Chairs: Jim Amsden, Jad El-khoury
Previous Minutes: https://lists.oasis-open-projects.org/g/oslc-op/topic/oslc_op_meeting_minutes_jan/110774994
Attendees:
Status summary:
Configuration Management 1.1 has a draft spec for change set delivery and history based on a proposal by David Honey. This needs review and possible contribution from implementors. The only known open issue is to ensure recovery from failed delivery is possible.
Eran and Martin have a PSD revision of Link Discovery Management. Publishing the revision is pending edits by Martin and vote by the PGB. SysML vocabulary and constraints have been moved to OMG, and the OSLC SysML spec now refers to those documents rather than forking them. FTF2 is about to complete and these documents will become OMG standards.
Eran has a draft Link Validity Service specification. This needs review and commitment from implementors. There is an open question of whether the LVS specification need to cover profiles that specify what properties contribute to link validity.
There is a working draft of a PLM specification based on a proposal by Siemens. This draft covers PLM as a (versioned) resource, and does not address PLM as a configuration or effectivity. Establishing a working group to take this forward is in progress, and is lead by Ed Gentry.
Eran did some updates responding to Jim's review. The spec is ready for further review.
The working draft looks good. There is an opportunity to reuse some existing predictes instead of introducing new ones. For example: rdf:predicate to specify the predicate, and dcterms:contributor to specify who updated the validity.
The spec needs to specify what servers that provide information to support LVS need to do including returning the version hashes for one or more resources, and to specify link validity profiles for what properties contribute to link validity.
We discussed the need for LVS profiles that describe in a standard way for how to specify the properties that contribute the calculation of a hash. Eran thinks this is not required. The servers can calculate their hashes any way the want. Link validity is only based on the equality of the hashes. The meaning of suspect is not informed by the hash.
Jim thinks that the meaning of link validity depends on how the clients are using the links, similar in concept with OSLC ResourceShapes for defining different constraints on vocabularies. LVS profiles could be created to specify which resource properties contribute to the calculation of the hash, and therefore what is the meaining of a potentially significant change. Different users of LVS could define different profiles for different purposes.
This would have the additional benefit that different servers managing the same resource and link types could calculate the value of the hashes the same way, ensuring that the user has similar meaning of validated or suspect links across different servers.
Another potential benefit is that a user exploring a suspect link could use the profiles to understand what properties might have changed.
We also discussed the need to engage with potential implementors, or vendors who already implement some form of link validity to get their input, and see if they are interested in contributing future statements of use.
Jim will review the proposed LVS specification against the existing MID link validity service.
We still do not have enough OSLC-OP PGB members to function as an open project. MID and Siemens are exploring becoming an OSLC-OP project sponsors.
OASIS members: https://www.oasis-open.org/tc-members/
For OSLC-OP project sponsorship, contact: Holly Petersen holly.petersen@oasis-open.org
An initial OSLC PLM draft specification has been produced and checked in. See https://github.com/oslc-op/oslc-specs/tree/master/specs/plm.
The current OSLC PLM draft specification covers the conformance clauses for PLM as a resource. It does not cover:
No canges this week on LDM. The following issue is still open:
The OMG FTF2 changes to SysML.ecore have been updated and the vocab and constraints documents have been regenerated. The vocab and shapes documents are now hosted on /Users/jamsden/Developer/SysML-v2-API-Services in the public/oslc folder. The OASIS SysML specification vocab and shapes HTML files will reference these OMG documents.
Need to remove SysML namespace reference from open-services.net.
No progress this week.
Open issues for Config 1.1 change set delivery:
Change set delivery failure:
We are monitoring the migration of open-services.net to OASIS and will deal with issues as they arrise.
We belive OASIS has taken on ownership of open-services.net. There is still an open question about the location and ownership of https://oslcfest.org/. The site is still active.
Useful links (DO NOT COPY TO THE RELEASED MINUTES)
Project task board: https://github.com/orgs/oslc-op/projects/2
Call details: https://meet.jit.si/oslc-op
Minutes: https://github.com/oslc-op/oslc-admin/tree/master/minutes/2022
Spec drafts: https://github.com/oslc-op/oslc-specs#quick-links
-
Any changes
Be notified of any changes
-
Mention me
Be notified of mention me
-
Unsubscribe
Subscribe