Skip to content

Latest commit

 

History

History
77 lines (54 loc) · 3.34 KB

2018-11-28-minutes.md

File metadata and controls

77 lines (54 loc) · 3.34 KB
Raw

Attendees

  • Raniere Silva
  • Tracy Teal
  • Kate Hertweck
  • Maxim Belkin
  • Sarah Brown
  • Francois Michonneau
  • Naupaka Zimmerman

Notes

Discussion on draft of lessondown feature proposal

(also realtime edits to file above)

  • will be a first step prior to other rmarkdown

  • additionnally, generate html (potentially replacing jekyll)

  • potential problems in running and autogenerating solutions for shell and git, but should work for R & Python

  • some downside to automated output whenit is very long

  • concern about how people can render lessons locally

  • people will not need to use travis but it wil require travis setup for each repository

  • potentially adding complexity in setup , because more components

  • Add explanations of why features are important

  • Evaluate dependencies of things like Ruby, Jekyll, RMarkdown, and make it easy for maintainers to install them. Consider not only ease, but also size of files.

    • R relatively easy to install and maintain, and bundle dependecies

  • Concern about R could be that we don't have enough features that we need.

  • If you add lots of complexity, then it makes it less accessible. For instance if you need to do Travis things, would make it less accessible.

  • Killer feature' from Raniere. Manage to implement something that shows where your syntax error is. Like in LaTex that shows you where to change that makes it so it doesn't compile.

Updates on Delivering Exercise Discussion session by Sarah

  • 14 people signed up for the meeting

  • Had some discussion about what other tools people were using. Sarah has notes from the session. Many of them aren't free or open source, so not sure how feasible, but worth looking at them.

  • Lots of support for being able to use the exercises in different ways in general. Good community support for this idea.

  • Marianne mentioned she thought that exercises were in separate files. More support for leaving them in the main lesson as we've discussed here, and then programmaticly extracting them.

  • Talked about people using them in a slide format. Render in large text format, so displayed on a projector.

  • Output in a format that's consistent with Socrative.

Update on translation infrastructure

  • main proposal is from David. He'll provide an update in the next week or so, and will be updates on the github issue

  • Still plans to try it on a lesson or two. Francois will coordinate with David.

  • At SuperComputing, Maxim saw another way of swapping out elements of lessons and keeping things in sync

  • A group in Europe working on translations using Google translate

  • Not yet much documenation of process around translation infrastructure

Francois and David talked about the translation infrastructure. They are going to recollect next week and see what they can do in the next few months.

At this point, most of the work is on manual translation. During BioHackthon in Paris, France, people played with machined translation. The question is how well can we train th machine and annotated the documentation to avoid some works being translated, e.g. "cat" the command and the pet.

Rayna mentioned that at the time that she try, there were not much documentation and was challenge to do the translation. The learnign curve of some tools aren't smooth.