Rust Is Hard, Or: The Misery of Mainstream Programming

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by WastingMyTime89 on Friday June 03, 2022 @05:45AM (#31606212)

I for one as someone who programmed professionally in both Ocaml and Ada still don’t see where Rust is supposed to fit.

If I was doing something high level, I would probably use Ocaml which has nicer features. If I wanted to write safe code, Ada offers a better experience with the availability of SPARK for parts I would probably end up wanting to prove. If I just want to write concurrent code with the certainty I could find someone to maintain it, Go seems the way to go.

Rust seems to want to be a replacement for Ada but seems to actually attract programmers used to high level language chasing the hype.

What’s the story with Rust and verification nowadays?

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by qznc on Friday June 03, 2022 @06:18AM (#31606445)

I think Rust has a good chance to take away market share from Ada and Ocaml. Its eco system (packages, documentation) is more comprehensive already thanks to its broader appeal and being in fashion. I don’t see why stronger verification should not appear at some point. Progress and need is there.

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by thesuperbigfrog on Friday June 03, 2022 @08:42AM (#31607447)

>> What’s the story with Rust and verification nowadays?

It is in progress but likely several years to go before a MVP is released. Ferrocene [1] is the primary effort that I am aware of and that seems to be making progress towards a verified rustc suitable for safety critical work.

Ferrous Systems is working with AdaCore and just released the Ferrocene Language Specification [2] to formally document the Rust subset that Ferrocene will use.

[1] https://ferrous-systems.com/ferrocene/

[2] https://ferrous-systems.com/blog/ferrocene-language-specific...

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by umanwizard on Friday June 03, 2022 @08:34AM (#31607369)

Rust is supposed to be a better C++, and resoundingly succeeds at that goal.

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by pjmlp on Friday June 03, 2022 @08:47AM (#31607476)

It partially succeeds at that, because while it has a much better story in being safe by default, there are plenty of C++ use cases where Rust still hasn't a story to sell.

HPC, HFT, GPGPU, LLVM/GCC infrastructure, GUI, console SDKs, drivers SDKs, security certification ,....

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by umanwizard on Friday June 03, 2022 @08:51AM (#31607519)

Ok, fair, you’re right. Rust is successful in a rather large subset of C++ niche, but not yet the whole thing!

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by adgjlsfhk1 on Saturday June 04, 2022 @01:42AM (#31617362)

imo, Julia is a better fit than rust or C++ for hpc/gpgpu

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by pjmlp on Saturday June 04, 2022 @05:03AM (#31618358)

It might someday, still lots of room to improvement regarding current workloads.

I not not seeing Fermilab or CERN doing real time beam processing in Julia anytime soon.

Or stuff like having graphical shader debugging tools for Julia.

[ Parent | on HN ]

Re:Rust Is Hard, Or: The Misery of Mainstream Programming by adgjlsfhk1 on Saturday June 04, 2022 @12:51PM (#31622583)

Julia already is starting to be used for high energy physics (see https://arxiv.org/pdf/2003.11952.pdf and https://www.youtube.com/watch?v=QlfAa-LN1SA for details). It's definitely not the default yet, but there absolutely is effort to use it for particle accelerators.

[ Parent | on HN ]