python and haskell for fun

[Christopher Browne]

Kendall Clark <kendall at monkeyfist.com> wrote:
> >>>>> "sandy" == Sandy Norton <sandskyfly at hotmail.com> writes:
> 
>   sandy> "Patrick" wrote:
>   >> This subset of PASKELL will be clean, beautiful and functional.
> 
>   sandy> I'm sorry I wasn't clear. I meant 'beget a pretty monster' of
>   sandy> an application, not a language. The last thing I would want
>   sandy> to do is waste time discussing a new hybrid 'PASKELL'
>   sandy> language...
> 
> I've been daydreaming for about a year of a Python implementation
> written in Haskell, which would obviously make writing Python
> extensions in Haskell easier (and more elegant, IMO, than writing them
> in C or Java).
> 
> However, John Paul Skaller did something similar with Ocaml, which he
> called Vyper, and that went over in the Py world like a leaden balloon
> -- actually it went over about as poorly as Stackless Python did;
> which does cause me to wonder what it is about Python culture, if
> anything, that's resistant to unusual alternative implementations?. A
> Python implementation in Haskell is probably a fool's task. Though
> given the various Haskell compilers, including some pretty impressive
> parallelizing stuff, it is an interesting idea.
> 
> (Ocaml is pretty much Caml + objects; Caml is INRIA's implementation
> of ML; in the neighborhood of Haskell, but less syntactially elegant,
> IMO.)
> 
> At least, a Haskell implementation of Python seems more practical than
> a Python implementation of Haskell, which would be rather toylike. But
> YMMV.

I'd love to see Vyper go further; it's a neat approach to the
"wouldn't it be neat to static type Python?" question.

The thing that seemed particularly neat was the way it added in
ML-style pattern matching.

I have a feeling that what has happened isn't so much that people
started farting in the author's general direction ("your mother smells
like elderberries!"), but that it was a one man show, and that nobody
really took interest in it.

In effect, it may be just like most of the SourceForge projects that
get started up by someone that thinks they have an idea, but who never
get around to any usable degree of completion.

Alternatively, it could be compared to SOAP implementations on Python.
The wonderful thing about SOAP on Python is that there are so many
implementations to choose from.  Unfortunately, most of them got about
half-way through, and authors' companies got lost in the wastelands of
the "Internet Bubble."  And so we have a whole bunch of
implementations, most of which are completely unmaintained.  (Compare
to Perl's SOAP::Lite...)
-- 
(reverse (concatenate 'string "gro.gultn@" "enworbbc"))
http://www3.sympatico.ca/cbbrowne/python.html