[cap-talk] "Same" key

Mark S. Miller markm at cs.jhu.edu
Sun Feb 4 18:53:44 CST 2007


Dean Tribble wrote:
> PS Ken Kahn first articulated this stretching of the object notion in
> a way that I "got it", though as with much of concurrent logic
> languages, Udi Shapiro probably had already internalized it :-).  I
> think it was in "Language Design and Open Systems" by Ken Kahn and
> Mark Miller.


Much as I'd like to take partial credit, I think the paper you're thinking of is

   Kenneth M. Kahn, "Objects - A Fresh Look," ECOOP 1989, 207-223

I recommend it, but I was unable to find an online copy by googling. Ken?


David Hopwood wrote:
 >> Thus, the notion of objects is confusing, but the notion of references
 >> remains coherent.
 >
 > I disagree. The notion of objects used in the obj-cap model is quite
 > straightforward, and can model all of the examples above easily.

Yes. I don't know about "easily", but otherwise I agree with David. Fred's 
SCOLL formalism is based on the object-capability model. Fred & I talked about 
how one would use it to model Oz-E logic variable, which is a case that's a 
least as confusing as Joule channels, and one that closely resembles the 
issues that Ken had in mind when he wrote "Fresh Look". Fred and I agreed that 
we needed to model the individual logic variable itself as an object. I don't 
remember whether logic variables in Oz have multiple facets the way they do in 
FCP, but if so, then in SCOLL, we'd again model each facet of each logic 
variable as a separate object.

Fred & I haven't talked about Joule, and you and I haven't talked about SCOLL. 
But I suspect that the right way to model Joule in SCOLL is again to treat 
each facet of a Joule Channel as a separate object.

Hopefully, this will all become clearer when Fred releases a draft of his 
thesis ;).

-- 
Text by me above is hereby placed in the public domain

     Cheers,
     --MarkM


More information about the cap-talk mailing list