Comment by jdw64 | Hacker News Reader

Comment by jdw64 | Hacker News Reader

Parent Full thread

What is that emotional longing, exactly? I don't really know.

First, to be honest about my own feelings toward Rust: as you know, Rust's traits feel like a mix of Haskell's typeclasses and OOP, and that mashup of multiple languages just didn't click with me. I'm not a fan of solving compiler puzzles either. Especially when I've used AI to generate Rust code, it produced a lot of bad code relying heavily on clone, so it's not a language I'm particularly fond of.

In that sense, I do understand part of what you're saying. I suppose this is exactly the "emotional longing that isn't being satisfied by the technical reality" you mentioned.

So then why does the community keep holding Rust up as this symbol?

That's the hard part. Rust's promise is solving undefined behavior. But UB has already been largely solved by GC languages too. So what is it about Rust that pulls people in? Is it because it replaces C and C++, the oldest legacy in programming? Or is it because it's hard for a new superstar to emerge within the legacy that C and C++ created, so people are drawn to Rust as a fresh language? I really don't know. It's a tough question.

The last time I remember seeing the same kind of behaviour was when document databases appeared on the scene. Content creators far and wide were writing about MongoDB vs Postgres, stories about rewriting an app to be "web scale", and whatnot. It grabbed reader attention because a lot of people were feeling unsatisfied with products like Postgres (this was before it added a lot of the features we expect today), MongoDB solved a problem for a niche set of users, and so the race was on to read everything they could to try and convince themselves that MongoDB would also solve the problems they have, even where their problems were not within the small area where document databases are appropriate.

The most likely reason for why people are drawn to want to read about Rust for much the same reason. Note again that they are spending their time wanting to read about it, which is different than using it. Rust solved a real problem for a niche set of users. Now everyone who has problems with their existing technology stacks are trying to read all they can to try and convince themselves that Rust will also solve their problems. The heart lusts after Rust on the promise that it will solve all problems, but the mind knows it won't solve all problems. Attention lies at the intersection of those competing interests.

Rust is a big deal, but that isn't a big deal. MongoDB was also a big deal. Look at where it is at now. The good news for would-be language designers is that there is a strong signal in the market begging for a solution to their problems. The only question is whether you can deliver before the incumbents catch up (see: modern Postgres).

Thank you, senior programmer, for sharing your valuable perspective. I'm Korean, so I've mostly studied Western culture through the curation of specific figures, which means I only knew broadly that there was a conflict between Postgres and MongoDB but I never knew the details like you do. In other words, for me it feels like reading a history book, whereas for you it feels like lived experience, so I imagine our perceptions are quite different.

'The heart lusts after Rust on the promise that it will solve all problems, but the mind knows it won't solve all problems.' I really like that sentence.

Personally, combining your thoughts with mine, I think this is also a matter of community belonging. In other words, I don't think the issue with Rust is that it's a solved problem. Rather, I think it's a process of burying anxiety about careers and professional uncertainty into community voices, as a way to project that unease. Learning a new language and all the libraries and frameworks tied to it is very demanding, and internalizing the conventions of a language takes time. So it becomes a question like, 'What if the skills I've invested so much in are suddenly no longer relevant?'

And while Rust's approach to problem-solving is attractive, as both you and I know, no single language can solve every problem. After all, every language has its own trade-offs and subsets. As you go lower-level, cognitive load increases significantly, which is why high-level programs are often written in low-level engines and scripts in high-level languages.

Anyway, I thought this place, where the world's best programmers gather, would be different, but I'm realizing that most programmers are quite similar.

Thanks for your thoughtful input. I hope I haven't taken up too much of your time. After all, this kind of question isn't usually encouraged in communities, and it's a difficult one to answer. Everyone thinks differently. But your explanation made the most sense to me. Have a great day.

Reply on news.ycombinator.com