Dennis E. Hamilton writes:
 >
 > One of the things about Miser for me is to get to such a primitive system,
 > even though it is sufficient in some sense, that one can start noticing all
 > of those assumptions as one enriches the system with more directly-appealing
 > objects.  I love it that in C char isn't even about characters, it is
 > defined to be about character codes, and the particular code isn't even
 > clear!  And then we disguise all kinds of characters as chars, and char
 > becomes a disguise for other things (unsigned integers, short signed
 > integers, or Boolean flags, etc.).
 >
 > I notice that Java seems to be more rigorous/rigid about this (and makes it
 > clear that its "characters" are Unicode codes), but there still seems to be
 > a lot of illusion in it.

I don't have the code in front of me, but when I coded the
encode/decode base64 functions for SIF in Java my Indian mentor made
me use a byte[] for the base64 alphabet. He said that Java chars would
be 16 bits and that *is not* what we needed for base64 encoding (we
needed 8-bit bytes)!

 >
 > So, with Miser, I have to start with something, but it is something one can
 > provide a simple theory for, and check that Miser implements an
 > interpretation of that theory.
 >
 > And then I can notice all of the assumptions that arise, and what it is to
 > disguise something in a representation of something else.  It looks like
 > coherence can be discussed rigorously in this context too.  Certainly on
 > what it is that objects are versus what they model, etc.

I would very much like to understand what you're getting at here. And
regarding assumptions I spent the train ride from NYC this afternoon
reading a pocket guide to XML. Yikes! It seems like a very tortured
web they weave to insure that everything is specified. Also
problematic are the examples chosen; they themselves are full of
assumptions about some imagined application world. I'm sorry I can't
be more explicit right now.


 >
 > I want to get into themes more when we talk on Friday.  I want to have
 > something for wider review over this weekend, especially for establishing
 > ideas for workshops in Japan and in the Eastern U.S. in the first half of
 > 2002.

looking forward to it!

And finally, do you remember Peter Naur's paper "Programming as Theory
Building?" -- I'll try to read it again before Friday.

Bill