Re: Some Comments on ID/Loc Separation Proposals

[Dave Crocker <dhc@dcrocker.net>]

Erik,

EN> Do you intentionally want the definition of endpoint to be open to
EN> the endpoint being any of
EN>  - (an instance of) the IP protocol processing on a node
EN>  - a IP service access point (i.e. the point where the transport protocols
EN>    attach to IP)
EN>  - a transport service access point 
EN>  - something higher up in the protocol stack

1.  I want the term to be precise.  Anything in the definition that
permits ambiguity is a bad thing, in my view.  So you should assume
that any of the text I offered that results in ambiguity is an error
in my text and needs to be fixed.

2. The difference in focus, between IP processing, transport
processing, and application processing is exactly what makes all this
difficult.  I am loathe to believe that we need lots of new terms.  So
I think we need to understand exactly how we want to use this one (or
two) new term(s) and then calculate the definition back from it.

And, yes, the fact that I am not giving you a more concrete response
is intentional. I've offered some specific text.  I want others to fix
it.  (Or, rather, I can't come up with anything better...)


EN> The reason I'm asking is because there are proposals that explore the
EN> first three in the list. If you want to keep the definition open
EN> to applying at different layers it would be good to make that very explicit
EN> in the definition.

I think it will serve us well to _start_ with reference to a local
instance of an application, and then derive text back from it.
Focusing only on the lower layers causes us to miss what is most
interesting and useful, I fear.



d/
--
 Dave Crocker <dcrocker-at-brandenburg-dot-com>
 Brandenburg InternetWorking <www.brandenburg.com>
 Sunnyvale, CA  USA <tel:+1.408.246.8253>