2018-10-01: Development meeting

Kacper, Ian, Tommy, Adam, Damian, Sebastian, Bryce, Craig, Mike

Updates

  • Kacper:
    • Finalized development environment PR
      • Can’t configure local source, but can use your own image (tedious but functional)
    • Worked on design and plan docs with Craig
    • DMS is deployed to dev
  • Craig:
    • Design and planning documents
  • Damian, Sebastian, Adam:
    • No tasks in this sprint
    • Need 1-week heads-up to incorporate work into sprints (probably 2 weeks away)
    • Sprints are organized on Monday mornings
  • Ian:
  • Mike:
    • Q. How are we on the DMS configuration? Seems to be working fine. Encourage people to run previous tales.
    • Initial version done
    • Next – finishing Globus integration
      • Recursively searching Globus for total size of dataset. Can we skip it?
      • KK: It’s not important, just displayed in the widget
      • CW Q. Do we always need to register the entire dataset?
        • MH: DOI is associated with the whole dataset, but within the whole dataset, we refer to individual files.
        • We could do a lazy import – you don’t have the whole size, but you could lazily browse through the source.
        • Today, we import whole datasets.
      • KK: Dynamically browsing endpoints/datasets would be a significant project.
        • Focus on registering the entire dataset now, dynamic browsing after 36 month review
  • Tommy:
    • Connected the job updates to the UI
    • Down to one broken unit test for adding the DOI field to the Tale (part of backgrounding the modal task)
    • Implemented the progress bar-down to adding some styling to it (http://recordit.co/y9P6l5wmG9)
    • Granularizing the file upload progress message
    • Deploy WT stack on probable-cattle and configure GitHub repo/flow to accomidate for this (making dev2 point to the publishing branches on the WT repos, and probable cattle point to my personal repos)
    • Comments on 36 months:
      • Review of deliverables: publishing, adding/removing data from the tale, workspaces, custom environments
      • KK: For publishing, you would get data from the session ID not the involatile data
  • Bryce:
    • RO paper
    • Will update tale specification in readthedocs