Page History

...

1. Update on issue of non-existent child resources (Kevin Ford): 
   1. Fixing this problem in the affected repository will require the following "open-heart surgery" process:
      1. making a Modeshape backup;
      2. determine which resource have non-existent children and delete them;
      3. restore the backup.
   2. But fixing it won't resolve the underlying issue in Fedora/Modeshape.
   3. Another goal of the investigation will be to determine when these objects first appeared.
   4. Michael Durbin: we need to have an automatic full stack trace for any 500 errors.
2. Updating server-managed triples change (

   Jira
   server DuraSpace JIRA serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 key FCREPO-2432

   , Michael Durbin):
   1. Currently working on some edge cases for updating the modification dates of linked resources when restoring from a backup.
   2. Plan is that when a child object is restored via import, the modification timestamp of the affected parent would only change if the last modified date is later than the creation date of the restored child object (thus mimicking how things would appear if the restored object had never needed to be restored).
   3. Changes are needed for this to be useful for import/export, but there are some thorny aspects to this and the goal is to create an intuitive way of dealing with those.
3. API Specification Working Group (Esmé Cowles):
   1. Latest developments can be found in the minutes.
   2. Next meeting will be in two weeks. All input and participation is are welcome.
4. Question of the ticket (

   Jira
   server DuraSpace JIRA serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 key FCREPO-2423

   ) for an upgrade utility to accompany the 4.7.1 release was discussed. 
   1. This utility would allow repositories that have already made use of blank nodes to update existing blank nodes (that follow the old "well-known" behavior) work the same way as newer blank nodes (using hash URIs). 
   2. Danny Bernstein is working on a feature that would allow well-known to continue to be used for those who desire continuing to use the old behavior.
   3. So far nobody has taken up this ticket, probably because nobody who is in a position to work the ticket needs it (none of those present at the meeting currently makes use of blank nodes).  It is thus unclear whether there is really a need for this upgrade utility.
5. In-Flight tickets:
   1. One ticket (

      Jira
      server DuraSpace JIRA serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 key FCREPO-2436

      ) was resolved by Bethany Seeger.
   2. Work on a number of other in-flight tickets will resume with the import/export sprint.

...