- Roll Call
- Scribe nomination
- Review minutes of previous meeting
- Change Management topics
- Configuration Management topics
- Any other business
Nick to scribe.
Chat transcript from room: oslcnccm 2016-08-25 GMT-08:00
[07:05] List of attendees: David Honey (Persistent/IBM), Martin Sarabura (PTC), Nick Crossley (IBM)
[07:05] Nick Crossley (IBM): Minutes of previous meeting are at https://wiki.oasis-open.org/oslc-ccm/Meetings/Telecon2016.08.11
[07:05] Nick Crossley (IBM): Minutes approved
[07:07] Nick Crossley (IBM): Meeting on September 8th canceled, since Nick will not be available.
[07:08] Nick Crossley (IBM): Martin to update proposal for issue 27
[07:12] Nick Crossley (IBM): David and Martin discussed and agreed on the need to separate authorization for a change vs. an actual change or change proposal
[07:14] Nick Crossley (IBM): For example, some of the fields / attributes of the authorization and the discussion leading to that may not want to be exposed on the final change request
[07:14] Nick Crossley (IBM): Authorization process may have a different life cycle from the change process
[07:19] Nick Crossley (IBM): Martin: Also need to address containership - how to associate a set of change requests to a 'parent' or container.
[07:19] Nick Crossley (IBM): Martin: Action Item: raise separate issue on this, splitting it out of issue 27.
[07:22] Nick Crossley (IBM): Issue 28: what should be the set of pre-defined relationship types between the types defined by the Change Management spec?
[07:23] Nick Crossley (IBM): Martin: V-model shows decomposition, and association between upstream artifacts and downstream artifacts.
[07:25] Nick Crossley (IBM): Classical levels of v-model should inform us in the set of pre-defined relationships, but should not mandate it (there are other development methodologies that would imply other relationships).
[07:29] Nick Crossley (IBM): Core OSLC should require that when trace relationships are defined by domains or implementations, those definitions should make it clear what the direction is, since direction is key in traceability.
[07:38] Nick Crossley (IBM): OSLC should define a wide vocabulary of possible/probable relationship types; domains can then add SHOULD and MUST implement requirements.
[07:39] Nick Crossley (IBM): Since many relationship types are common across domains, or used across domains, many of these relationship types should be defined by Core, so the definition is common.
[07:39] Nick Crossley (IBM): Nick: one issue with the above is that we have the existing OSLC 2 relationships defined at the domain level.
[07:42] Nick Crossley (IBM): Martin: we could add new Core vocabulary terms and deprecate the domain-specific ones.
[07:43] Nick Crossley (IBM): David: this is challenging for compatibility - queries would probably end up asking for a union of possible relationship types - or we might rely on owl:sameAs.
[07:44] Nick Crossley (IBM): Martin: apps could also use OSLC versioning to determine whether to use old or new terms.
[07:45] Nick Crossley (IBM): David: namespace also determined by the probably source of the link - so a link from a test case to a requirement typically uses the oslc_qm namespace.
[07:47] Nick Crossley (IBM): Action Item: Nick & Jim to review IBM change management link types in commonly installed process packages, and add terms and desirable support levels to issue 28.
[07:47] Nick Crossley (IBM): Action Item: Martin to do the same for PTC.
[07:48] Nick Crossley (IBM): Martin: this is an important issue, since much of the value of OSLC is in determining the links between artifacts and the meaning of those links.
[07:51] Nick Crossley (IBM): Martin: going back to issue 27: ChangeNotice vs. ChangeAuthorization. Most appropriate term may be ChangeNotice, since if we have that we can add authorization to it if desired, whereas if we define just ChangeAuthorization it is less clear how to handle notices without authorization.
[07:52] Nick Crossley (IBM): Or we can define both!
[07:53] David Honey (Persistent/IBM): https://issues.oasis-open.org/browse/OSLCCCM-29
[08:00] Nick Crossley (IBM): We discussed issues with representing ordering in RDF.
[08:01] Nick Crossley (IBM): Meeting adjourned.