Design-by-Committee

Paul Prescod paulp at ActiveState.com
Wed May 9 09:51:17 CEST 2001


Paul Foley wrote:
> 
>...
> 
> John: Blatant hype.  You could already use Lisp's S-expressions; it's
> exactly like the XML that some clever-clogs will reinvent 40 years in
> the future, but easier to use, better in various ways, and available
> today...

I think it is pretty accurate to say that XML is available to day.
Anyhow, this argument goes round and round (on this newsgroup and every
other one) because people are using the term "XML" to mean two different
things. Paul is using it to mean the use of angle-brackets to describe
elements and attributes.

Most other people use it to describe a family of technologies which are
*data-centric* rather than *code-centric* and solve some problems that
the code-centric technologies that we have used for years failed to
solve.

Lisp treats code as data and data as code. That's great for solving some
problems. Lisp s-expressions are great, but what standard,
non-turing-complete language exists for declaring vocabularies of them?
What standard, non-turing-complete language exists for querying them? I
stress the importance of non-turing-completeness because that property
is important for reasons of proof, performance and security.

What is the Lisp "vocabulary" for hypertext documents? Or technical
manuals? Or running tables? What is the standard mapping from EDI
messages to S-expressions? What is the canonical form of S-expressions
for purposes of digital signing?

S-expressions hardly even start to address the issue. They aren't even
as good as the angle-brackets as a syntax ((((who)) can (keep) track) of
(them))())? And they are no better at addressing any of the deeper
issues.


-- 
Take a recipe. Leave a recipe.  
Python Cookbook!  http://www.ActiveState.com/pythoncookbook




More information about the Python-list mailing list