Time/Place

This meeting is a hybrid teleconference and slack chat. Anyone is welcome to join...here's the info:

Time: 11:00am Eastern Daylight Time US (UTC-4)
Audio/Video Conference Link: https://duraspace.zoom.us/my/fedora
- Dial-in:
  +1 408 638 0968
  +1 646 876 9923
  +1 669 900 6833
  Meeting ID:
  812 835 3771
Join fedora-project.slack.com on the "tech" channel

Attendees

Agenda

Announcements
1. Java Client Release
2. 5.0.1 Release Notes
5.0.2:
1. issues

1. Call for release manager
2. Date?

CTS appears to have three new failures : https://github.com/fcrepo/Fedora-API-Test-Suite/issues/319

Unanswered mailing list messages
1. [REINDEXING SERVICE] Question about how it works
Status of ecosystem tools:
1. Java Client Release
2. fcrepo-camel
3. camel toolbox
Next steps for import/export/4→5 migration
1. Initial investigation
2. Volunteers for a task force
5.1.0
1. Ready for review
  2. "Check if metadata is available in commons.js" : https://github.com/fcrepo4/fcrepo4/pull/1499
2. Issues that are ready to be worked:
<your issue here>
Please squash a bug!
Tickets resolved this week:
Tickets created this week:

Minutes

Announcements
1. fcrepo-java-client released and ready for use
5.0.2
- The two bug fix tickets should now be merged in
- There is still a lingering issue with C3P0
  - We could try to remove it, or wait for a fix on their end in the next few weeks.
  - The maintainer has indicated they will probably resolve it in the next week or so
    - https://github.com/swaldman/c3p0/issues/123
- Fcrepo-2977 is a fairly significant issue, dbernstein suggests not waiting more than a few weeks
  - whikloj: commons-dbcp may be a potential replacement http://commons.apache.org/proper/commons-dbcp/
- Looking for someone to do release of 4.7.5
CTS new messages
- 2977 and 2975 were both merged. 2977 was tested against the CTS, 2975 was not
  - Could be a timing issue, or the number of messages. It was expecting 2 messages, there might be 3 now because of the immediate creation of timemap
  - Bethany: isn't this a fedora-ism? Would other implementations pass? Maybe it should just test for at least one message.
    - Would be good to know how many messages are being received
    - Don't want to change the test to match the community implementation, needs to work for all impls.
  - Need to determine what the issue is first
  - Do we need to get messages for creation of timemaps?
    - Ben: for indexing purposes this is helpful
  - Whikloj: should we be able to provide the number of expected messages by implementation?
    - Bethany: maybe do this later, in short term could do the more minimal route
  - Bethany: happy to take a look to confirm that the issue is the number of messages
Mailing list question - reindexing
- Peter was going to follow up, ask how they are triggering the reindexing service
Ecosystem tools
1. fcrepo-java-client - completed
2. camel
  1. Peter wants to take a look at Aaron Birkland's camel work, but has gotten bogged down
  2. dbernstein: would it be useful to gather a team for this?
    1. Peter: Could be useful to have a mini-sprint to devote to it. Is guessing its not huge, but larger than we suspect it is
    2. AaronB: There are processors that create headers, so the amount of work that depends on the headers vs the body
      1. If it just looks at headers, might not be much work. If looks at body, that'll be more work.
      2. Guessing translating the headers is most of the work
    3. camel toolbox has its own integration tests, which should be helpful for testing it
      1. need to get fedora setup for these tests
    4. Need to get PR for updates to fcrepo-camel merged before we will be able to do testing of the toolbox
  3. Peter will talk to his supervisor next week about carving out a week to devote to this
  4. Danny will dig in tomorrow to see what happens if runs Aaron's changes through
Import/export
- may need to gather a group to look at the scope of the issues.
- Danny doing a bit of testing to see where things stood
- Versioning
  - Import of versions is not present for 4.x
    - Do we need to support this? No one has clammored for it
    - There is an existing PR to add this feature, but likely wouldn't agree with
  - Export and import of versions does not currently work for 5.x
  - Migration from 4.x to 5.x is higher priority than 4 to 4
- The changes do not seem that big, a small task force might be able to get that done
  - Bethany and Peter interested in working on it, but have very limited time at the moment
  - Danny would have time, but needs funding
  - Ben P would need to talk to supervisor about it
  - Danny will call out to community for support
5.1.0
- Danny is guessing 3-4 months out
- Close some of the gaps on the CTS tests that are not passing, particular the MUSTs around containers
- There are issues available to be worked
  - its possible 2976 (state tokens) may need to be broken into smaller tasks
    - There are a number of actions, like GET and PUT that need to be handled, as well as multiple cases
- Ben: Where will https://github.com/fcrepo4/fcrepo4/pull/1501 be included? 5.1.0 or 5.0.x?
  - It is going to go into 5.1.0
Checking in with Peter about UMD fixity checks being stored in fedora
1. Peter going to add use case results somewhere in fedora documentation. Is there a place for this kind of thing?
2. Danny will find a spot for it and pass along a link

Actions

Peter Eichman will talk to his supervisor next week about carving out a week to devote to fcrepo camel toolbox
Danny Bernstein dig in on camel toolbox friday to see what happens if runs Aaron's fcrepo-camel changes through it
Peter Eichman is planning to work on documenting UMD's fixity check system on the Fedora in Production: Case Studies wiki page.
Bethany Seeger will take a look at https://github.com/fcrepo/Fedora-API-Test-Suite/issues/319
Peter Eichman will respond to mailing list question: [REINDEXING SERVICE] Question about how it works
Danny Bernstein will reach out to Yinlin Chen to see about a fcrepo4-docker release with the 5.0 architecture.