It reminds me of my favorite thought experiment: imagining a world where organizations with more than 150 members (Dunbar’s number) are simply not allowed.

It’s an interesting exercise to think through what the world with such a rule would look like and what the gains and losses would be.

The above is really just the first approximation: many other factors can sway the decision in one way or the other. But I still think it’s a good starting point for making the choice.

If you’re not in such a situation, but still need speed, Go is usually a better choice (unlike Rust, it can be learned in a week). In all other cases, it’s probably best to go with the language your team feels most productive in.

The main issue is that Rust is quite complex and takes time to learn. I believe this complexity, which is a result of making a fast language secure, is warranted only in specific situations.

Your decision tree should look like this:

▪ If you need raw speed and absolute control over memory (e.g., can’t tolerate GC pauses and/or memory overhead), go with Rust.
▪ If you only need speed, but can live with GC, use Go.
▪ Otherwise, use Python, Node.js, or whatever else your team prefers.

Omlouvat se netřeba 🙂 Doufám, že jste si to užili.

Jasné, chápu. A díky za prvotní impulz k té přednášce (přenesený následně přes Janu)!

Po přednášce jsem mimochodem od účastníků několikrát slyšel stejný příběh o tom, jak se Rust začali učit, ale z různých důvodů toho po čase zanechali. Zdá se, že Rust vyžaduje větší motivaci než jiné programovací jazyky.

I feel the same, some of the joy is definitely gone for me. I’ve decided not to use AI for personal/learning projects precisely for this reason.