Skip to content

May 11, 2016

Jump to bottom Edit New page
Nick Ruest edited this page May 11, 2016 · 9 revisions

Time/Place

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

Attendees

  • Nick Ruest
  • Melissa Anez 🌟
  • Bryan Brown
  • Don Richards
  • Jared Whiklo
  • Diego Pino
  • Christina Harlow
  • Ed Fugikawa
  • Marcus Barnes

Agenda

  1. Discuss: rdf namespaces and properties: dc, dcterms, schema.org and DPLA alignment
  1. #182
  1. May Sprint, 16th-27th
  1. ... (feel free to add agenda items)

Minutes

  1. Data modelling. Lots of different options. How to reconcile. This conversation started before LDCX, where Nick and Christina were in a group that also discussed the topic. Time to talk it out amongst the people of CLAW.

Diego: We've been working with PCDM for a while now. We've had some discussion about how close we can match our idea of what we need to PCDM, and if it really works for us. This is related to how much alignment really matters to PCDM. If we're going to RDF/linked-data road, not only is modelling a concern, but making our data understandable to other system. Also, making the way we display our assets as close as possible to the model. So IIF comes in there. For instance: ordering in PCDM is different from IIIF, which uses JSON. Much simpler to manage. Major concern right now is finding the best approach. Glad to have Christina here, since she has a lot of expertise on this.

Christina: LDCX session is linked in the agenda. Within that session you see detailed descriptions and diagrams, which link to a spreadsheet she started before LDCX. Omits IIIF, which she is aware of but not involved in. This gives us a place to start. You can see where people have started to use PCDM, now PCDMworks in the Hydra world. Some parts of PCDMworks will be integrated into PCDM. Not sure what the governance model is for PCDM. Or how these choices will be made.

Nick: There are a set of Committers, so if a PR is proposed it goes through review by them, such as here.

Christina: Time to get rid of the concept of "work" from Hydraworks because it's not being used for anything special. Work may be on its way out. Curious how we get all the descriptive metadata from all the various parts of the models. How to link PCDM objects to other domain models? Bibframe, etc? Rob Sanderson proposed pulling something from Europeana. They make more of a split between web resources and source resources. Are there questions for CLAW implementation of PCDM? Will you use some extension like Hydraworks?

Nick: A few weeks ago we were worried we weren't quite understanding what was going on with PCDMworks and making sure that both communities implement PCDM in a consistent manner so we keep the dream of interoperability alive. We're still trying to get to the point where we actually implement. We've done a little with PCDM Collections, but otherwise it's a lot of models and ideas.

Diego: Right now PCDM behaves more like a vocabulary than a real ontology.

Christina: Looking at NYPL implementation, they focus very much on the capture side of things. At the same time, University of York and Hull are trying to use PCDM for digital preservation coming out of Archivematica. Locally, with LDFL project, there are questions about whether PCDM can represent FRBR classes. Doesn't think we want to put all of these different models into PCDM, but there's noting in PCDM that says we can't. Her issue with PCDM is where do we expect to find these descriptive metadata points? Waiting for someone to try to use PCDM as a replacement for Bibframe to see what happens.

Jared: Thinks what we'd like to do is stay close to the PCDM vanilla so we have interoperability, because that's all PCDM really has to offer. We'd also like to add more actual data modelling to the ontology. Things that everyone says you have to do, but that aren't stated in the ontology. Things we understand as a community. We might end up with some sort of CLAWworks to show how we think it should work And then some discussion with other PCDM implementers to see what in the things we're asking for make sense. Our goal is to not just walk away with something new, but to keep going back and refining.

Christina: Hydra-in-a-box seems to be guiding some modelling discussions. Seeing their GitHub issues has been enlightening. For comparison, ld4l ontology.

Diego: Right now PCDM doesn't say anything about the types of objects. More complex concepts.

Christina: One consideration is how we connect this to meaningful concepts, and maybe Fedora isn't the place to do that. Not sure if Europeana is interesting here, but trying to parallel their approach: repo object, descriptive metadata domain object, etc. How many people have meaningful links between Islandora repos and their catalogues or local authorities?

Nick: Depends. If you are using Scholar and Entities, you can using local authorities using MADS. Stored in Islandora. We have an idea to use ATOM to create authorities and integration between that and Islandora.

Christina: At Cornell we're testing having a separate Vitro system (VIVO) to manage instances of context class. So if you wanted to have a bunch of FOAF agents and searches and link out to them, those uris would be pulled in.

Nick: at iCampFL there was a great discussion about what we do with MODS as we move to LDPland. Nick pushed to blow up the record and have everything be RDF. It would simply a lot in Drupal/fcrepo4(RDF) (no more XML forms). Drupal speaks RDF natively.

Christina: Doesn't think getting records isn't going to be as big a deal as we think. It requires the metadata to be so weird because we want to say things about things that are related to what we describe, not the thing we describe. We end up with a lot of misused fields. Once people realize that more granular descriptions will be allowed, it won't be so bad.

Nick: It's a challenging discussion because participation requires participation in a lot of different channels, which is too much fora very small group.

Christina: Some resources and documentation:

  1. Will be the focus of the sprint. One Repo to Rule Them All is not longer viable, so we need to decide (and document!) how to pull it all apart.

  2. Sprint: Next sprint runs May 16 - 27. Priorities: Restructuring and continued work on microservices. Check out milestone tickets. New sign ups are very welcome - if you are interested, you can chat with Nick and he'll get you set up and find a comfortable place for you to join in. Documentation, dev ops, discussion - lots of roles for this one.

You may be looking for the islandora-community wiki · new to islandora? · community calendar · interest groups · roadmap

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

⚠️ ARCHIVED Islandora Tech Calls

⚠️ ARCHIVED Islandora User Calls

Clone this wiki locally