Re: Bidirectional Binary Self-Joins

On Mar 31, 10:19 pm, "David Cressey" <cresse..._at_verizon.net> wrote:
> "Marshall" <marshall.spi..._at_gmail.com> wrote in message
>
> > Well, uh, I guess. I expect we can certainly agree that sets and
> > lists are different things and that syntax should reflect this.
> > However
> > if the syntax is unambiguous, (which it darn well better be) and
> > there is still confusion, we have to put that at the feet of
> > the reader.
>
> Fair enough. And I guess the braces in your syntax enclose
> a set and not a list. My bad.

Well, it's not like we have a common formal language. Much of the stuff we write here is notation-by-convention, and it's not necessarily even unambiguous. However if we imagine a language that includes sets and lists, we would expect that it would be unambiguous. Whatever it was.

> > Note that syntax is fundamentally ordered. Any sentence in any
> > language is a finite sequence of symbols, and by "sequence" I
> > mean exactly "list." (The two terms refer to the same thing,
> > mathematically.) Therefore any syntax we have for sets must
> > necessarily have some order to the symbols, and we explicitly
> > distinguish when certain sequences of tokens have meaningful
> > order and when they don't.
>
> I think "any language" in the above applies to data structures in a computer
> as well as text languages. Sigh.

You only say that because it's true in all existing computer architectures. :-)

It's not *necessarily* true.

Marshall Received on Sun Apr 01 2007 - 10:11:22 CEST