[OSGeo-Standards] Comments on revised OGC 08-085r2, OGC GML in JPEG 2000 (GMLJP2)

[Even Rouault]

1. Evaluator:
Even ROUAULT (Geospatial Professional Services)

2. Submission: [OpenGIS Project Document Number, Name]
OGC 08-085r2,
OGC GML in JPEG 2000 (GMLJP2) Encoding StandardPart 1: Core

========================================================
1. Requirement: General  - Format of document to comment

2. Implementation Specification Section number: General

3. Criticality: Editorial

4. Comments/justifications for changes:

Please avoid using proprietary formats for documents that deal with open 
standards... I'm not sure to have the right pages and section numbers in the 
below comments...

========================================================
1. Requirement: Coordinate reference systems 

2. Implementation Specification Section number: Section 16, Page 27

3. Criticality: Major

4. Comments/justifications for changes:

It is mentionned at page 26 that the geometric properties of GML features 
[...] are interpreted within the context of a coordinate reference system 
(CRS)", but example at page 46 (section 31) shows coordinates of the Road 
feature in the range 200-400, whereas the Envelope is in EPSG:4326. And the 
releated GMLJP2RectifiedGridCoverage has no CRS defined. This is confusing.


========================================================
1. Requirement: Req 19 /req/xml-box-signal

2. Implementation Specification Section number: Section 25, Page 39

3. Criticality: Major

4. Comments/justifications for changes:

It is annoying that writer of images must now write that Reader Requirements 
box. Is there a real need in making that a requirement ? It is easy for 
readers to explore the JP2 boxes to discover if there's GML content or not. 
All software handling current version of GML JP2 must already do that. And, 
anyway, they must parse the other JP2 boxes for other useful information.

========================================================
1. Requirement: 36.Multiple codestreams case

2. Implementation Specification Section number: Section 36-37, Page 50

3. Criticality: Major

4. Comments/justifications for changes:

It is not clear at all how to write the RectifiedGrid of - let's say - a RGB 
image where the 3 components share the same georeferencing. I guess the 
domainSet at GMLJP2CoverageCollectionlevel should be used, and then referenced 
in each GMLJP2RectifiedGridCoverage member for each codestream ? A more 
meaningful example than the kind of empty one would be helpful.

Best regards,

Even Rouault

-- 
Geospatial professional services
http://even.rouault.free.fr/services.html