[xvilka]

Article mentioned Apple with their own mainstream language, I suppose Swift. But it's far from being mainstream, didn't gain the desired attention and failed miserably with TensorFlow. Thus, it's a niche language making no sense outside of the Apple ecosystem. Is not a serious contender to Haskell.

[mark_l_watson]

While my preferred languages are Common Lisp, Clojure, Haskell, and Scheme, I must say that Swift is pretty nice.

Apple’s supplied deep learning models are useful and easy to use, and compiling on an IPad Pro or M1 Mac is fast, even with the type inferencing. Right now I am slogging through a 50 hour SwiftUI 2 class.

I am disappointed that TensorFlow for Swift has been mothballed, but Julia and Python are probably much better options for deep learning. BTW, Julia is a fairly good general purpose language also. I have played with little bits of Julia code for text manipulation, REST, SPARQL queries, etc.

EDIT: since this is a Haskell thread, pardon the plug: you can grab a free copy of my Haskell book at https://markwatson.com

[apetrovic]

Swift is mainstream in Apple ecosystem, like VB (and C# before dotnet went open source) was in MS ecosystem back in the day. I'm not questioning Haskell influence and importance, but, what you think, how many developers use Swift and get paid to work in Swift every day, compared to Haskell developers?

[auggierose]

Swift TensorFlow wasn't an Apple thing, I think. And it never made sense in the first place ...

[frou_dh]

I would love to know some actual numbers. It does seem like Swift adoption on server/Linux by people who aren't already Mac/iOS developers is horrendously low.

[pjmlp]

Swift is mainstream on the context of Apple ecosystem.

Not everyone cares about cross platform.