A great feedback

All engineering is reverse engineering if you document things poorly enough.

I also agree: if you act alone on your free time, then go for simplicity. You have a limited time budget and it should be fun.

Trying new ones when they are needed makes also sense, but only when they make sense.

The article is well written and connects multiple topics: line of code and care work to the software, computer architecture and speed, its industry and more.

Debian 12, for comparison, is 1,341,564,204 lines of code. For comparison, Google Chrome is about 40 million lines, which is in the same ballpark as the Linux kernel these days. No one, even a team, can read these entirely.

Computers aren't much faster now than they were a decade ago, and they will probably never again return to the rate of performance improvement they had for 60 years up to the mid-noughties.

The thing is, that doesn't scale very well. On the desktop we have four-core machines and now we're moving to eight-plus cores, but a single person can't use that very helpfully, so instead, we're getting computers with a mixture of high-performance but hot, power-hungry cores, and lower-performance, cooler, but more electrically-efficient cores.

A limit to multiple cores is the Amdahl's law: even if a program can be made 95 per cent parallel, the maximum speedup you can get is about 20 times, no matter how many processor cores you throw at it.

What is the maintainable way to build things?

For CSS, how do you structure tokens?

1. Naming things: --umap-color-darkBlue?
2. Give a meaning to names: --color-primary: var(--umap-color-darkBlue);

Lorsqu’on envisage un commun sur ces 10 prochaines années, comment trouver une stratégie maintenable qui s’inscrira dans la durée avec enthousiasme ?

Le contraire de Tailwind :/

In honor of Valentine’s Day, here’s a thread of pickup lines for the programmer in your life♥️

1. Are you a CLI benchmarking tool? Because you’re looking hyperfine today
2. Do you write Rust? Because you clearly care about my safety🦀
3. Do you write JavaScript? Because I bet you know what this could mean
4. Has anyone ever called you pnpm? Because you’re my favorite package manager 📦
5. Excuse me, is your name @vite
6. Because you’re the best development I’ve seen in years.👀
7. Let’s trade—you buy me dinner and I’ll buy you a new domain name.💰

Parsing increases the information in the type system. A list can be of type NonEmpty, i.e. there is at least one element.

Use a data structure that makes illegal states unrepresentable.
Push the burden of proof upward as far as possible, but no further.

and awesome guidelines to follow.

So parse "data" and return the closest type instead of only validate them.

Un retour sur un dev

How a startup serving files can make this wrong

Wizard programmers prefer up-front design. They apply reason and logic to divide and conquer a large problem, they rely on building blocks like design patterns and algorithms. Wizards rely on explicit knowledge.

D&D sorcerers have an innate connection to the magic. They wield tremendous forces that they sometimes don’t quite understand, it’s wild and unpredictable. It’s not something they’ve learned to do, but something they’ve discovered in themselves, a talent.

I like the comparison because it seems right.

It’s a massive differentiator to have launched something. To have had an idea, and put it into the world.

I think it’s worth noting that there’s a certain luxury of time that affords side projects that comes wrapped up with privilege. [...] So it can be very small.

Classic.

If you’re writing code for people to use, and you’re working on it with people, not having people skills isn’t really an option.

It's the dev behind atuin 👏

And she's going full time on it https://ellie.wtf/posts/i-quit-my-job-to-work-full-time-on-my-open-source-project 🚀

It ensures the expected meta tags (OG and others) are set as expected.

“The best time to start a weblog was 20 years ago. The second best time is now.”

These bugs highlight the risk of unknown unknowns, i.e. stability issues in our stability monitoring itself that we are blind to