Re: Experimental languages.

Thomas M. Farrelly (s720@ii.uib.no) wrote:
: I'm looking for references/links to highly experimental programming
: languages with focus on both syntactic and semantic consistency.


To which the moderator remarked:


: [The urge to mention INTERCAL is almost overwhelming, since it was
: carefully designed to have none of either. -John]


And, if it passes his "relevance filter", I might add that
the "Usual gang of idiots" who keep the Intercal flame burning
came up with a a language in which control-flow is two-dimensional.
I do not at the moment recall the name, but it exhibitted some
interesting properties IIRC. No, really. It was just a little
bit too wired for me. I still looking forward to re-targetting
gcc to my RSSB single-instruction computer :-)


On a more serious note, the only languages I've seen which
are _highly_ consistent both syntactically and semantically, and
yet useful, are pretty stark. Examples are Forth and Lisp. Each
has a very simple conceptual model that "fits" well in a user's
(well, _some_ users :-) brain. The complexity and richness derive
from the layers of domain-specific extensions that both languages
allow.


That said, a really powerful and "comfortable" system in
either is also often highly personal. To a user approaching it for
the first time it can be as intimidating as MicroSoft Word. Well,
maybe not that bad but generality and fitness for a particular use
are in opposition to each other. Again I make a plea for everybody
reading this to read and digest Gerald Weinberg's "Psychology
of Computer Programming", and perhaps Weisenberg's (?) "Computer
Power and Human Reason". Language design is a strongly _human_
activity and is as much about communication with other people
(including later stages of the author's life) as with machines.


Mike
| albaugh@agames.com
[I'd say that Common Lisp resembles MS Word in its kitchen-sink-ness, but
Scheme remains one of the cleanest and most productive languages around.
-John]




--

Return to the comp.compilers page.
Search the comp.compilers archives again.