2020.01.21 Editors Meeting

Call-in

https://emory.zoom.us/my/rmetz

Attendees

• Julian Morley (Stanford)
• Rosy Metz (Emory)
• Simeon Warner (Cornell)
• Andrew Woods (Lyrasis) *
• Andrew Hankinson (Bodleian Libraries)
• Neil Jefferies (Bodleian Libraries)

(* indicates the notetaker)

Agenda and Notes

1. Open 1.0 Spec Issues and Open Pull Requests
   1. Issue #411: We need consensus on this issue. Julian, Rosalyn, and Simeon think there should be no restrictions in the logical path.
   2. Issue #403 and WIP #404: What needs to be done to close this?
   3. Issue #413: what's the status on this?
2. Releasing 1.0
   1. Status of validators
   2. Status of fixture objects
   3. Status of test suite
3. Review Previous Actions

Notes

Open issues

1. Issue #411 updated with consensus on minimal limitations related to logical paths
2. Issue #413, Rosy is working this.

Extensions

Related pull-requests/issues:

• PR-404
• Issue-403

Related to extensions, what are the requirements/expectations/limitation for:

• Storage Root?
• Object Root?

What are the scope and limitations of extensions?

Storage root

1. Storage-root-level extensions should be isolate to a top-level "extensions" directory
2. This directory may include:
   • JSON extension descriptions
   • Whatever an extension defines
   • Individual directory per extension?
3. Could be potentially more clear if the structure/requirements for Storage root "extensions" directory paralleled that of the Object root "extensions" directory
4. Each extension must have an extension definition/specification
   • Neil is working on defining a parameterized way of referencing an extension definition (issue-2)
   • Or... do extension definitions belong in the Storage root
5. Examples of Storage root-level extensions:
   • Pairtree
   • Deposit dir

Object root

1. For objects that use extensions, there must be an 'extensions' directory
2. Regarding WIP Extensions PR-404
   • The term "registering" needs to be defined: official registry? local registry?
   • PR should add explanation for why MUST/SHOULD
   • Only files/directories inside 'extensions' should be extension subdirectories
   • Principle: extensions be identifiable by extension sub-dir name
   • Agreement that:
     • "Extensions SHOULD be a defined/registered"
     • "Extensions MUST be in their own sub-directory"

Actions

Previous Actions

• Neil to create issue for discussion of ocfl_layout.json options with respect to keys and optional parameters