Spaced repetition has been around for a long time. If you’ve never heard the term before, it’s best described as flashcards on timers. an algorithm such as SM-2 or the more recent FSRS keeps track of how you did on the flashcard, makes a guess as to how long it could possibly wait to show you the flashcard again before you below, say, a 90% chance of getting it right the next time, and then schedules the flashcard for a new day. ...
A while back I stirred up some controversy on Hacker News by talking about why I liked it when tutorials take you from clean VM to working, installed software. I’ve since taken to calling this the “tutorial-in-a-box” method. When I write them myself, I usually put them under the header Slowstart, a riff on the proverbial Quickstart. Two examples: A gentle introduction to reposurgeon. The Slowstart for selkokortti, some flashcard generating software based around my Finnish language news archive. The point of a Slowstart is to make it so that even absolute beginners can start to pick up some pointers about how people “in the know” of your chosen software ecosystem actually get things done. Instead of a Dockerfile or a shell script, you take them by the hand, spin up a totally fresh virtual machine from the ground up using something like Vagrant or virt-builder, and walk them through each and every command they need to execute in order to get to a working install. ...
There are few things I think about more than the essays on gwern.net, and there are few with as satisfying a theoretical payout to contemplate in my orb as his essay on “leaky pipelines”, aka log-normal distributions. The skulk: Say you’re working on a Laravel web app. You’re about 90% sure you know how to start the app. You’re 80% sure you know how to handle the infra you’ll need to get it online. And you’re 70% sure you know how to get your first customer. What is your chance of successfully going from zero to first customer? 0.9 * 0.8 * 0.7 = a little over 0.5. That’s … a lot less encouraging than any of the previous numbers, if you buy my multi-step modelling. ...
I Robert Anton Wilson was, is, and always will be a fascinating and hiliarious writer to me. I first read The Illuminatus! Trilogy when I was 13, and while it was coincident with a total and suffocating blackout of meaning, I no longer think reading it actually caused that to happen in any significant sense. Au contraire: Teen me found refuge in his absurdity - it felt bedrock nihilstic, sure, but a far more artfully and deeply buried nihilism than I was able to find elsewhere at the time. (My words, not his. RAW wouldn’t describe himself as anywhere close to a nihilist. I think the glove fits.) ...
1 wall "Hello, world!" will send a message that looks like 1 2 3 Broadcast message from root@localhost (pts/0) (Sat Feb 1 14:50:14 2024): Hello world! to every user currently logged in to the system. When might this be relevant? When you’re working in a small team, by remoting into custom hardware, and constantly reflashing the device, to give everyone a heads up, for instance: ...
A communication heuristic, optimized for asynchronous communication of detailed concepts between human beings. Not really a TIL, I’ve been sitting on this one for a while. Rationale Existence “proof” Our short term memory can hold at best about 7 +/- 2 chunks of information at a time. Start with the lower end of that: 5 chunks. Assume that whatever someone is “really” trying to do takes up 1 of those 5. We’re left with 4 chunks. Conclusion Organize your work processes as much as possible so that they can be effectively understood using at most 4 chunks of short-term memory. If you feel a process is too complicated to be held like that, find conceptual fault lines to hack against until it’s true. ...