[geos-devel] M-coordinate support

[Sandro Santilli]

Luca: I confirm you need to modify the library in order to get M support.
I guess it would increase memory requirements of operations by about 30%
(from XYZ to XYZM) -- I think Martin Davis would be the one in the best
position to answer this, GEOS being a port of his JTS.

--strk;

On Tue, Jul 10, 2012 at 02:37:42PM +0200, luca paganotti wrote:
> Hi Mike,
> 
> I agree with you about the M-coordinate support in geos, there are some
> cases in which this support is a must.
> 
> For example I'm using geos to represent 3D points changing their position
> in time so to have M seems really necessary, but I can be wrong, this
> support is already built in.
> 
> To be able to use geos with ZM points I've lightly modified and rebuilt
> geos tailoring it only to my needs. I've not done any kind of analisys or
> deep thinking about it so it's only a very raw support for M, but that was
> I needed.
> 
> P.S. I made also a little post to this list some time ago saying that, but
> I received no answer, maybe it was not so interesnting or I was missing
> something that could give me M-support without rebuilding the modified
> library.
> 
> Best regards
> 
> lp
> 
> 
> 
> On Tue, Jul 10, 2012 at 1:41 PM, Mike Toews <mwtoews at gmail.com> wrote:
> 
> > Hi devs,
> >
> > I'd like to know if "measure" or M-coordinate support is somewhere on
> > the development horizon for GEOS. I could see the benefit of opening
> > client software to support ISO 19125 compliant geometries, such as in
> > OGR or Shapely.
> >
> > Possibly the first widespread use of M-coordinates started in the
> > 1990s with the Shapefile file format, as detailed in the 1998 ESRI
> > technical specs. With open source software, Shapelib and PostGIS has
> > supported M-coordinates for quite some time (possibly close to a
> > decade?). More recently, Simple Features Access Standards published by
> > OGC and later ISO 19125, detail the use and storage of of
> > M-coordinates in various geometry types.
> >
> > Are there any limitations to adopting M-coordinate support for GEOS?
> > Potential performance loss? Other limitations or incompatibilities?
> > I'm curious, as I often run into instances where this dimension is
> > useful, but am a bit stumped at it's lack of support in some software.
> > Thanks.
> >
> > -Mike