commercializing ocaml

commercializing ocaml

[miles]

I've been eagerly following the success of Bluetail and Erlang and wondering if
something similar might be possible with OCaml.  The model seems simple:
assemble a small team of first-class programmers and take advantage of the
productivty gains afforded by a good functional language to compete on quality,
performance, and time to market.  

For example, I see potential opportunities in the emerging ASP/hosted
application market, where most interfaces are simple network text or XML
protocols and fpl's advantages in dealing with complex logic could be critical.

Has anyone considered such a venture?  Perhaps the folks at INRIA?

(sorry, my French isn't up to the translation)

-- 
miles at caddr dot com

Re: commercializing ocaml

[Xavier Leroy]

> I've been eagerly following the success of Bluetail and Erlang and
> wondering if something similar might be possible with OCaml.  The
> model seems simple: assemble a small team of first-class programmers
> and take advantage of the productivty gains afforded by a good
> functional language to compete on quality, performance, and time to
> market.

I did not reply immediately to your message, because I hoped members
of this list with more experience in start-up companies would comment.

The main issue, in my opinion, is to have a suitable application area
to target.  You simply can't sell a new language alone, even with a
good implementation.  Several companies were created in the '80s and
'90s to sell functional or logic languages and implementations, and I
think all of them went down.

The remarkable success of Bluetail is mainly due to them being experts
in a hot domain (telecom software).  The additional productivity and
reliability brought by Erlang over more conventional languages helped
them a lot, of course, but Erlang by itself would not have allowed
them to make such a big hit.

> For example, I see potential opportunities in the emerging
> ASP/hosted application market, where most interfaces are simple
> network text or XML protocols and fpl's advantages in dealing with
> complex logic could be critical.

That's one possibility -- although you should not disclose your
business plans on a mailing list :-)

> Has anyone considered such a venture?  Perhaps the folks at INRIA?

We didn't, really, by lack of a hot application area.

- Xavier Leroy