Re: Joins on capabilities that have passed through different membranes

Kenton Varda <[email protected]> wrote:

>I'm interested to know if you've thought about the following problem...

It's not obvious to me that this one has a correct answer.  Rather, it seems to
me that what you want to happen varies depending on use case and intent.  You
might want the semantics to be "This revocation takes effect regardless of
which path to the ultimate object the holder of the cap uses".  You might want
it to be "as long as someone has some path to the object, then they still have
access to it".  These feel like joins with AND vs. OR like behavior, but that
may be pushing analogies too far.  It possible there are other combinations
that make sense.  As Bill pointed out, to the degree that the various objects
involved have various kinds of under-the-hood relationships with each other,
all manner of behavior is plausible.

Chip