amen! 

> Unfortunately such committees, being almost entirely political animals,
> usually find ways of weaselling in their own totally new features and
> making unnecessary and incompatible changes to existing features instead
> of simply fixing "bugs" in the existing standards, while at the same
> time trying to claim that they're still defining a language called "C"
> -- in realitly they're defining something that may not even accept code
> written in the real "C" (i.e. K&R C by the time it was truly publicly


yes! and personally i would only really accept something to be 'C' if it
was created by or at least approved of by the ones who _created_ C in the
first place! K&R are the only real authorities on C, and unless _they_
change their language, the result just ain't the Real Thing. 
Even ANSI-C has oogy things and restrictions imposed on it that had no
place ever inflicting themselves on the C world, but at least ANSI-C is 
still close enough to Real C that one can feel more or less comfortable
with it and still get away with most K&Risms legally. Real C is so much
more flexible and while certainly allows one to write nonportable code, 
doesnt have anything terribly nonportable in it intrinsically. All the 
subsequent 'revisions' of the language have tried to impose restrictions
and structure and stuffy rules to 'enforce' the 'portability' of things.
that's just not the bloody way C was meant to be! Languages that hold your
hand all the way exist, and people who just can't stand the idea of being
able to play pointer math or take advantage of some feature of their
machine to cleverly do something, those people should use a different
language more to their suiting instead of try to make C into such a beast! 

> available sometime shortly after the book was published).  The C9X
> standards committee had the chance to undo mistakes made in ISO-C (while
> of course at the same time still creating an accurate and complete
> definition of the language), but so far as I've been able to tell from
> their drafts they've continued down the same path of invention rather
> than standardisation.
> 
> DMR's approval of ANSI/ISO C was only given very begrudgingly and was
> mostly due to the fact that he didn't have the time or energy to spend
> moderating such a standards committee (I'm paraphrasing what I remember
> of him saying about this in both public talks and personal conversations
> and I hope I'm not entirely mis-"quoting" him).  I suspect he's
> personally almost entirely divorced himself from "Modern C", especially
> since the languages he no doubt uses for what day-to-day programming he
> still does are far more interesting and far more "advanced" than the
> mess C9X is.
> 
> Software standards groups still have a *lot* to learn from true
> engineering standards.

And i'm keeping my pre-ansi C compilers warm! 

happy hacking,
isildur