> -----Original Message-----
> From: Christopher Schmidt [mailto:[EMAIL PROTECTED] 
> Sent: Monday, July 16, 2007 4:53 PM
> To: Jeff Dege
> Subject: Re: [OpenLayers-Users] Problem with GML coordinates?
> 
> On Mon, Jul 16, 2007 at 04:47:27PM -0500, Jeff Dege wrote:
> > I'm no expert on the GML specification, but my expectation 
> would be that
> > if the ogr tools can create GML files with three-dimensional
> > coordinates, then the GML specification probably allows for it.
> > 
> > In which case, OpenLayers.Layer.GML should probably cope 
> with it, even
> > if it's only throwing the third dimension away.
> 
> The GML format class is meant to handle 3d data. IT looks for an
> srsDimension attribute on the gml:pos or gml:posList elements. 

I think that's the problem.  It's assuming that if it uses
gml:coordinates, it's GML2, and if it's using gml:pos or gml:posList,
it's GML3, and that if it's GML2, it only has two dimensions.

Now I'm not sure whether ogr is incorrect in using three dimensions in a
gml:coordinates element, or whether the GML format class is incorrect in
assuming that the lack of gml:pos or gml:posList indicates two
dimensions.  But if it's a bug in OGR, it's going to be a widespread
one.

> Can you
> put up a full GML file generated by ogr2ogr in this way so that it can
> be examined by developers who might be able to improve the parser to
> handle this case? 

I'm using the ogr2ogr bundled in MS4W version 2.0 - which seems to be
gdal/ogr version 1.3.2.

My commandline:

: ogr2ogr -f "GML" -s_srs EPSG:4326 -t_srs EPSG:102004 testoutput.gml
testinput.gml

I've attached the input and output files, and copied the email to you
since I'm not sure the list will accept the attachments.

Attachment: testoutput.gml
Description: testoutput.gml

Attachment: testinput.gml
Description: testinput.gml

_______________________________________________
Users mailing list
[email protected]
http://openlayers.org/mailman/listinfo/users

Reply via email to