- Sprint Progress
- no blockers, Andrew Woods wondered where remaining gaps may exist and where help is most needed (possibly reallocate efforts to ensure completion)
- closing in on the test suite as well (almost all but messaging is complete)
- Completing and passing the compatibility test suite should be prerequisite for a release
- Differing opinions about how close we are to having a shining example of a Fedora Spec implementation were expressed.
- WebAC is getting close as well
- Issues to Discuss
- There are a few PRs in Compatibitliy Test Suite that should have straightforward reviews that should be reviewed and merged
- FCREPO-2854, FCREPO-2855: Jared Whiklo doesn't see any purpose in accessToClass given that we don't have defaults and it only applies to a resource anyway. Peter Eichman considers it useless but not harmful, and since maybe someone might find a reasonable way to use it in the future.
- Peter Eichman suggested that some sort of best-practice-auditing tool might help (to warn people that they might be doing something that doesn't have any real effect in the repository and ensure they're not making a mistake), but didn't go as far to as to recommend anyone to write such a tool at this time.
- FCREPO-2852: are we satisfied that mementos should return original URIs and subjects? (Indexing mementos in a simplistic way will result in historic triples included with the current ones)
- Kevin Ford suggested the extraordinarily clear documentation is likely the best approach now.
- Danny Bernstein said that technically it's be just reverting a commit to change this behavior
- Bethany Seeger suggested that it might be implied in the spec that the memento use original URIs
- Modeshape snapshots had references to the other elements in the snapshot. Since some references point out from the snapshot to unsnapshotted resources, there wasn't any way to point to the memento because it doesn't exist. They can only be original URIs.
- There were strong arguments for the current approach, but an acknowledged need to update the camel indexing routes to support better handling (probably dropping) of memento triples.
- Danny Bernstein will close the ticket, but create a new one about the documentation/camel toolbox
- FCREPO-1402: Aaron Birkland doesn't remember when/why we removed messages emitted during changes to linked resources. Many current users might depend on it now. While it's an easy change to make, do we have to, and who will it harm?
- Jared Whiklo has a workaround that morphs events into something else rather than throwing them away.
- No conclusion, will discuss further.
Future Agenda Topics:
- Amherst College's acrepo apache camel services: https://gitlab.amherst.edu/acdc/repository-extension-services
- Reviving OAI