ALU capability (was Re: [E-Lang] Authority -- what is its dual?)

[Mark S. Miller]

First, Dean, thanks for saying well just about everything I was trying to 
figure out how to say.  I feel I only need to add some emphasis to one point.


At 09:13 PM 10/18/2001 Thursday, Jonathan A Rees wrote:
>- It forces you to always decide who's on top in any interaction,

*THIS* is precisely the main reason I'm such a strong oo partisan, and why I 
find approaches like CLOS so abhorrent.  Hewitt got this precisely right and 
for the right reasons -- the "Open Systems" oriented constraints.  For 
Hewitt's Open Systems computation, of which distributed secure computation 
is a major facet, every action is taken by some entity, and it is important 
to be clear about which entity that is.  Just as there shall not be any 
ambient authority, there shall not be any ambient actions or knowledge.

When a computational model allows computation to be expressed in which no 
one is on top, then no one is taking responsibility, no one is to be or not 
to be trusted to compute a sensible answer, etc.  Such systems seem to be 
built by people who believe in objective and accessible truth.

OO systems, on the other hand, seem to be built by people who believe 
neither in the naively objective nor in solipsism, but in the 
inter-subjective.  My own statement of the intersubjective (from 
http://www.erights.org/talks/pisa/for-don/siframes.htm , derived from a
paper by Don Lavoie):

>We each work to partially understand and value the world and 
>each other, but each starting from and in service of our own unique 
>endowment of knowledge, situation, relationships, purposes, atitudes, and 
>aptitudes.

Although this may state it better than Hewitt, it's essentially what Hewitt 
means when he talks epistemologically about Open Systems.  Although Alan Kay 
has been less clear, I think it's also reasonably representative of his 
view, and is among his motivations for objects.

(Note that the above quote is from a context where I'm talking about humans 
as the agents.  Were I talking about computation, I would have said it 
differently.)

I know that the above may seem vastly more touchy feely than anything said 
so far in this thread, but that's the nature of the subject we're dealing with.

At 02:36 AM 10/19/2001 Friday, Dean Tribble wrote:
>>I stayed up too late!
>
>Tell me about it!  Look at the time :-)

Indeed.


        Cheers,
        --MarkM