My https://til.simonwillison.net/til site acts as a sort of off-site backup - but I should probably run a cron somewhere which pulls my repos. I have a Digital Ocean box I could do that on. Good idea!
Nice! Like the idea of tracking TILs with git. I've also made self-rewriting READMEs. I did something similar but using plain git hooks (post-commit) and sed. I had to run 'git --amend' in the hook though (to avoid generating double-commits) which is probably a bad idea.
Author here. I was really proud of this piece - it's a very detailed tutorial on how to use GitHub Actions, which for me fulfills exactly the kind of content I want to see on Hacker News.
It's the article I wish I had been able to read when I started working on this project.
Also inspired by the same post, I set up a TIL repo. I went with a shell script where you pass the category and the filename and it creates the file and recreates the README:
https://github.com/acdibble/til/blob/master/create_til.sh
Thanks for sharing. I also recently started a TIL repo from the same inspiration and found it lame to write my first README entry. Will look at setting up what you described.
If there was a public place for posting these with peers, that was built for TILs rather than just a repo, would you use it? Especially if it included an easy way to export all to markdown for backup?
This would allow folks to see each other’s TILs, search them, browse by tag, like/favorite each other’s TILs, and some other fun stuff eventually
Actually that's wrong, in the morning I convinced myself that given a sequence of code points which forms a valid Unicode character (an extended grapheme cluster), any non-empty prefix of it also forms a valid Unicode character.
