Part 2:

Agenda

...

Beginning to sketch the DB schema
Interest in tying DB updates into internal message-bus
Details
- If txn is acting on a resource that is updated outside of the txn, the txn wi
  ll fail on commit
- Containment index: potential simplifying modeling assumption: children have only one parent container
States of the index
- In a txn
- head
- Within a version
Need to define the expected behavior within a transaction
- Are updates happening outside of the txn visible within the txn?
- or, the repo state is a snapshot from the point when the txn was started
ACTION: Peter to investigate approaches for txn behavior
- ideal (complete isolation within the txn)
- F5
- most practical
Conversation will continue at 1pm ET
- Should also include resolution of "Open Questions"

Open Questions:

We went through the list of open questions. Many items moved into the decided list. A couple moved to nearly decided. A few questions remain.
See Decisions and Open Questions for details.

Actions

Aaron Birkland to look explore notion of OCFL client with database as authoritative metadata source + asynchronous writing of the inventory.json file
David Wilcox will review the NDSA matrix and pull out the concrete technical requirements that could be considered during the Fedora 6 development.
Call for comments on https://docs.google.com/document/d/18rSFqqoxixmozZrgPKON59Ojpg5iOu4lsHDNMm23soY/edit# till tuesday
Clarify in in documentation that multiple simultaneous writes to OCFL are not supported

...