Re: Programming language specification languages

On Date: 20 Sep 2001 00:30:22 -0400 the original poster, Nick
Maclaren, nmm1@cam.ac.uk, indicated
<<
I want something that enables me to say what I want to say, but precisely.
>>


and got a response that centered on several specification tools that
are ASCII based. One hesitates to gravitate against other sentiments
expressed by the original poster relative to preference for proven
existing technology, and an adversion to too much developmental
adventurism (my phrasing), however, ...


It seems worth questioning whether professionals nowadays ought to be
oriented to 8-bit foundations.


To the exact point of precision in specifications, do we not constrict
precision if the core tokens are so coarse?


And the substantives that will be operated upon, are we not interested
in a much larger symbol set in the world market, ... regardless of
whether the tools known as markup language are utilized in language
development?


Also deep in there is the need to handle non-contiguous alphabets (or
cipherset subsets) and non-orthogonal uppercase/lowercase masking
... atleast for modern international systems. So can an ASCII core be
the best for a new approach?


Out at the surface, it seems like there is a real strong need for lots
of symbols for a new language that would permit easy expression of the
marks for parallelism as well as all the traditional operators and
delimiters. An 8-bit foundation leaves little room for single cipher
symbols for new markers.


At any rate, there seems to be a distinct possibility that the 8-bit
encoding facilities are actually clogging up creative
channels. Sixteen bit foundations do not necessitate markup language
tools, but they do open up a realm that has far too few language
development tools.


Robert Rayhawk
rayhawk@alum.calberkeley.org

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