[Shootout-list] dispatch

[Jon Harrop]

I just noticed something more surprising: with no implementations in 
mainstream languages (only Eiffel, D, Oberon-2, C# and Smalltalk), the 
dispatch test is on the main page instead of the sandbox.

On Saturday 30 July 2005 07:58, José Bollo wrote:
> Well you are right when you say that my proposal is not enough because the
> program can still be reduced to some formula of the arguments!

Yes.

> It is great to know that a such thing exist!
> I would like to see the Ocaml / dispatch bench with a good explaination of
> such compiling feature.

In OCaml, you would have to do the simplification yourself (as Xavier says, 
the OCaml compilers are designed for compiling well-written code) but there 
are several other languages that would let you do an arbitrary amount of the 
simplification at compile-time, e.g. Lisp, MetaOCaml and any CAS programs.

> IMO raw results are not enough and it would be good to have more (short)
> comments on the tests (when usefull). Comments are actually at the bottom
> of the pages. They would better be near the top.

It is also interesting that the ML implementation of the dispatch test is 
shorter than the textual one. :-)

> About your opinion on OO i can not see defect in it. I will not recall you
> things about program's life cycles that are out of the shootout field. For
> the shootout your challenge is a good challenge even if impossible.

Yes. Perhaps OO is better on huge programs (>100kLOC) but I see no evidence of 
this...

-- 
Dr Jon D Harrop, Flying Frog Consultancy Ltd.
Objective CAML for Scientists
http://www.ffconsultancy.com/products/ocaml_for_scientists