Doing code archeology to figure out how obscure the bugs I was hitting were.
This post was delayed thanks to: something in my Neovim install or environment that jacked up the file recovery process somehow.
"Most people can learn how to kill someone barehanded, but it's not really practical, so this doesn't change much."
I mean, it'd be weird to have Clydesdales without a Clydesdale.
It's kind of hard to read the output from the nameless-actions version of MOTR, but, like, oh well.
I have no idea what flit build was doing when I had uncommitted work, but I extremely did not like it, and it makes me wonder if I can trust flit build output in general during development. (Maybe it was because I wasn't bothering to pass --wheel, so it was making the wheel from the sdist instead of directly? Eugh.)
A little bit slower work than I would have liked, but there was some stuff in there where I needed to be methodical.
"I cannot think of any problems with this. [shortly thereafter] I have found a problem."
This does involve a bunch more I/O than the noxfile, but I don't think it's a perceptible difference.
After this entry was mostly written, I began to fail to avoid grappling with technology.
« Page 30 / 154 »