Ask HN: What ML dialect do you recommend for someone to start with?

9 points·cartesius13··3 comments

Ocaml, Standard ML, F# or maybe some lesser known one? Right now I'm mainly using Go because of its simplicity, rich ecosystem (many libs), great docs, great tooling etc.

Which ML is good in all these points?

View on HN

I like OCaml and I like F#. The choice between them is really "do you like .NET?"

As a Java programmer I've been amused at how ideas have diffused slowly from ML to Java such as the hotspot compiler & generational runtime (Sun hired somebody who did an academic project for a high-performance ML-famly runtime) as well as pattern matching, sealed classes, records, etc.

I'll second that as "I like F# and OCaml".

Much earlier on F# was hard to get started with especially if not on Windows. C# still gets a lot more coverage than F#, but it pretty much works out of the box with free Visual Studio Community or Visual Studio for Mac (not VS Code) and some setup tutorial instructions.

And as a (former) Java programmer, I've been frustrated at how slowly ideas have come. Union types? Only in multi-catch expressions.

I don’t know, but I find the book “Purely Functional Data Structures” presents code using Standard ML in a way that is completely elegant and beautiful. I haven’t coded much ML myself, but the way it’s presented in the book is some of the cleanest code I’ve ever gazed upon.

Reply on news.ycombinator.com