Ask HN: Language After Python and C?

9 points·greenfield101··11 comments

For someone in the DS/ML sphere is there anything beyond Python and C/C++ (and SQL ofc) that would be valuable to learn?

View on HN

If you really want to learn another language specifically instead of something else, then getting well acquainted with bash and awk wouldn’t be wasted effort.

Especially awk, it's a DSL for quick text manipulation. Simpler than Perl and a better API than Python for that specific use case. It's essentially a filter that applies to a specific region identified by a regex.

Mojo has an interesting mix of things to investigate. Julia is also interesting things that make CUDA/GPU programming easier than Python or C++.

Agree for DS, Julia would be very valuable.

Elixir (https://elixir-lang.org) and its Nx stack (https://github.com/elixir-nx), maybe ?

Rust is the trendy language.

Otherwise learn the Bootstrappable Builds stack from the bottom up; commented hex of machine code, that with short labels and offsets, that with long labels and label addressing, simple assembly, a subset of C, larger subset of C including a preprocessor, various larger C subsets, make, sed, parser generators, bash and so on.

https://bootstrappable.org/ https://lwn.net/Articles/983340/ https://github.com/fosslinux/live-bootstrap/blob/master/part...

Do you feel comfortable with array based programming and using tensors? Even though it’s still python, really understanding numpy and torch, are basically learning new programming paradigms. And very relevant to DS/ML

Alles ist gut.

Aber Yava über alles!

https://www.howtopronounce.com/german/java

JS/TS could be useful if you plan to do any front end work, even just for small demos of your DS/ML work.

only rust

not really

Reply on news.ycombinator.com