[gdal-dev] Call for discussion on RFC 64: Triangle, Polyhedral surface and TIN

[Even Rouault]

On mercredi 7 décembre 2016 02:01:54 CET jratike80 wrote:
> Even Rouault-2 wrote
> 
> > Hi,
> > 
> > This is a call to discuss the proposed RFC 64: Triangle, Polyhedral
> > surface and TIN
> > 
> >     https://trac.osgeo.org/gdal/wiki/rfc64_triangle_polyhedralsurface_tin
> > 
> > ~~~~~
> 
> I am wondering this part:
> 
> "GEOS methods are still used in some cases, but with the following
> limitations - a Triangle is converted to a Polygon with one exterior ring;
> Polyhedral Surfaces and Triangulated Surfaces are converted to
> MultiPolygons?. (each Triangle in a Triangulated Surface is converted to a
> Polygon as described previously)"
> 
> Does it mean presenting a TIN with two triangles as something like
> 

Something missing here, but I guess you meant as a multipolygon made of 2 polygons ?

> 
> 
> As far as I know such multipolygon is not valid accorfing to the Simple
> feature specification where "boundaries and may touch at only a finite
> number of Points". Valid presentation would probably be to wrap polygons
> into GeometryCollection. Is GEOS used in a relaxed way so that shared parts
> of boundaries in MultiPolygons do not throw an error?

That's a very good remark. Indeed for GEOS purposes, a MultiPolygon with Polygons that 
have shared boundaries is invalid. I've changed that to a GeometryCollection. But anyway, TIN 
& PoyhedralSurface in GEOS world are degraded cases where behaviour will always be 
suboptimal.

Thanks for the feedback.

Even

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20161207/4b78a063/attachment.html>