search

temporalecologylab / labgit

Where we keep general (non-methods, non-private) info. Look here for lab meetings, expectations, data management and so much more!
5 stars 3 forks source link

discuss GitHub (and git) #16

Open lizzieinvancouver opened 6 months ago

lizzieinvancouver commented 6 months ago

Let's keep a list of things to discuss about git and github at a lab meeting this summer!

  1. @DeirdreLoughnan already gave a good one: how would you make a repo public that uses a combination of our own and conditionally shared data?
  2. When to make a repo private.
  3. Things you might worry about with public repos (please reply!)
  4. Things to worry about with private repos.

From @DeirdreLoughnan regarding (1):

In thinking about this yesterday, my plan for my Synchrony meta-analysis repo, which includes data we had to request that is not openly shared, is to remove any raw data files that have the phenology data from those sources. But to leave data/input files that have summary statistics based on that data, such as saved files with slopes or intercepts. I was only going to leave in the repo the data that I scraped from papers or downloaded from open sources like Dryad.

I should probably update the data management policy also to outline what I prefer (and sometimes require) and why ... This includes:

  1. Using terminal instead of a GUI (because you learn some basic coding, GUIs obscure what is happening and I see this lead to issues, and if you have an actual issue in a GUI it usually just sends you to terminal and then you are in terminal because you have a problem, but lost -- a bad place to be)
  2. Making repos public when possible (why? To be open about our science, to pay it forward and back to other lab members, so we can all see each others's repos -- especially me, who sometimes goes back years later to find them without know the name and if they are private I will never-ever find them and this is bad for the lab).
lizzieinvancouver commented 6 months ago

Also:

  1. How to handle using GitHub to make code public for papers etc.
christophe-rd commented 6 months ago

I am adding @FrederikBaumgarten's questions that we could cover:

  1. Does this mean, anyone can download the current state of any manuscript you are working on?

  2. How would you deal if someone uses your dataset before you published the first paper out of it?

lizzieinvancouver commented 6 months ago

Thanks @christophe-rd for adding @FrederikBaumgarten queries!

If we have time I would love to think about how to organize our git resources better too. See also issue #17