Hi all, 

	About a year and a half ago I asked something like why Class was not
a subclass of Collection. Around that time I also asked about "formalizing"
OO. It turns out that my English betrays me from time to time (I am not a
native English speaker), especially when I try to communicate more difficult
concepts. I will try to use more words to get around the communications
barrier, even though I dislike redundancy. Please bear with me.

	What I really meant was not some academic (pejorative connotation
here) lambda calculus or correctness proof, but a higher concept that surely
has to still hold true even in this "new kind of mathematics" (approximate
quote from Alan Kay) that is computer science, namely rigor, attention to
detail. A mathematical or logical proof can be valid even if expressed in
informal terms. IMHO Smalltalk's core library deserves a very careful
scrutiny and possible reconsideration (even after 20 something years of
successful use - actually that's why it "deserves" anything) with solid
justifications for all the (more or less explicit) decisions that are
embedded in it. 
	Mathematics model reality. The class library also models reality. I
think it is a worthy goal to pursue the beauty and simplicity of axiomatic
systems, with the same care for consistency and minimality (and minimizing
the implementation issues' impact). Intuitively this would lead to better
software too.
	And if there is a place to do it, that place is obviously Squeak.
Maybe not the Squeak envisioned by Disney, so this is where I would like to
ask your opinion: is there place in the main Squeak for pursuing such an
ideal that could mean a departure from the original Smalltalk (core library
and possibly even language syntax), and how would such an effort take place
?

	On a more concrete aspect, I have played lately with a numbers
hierarchy. IMHO this is one of the important parts of the core library that
would considerably benefit from a reevaluation (from a math purist's point
of view) and from a synchronization with the advances in computer science
(i.e. IEEE 754).
	Some details about this in (PART II) - this one's a real
cliffhanger, isn't it? ;-)

Florin