Feedback on new section 3

[svenvanderburg]

I think it is useful to share our experience teaching the new material in section 3. In summary: our participants (and we) really liked the content! It is super relevant for and directly applicable to participants' own work. So well done! 👍

There were also quite some rough edges, like broken links, missing guidance, pieces of information missing to do the exercises. I will create separate issues where applicable.

I am mostly concerned about people mentioning that the learning curve was perhaps too steep compared, I hope this will be better when we add a bit more guidance and have a better flow between the different parts.

👍 What went well?

• Much of this section is directly applicable to my project, so I took a lot of notes and am going to start applying these today!
• I finally understand what I can use classes for :) x4
• Everything was new for me, it was very very nice, I learnt a lot!x2
• This section really makes you think about how you can rewrite your own code in smaller, more understandable segments. Making it more intuitive for yourself and the eventual user.
• I enjoyed learning about how functions can be broken down into pure functions and why we do this; when I started to applying testing from last weeks lesson to my own code, I quickly saw how "inpure" my functions were, but since this wasn't introduced until this week, I didn't know how to describe this.
• The content is really relevant :) x4
• I already have quite some code, and this lesson has taught many ways on improving it (as opposed to writing new 'good' code from scratch).

👎 What can be improved?

• Part of this section required implementing tests which would automatically fail, as the functions they apply to were deliberately left unfinished or imperfect. This is not a problem by itself, but it caused some confusion as this was not mentioned in the course.
• Today's lesson was much harder. Almost feels like two weeks/sessions worth of content in one lesson.
• If you try running the analyse_data function it freezes. I guess this is part of a later excerise but it confused me a bit when i was trying to run it
• First URL in the https://carpentries-incubator.github.io/python-intermediate-development/32-software-architecture-design/index.html#reusable-patterns-of-architecture section doesn't work
• I feel like the learning curve skyrocketed today... many things were not clear but that should of course be fixable by studying more this week x2
• Quite a few typos and mistakes in today's lesson, which detracts from the learning experience x3
• It would be great to have some examples of research projects that have a Git repo that we can look at to understand how architecture design can be implemented in real-life examples that are closer to our work.
• A lot of reading in today's lesson, and not so many exercises - especially in the beginning x2
• Maybe not relevant, but I'm not really sure what the code in the main controller is doing with the args Parser and stuff like that, although it's becoming relevant as we're writing code that gets called there. So might be nice to have a small explanation at some point as to what's going on in there.
• I feel like this section made quite some jumps in reasoning, explanation or steps. For example not mentioning the expected error when running the test on analyse_data, not explaining inflammation-analysis a bit more, not telling how you could run inflammation-analysis through the terminal with the new command line parameter --full-data-analysis, etc. For me it caused more confusion than I anticipated.
• The section "Test Using a Mock Implementation" in the "Code Decoupling & Abstractions" chapter is a bit confusing and felt out of place. Especially since the chapter right after goes deeper into testing. For example, I wasn't sure where to put the test function for this, ended up adding it to the test_models.py file (next chapter we are instructed to make a test_compute_data.py file, which makes more sense, why not guide us to make that earlier?)
• Sometimes the code snippets require us to import specific packages, eg. pathlib.Path in excercise "Write Regression Tests". Only the solution code will show you the package. If you aren't familiar with common python packages it might be hard to know from the skeleton code what packages need to be imported.
• The links for extra episodes on procedural programming, functional programming and object-oriented programming are broken