On September 15th, the mentoring subcommittee held the 18th round of debriefing for instructors. At this round, we have a great discussion about teaching (or not) branches during the Git lesson and about Python packages used in examples.
Can we hit zero installation issues?
Instructors reported that they still encountering some installation issues during the workshops that they or the helpers solve in a couple of minutes. The most common are:
- nano was not installed, had to use sublime text or notepad
- old version of OS X, Windows or a GNU/Linux distribution
- wrong version of Python (that should be reduced know that Continuum Analytics has updated their download page).
Along this lines,
there is also the problem of
man not be installed in Git Bash on Windows.
In one of the workshops, instructors had problem due cyrilic characters.
Learners from one of the workshops suggested in their feedback that instructors provide a cheat sheet at the begin of the workshop so the learners could consult what they learn easily.
The "Reference" part of our lessons is close to a cheat sheet and can be provide to students.
You can print the "Reference" directly from your web browser since it has a special special style sheet for printers.
Jupyter Notebook and scroll
Instructors reported that they are having problems to show enough code for learners when using Jupyter Notebook, with code scrolling above the top of the screen. Unfortunately we didn't settle on a good solution to this but is was helpful having a helper translate all the code onto the etherpad.
We discussed if we should or shouldn't teach branching during the Git lesson. Branch is a powerful tool that enables one to experiment with different changes, while avoiding the problem of different copies of the files. Unfortunately, it is a lot for learners to take in and understand during the workshop and it can confound some beginners. However, it should also be noted that git made the made the eyes of the more experienced shell users in the group light up quite spectacularly.
Packages for examples
We also discussed that some packages are more suitable for some audiences. For example, Pandas is more suitable for learners that need to do data analysis and NumPy is more suitable for learners that need to process/create data.