On Fri, May 11, 2007 at 12:16:46PM -0400, John Cowan wrote:
> Uniformly legible to whom? If Scheme programs are limited to ASCII,
> they are also implicitly limited to those who read and write American
> English.
Programming languages are not related to natural languages. Thinking
that they are leads to COBOL.
> I am told that one of the early attractions of Java in Japan
> was the ability to write programs with identifiers that were meaningful
> in Japanese, thus liberating Japanese programmers from the burden of
> deciphering either English or their own language in transliteration.
There is no English in a computer program (except possibly in
comments). Some identifiers and keywords may be superficially similar
to English words, but their meaning is only vaguely related to the
meaning of the corresponding English words. I think
non-English-speakers are actually better off not having any
presuppositions about what an identifier is supposed to stand for.
I learned BASIC before I learnt English, and I don't think knowledge
of English would have been of any help.
> ASCII is just another arbitrary limitation, and one it's time to let
> go of.
What fine company you are in when making such a proposal:
http://www.research.att.com/~bs/whitespace98.pdf
Textuality is another arbitrary limitation. Let's let go of it,
too. Why shouldn't we allow arbitrary images, sounds or colors to be
identifiers?
I'd better stop here, as I'm just repeating myself:
http://google.com/groups?q=insubject:sensitivity+author:la at iki.fi&filter=0
Lauri
Received on Fri May 11 2007 - 12:58:47 UTC