Page History

...

1. Announcements
   1. Fedora API specification was just released.
      
      1. Hoping in the February timeframe to have 2 verified implementations with the test suite.
      2. What are the 2 implementations might be, the Modeshape one, the other is possibly Cavendish there is a Python implementation that Simeon Warner is working on.
      3. Esmero is a possible PHP implementation, not much movement lately.
      4. Northwest Northwestern was doing implementation in Python with an AWS lambda backend but had to set it aside.
      5. Derby is another possible contender to implement the API.
      6. Lambdora - an AWS fedora 
      7. Trellis - supports all of the underlying specs, but may not conform to all of the fedora-specific requirements
   2. Amazon now supports a graphstore that support SPARQL query.
   3. Call tomorrow at 11:00 ET for the Oxford Common Filesystem System.
2. Activities
   1. Postgres clustering configuration, Andrew spoke with Rob Kaufman and he has implemented this at UCSD. He says it supports multiple masters for writes, which would be very interesting. Can we get someone to stand-up a cluster and test this. Carrick Rogers might have some resources to stand up some instances to test this.
   2. Anyone going to Code4Lib on helping Esmé Cowles with a workshop. Michael Durbin will be there and can help out.
   3. Digital Initiatives Symposium needs someone to help facilitate a Fedora user group meeting, Danny Bernstein will be there he thinks, Seth Shaw is going for Islandora and might be able to help.
   4. Fedora exhibitor at the Southern Miss IR conference, do we have anyone to do this and/or possibly a proposal for the topics? Anyone in the area? Contact David Wilcox if you might be interested.
   5. There is a Fedora slack channel, but there is a pattern for creating an auto-invite. Samvera has setup this type of work. Declan or Mike Giarlo have set it up or are maintaining it.
3. 4.7.5 Release
   1. Shooting for a release in January. Any reason not to do it now? Just the time of year?
   2. Any other bug fixes?
   3. Any volunteer to do the release?
4. Single Subject Restriction
   1. Would this allow you to persist subjects that are not within their repository.
   2. Relaxing the single subject could allow several bad patterns, all descriptions on one resource for example.
   3. They want to have various stuff in their resource, some is of a real world object that is published elsewhere.
   4. Is it Fedora's job to restrict people to keep them from making bad decisions.
   5. There are easy ways to do this by persisting these statements in a serialized XML.
   6. If we are going to maintain the SSR then we need to ensure that we well document the principle that leads to that decision.
   7. In regards of named graph provenance you might want separate audit statements held in different places.
      1. a resource in the repository is a named graph, and it would be plausible that a resource could contain statements that describe the provenance of other resources in the repository.
      2. If you want to manage the above as a separate resource, that could not be done well.
      3. Storing ontologies as a RDF source is not able to be stored in the current repo based on SSRs and blank nodes.
   8. No ideological attachment, but there might be real world reasons for having it.
5. 5.0.0 release
   1. Have a look and please add any notes you might have for future discussion
6. Beyond 5.0.0 - Areas for improvement
   1. Have a look and please add any notes you might have for future discussion