Coding 2021-11-20
I was tired today, so I picked some easy stuff to work on. I fixed some mypy errors, some flake8 errors, and got coverage up, but not all the way up. Then I rewrote the imports to be shorter, and this actually uncovered an issue that had been concealed by the previous format, so I think this has already been a positive change.
Of the remaining thirty lines to have coverage added, some of them need further design work to be able to write the code that produces the data that the functions need to execute properly.
So, for now, I'm going to do something I realized that I could do earlier today, and get rid of one of the actions that I don't need. It shouldn't seriously change things, but getting rid of it makes things make that much more sense. Basically, it was an artifact of how I was doing stuff in the noxfile that I derived the motrfile from, that I was doing coverage report after coverage html, and passing the --fail-under argument to the former instead of the latter. See, because I wanted the report to be at the bottom for when I scrolled up, but that meant it was the last thing to run, so it was the thing that needed to be failing if coverage was too low. But now that everything happens at once and I don't see the coverage report output ever, it makes more sense to put all of the responsibilities on coverage html.
So, I accomplished some stuff, and now I've got to put in the work to do stuff like enhance the "registry" types, and plan out things like "How do I specify different .coverage files, in the context of stuff like having multiple packages under test, and potentially using limit-coverage?". That's a good thing to look into for tomorrow, because I'm done for today.
Good night.