FAQ.md

FAQ

Why is it important to have F# run on the JVM?

I want a high-performance, modern, no nonsense ML for the JVM. I think F# fits the bill.

Have you considered cross-compiling the existing compiler?

I have. It's an interesting project for someone else.

Have you considered cross-compiling .NET bytecode to the JVM?

I have. It's a world of pain.

How are you going to handle tail calls on the JVM?

With branches. (It's not a perfect fit, I agree, but it's not an unsolvable problem either.)

How are you going to handle value types?

Turn them into classes. Not a perfect fit but not a huge problem either.

How are you planning to deal with inferior generics of the JVM (most F# libraries rely on the JIT specializing the generic code for efficient execution)

To be honest, I haven't thought about this too much. AFAIK, JRuby uses annotations to carry similar information so it's I think it's a solvable problem. But I completely agree, it's not a nice fit either.

Why are you using Java as an implementation language?

I'm following similar design to JRuby, Clojure, and Smalltalk Redline, for example. The main objective is to be a minimal standalone core that doesn't require bootstrapping (see 'the other project').

Why start from the parser (the least important and interesting component)?

You have to start somewhere.

Why focus on F# instead of a better defined (smaller) functional language in the ML family like SML?

Because I'm interested in F#, not SML. See answer to the first question.