Skip to content
Nick Ruest edited this page Nov 16, 2016 · 8 revisions

Time/Place

This meeting is a hybrid teleconference and IRC chat. Anyone is welcome to join. Here is the info:

Attendees

  • Nick Ruest
  • Jared Whiklo ⭐
  • Bryan Brown
  • Marcus Barnes
  • Amanda Lehman
  • Natkeeran Ledchumykanthan
  • Diego Pino
  • Melissa Anez
  • Danny Lamb
  • Jon Green
  • Martin Dow
  • Aaron Coburn

Agenda

  1. Metadata IG - follow-up
  2. Can a third person test this? https://github.com/Islandora-CLAW/CLAW/issues/411
  3. Another easy ticket to test: https://github.com/Islandora-CLAW/CLAW/issues/420
  4. https://github.com/Islandora-CLAW/CLAW/issues/421 Versioning, versioning, versioning!

Minutes

  1. Outcome/Follow-up of Metadata IG call.

    Not sure if anything solid came out of Metadata IG call. They are going to meet to get people energized and talking about the issues around CLAW. Will come back with more next month or possibly January. Some of the work that the IR IG is doing around finding common ontologies for scholarly predicates. So hopefully Bryan will attend future calls.

    Diego has a need of mapping MODS to RDF as there isn't currently. So if we accept MODS we still need to map to RDF.

    Steve Anderson has the Hydra MODS -> RDF group and was on the Metadata IG call. There is a lot of work have done a lot of work around these mappings. Amherst has done it, York is starting it.

    One of the big challenges is in Fedora 3 you just store MODS, very easy. For a lot of people the question is how do we store MODS in Fedora 4, but that is the wrong question. You want to be describing your resources, the ontology is less important. Separately from that you should be able to say using this resource how can I generate a MODS/METS/etc record for this resource.

    So each institution needs to know what information do you "really" need to have. Some of the metadata may be dropped if it is no longer appropriate.

    Some of the metadata that gets stuck in MODS because it is easy to edit, where Technical metadata would be better.

  2. Environmental problem, we need a third to test. Nat says he'll do it. Fairly simple test. Remember to rebuild your cache.

  3. Small bug in the Views UI which is not documented. Just to have someone test it. Simple and just need a test.

    Nick volunteered to test

  4. Versioning

    Open conversation about versioning and snapshotting and how we interact with Fedora through Drupal. The way Fedora4 does snapshotting is different from Fedora 3. Drupal has a defined way of doing versioning, and the question is do we want versions sync'd back to Drupal or can they live only in Fedora. They we just need someway to browse the versions from Fedora in realtime and revert on demand.

    Danny: It would be nice to get more input from people on versions. But we it is necessary to manage the complexity of the situation. Would be good to get some versioning use cases.

    Diego: Maybe a way to engage more end user (community) input is to move the discussion from APIs, etc to UI and how people want to use the software.

    Jon: It's sometimes hard to get use cases out of people. Which can make building and iterating the way to go.

    Nick: I see versions in an Islandora site like Glacier. They need to be there, but we want the front end of the Islandora site to be fast. So if you need to revert that will be a costly (in time) operation, and you'll need to allow for that.

    Danny: If we store all the versions information in Drupal then using Drupal versioning pre and post create call to Fedora.

    Diego: Right now versioning in the LDP tree, it does not only happen on the resource. But on the tree.

    Aaron: Versioning of sub-trees is truly bizarre, but look at what Memento is doing. The future should be that if you are versioning a resource you are only versioning that resource and not the entire sub-tree.

    Named versions that are created at a point of time. Remove that. If you have a resource, you should be able to fetch that resource as it was at a particular point of time. Then you use the Memento header to try to get that view (if it exists).

    Need the true separation of Fedora API Spec from JCR Spec and Modeshape's implementation.

    Please review the proposed Fedora API Specifications and make your voice heard. Even if it is just to say that you don't understand.

This is an archive. For new Tech Call notes, click here

⚠️ ARCHIVED Islandora Tech Calls

⚠️ ARCHIVED Islandora User Calls

Clone this wiki locally