There is a new programming language initiative striving to combine XML
syntax with LISP parsing and lambda calculus evaluation. The public
documentation at http://www.xmlive.org still looks a bit sketchy, but
it has some really determined (and nice-looking) supporters, judging by
http://coveries.blogspot.com/2005/03/support-xmlive.html :-)
BR,
Jacques
··················@hotmail.com wrote
> There is a new programming language initiative striving to combine XML
> syntax with LISP parsing and lambda calculus evaluation.
Judging from the code on the front page, neither XML syntax nor
Lisp-style parsing is involved.
I think it looks somewhat suspect:
* You can "bind" a symbol without supplying a value. I suppose this
is a kind of declaration, but binding without a value sounds like
knocking without a door.
* The same operator ("bind") can be used to bind symbols, define
functions (shades of Scheme?), and declare a type for a symbol.
* Expressions use infix syntax.
* The assignment operator seems to be =!, which is probably the least
intuitive variant I've ever seen. Possibly it's a combination of
C-style "=" with Scheme-style "!" for a destructive operator.
* Function definitions look like Scheme, only a bit more awkward.
* It seems that in order to evaluate a symbol in the "REPL" you call
it like a function. Inside expressions, symbols are evaluated
directly.
* It's not clear if the angle brackets surrounding most forms are list
delimiters. When a range of integers is to be produced, curly
brackets are used instead of angle brackets.
* When a symbol is already bound, you need another operator to bind it
to another value.
* When defining structures, a fourth syntactic variant is introduced:
"{{A, B, C}}" where A, B, and C are member declarations.
* Finally, square brackets are used in a fifth syntactic variant; I
can't discern why it is used where it appears.
"Peter Lewerin" <·············@swipnet.se> wrote in message
·································@posting.google.com...
> ··················@hotmail.com wrote
>
>> There is a new programming language initiative striving to combine XML
>> syntax with LISP parsing and lambda calculus evaluation.
>
I think the the only reasons they used angle brackets
was to add that 'modern air'. It would never pass
as XML. You've got to give points for verbosity,
reminds me of COBOL with <points>.
--
Geoff
I believe Gary Sussman often refers to what a C programmer would call
'declaring a variable' as 'binding a symbol in a frame', so the bind
declarations might make some twisted sense... I really liked the girl
though. Her blog postings were very...unusual.