote:
>
>
>> Thanks, that would match nicely, as rgdal already has bindings for
>
>> GDALVersionInfo() but only uses "--version" and "VERSION_NUM" so far; our
>
>> test for GDAL built with GEOS tests a fake union to detect success or
>
>> failure.
>
>
> Thanks, that would match nicely, as rgdal already has bindings for
> GDALVersionInfo() but only uses "--version" and "VERSION_NUM" so far; our
> test for GDAL built with GEOS tests a fake union to detect success or
> failure.
Was available through the C++ API with OGRGeometryFactory::haveGEOS()
On Fri, 24 Nov 2017, Even Rouault wrote:
On vendredi 24 novembre 2017 22:19:08 CET Roger Bivand wrote:
On Fri, 24 Nov 2017, Even Rouault wrote:
I would have (perhaps) hoped that
pj_release.c is the runtime version, but mileage may
vary ...
It is. I just missed pj_get_release()
Thanks, I g
