Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  1. Please squash a bug!

    Expand

    Jira
    serverDuraSpace JIRA
    columnskey,summary,type,created,updated,due,assignee,reporter,priority,status,resolution
    maximumIssues20
    jqlQueryproject = FCREPO AND issuetype = Bug AND status in (Open, Received)
    serverIdc815ca92-fd23-34c2-8fe3-956808caf8c5

  2. Tickets resolved this week:

    Expand

    Jira
    serverDuraSpace JIRA
    columnskey,summary,type,created,updated,due,assignee,reporter,priority,status,resolution
    maximumIssues40
    jqlQueryproject = FCREPO AND status in (Closed, "In Review") AND created >= 2016-05-05 AND created <= 2016-05-12
    serverIdc815ca92-fd23-34c2-8fe3-956808caf8c5

  3. Tickets created this week:

    Expand
    Jira
    serverDuraSpace JIRA
    columnskey,summary,type,created,updated,due,assignee,reporter,priority,status,resolution
    maximumIssues40
    jqlQueryproject = FCREPO AND created >= 2016-05-05 AND created <= 2016-05-12
    serverIdc815ca92-fd23-34c2-8fe3-956808caf8c5

Minutes

  1. ModeShape5 update. There is a branch that is available and ready for community testing. The branch appears to be ready to be used, though it will not be merged into master until after a particular Modeshape5 bug is fixed in their 5.1.0 release.

    1. Modeshape5 clustering has been demonstrated to work, but with caveats (it uses a single central database for objects and shared filesystem/database for binaries). Federation-related bug has been resolved with Modeshape5.

    2.  There will need to be tooling for migrating content, possibly with fcr:backup and fcr:restore or with some other utility.

    3. Release timeframe for Modeshape 5.1.0: it appears to be scheduled for mid July. There is keen interest from the Hydra community with the release of Sufia7 w/r/t minimizing content migrations. With a three-week release candidate cycle, a Fedora-on-Modeshape5 release wouldn't happen until August at the earliest.

    4. Clear communication and good tooling will be important.

    5. There is interest in using S3 as a shared binary store in a clustered context. Bill Branan is working on a modeshape-based binary store connector supporting S3 integration. This may also be very useful for those interested in clustering.

  2.  An RdfSource that is not a NonRdfSourceDescription currently describes itself. In the java code, the NonRdfSource class has a getDescription() method. By moving that method into the FedoraResource class, it could tidy up the code considerably (e.g. for example, this snippet would be much shorter). It may also open up possibilities for specialized features related to provenance. There were no objections.

  3. Open Repositories tech meeting: if you plan to attend, please add your name to the agenda. Everyone is encouraged to add agenda items. The focus will likely be related to the API specification, Fedora data migrations, etc.
  4. Message serialization format (see proposal). At some point it would be good to align the audit module's use of PREMIS and this use of PROV. Given that audit and messaging have similar but ultimately different goals, it may not be possible. The proposal uses JSON-LD, and if we use that serialization, we need to ensure that this JSON-LD must be parseable by real JSON-LD parsers. Everyone is encouraged to review and comment on this proposal.
  5. Filesystem Federation Use Case & deprecation from core – what are the reasons for deprecation?
    1. In the process of specifying the Fedora API, this is not something that would be part of that specification; that is, alternate Fedora implementations would not necessarily implement this feature.
    2. In the context of digital preservation, when a Binary is "in" Fedora, it doesn't necessarily mean that Fedora is managing that Binary (e.g. via external content or the filesystem-connector).
    3. Filesystem federation has been seen to be unstable at scale
    4. Penn Use Case is to use the filesystem metadata dynamically (e.g. preserving access controls), which isn't addressed by the external message body construct.
      1. API-X may be an answer to this.
      2. More to be discussed next week
  6. Multi-tenancy - needs from Hydra-in-a-box: please read and comment on the linked document. This will be discussed next week, too.