CS in Review
A weekly roundup of my favorite papers/articles

The Software Ark: Issue 17

Mon 30 January 2023 / the-software-ark

Quote of the week

A year from now you may wish you had started today. — Karen Lamb



This is a brilliant article about how to debug an issue. And debugging is something that you can always get better at. It's why I love reading war stories about when something went wrong: to build that mental model of the sorts of things that can go wrong and how people figured that stuff out.

At the end of the day, debugging is all about constraining and validating hypotheses. You gain context, you pick what you think can go wrong, and you tweak it and see if that was actually it. The framework that Phil provides really frames the scientific method in a coding context.

And this applies beyond just debugging. The same approach applies to coding as well. It's why you always want to start with the hardest things: you'll either figure out you can't do it. Or you'll figure out you can, and the rest of the project feels more feasible. Very "fail fast".

This sort of work is not unique to coding. But it is more common when coding for sure. This takes up mental bandwidth, and you need some free time to think it through. Hence the push for a maker's schedule.

Some fun war-stories: - The heisenbug server - Debugging under fire

Open AI is scary

I used GPT3 to test out some webRTC stuff for a person project. Just to see if I could get started faster using it. Kinda scary how useful it was. It output code where all I had to do was change the import URLs, and debug a few errors. What would've taken me a few days took me 6 hours.

No wonder Big Tech is worried


Your moment of Zen

A satirical layoff letter