datalad-handbook / datalad-course Goto Github PK

@mih and I will be going to Dresden to teach a 2-half day DataLad workshop.

Important infos at a glance:

• Slides are created in this repo, branch ws-mpsc
• Audience response system (for MC questions or polls) at https://www.directpoll.com/c?XDVhEt8n6fLejElM1ruTg4PqVxQiKbA
• Zenodo deposition for slides will be at https://zenodo.org/deposit/6827086
• Attendee information is at https://docs.google.com/forms/d/e/1FAIpQLSd3V1v4gy0jXbpc58s7VttuJO3Fr4wOJRr8Tl2LLy7wWDJpVw/viewform
  • barely anyone used datalad
  • 2/3 used Git
  • of those that used Git, less than half did so collaboratively
  • 50% did not share code or data yet, 35% shared both code and data already

TODOs:

• fine-tune content and exercises
  • @mih adds "Demo of core features" to the slides at mpsc-introduction.html
  • @mih adds "Collaboration" part to the slides at mpsc-datapublication.html
  • datalad ML example tested
  • datalad create-sibling-osf example tested
  • datalad create-sibling-gin example tested
  • datalad create-sibling-webdav example tested
  • @mih completes block 2 start exercises
  • @mih contributes "outro" slides
  • @mih core feature set demo compiled and tested. We will use http://handbook.datalad.org/en/latest/code_from_chapters/yale.html#using-published-datasets and prepend the final rerun with a drop of the brainmask, a failing get, and a parameter-less rerun, which will bring the file back, albeit in a different version. We can remove the compression in the script to make it reproducible (and show that).
  • @mih release datalad-next with configuration update
• get train tickets
• Set up a Jupyterhub, map to datalad-hub.inm7.de
• Create around 20 user accounts; we don't have participant names, we can use random words, print them out, and hand them to people as their username
• Bring a stash of post-its for participants to indicate support requsts
• include some charts from attendee survey (e.g., Git usage) into the slides
• Deposit PDFs of slides on Zenodo and publish the draft
• Render slower version of webdav walk-through and post on youtube