[Geojson] GeoJSON as OGC standard?

Howard Butler hobu.inc at gmail.com
Mon Jul 25 14:19:34 PDT 2011

On Jul 25, 2011, at 12:51 PM, John Herring wrote:

> Chris, 
> The major thing a fast track needs is an extant set of conformance classes; what essentially is a complete abstract test suite for the proposed standard.

Haha, but serious, doesn't the GeoJSON 1.0 spec already contain that? :)

> This is the principle gotcha that force the GeoServices REST API into an SWG. 

The problem I have with OGC advertising "Come here, we'll be your home for your defacto/community-developed specification" is while it might be nice to have a stamp of approval, it seriously clouds who is now the authority with regard to curating the specification.  An excellent example of this is Arlo coming to the this list asking about empty and 3D/M geometries.  There hadn't been any substantive GeoJSON list traffic in a year+ before that.  I don't know that Arlo is satisfied (yet), but we definitely have a straightforward and open way to participate.

On Jul 22, 2011, at 09:16 AM, Chris Holmes wrote:

> So the spec can continue to
> evolve on this list (though I feel it's been quite stable for awhile, so may
> not need to), with this community having control of it.  After it's updated
> then can be submitted to OGC.

How exactly does this work?  How are new revisions of the document submitted and approved if GeoJSON were to ever have another?

It is clear this list is the GeoJSON authority until such a time that it either becomes disinterested or defunct (GeoTIFF is in this boat -- widely used but no community authority anymore to shepherd it along).  OGC(R) stamping will cloud that. 

We (well, not me so much but all the rest of you) made this specification successful by developing the thing openly on a public mailing list, and then going back to our respective sandboxes and actually implementing it.  Document + implementation (chicken(s) + egg(s)) is what makes these things gain traction.  OGC doesn't say to much about the implementation side of things, but it was a major component in GeoJSON's acceptance.  How does an OGC(R) stamp change or improve the current situation of GeoJSON at all?


PS, I'm not entirely opposed and mostly indifferent, even though I have spilled plenty of bytes about OGC. The GeoTIFF story is probably instructive here.  There's no one to really go back to anymore.  

More information about the GeoJSON mailing list