lastest model revision

[Greg Pavlik <greg.pavlik@oracle.com>]

Mark, I've reviewed your changes and I have done two things: some basic 
reordering and I deleted the first paragraph under Why Nesting? The 
problem I have is that it shows contexts inside of contexts in a way 
that is different from previous email discusssions. So in the interest 
of moving forward, I suggest we work on that issue separately without 
holding up the model in other respects.

At one point, I thought we were aiming at:
<ctx>
    <protocol>foo+bar</protocol>
    <foo data/>
    <bar data/>
</ctx>

though in the last draft we have
<ctx>
    <context1>
    <context2>
</ctx>

Not saying it's wrong, but I don't know how to interpret that.

also, to answer your three comments:

1) no, I don't think that the text implies only one ALS per activity. It 
just gives an example of two separate activities that overlap.
2) agreed, ALSes are treated as optional.
3) for now, yes, I think we should restrict nesting to matching protocol 
types if possible (KISS). Otherwise, we're going to eat alot of cycles 
in the short term.

CAF model clarification v5.doc