I'm not a traditional programmer, but I've been doing a lot of scientific computing in my studies and research, mostly using Python and Julia. Although I'm pleased with those two, I find the coding experience getting a bit stale. I'm looking to learn a new compiled language for scientific computing that could be beneficial in the future. Currently, I'm torn between C++, Rust, and Fortran, and I'd appreciate any insights on which one might be the most useful down the line.
5 Answers
It really depends on your field. Rust is still young in the scientific community, and if you'll need to deal with legacy code, C++ might be a more practical choice. Fortran is popular but less common compared to C++ nowadays. Just my two cents as an RSI consultant!
I hobby develop scientific libraries in Rust, but its ecosystem needs more maturity before it can be a go-to.
C++ and Fortran are pretty much the go-to languages for scientific computing when performance is key. Rust is emerging but still a bit rough around the edges in terms of ecosystem. It's worth learning, but if you want something solid, go with C++ or Fortran. Rust could be a fun side project!
You might want to look into Go as well, though I'm not sure how robust the libraries are for scientific computing yet.
Just curious, why are you specifically looking for a compiled language? Are you aiming for better performance or type safety?
I've seen a presentation claiming that modern Fortran is surprisingly powerful compared to the older versions. It’s evolved a lot, so it might be worth exploring if you’re curious about it!
Julia gives you a Fortran-like performance with a Python-syntax feel; I'd prefer that over Fortran these days!
Yeah, I didn’t realize it was still getting updates, either!
Exactly! Rust doesn't have a stable ABI, which complicates using it with non-Rust projects.