Improving Unix-fu with Anki

Anki has helped me maintain and improve my Chinese for years, but I’ve always been intrigued by using it to develop other kinds of knowledge. At university I used it heavily for East Asian history, our modern Chinese literature course and classical Chinese. I’ve now started using it to improve my Unix command-line fluency, and the results have been quite dramatic.

It’s not really possible to learn programming with an SRS system like Anki, but you can certainly use it to reinforce tokenized knowledge (like [this]( spaced-repetition/)). Whilst programming ability doesn’t tokenize very well, a lot of related computing skills certainly do.

The first thing I started reinforcing with Anki was bash and the Unix command- line in general. It’s easy to get this started with a simple rule: any time you use man or Google to learn about a command, make sure to add an Anki card for it once you’ve found the information you’re looking for. In this way, I’ve ended up with a lot of cards for the flags and options of various Unix commands. Man pages lend themselves very well to this, as do concise answers from the likes of Stack Overflow.

After that, I started seeking out more interesting techniques and approaches from sites like [Commandline Fu]( /sort-by-votes) and general searches for Unix one-liners. This quickly created a rabbit hole effect, and everything I found interesting ended up in my Anki deck. With daily reinforcement of an increasing range of commands, I’ve found that my Unix fluency has improved far faster than when I was trying to acquire the knowledge organically.

From that starting point, I now regularly add cards for every tool I use, such as Git, Vim and all the usual Unix suspects. As well as command line tools, SRS cards also lend themselves well to snippets of programming languages. SQL fits nicely into this model, and I’ve been able to commit various handy SQL phrases and techniques to memory and avoid searching for them the next time I need them. The amount of time I spend looking up documentation has gone down a lot, and I use that as a rough metric of how succesful this approach is.

So far I’ve used two types of card for this knowledge: a traditional prompt and response card type, and [cloze deletion]( deletion) cards. The prompt-reponse cards have some simple styling that suits code – monospace font, left alignment etc – and usually start with a one word topic to indicate where the knowledge applies (e.g. ‘bash’). That format covers the majority of commands, keyboard shortcuts, code snippets and language phrases, whilst the cloze deletion cards cover items of knowledge that work better as short sentences (such as this kind of guidance).

Since reinforcing my memory this way, I feel far more productive, particularly in bash and vim as those get the most constant use. I also feel more confident approaching new tools or languages, as I trust that I have this system backing me up as I learn them. Further, programming and computing books are now an even more valuable resource, as I have a realistic chance of retaining the information that I find useful or interesting as I read through them.

Tech mentioned