Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I love building a startup in Rust but wouldn't pick it again

propelauth.com

Post reply on HN
Print view View original

141–150 of 496 posts

Re: I love building a startup in Rust but wouldn't pick it again

#141

Post by cneu »

synergy20 wrote: post #83
pjmlp wrote: post #62

Earlier quoted context omitted.

Nowadays? That is how AOLServer used to be, and all the other scripting languages developed as Apache plugins, back in the 2000's .com wave, like mod_perl and PHP.

Maybe scripting language was overused down the road? i.e. to use it for everything, and use them like a compiled language(ruby in rails, php framework, django,etc) that made things slow? point here is that to restrict script languages to glue logic for the most part, and always remember to use ffi for heavy lifting, not sure how to balance both yet.

The main reason why server-side stuff is slow is poor use of the database. Doing e.g. nested loops in a compiled language uses way less CPU than a scripting language, but it should be done in the DB in the first place.

Re: I love building a startup in Rust but wouldn't pick it again

#142

Post by estebank »

I find it surprising that so many people are arguing about the benefits and drawbacks of `?`, when in my experience the handling of Result and Option haven't been an issue in practice on the consuming side (`?`, `.unwrap()`, `.map()`, `.ok()`, if let, match, let chains, let else, etc. help a lot), but where all the pain comes from is having to declare the appropriate error type itself. Libraries like `anyhow` takes some of the pain away, but declaring an appropriate struct or particularly an enum in the right places, and the boilerplate for all the type conversions (From/Into impls) are where, during development, I have frustration. What I do then is either use Result or a single `struct Error(String);`, and go back once I have all the scaffolding in place and pry the implicit error tree back into the type system. Anonymous enums like typescript (`A | B | C`) could presumably help here.

Re: I love building a startup in Rust but wouldn't pick it again

#143

Post by Ylmaz »

I like this quote from 'The art of Unix Programming' published in 2003

"While it still makes sense to write system programs and time-critical kernels of applications in C or C++, the world has changed a great deal since these languages came to prominence in the 1980s. In 2003, processors are a thousand times faster, memories are a thousand times larger, and disks are a factor of ten thousand larger, for roughly constant dollars.

These plunging costs change the economics of programming in a fundamental way. Under most circumstances it no longer makes sense to try to be as sparing of machine resources as C permits. Instead, the economically optimal choice is to minimize debugging time and maximize the long-term maintainability of the code by human beings. Most sorts of implementation (including application prototyping) are therefore better served by the newer generation of interpreted and scripting languages. This transition exactly parallels the conditions that, last time around the wheel, led to the rise of C/C++ and the eclipse of assembler programming."

Re: I love building a startup in Rust but wouldn't pick it again

#144

Post by a5huynh »

estebank wrote: post #142

I find it surprising that so many people are arguing about the benefits and drawbacks of `?`, when in my experience the handling of Result and Option haven't been an issue in practice on the consuming side (`?`, `.unwrap()`, `.map()`, `.ok()`, if let, match, let chains, let else, etc. help a lot), but where all the pain comes from is having to declare the appropriate error type itself. Libraries like `anyhow` takes s…

I've run into similar issues and found that the `thiserror` crate (https://crates.io/crates/thiserror) combined w/ anyhow makes a lot of that pain go away

Re: I love building a startup in Rust but wouldn't pick it again

#145

Post by steveklabnik »

ThatGeoGuy wrote: post #140
steveklabnik wrote: post #65

Oxide is a startup and we use Rust for everything except the front end of websites (where we use TypeScript.) In some cases that’s due to hard requirements (embedded) but we use it for web backend cases as well. Iteration time hasn’t been an issue, but compile times can be annoying. Though obviously compile time is related to iteration time. Of course, all of these things are anecdotal. Collecting anecdotes is how yo…

Tangram Vision [0] is also a startup and we also use Rust. We're using it to develop robotic / autonomous sensor calibration tools that would normally be written in a variety of C / C++ libraries. For context: most if not all of our team has developed calibration tooling similar to what we're doing now in the past, just at different startups and very specific to certain robotic or sensing configurations. If anything,…

Very cool!

And yeah while I've had similar experiences, I'm glad to hear that it's so good for others. And, for example, sccache isn't something I have much experience with, so repeatedly hearing "it works well" is nice.

(Also, secret of replying to people on the internet: you are often replying to the anonymous reader of the discussion as much, if not more, than to the person you're replying to.)

Re: I love building a startup in Rust but wouldn't pick it again

#146

Post by FpUser »

ntonozzi wrote: post #82
richardwhiuk wrote: post #72

Earlier quoted context omitted.

Not sure I agree with Go vs Rust. I think if you would choose Java or Python or C#, then Rust might not be the right choice.

Go belongs in the exact same bucket as Java and C#.

Performance wise yes. As for language features - not really. Code in Go would likely have way more LOC

Re: I love building a startup in Rust but wouldn't pick it again

#147

Post by vore »

gpderetta wrote: post #117
flohofwoe wrote: post #49

Earlier quoted context omitted.

The problem with exceptions isn't the syntax, but the hidden control flow (they are essentially a goto in disguise). Error union return values make a lot more sense, the rest is just syntax sugar details (and that's where opinions differ I guess).

Exceptions are not a form of gotos, they are both less powerful as they are structured and more powerful (as they are nonlocal). They desugar to continuations, but so does rust option type handling and ?. In fact they are pretty much equivalent. I'm not terribly familiar with either language, but I don't see any particular difference between swift and rust error handling for example, swift will also mark fallible fun…

I think the nonlocal part is the scary part: it becomes very scary to figure out which parts of the code can fail and how, especially when failures can come from an arbitrarily deep call stack.

Maybe checked exceptions could be more useful to explicitly annotate allowed failures, but at the same time we all know how that's going in Java world.

Re: I love building a startup in Rust but wouldn't pick it again

#148

Post by adastra22 »

bayesian_horse wrote: post #21

How much more performance do you need to get from Rust over Python (even Cython, PyPy, Numba, etc) to justify the extra development cost? A 2x gain is certainly not worth it. A 10x gain? Maybe. But that is hard to achieve when much of your "compute" is spent on the DB side of things. How many startups actually scale out of needing a few non-db instances?

Rust easily gets 10x improvement in performance over Python in a lot of applications. This is absolutely my experience.

A better statement is that it doesn’t get 10x improvement over Go, or other ergonomic compiled languages.

Re: I love building a startup in Rust but wouldn't pick it again

#149

Post by theptip »

hinkley wrote: post #90
rom-antics wrote: post #28

Earlier quoted context omitted.

The mistake was not "explicit errors". It was having a mix of error types, some explicit and some implicit, with no convenient way to combine them, plus the interface complications. Note that most newer languages are choosing explicit errors. This includes at least Go, Rust, Swift, Zig, and Odin.

The second mistake was only flirting with Bertrand Meyer’s work until the Gang of Four showed up and wrecked Java forever. Meyer + functional core nets you a great deal of code with no exception declarations and an easy path for unit tests. If it hurts to do stuff it might not be the language that sucks, it might be you. Pain is information. Adapt.

Don’t know Meyer’s work - any suggestions for good starting points?

Re: I love building a startup in Rust but wouldn't pick it again

#150

Post by adastra22 »

RcouF1uZ4gsC wrote: post #38

Question for HN, all things being equal (you are not more familiar with one language/framework) what language would you choose to build a startup in?

Go. Swift if it was Apple ecosystem. Rust only for very specific tooling as required.
Post reply on HN