On Sat, May 05, 2007 at 01:34:38AM +0100, Jon Harrop wrote:
> On Friday 04 May 2007 22:57, Richard Jones wrote:
> > I wouldn't recommend that people tie themselves into Windows as a
> > platform however.
> 
> F# runs under Linux and Mac OS X as well.

... 10 to 30 times more slowly under Mono, according to your posting here:

http://caml.inria.fr/pub/ml-archives/caml-list/2007/04/83072c819929adb7612aecd5286c643d.en.html

Do Microsoft produce a free (as in speech) .Net framework with
equivalent performance which runs under Linux?  Will they continue to
make F# work with Mono?  or indeed do they support it right now?

> > And somehow F# seems to be licensed under a very restrictive license
> 
> In what way? I can sell programs that I write in F#.

The license, "Microsoft Research Shared Source License Agreement"
(http://research.microsoft.com/fsharp/fsharp-license.txt) explicitly
prevents commercial redistribution of the F# sources, and that means
for example that it could never be part of a free operating system
like Debian or Fedora.

In practical, day to day terms I'd better hope that Microsoft Research
continue to produce and support new versions of F#.  If one day they
just abandoned it, or if they decided to drop support of Mono then I'd
be up a creek without a paddle (but with a lot of software which I'd
have to rewrite in another language).

> > (how did they get away with that one? - Did they pay INRIA for a separate
> > source license?) 
> 
> Besides legacy, F# has nothing to do with OCaml and INRIA.

So F# was a separate implementation?  I thought it was derived from an
older version of OCaml - but with the licensing issues above I'm not
about to download the source to check that.

Discussion moved to caml-list.  It's important to get any
misunderstandings about F# resolved.

Rich.

-- 
Richard Jones
Red Hat