As discussed in the PCA MMT SIG meeting, here are ID spec examples (in triple structure):

1. Items with both RDSnnnnn and Rnnnnn (existing today in PCA RDL):
http://posccaesar.org/rdl/Pump pca:hasIdPCA "RDS327239"
http://posccaesar.org/rdl/Pump pca:defaultRdsId "R20735180747"

2. Items with RDSnnnnnn but no Rnnnnnn (existing today in PCA RDL)
http://posccaesar.org/rdl/DrillBit pca:hasIdPCA "RDS1707471151"
http://posccaesar.org/rdl/DrillBit pca:defaultRdsId "RDS1707471151"

3. Items with no RDSnnnnnn but existing Rnnnnnn (existing today in sandboxes but not in PCA RDL yet)
http://posccaesar.org/rdl/ActualLength pca:defaultRdsId "R70038186785"

4. Items with no RDSnnnnnn and no Rnnnnnn (non-existing today - to be added to the PCA RDL in the future)
http://posccaesar.org/rdl/MannedContainer pca:defaultRdsId
"R-2a9583ba-6260-4c0c-92ad-ef6a1b9e78e2" (fictitious)


For evaluation, these have been implemented in PCA development environment - human access http://posccaesar.org/development , for machines http://posccaesar.org/development/sparql . Note that the rdsWipEquivalents are also kept in these examples. In addition, there are sameAs-relations from the old scheme to the new one. Query from development endpoint or see:
http://posccaesar.org/dev/Pump
http://posccaesar.org/dev/DrillBit
http://posccaesar.org/dev/ActualLength
http://posccaesar.org/dev/MannedContainer

The spec itself is available here: https://www.posccaesar.org/raw-attachme ... ec-v41.doc

Hi Lillian,

1. Of the above example, the namespace is http://posccaesar.org/rdl andthe rdf:ID is "Pump" correct?
2. From your words it is not clear if it is a namespace or an endpoint address.
3. Of the last example, the rdf:ID is "MannedContainer" correct?
4. I guess the rdf:ID should rather be the R-UUID and the rdfs:label MannedContainer
5. What is the difference between
   
   • http://posccaesar.org/rdl/Pump
   • http://posccaesar.org/dev/Pump
   • http://posccaesar.org/development/Pump
6. Should the namespace rather be
   Code:

   http://standards.tc184-sc4.org/iso/15926/tech/reference-data#

*edit*
The Pump is in http://standards.iso.org/iso/15926/-4 (spreadsheet is rotating equipment)
Its ID is

1. I should add that the examples chosen are PCA RDL examples, and they are in http://posccaesar.org/rdl namespace. And yes, it is proposed to change to human-readable ID as main URI.

2. PCA has several endpoints. The development endpoint that has implemented ID spec examples is currently available from: human access http://posccaesar.org/development , for machines http://posccaesar.org/development/sparql . The production endpoint for the PCA RDL is accessible from http://posccaesar.org/endpoint/ (human access) and http://posccaesar.org/endpoint/sparql (machines) and resources have not been changed.

3. Actually, you can view data about the MannedContainer resource from http://posccaesar.org/dev/data/Pump?output=xml A different seralization is used, but yes, MannedContainer will be the ID of that resource.

4. No, the spec proposes to change to human-readable ID as main URI and have the RDSnnnn/Rnnnnn/R-UUID at the end of a predicate

5. An implementation of the JORD ID spec on the resource http://posccaesar.org/rdl/RDS327239 existing today in the RDL would result in http://posccaesar.org/rdl/Pump . http://posccaesar.org/dev/Pump is an example resource in a different namspace than the one used for the PCA RDL, and should not be referenced to and can not be considered stable. http://posccaesar.org/development/Pump is not a resouce and is not planned to be a resource. http://posccaesar.org/development/ is the endpoint address and to avoid confusion a different namespace is used.

6. Currently there are no ISO references as such in the PCA RDL, but that might change.

Point 6 (ISO IDs) is an important issue. If we're going to treat ISO identification along with other identification types, it will require introduction of an additional predicate to form the triple:


and equivalence statement

Onno, you are saying:



As far as I understand, there was only a human-readable ID, no alpha-numerical UUIDs?

As for the rest of your last comment, looks like everything now is dependant on ISO position. Whether they can host any software, or they can delegate the right to host anything in some subdomain. Even the right to use iso-namespace in URIs should be a subject of some agreement between ISO and PCA, I assume.