3 ms·
I’m trying to make the same decision right now. I’m the first product engineer at a AI startup and started using node.js but... meh (i use crystal for stuff and
by devmunchies 6y ago
I’m trying to make the same decision right now. I’m the first product engineer at a AI startup and started using node.js but... meh (i use crystal for stuff and have exp in ocaml, elixir, ruby)
I’m leaning towards f# or Scala. I love static typing.
Since it’s a data science, machine learning company Scala makes more sense.
- siwatanejo 6y agoBefore choosing Scala I think you should watch this talk: https://www.youtube.com/watch?v=uiJycy6dFSQ https://www.youtube.com/watch?v=uiJycy6dFSQ (note: he's not a troll, he worked 5 years with Scala)
- didibus 6y agoWoa, ya those initial choices for an AI startup seem so wrong. Scala sounds most reasonable, I'd say even consider Java to be honest.
- cultus 6y agoI honestly don't see a huge value proposition in Scala, now that Java has reduced pain points around FP. Scala has higher-kinded types, but you see people tie themselves in knots with those as often as not. It is just such a huge language. Scala 3/Dotty sounds like it will be a big improvement.
- yawaramin 6y agoJava doesn't have one of the fundamentals of typed FP--algebraic data types with exhaustive pattern matching. And just in general--Scala, despite its warts, can be very succinct and packs a lot of power in simple-looking code.
- devmunchies 6y agoNo, sorry. None of the languages I listed are being used at the startup (except node). These are just what I’ve used previously. Everything is in python but I’m building out the main user-facing app to interface with all the models.