Hacker News new | ask | show | jobs
by jplayer01 2436 days ago
After experimenting with everything under the sun from Evernote to OneNote to TiddlyWiki and everything inbetween, I’ve settled on plain and simple files in a deep folder structure. The whole folder (now 80GB in size) is permanently kept in sync with SyncThing across my Android, laptop and desktop.

Using normal files allows me to store anything I need, whether it’s webpages as html files saved with SingleFile (FF extension), videos downloaded from YouTube, notes made with emacs orgmode, podcast MP3’s, eBook PDF's, etc.

Folders are deeply nested according to field/topic, and I have a git repo that ignores all non-org and non-html files. This lets me use ripgrep or emacs Helm to immediately search text for whatever I’m looking for. z allows me to traverse the tree without double-clicking through a deep tree of directories or cd'ing and typing crazy amounts.

So tools can be anything - Firefox with extensions, ripgrep, emacs, vim, git, z, or even whatever Python script I write to fill a unique use case that I discover for something that feels tedious. Normal files mean that if I find a cool program that does something useful with files, I can easily integrate that. I'm also working on ways to give me easier/quicker access to the metadata like the most recent files of a subtopic, or even add my own metadata like ratings and tags.

Ideally, something like Memex would provide some sort of api from which I could automatically query for all the browsing/history and text data, so I could potentially add it to my knowledge base in some way. Or maybe if Memex synced automatically to a DB file or some other file(s) that are well-documented that I could easily access, parse and sync.
1 comments
kevinslin 2433 days ago
reading your comment, I feel a sense of dejavu because its almost the same as the response I would have written. I've gone through a similar path of trying every knowledge base under the sun and settling on a text based, git managed custom knowledge base with plain text files (currently managing +10k notes this way).

currently building a service that can index and query across text based knowledge bases. you can find demo here: http://demo.alphacortex.io

would love to hear your thoughts and talk further about organizing knowledge :)

link