RE: [emergency] Circle and Polygon

["Bullard, Claude L (Len)" <len.bullard@intergraph.com>]

As long as the representation of ontological class models 
are not mixed into the presentational object models of the 
viewer systems, this is fine.  Otherwise, we build 
toolkits for graphics and an application per ML.  The 
economics of that are not good.  The lesson of the 
web infrastructure is to separate the presentation system 
from the content to improve the rendering and behavioral semantics 
of the presentation systems.  If GML and its profiles enable 
that, we should be able to pull and push the 
semantic information into and out of the database and 
to the thin or rich client presentation system JIT.

It is the implementation of the rendering systems that 
concerns me, so yes, we can take the offline.  I am 
concerned that the papers you referenced are inaccurate 
with regards to standard 3D systems widely and cheaply 
available.  It is bad design to mix 'thematic, attributes 
and interrelationships' into the rendering tool. LOD is 
supported in every version of VRML and now X3D since 
the min-nineties.

len


From: Carl Reed OGC [mailto:creed@opengeospatial.org]

Folks -

This discussion on GML is way beyond the needs and requirements for 
representing CIRCLE and POLYGON geometries in EDXL. I would suggest that Len

and I continue this dialogue "off-line".

Cheers

Carl

----- Original Message ----- 
From: "Bullard, Claude L (Len)" <len.bullard@intergraph.com>
To: "'Carl Reed OGC'" <creed@opengeospatial.org>; "Ham, Gary A" 
<hamg@BATTELLE.ORG>; "Renato Iannella" <renato@nicta.com.au>; 
"Emergency_Mgt_TC TC" <emergency@lists.oasis-open.org>
Sent: Wednesday, June 08, 2005 2:35 PM
Subject: RE: [emergency] Circle and Polygon


> Having taken a look at the referenced documents:
>
> 1.  By claiming geography, GML claims geometry.   I am
> curious why you say this is not a mapping application.
> It appears to be the geometric component which
> can then be mixed via namespaces with other components,
> but the primary application is mapping for information
> management purposes.  It is clearly GIS.  It is not
> the whole story on geometry.
>
> 2.  GML application languages incorporate GML geometry.
> Are they free to incorporate geometry from other vector
> formats?
>
> The problem here is that not all elements that are spatio-temporal
> and application specific are best described by this model.  I
> recognize the XML means and that is sound.   However, the issues
> of object-model behavioral fidelity and rendering fidelity are
> not addressed AFAICT.   Where GML is an excellent basis for
> GIS, GIS is not the whole story of spatial management because
> the most important aspects of distributed simulation are not
> described.
>
> So there is a case to be made for X3D systems that reuse GML
> information.  I suspect however, there will be clashes in the
> object model fidelity as rendered into the simulation systems.
> It seems odd that the OGC works with SVG yet does not work
> with X3D given that the other important standards (XSBC,
> XMSF) work with X3D.  What is the story for the use of GML
> in real-time distributed simulation and visualization of
> real world objects?
>
> len
>
>
> From: Carl Reed OGC [mailto:creed@opengeospatial.org]
>
> Claude -
>
> GML is not a mapping standard! GML is a standard, grounded in a variety of
> ISO 19*** series standards, for
> encoding/expressing/communicating/transporting
> geographic/location/geospatially enabled content. Now, in terms of 3D, I
> would strongly encourage you to check out LandGML
> (http://www.transxml.org/GML+Experiment/Resources/261.aspx and
> (http://www.opengeospatial.org/initiatives/?iid=133) and CityGML
> (http://www.ikg.uni-bonn.de/sig3d/docs/Gi4Dm_2005_Kolbe_Groeger.pdf), both
> application profiles of GML 3.1.
>
> In terms of SVG, there are any number of tools that do GML to SVG for
> rendering. Some are free, some are for fee.
>
>