> If we expand the definition of standardized languages a bit, we
> actually find some influential committees where companies only pay
> expenses - they don't buy voting memberships.

You are right.  But those are the ANSI rules.  Also, the Smalltalk-98 
standard is (C) ANSI so unless we want to rewrite *everything*, it is 
not so easy to pick a different standardization body.  I am not saying 
that rewriting everything is necessarily a bad idea, but it would be 
quite a big endeavour.

> I can't think of another language that ratifies a new standard every
> 18 months. 5 to 10 years is more like the ones I follow.  I wonder why
> Smalltalk will be different.

Language standards like C are reviewed every 5-10 years, but their scope 
is much less broad and deep.  C++ is finalizing the next version, which 
also puts them in the 10 year range, but the amount of changes between 
C++98 and TR1 (which was out around 2003) was already huge.  In the end, 
  for many languages the sheer cost of implementing the standard makes 
it very hard to have a short period between successive revisions

The differences between successive Smalltalk standards would be more 
comparable to the differences between successive versions of 
Java/Python/Ruby (additional modules for different class libraries, 
etc.).  The release cycles of those languages is indeed around 2 years.


Paolo