Talk for e-Rum 2020

[sgibson91]

Summary

This PR adds PDF slides for a talk at e-Rum 2020 to the repo.

• Related to #25
• Abstract: https://github.com/sgibson91/ssi-fellowship/blob/master/talks/abstracts/eRum2020.md
• Slides in pptx format: https://drive.google.com/file/d/1YdIcuYVRiYNZlJ1GIdp1iCtJpgDir0s0/view?usp=sharing

What's changed?

• talks/README.md
• Added talks/presentations/SGibson_eRum2020_Binder.pdf

What should a reviewer concetrate their feedback on?

• [ ] Everything looks ok?

[choldgraf]

quick thoughts!

• On the timeline plot, you may wanna crop the right side so the "drop" doesn't show up. That's just there because the month wasn't yet over.
• It gives a little bit more credit to eLife than is due, I think :-) basically they let us post a blog post, but then decided to make a contract with Stencila instead of developing with Binder
• I like your timeline-of-a-launch slide
• I think we'd be remiss not to mention the Rocker project as well as Holepunch in talking about R support, because those are both "R-community-driven solutions" to supporting R (as you mention in the slides, it's important to meet the community where they are rather than asking them to use conda for everything haha)

[betatim]

Nice slides!

It took me a moment to realise that the picture of "motivations" was a picture (from behind) of group of people linked together. Not sure why it took me a moment but at first I thought "towels? What do towels have to do this this??" (might tell you more about my mind than about the picture...)

Do you know or could you find out from people at eRum what packaging technology/approach renv uses? Is it based on CRAN packages or a different format or ...? The reason to dig into it a bit is that as long as we rely on CRAN packages it will be slow to install, because they need compiling from scratch on linux. On OSX (and Windows?) there are binaries, so they are quick to install but on linux not :(

If we want to solve the "slow to build" problem we need binary packages for linux. I've not yet been able to find someone in the R community who wants to setup the build infrastructure for doing this. It seems like the conda-forge community is the only one crazy enough to do this :D

So, learning more about what new tools like renv use under the hood would be interesting.

[sgibson91]

Thanks both! I've updated the slides incorporating your comments