srid

[ibastrak]

1. The sridType may be a misleading name to identify the SRID. At the moment the element has no definition and referencing the srid code, not its type. The element name can be changed to sridCode. A new element may need to be introduced to define the registry/owner of the srid code list, e.g. EPSG
2. The list of the spatial reference system needs to be expended to enable people to select the best coordinate system for their region. Although the WGS84 (the only allowable value now) provids common projection/datum for location interoperability, it is too generic if users need to define the sample coordinates more precisely. The vocabulary and service provided by the EPSG can be used (http://www.epsg.org/) as a list of SRIDs

[smrgeoinfo]

SRIDCode seems kind of redundant. Why not just SRID?

I strongly suggest that there should be a convention for SRID; allow multiple location elements with different SRID, and perhaps different geometry encodings (gml, geoJSON...), but require at least one with SRID = epsg:4326 and a standard, simple point-encoding scheme for interoperability.

[ibastrak]

If we agree that the IGSN metadata allows only one catalogue (EPSG) defining reference systems, then I agree with Stephen's comment. However, if we want to use multiple reference system catalogues, the metadata needs two elements: reference to the catalogue of the reference system and the value/code.