Open DavidGrant-NIWC opened 1 year ago
agree this is poorly specified and the exchange set has a couple of different examples. PArt of this is because some of the datasets metadata has been imported (And converted) from edition 4.0.0 metadata implementations. The new version (under 2.0/IHO_V12) only encodes productIdentifier and number. but number is always = 1.
There's no way, currently (other than matching arbitrary strings to uniquely identify which catalogue matches which dataset. This needs some agreement - I'm happy with mrn for this but possibly needs a concrete proposal. It also needs clarification in Part 17 how it's supposed to work (and possible clarification in S-98 specifically for ECDIS).
Currently, the design requires associating each catalog with a version of the product specification. Catalogs updated as the result of corrections or minor changes not affecting the dataset encoding would have the S100_CatalogueDiscoveryMetadata.editionNumber incremented.
This doesn't work very well for a variety of reasons:
For the FC/PC, the ECDIS just needs a way to associate one of these with each dataset. It doesn't really care about the product spec. Each PC targets a single FC. Each dataset was generated from a single FC.
This could be moved to S-164 for any further discussions. Probably not required in S-101 dataset repo.
Some datasets:
Others:
If name is encoded, it should be "Electronic Navigational Chart" for S-101 datasets. number should be the GI registry index value (although there is currently no easy way to determine this...)
Given that the goal is to associate a FC with each dataset in the XC, we should consider updating the schema of the XC, FC, PC, and IC to support specifying either the FC or product (or both) via mrn:
urn:mrn:iho:prod:s101fc:x:y:z:whatever
orurn:mrn:iho:pub:spec:s101:x:y:z:whatever