[Nml-wg] Teleconference Dec 15th

[Jeroen van der Ham]

Meeting minutes:

Attending: Freek, Aaron, Jeroen
Excused: Jason, Roman and Chin

- There was some slight confusion about the time of the call. It should really be:
UTC 16:00 = 8:00 Pacific = 11:00 Eastern = 17:00 CET
Next call will be on January 12th.

- Number of sessions for OGF34, this really depends on the expected attendance of the group members, otherwise we will request 2 1/2 again.

- Discussion on URN identifiers, to ensure uniqueness use <DNS name>:<year> as part of the prefix, where <year> is the year of assignment of that identifier. This should ensure uniqueness in almost all cases.

- We discussed the subPropertyOf, problem may be that the mapping between XML and RDF syntax may not be apparent because of different way of defining relations. Jeroen will send an example to the list.

- Capitalization and URN ids of the XML have been fixed by Jeroen, but committing is not possible at the time. After the move of Gridforge the SVN repositories have become broken. Jeroen will take this up and try to get it fixed.

- The symbol restrictions should be put into the schema document as well. It is up to the organizations to fix this.
Aaron asked whether it was possible to encode them somehow. Freek replied that URNs can allow %-encoding, but this has not been allowed for the urn:ogf: prefix to prevent problems with normalization.

- Jeroen is busy with discussing topology descriptions in NSI. He will keep the NML list posted on the progress of this as well.

- Jeroen brought up a potential issue with the "next" relation in ordered Groups. If there are overlapping Group instances, this may cause problems. Freek will write up an example and send it to the list. It should also be documented in the schema that the "next" relations should always be considered with the context of the specific Group object in mind.

- There is no real reason not to use hasInboundPort and hasOutboundPort for Nodes also. Then again, it would also be possible to use a Group concept for that (Port). We should make an actual use-case with associated code to figure out the best way to approach this. We should define the directionality, the relation to the Node, but also the relation to each other.



Action Points:
- Jeroen will send an example of subPropertyOf relations to the list to show mapping between XML and RDF
- Jeroen to chase SVN fixing on Gridforge
- Jeroen will forward relavant emails about NSI topology discussions to NML
- Freek will write up an example of overlapping ordered Groups to expose potential problems, also with regard to idRef XML notation.