Pascal Costanza  <costanza@web.de> wrote:
+---------------
| The good thing about Common Lisp and its ANSI standard is that it is 
| mainly derived from decades of actual programming experience with Lisp 
| dialects. So this standard is a unique example of things that weren't 
| designed and then forced upon a community without actual experience, but 
| just the other way around: Lisp programmers identified common idioms and 
| agreed upon compromises between mostly superficial differences.
+---------------

<RANT>
Actually, this is the way *ALL* standards used to work, up until
around 1980 or so!!! Until then, the assumption was that you didn't
try to "standardize" something until there was enough experience to 
know what was good and what was bad, and then you tried to standardize 
the "best current practice" or some reasonable compromise thereof.

But then something happened, I'm not quite sure what. Suddenly the
default modus operandi became "design by standards committee", with
the principle of "maximum mutual disadvantage" being enforced so
that no-one who *already* had a decently-working system would have
an "unfair" advantage over others on the committee. And as a result,
we got monstrosities like FDDI/LMT/CMT/SMT and ATM/UNI/LANE and the
growing alphabet soup of web "standards" (which mainly seem designed
to keep anyone with less than a billion dollars out of the game
entirely).

The last decently-spare standards based on *working* technology,
at least in the networking protocols area, were probably Ethernet
(and it just *barely* escaped being broken incompatibly by IEEE 802.3!)
and HIPPI, and maybe IEEE-488 (the standardization of HP-IB).
</RANT>


-Rob

-----
Rob Warnock			<rpw3@rpw3.org>
627 26th Avenue			<URL:http://rpw3.org/>
San Mateo, CA 94403		(650)572-2607