[kaizendad]

This is good, right? There are better languages for most (although certainly not all) purposes, whether you measure "better" through features like memory-safety or through developer happiness or anything in between.

Most of the systems that were written in C++ in the past didn't have to be written in C++. Now they'll become hard to support, because they are in fact hard to support, and eventually they will be refactored or rewritten.

Same thing happened with Assembly and COBOL programs being written into C++.

[jleahy]

Unfortunately high frequency trading is an area where C++ is normally the right choice, even for a new project.

[kaizendad]

Fair, I hope I was clear in my post that I don't mean to say C++ is never the right choice, but... rarely.

[varajelle]

Why not rust?

[jleahy]

Try doing (serious) high frequency trading in rust, then come back to me.

[generichuman]

Jane Street is doing it in Ocaml (which is pretty similar to Rust + GC) and they seem to be pretty serious.

Are you working at a more "serious" company where they only use languages with the safety level of Assembly?

[jleahy]

a) Yes I run technology at a more serious HFT firm than Jane Street. b) Jane Street also use FPGAs, so you cannot exactly say they just use ocaml, it’s more nuanced than that, they are using a mixture of technologies. I think they made a very unfortunate choice early on and are still paying for it.

[varajelle]

Do you have actual technical explanation? (other than argument of authority)

[epinephrinios]

Two Sigma launched new Rust teams to replace their core systems.

[varajelle]

HFT is not my field at all, so I don't know. Why is rust not good for it?