About a new HLL

Thu, 4 Jan 1996 02:33:49 +0000 (GMT)

> Hi Fare',
>           I appreciate your attempt in order to make an ideal language.

> To do this must be considered why the others languages have not meet
> the objective:
[I put your first suggestion last, because my reply to them is quite longer]

> were the people less intelligent than yours?
   Sure sure not !

> were the forces (# human hour) less than yours? ... sure not!

   Those two points have long inspired me fear and doubt, which up to now
has prevented me from advancing more on my language design.

> were the ambitions fewer than yours?
   I dare affirm that in most cases, yes.
   Most languages (FORTRAN, BASIC, C, Pascal, Prolog, Smalltalk, C++, {,s}ed,
awk, {,ba,c,k,z}sh, Perl, etc) began as a quick and dirty hack to get things
running with some programming style in mind, with more hacks added as
intended programs needed some, sometimes ending in a huge kludgy bloat.
Though seemingly very efficient in the limited scope of programming they were
developped for, these languages are horrid monsters that are unable to do
anything with any efficiency outside of this limited scope, and tie
miserably the problems they process to the ad-hoc representation
that was used at the time they were designed.
   Some languages (Algol, Ada, Common LISP, *ML, etc) had the ambition
of being universally expressive, with varying success as compared to the
previous class of languages. Algol died of creeping featuritis and formal
snobbism, as an attempt to make a language from a soundtrack of the tower
of babel. Ada lives by military paranoia and design centralism, in an attempt
to fix computing technology into the state a clever team reached, that could
please the military. Only LISP developped freely, and could reach the
state of non-trivial reflectivity, but still of a dirty kind, by
lack of a developped type system. ML developped such a type system that
LISP lacked, but voluntarily despised reflectivity, and made the type
system poorly expressive enough so as to achieve easier computability results.
   The TUNES HLL ambitions to be a language of choice for reflective
programming: writing programs that produce anything, including programs.
It also ambitions to provide semantics clean enough for a proof system
to be included as part of the language development system, and to
actually become (by means of reflectivity) a very actively used part
of the language itself (with programmable types including proof-based types).
It ambitions to provide dynamically programmable syntax, and programmable
code generation, that would allow any class of problems to be *efficiently*,
yet *securely* represented as a sublanguage. To conclude, it ambitions to end
the days of ascii text and offer semantics-based programming instead of
syntax-based development.

   Perhaps I may fail for the two last reasons you gave (that I quoted first
in this reply), but I sure am more ambitious than those who achieved
currently available languages. Perhaps I will fail because of this oversized
ambition. Surely other people more ambitious than I failed in the past,
because they realized their ideas were but dreams, or were just
unimplementable on hardware available to them in that past. But be sure that
the TUNES HLL, if it is ever completed, will be much more than what currently
is available, and for deep reasons.

> As an example recall the Ada history ...
> A better HLL could certainly be found but no strict timeline must be setted
   Are you suggesting more than I see in your message ?

> and the reason of the others languages should be taken with care!
   I hope I am not missing much. Please enlighten me !

>                                         Yours faithfully,
>                                         Stefano Marago`
>                                         (marago@di.unito.it)

--    ,        	                                ,           _ v    ~  ^  --
-- Fare -- rideau@clipper.ens.fr -- Francois-Rene Rideau -- +)ang-Vu Ban --
--                                      '                   / .          --
Join the TUNES project for a computing system based on computing freedom !
		   TUNES is a Useful, Not Expedient System
WWW page at URL: "http://www.eleves.ens.fr:8080/home/rideau/Tunes/"