[OSGeo-Discuss] Would you be concerned if the "GeoServices REST API" became an OGC standard?

[Adrian Custer]

On 5/4/13 6:21 PM, Cameron Shorter wrote:
> Adrian,
> Thankyou, I was hoping that someone such as your self with insights into
> the standard would explain the details. You email has been a great help.

Cheers.

>
> I'm also hoping that someone will provide a more detailed comparison of
> the similarities / differences, to help the rest of the community
> understand what is happening.

I have not taken much detailed interest in these services ever since it 
became clear that they had no interest in working with the WMS folk and 
that nothing could be changed to improve the standard. (They couldn't 
even fix dates to be in the unambiguous ISO 8609 format YYYY-MM-DD since 
that would break 'backwards compatibility'!)


My understanding is that these services are built on what I call the 
"Flat Feature Model" which are Features with single geometries made up 
of 2D, linear structures (points, lines, or polygons) and a list of 
primitive value attributes (the shapefile data model).

     ("Simple Features", it turns out, are not so simple; they can only
      have a single geometry but that geometry can be multidimensional
      and complex while the attributes can be arbitrarily complex and
      in various namespaces. So 'Flat Features' are what I used to think
      'Simple Features' were.)

There is surely a need for very simple geospatial services which are 
limited to the Flat Feature Model. That is why we have all been working 
on rewriting the W*S standards in modular form to allow for very simple 
implementations while also enabling more complex implementations, 
experiments, and easier fixes. The ESRI effort, had it been designed to 
help users, could easily have plugged into the efforts going on in the 
various W*S groups. Instead, it has so far been a complete drain on my time.

~adrian