FW: specs

RE01 Rice Brian T. EM2 BRice@vinson.navy.mil
Sun, 3 Jan 1999 17:38:01 +0300

> ----------
> From: 	RE01 Rice Brian T. EM2
> Sent: 	Sunday, January 03, 1999 6:19 PM
> To: 	'wtanksle@ucsd.edu'
> Subject: 	RE: specs
> First of all, I see that when you say 'N dimensions' and 'N slots' you're
> saying the same thing.  Wrong?  Right?
> right.
> Okay.  And the 'type' of an arrow is essentially the same as its dimension
> (sure, it's a different number, but big deal).
> yes.
> Unless all you want is a gesture -- well, here you are: (define (arrow
> head tail) (cons head tail)).
> ok, but i was most interested in removing the text-based nature from the
> picture.  (i.e. a:=0 is different from zero?)
> Of course.  A pity ... I've been looking for a model of computation which
> actually accounts for the fact that computers are finite (no current model
> takes it into account), and all I see is a new one even worse (in that
> regard) than all the others.  No insult intended.
> none taken.  maybe we can work past it with infinitary notations.  you
> think?
> >>  - what's the use?
> >well, i could mention object-orientation and lambda-abstraction, but
> those
> >uses (i believe) are obvious to you.
> At the risk of seeming an idiot, I'll say that they're not.  I can see how
> this system could implement objects, but since this system is itself
> implemented using object-oriented design I don't see that as very special.
> I don't see how it would implement objects in a way in any way superior to
> any other system in current use (just my statement of ignorance and plea
> for help).
> that's ok.  i'm pretty humbled by you all.  the point of this system is to
> create a space where the fact that an object is not postulated does not
> preclude its use in computation.  this, i think, goes along with
> specifying an object by formal specification and not declaration.  yes, i
> know that they seem the same, but there is a difference as i see it.  (i'm
> not sure how to continue this, but maybe you have thoughts about this).
> As for lambda abstraction -- that doesn't make sense to me.  Lambda
> calculus is about attaching names to things, and your description of
> arrows hasn't mentioned names.  Unless you're going to spell them out with
> arrows (but you didn't mention fonts).  Sorry; that was sarcasm, not a
> request.  Perhaps it serves to illustrate my fundamental confusion.
> ok. well i'm looking at an 'abstract' version of it that works like
> category theory or set theory, just like i've been looking for an arrow
> language operator to postulate arrows, much like 'there exists' and 'for
> all' operators.
> >the main new benefit is arbitrary
> >cross-cutting of ontologies _without_ the invocation of any special
> >reflection method.
> In the vulgar tongue, you can slice theories of knowledge in half.  Sounds
> ... unique.  Doesn't sound useful, unless you're a really agressive
> debator.  I'm joking, but I'm serious that this sentance doesn't mmake any
> sense to me.
> umm.  it's sort of like having everything open for inspection and
> expressed in a language of concepts that is consistent and easy to perform
> transformations on.  (i need more time to put something together to
> explain this, so you'll have to wait, i guess.)
> >in other words, this language reflects "right out of the
> >box", so to speak.  aspect-orientation will emerge as a property of the
> >kinds of systems that Tunes requires.  (obviously this last statement
> begs
> >explanation.  this will follow soon.)
> Love to see it, although strangely enough I grok this more fully than most
> of the rest of what you're saying.  I don't know how your system will
> "implement" it, but I suspect I understand it anyhow.
> the language just sees all the things about the computer and the
> information on it not as described by arrow structures, but as the arrow
> structures themselves.  (to be continued...)