...
Time | Item | Who | Notes |
---|---|---|---|
2 min | Housekeeping and updates |
| |
20 min | Timeline | Please discuss and comment. Questions: When will know? With 2-month extension - more time to comment and review. Oct 28 group meeting: Review of updated UI diagrams Nov 1: Final UI documents posted on Confluence Nov. 15: Specs due for Advisory Board review Jan 15, 2020: Comments due from the Advisory Board Feb 28, 2020: Comments integrated into the specifications and posted March 30, 2020: Final report and deliverables due to Mellon Without 2-month extension - Less time for comment and review. Nov 8: Specs due for Advisory Board review Dec 13: Comments due from the Advisory Board Jan 1, 2020: Comments integrated into the specifications and posted Jan 30, 2020: Final report and deliverables due to Mellon (UI deadlines stay the same) | |
Outstanding questions from User Story Activity | Goal here is to clarify that these questions are slated to be answered in the specifications.
Are there other outstanding questions from the in-person meeting in need of addressing? Audit format - We came up with a set of items to be included. Not comprehensive but instructive. | ||
Responding to community comments | Will one of the specification writers respond to the comments left on the Preservation Flow page? From in person meeting: 1) Expectation is that deltas are sent, not the entire object. Send everything the first time, deltas later. Full manifest sent from Gateway- Bridge should know if it already has particular files or not? Send OCFL manifest? 2) Audits are not triggered on request. Audits may cost the DDP money (if stored in Glacier, etc). Audit frequencies are part of the SLA between the institution and DDP. 3) Relationship between objects and file group IDs is intentionally abstract because the repository and gateway know about object, whereas the Bridge and Gateway do not. Filegroup IDs represent a grouping of files ingested into DDP. Abstraction is injected so that we have a well documented way of taking files on disk and mapping them back to objects in repository. Filegroup is a serialized object. To the DDP it is a grouping of objects. Yes, we should document this better. Don’t necessarily want to use the word “work” to limit the spec to Hyrax.
|
Action items
...