2019-11-18: Development Planning

Craig, Mike L., Kacper

  • v0.9

    • User settings
      • DataONE + Zenodo
    • Register data from Zenodo (merged)
    • Publish to Zenodo (review)
    • Publish refactor (UI)
    • Run/import from Zenodo
      • Detect whether tale already exists (don’t import blindly)
  • Three different things

    • Registering a DOI, even if it is a tale, will always come in as data
    • For DOI that is not a real tale
      • asTale = true puts data in workspace
      • asTale = false puts in /data
    • For DOI that is actually a real tale
      • There is no asTale
      • It’s just an import
      • If I really want to use your tale as immutable data, I don’t want a zipfile (probably)
  • User flow

    • You found a tale in Zenodo and I have the DOI

      1. I want to run the tale as a tale
      • You don’t need the modal, you just do it
      1. I want to analyze the tale as data using Jupyter
      • Registering the DOI today (for Zenodo) would give you a zipfile
  • Discussion

    • Detecting a “binder” in Zenodo (it’s a zip) versus a tale (it’s a bag)
    • Differences in importing tales that are bags (Zenodo) v bags that are not tales (DataONE) v tales that are not bags at all?
      • DataONE has fairly specific dataone metadata that can be used to tell it’s dataone
  • How would we import from DataONE?

    • Integration converts to BagIt or Zip
    • Import could allow provider to specify the job
  • Can I publish a non-public tale?

    • All published tales must be public?