[Geotiff] GeoTIFF and the OGC

[Carl Reed]

Well, Howard, there is a GeoTIFF user community out there that 1.) does not 
want to see the GeoTIFF grassroots community "destroyed" but 2.) does want 
GeoTIFF to be an international standard that can be referenced as such in 
policy and procurement. What would you suggest we do??

That said, I am sorry you have issues with the OGC process. But based on 
what you have written, I believe you may have some misconceptions about the 
OGC and OGC processes.

The OGC process is much more open to non-Member participation than you 
believe.

1. OGC Working Groups can be made open to participation by OGC Members and 
non-Members. For example, the OGC Mass Market, Hydrology, and 
Meteorology/Oceans Domain Working Groups are open to anyone who wishes to 
participate.
2. The entire change request process is publicly accessible. Anyone at any 
time can submit a change request using the public change request proposal 
web application. All change requests, whether Member or non-Member in origin 
are publicly accessible.
3. OGC Interoperability Experiments can have non-Member participation. The 
only stipulation is that the non-Member must sign a participant agreement. 
Both the Galeon and the Oceans Interoperability Experiments have non-Member 
participation.
4. OGCNetwork is a public resource that anyone at anytime can provide and 
maintain content. OGCNetwork is a community resource.
5. Any draft/candidate standards are required to go to a 30 day public 
comment period. Anyone can provide comments (Members and non-Members).
6. All OGC face to face minutes are made publicly available (this was at the 
suggestion of the OSGeo community!)

As you can see, there are many points of opportunity for continued 
collaboration between the GeoTIFF community and the OGC community.

Perhaps even more importantly, the OSGeo and OGC Memorandum of Understanding 
provides the mechanism for providing free OGC membership to OSGeo 
participants. Specifically: OGC provides OSGeo with six one year Individual 
Memberships in the OGC at no charge, to be awarded to qualifying OSGeo 
members based on a selection process to be conducted by OSGeo.  So, if you 
are willing to help, there is a mechanism!

As to reference implementations and so forth, this is great!  One of the 
requirements for a document becoming an OGC standard is that there be 
reference implementations.

As to a community standard surviving the OGC process, well there is no 
intention on the part of the OGC to stop the community process. And if 
individuals such as yourself particpate and help in the process then we can 
insure that the communtiy process does survive. An excellent example is 
CityGML (www.citygml.org). CityGML started outside the OGC process, was 
brought into the OGC, and became a standard. The community process is still 
strong, vigorous, and the source of most of the change requests for a 
revision to the current OGC CityGML standard. The community process is an 
excellent source of expertise, development, testing, and creative ideas. No 
one wants such processes to stop. If they do, it may be because the 
community process was not viable to begin with (my perception).

As to the OGC statement on axis order, the OSGeo community was included in 
the dialogue. The basic guidance is that the schema, documentation, or 
encoding shall state the access order. Specifically from the guidance, ". . 
. any documentation, encoding, payload, or service interface MUST state how 
the coordinate axis order is actually encoded in the coordinate strings." 
What is so onerous about that? If one does not honestly state how 
coordinates are actually encoded in a payload, then interoperability is 
broken.

Finally, where in my email did I mention "evolve"?? Revisions to any OGC 
standard only happen if change requests are submitted and the community 
wishes to evolve a spec. Period. And the evolution can happen entirely 
outside the OGC and then change requests submitted into the OGC process. 
This is how KML is evolving.

I may not allay your fears, but hopefully I have provided more food for 
thought.

Regards

Carl



----- Original Message ----- 
From: "Howard Butler" <hobu.inc at gmail.com>
To: "Carl Reed" <creed at opengeospatial.org>
Cc: "GeoTIFF" <geotiff at lists.maptools.org>
Sent: Monday, October 26, 2009 2:07 PM
Subject: Re: [Geotiff] GeoTIFF and the OGC


>
> On Oct 26, 2009, at 12:59 PM, Carl Reed wrote:
>
>> Dear GeoTIFF list -
>>
>> At the recent OGC meetings in Darmstadt, several OGC Members (large  user 
>> organizations) asked whether there is any interest on the part  of the 
>> OGC to move GeoTIFF into the OGC standards process and  eventually make 
>> GeoTIFF an international standard. As we have done  with other de-facto 
>> standards submitted into the OGC standards  process, I would see very few 
>> (if any) normative changes to the  document as it moves to a version 1.0 
>> OGC standard. I know that the  implementation community would not 
>> appreciate any "normative  changes"! Further, as with KML, CF-NetCDF and 
>> other documents, we  believe that maintaining a strong collaborative 
>> relationship with  the existing GeoTIFF community is critical.
>>
>> As part of this process, the OGC and OGC Members would be  responsible 
>> for reformatting the document into the proper template,  marshalling 
>> volunteer resources to move the document through the OGC  process, and to 
>> insure proper communication and engagement with the  current GeoTIFF 
>> community.
>>
>> Your thoughts regarding this proposal are appreciated!
>>
>
> My concerns with the approach the following:
>
> - OGC subsuming the current GeoTIFF specification and becoming its 
> ongoing authority effectively destroys any grass roots community that  has 
> sprung up around geotiff because of OGC's membership  requirements. 
> Casual or short-term-to-complete-this-job interest in  geotiff is 
> effectively shut out from participating, though you could  argue there 
> isn't much of this anyway.
> - OGC hasn't demonstrated that it can pull community-developed 
> specifications into its world and have the community that birthed them 
> survive the process.  Does OGC's sausage taste that much better than 
> sausage made outside the OGC?
> - GeoTIFF already has a widely used and effective reference 
> implementation -- libgeotiff.  That's more authority than a document 
> stamped with an organization's letterhead can ever hope to be.
>
> What benefits would this development be to the GeoTIFF community other 
> than forced acceptance of OGC-style axis order discipline? ;)  Is the 
> main benefit is to be the ability to make changes to the specification 
> that implementers will have to implement to be compliant? Why try to 
> evolve a specification that has been widely used and unchanged for  nearly 
> 10 years?