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

1. Danny Bernstein
2. Peter Eichman
3.  Jared Whiklo  
4. Andrew Woods  
5. Ben Pennell  
6. David Wilcox 
7. James Silas Creel  
8. Jon Roby
9. Yinlin Chen

Agenda

1. Announcements

2. Release Candidate
   1. State Tokens are missing: what do we want to do about it? 
   2. HashURIs in mementos

   3. Remaining work

      1. Tickets

      2. CTS

      3. Testing

3. Any technical items to add to the newsletter?

4. <your agenda item here>

5. Please squash a bug!

   
   

6. Tickets resolved this week:

   
   

7. Tickets created this week:

   
   

Minutes

Sprint is wrapping up – Danny and Andrew added a few final cards.

Announcements:

Release candidate not going out this week due to dangling issues.

SCSFUG in January: https://wiki.duraspace.org/display/FF/South+Central+States+Fedora+Users+Group+Meeting%3A+16-17+January+2019.  David will be able to attend and present on Fedora 5

Release Candidate:

• State Tokens are missing: what do we want to do about it? Ref. 3.10.1 in the spec.  They allow you to get a unique token for a state of the object so you can confirm or disconfirm that changes have happened to the item since you fetched it and before you might modify it.  Existing HTTP mechanisms don’t handle this exactly the way the community needs, especially Samvera and possibly Islandora.  For example, RDF triples are unordered, and we have to use e-tokens.  And HTTP last modified tags have only 1-second resolution, and we need it finer-grained.  This is a new arrival in the spec, not currently supported.  Should it go in 5.0 or 5.1?  They are specified as “may”s.  If an impl doesn’t satisfy a “may” requirement, just skip it in the compatibility test suite.  It’s not a breaking change, so could easily be added as a feature later.  Consensus is to punt it to 5.1 at least.

• HashURIs in mementos- How should HashURIs be represented in mementos?  General consensus is that when you make a request to a momento followed by a HashURI, the subject in the retrieved RDF should be the original resource followed by the HashURI, not a memento.  A related card is in progress - https://jira.duraspace.org/browse/FCREPO-2929:  If a resource has HashURIs associated with it, when you try to create a memento from it, it is failing.  Ben is wrapping that up.  

• Remaining ticket is 2928: UI: Misleading link title under "Other Resources".  Danny will likely take it on.

• People are feeling good about the PR for 1889: Direct/Indirect Containers allow writing arbitrary triples to any resource, despite access controls.

• How to proceed with the release candidate when the time comes: Create a RC branch, tag it, create the pre-release release (all in GitHub), and send out an email announcement.

• Regarding some particular CTS failures - No version of Fedora has supported naked RDF containers, so let’s not support worry about the ability to add them or subsequently update them in a way that changes their interaction models.

Release testing: https://wiki.duraspace.org/display/FF/Release+Testing+-+5.0.0

Testing should occur off of master branch prior to the release candidate.

The progress toward the release candidate will get a mention in the newsletter.