Attendees

Agenda

  1. Updates/progress
  2. Recent API-X Workshops and Presentations
    1. DC Fedora users group March 22/23.  (Aaron Birkland): API-X update talk and hands-on workshop
    2. LDCX (Rick Johnson) quick talk Try API-X
  3. Demo 2 TODOs 
    1. Upgrade Amherst services to latest from Maven Central
    2. Use Amherst's named graph indexer for indexing objects.
    3. Write triple store demo
    4. Write package ingest demo
    5. Write intercepting extension demo 
    6. LDpath demo?

Minutes

  • JHU finished package ingest extension, Aaron Birkland has hot had time to write demo images.  The next 2-3 weeks are unpredictable as far as availability
  • Islandora has been advancing their Crayfish php microservices, they are just about ready to try deploying as API-X extensions.
    • Working on Mid-may deadline for upcoming Islandora conference
    • Islandora demos use Vagrant.  Porting to docker not on immediate agenda, but would make it easier to use, especially as part of API-X demo suite.
  • Islandora working on tesseract extension for OCR derivatives use cases
    • UMD has immediate use cases for this
    • Would be a great example to show the broader community
    • Microservice-architecture has been key to moving Islandora forward, aligns well with API-X 
  • Scalability important for tesseract use case - launch N services since it's resource intensive
    • High-level goal of APi-X to demonstrate this
    • Explicit IMLS grant outcome - JHU will be working with the docker images to demonstrate scaling in May, this aligns will with tesseract needs
  • FUG workshop went quite well, people were able to understand API-X by trying it and seeing what does
    • Attendees thought API-X makes Fedora easier to justify as a viable solution, especially by gov't agencies
    • Fedora itself doesn't have all necessary features out of the box, API-X is a formal way of extending its functionality so it meets relevant requirements.
  •  LDCX audience seemed generally not interested in topic of Fedora repository infrastructure, let along extending it.  
    • Engaging Hydra has been difficult; they're sort of heading in the opposite direction of API-X; consolidating functionality inside of hydra app, not using features such as messaging, using Fedora as a blob store.
    • API-X is really only of interest in microservices or repository-centric models
      • UMD, JHU, others planning in seeing if it's feasible to use Hydra apps as part of a broader repository-centric infra that includes API-X, some details may be problematic, e.g. WebAC differences.  
  • Action items:
    • Aaron Birkland to write demo exercises.  Next 2-3 weeks are iffy, so timeline cannot be predicted
    • Daniel Lamb and Islandora team working hard on finishing their crayfish services by mid-may.
  • Next API-X meeting May 11th
  • No labels