I just use a Git repository with plain text (and some custom marking up) files, as my knowledge repository, that also includes copy-pastes of the important info

I just use a Git repository with plain text (and some custom marking up) files, as my knowledge repository, that also includes copy-pastes of the important information that I found in the articles in question, thoughts etc. My custom markup consists of: - a number of '*' at the beginning of a line, for important things (more stars = more important; usually just 1 or 2, sometimes 3, if I'm super excited some more.) - '#' for comments from me between copy-pasted text - '[' .. ']' on separate lines to group things together, possibly nested (Emacs allows me to easily jump over them, and down or up nested groups). - '-' on a line between a copy-pasted comment and its reply; '- -' on a line between two comments (or sub-threads grouped with '['..']') on the same level. - '\SOME_UPCASED_CATEGORY_NAME' for marking something as belonging to that category I'm organizing the files into a hierarchy of folders. In case of ambiguity, I use symlinks. I've written a number of tools to handle such a repository well (search names by regex, search names and content by regex, cd into a directory matching a regex, etc.), currently mostly written in Perl[1]. I've recently started using Rust and will probably move those over to Rust over time. One tool I still want to write is for automatically updating symlinks when they break, by looking up moves in the Git history; I'll definitely write that one in Rust. It's got more than 17K entries now (including symlinks) after 7 years. The size of the .git folder after `git gc` is 51 MB. I synchronize it (simply via Git) between various machines. I use it to find information again pretty regularly. Even though I call the repository "bookmarks", I'm now also tracking projects in it (ideas, plans, sources etc.), as well as job postings and my handling of them. So the only bookmarks I keep in the browsers are for pages that I use regularly. [1] in my big pile of scripts, currently not well grouped for consumption: https://github.com/pflanze/chj-bin https://github.com/pflanze/chj-bin