I learned quite a bit from that book. I think Flanders may still have a site. I was on his mailing list, but I haven’t heard anything for the last decade or so.
Nowadays everything is so optimized and efficient, I've become nostalgic for the days when webpages sometimes sucked. At least they had personality, even if they were hard to use. It's like cars, I like looking at super old old cars in museums and wondering what all those pedals and levers do, even if I'm happy to not drive them.
I'm always wary of people who spend too much time in the world of inspirational books. It's healthy to read a little, and to "sharpen the ax" every so often, but reading too much of this stuff is mind-numbing outside of historiography reasons.
Deleted Comment
Apple Intelligence isn't it - it's just playing catch-up with a market that tries to slap AI onto everything it can think of.
The hardware upgrades are always nice but there's nothing 'out there' like a touch bar or even a 'dynamic island'. Just more safe iterations.
As a result, most of the non-memory issues listed here are things I solved once in those scripts and rarely have to think about again. Most of the simple memory issues can be solved by running `cheat x` and/or `tldr x`, where x = the program I'm struggling to remember the basic usecase of, both installed by those scripts of course. Most of the more obscure memory issues I can deal with with either ChatGPT or Anki these days.
I originally considered writing my scripts as a `host: localhost` Ansible playbook instead, but then I remembered that whatever I might gain in idempotency I would lose tenfold in hackability. I am guaranteed to be doing shell for the rest of my life; I am decidedly not guaranteed to be doing Ansible, or any other config management tool for that matter.