- Published on
How I Learn Tech Stuff
- Authors
- Name
- Teddy Xinyuan Chen
Table of Contents
RTFD
Established technologies often have high quality docs online, I look for the following when getting to know it at the beginning:
- getting started guide / tutorial / interactive tutorial
- non-official tutorials / people's blog posts
Some docs are really arcane, for example, bash / zsh's doc. I tried to read them, but I guess I don't really need to read them, so I stopped.
Someone once told me I was reading docs religiously, and I agree to a degree.
Document browsers
Dash (with Alfred integration), Zeal, devdocs.io.
Get your hands dirty
Set up everything and run it.
But I don't have the hardware! And cloud computer is expensive!
If you need GPU, use your company's!
If you want to learn Kubernetes, you don't need to buy 10 computers, just use VMs or some similar drop-in replacement.
Find an excuse to use the new tech
This is one of the most effective way to actually learn something, also the hardest one.
I want to try to do this more.
Books
My favorite publisher is Manning. They're not as well known as O'Reilly, but their books' covers are definitely (way) better.
AI
Everything changed after the end of 2022. I now know 100 ways to split up a docs / epub / pdf and feed relavent content into the models, and that's really helpful.
Now I noticed my tendency of getting things to work without understanding them, and that's kinda bad. When I have capacity, I try to read the docs after getting things to work.
What about tar? I can never remember the arguments.
I use tealdeer
(Rust port of super fast tldr
cheatsheet) for that, so I don't need to worry about it.
tar
's command line args are just weird, jvns.ca wrote a blog post about it and most HN people would agree with me.