[AFS3-std] Re: XCB union decoding & IPvN address Conversation

[Andrew Deason]

On Mon, 14 Feb 2011 17:28:22 -0500
Jeffrey Hutzelman <jhutz@cmu.edu> wrote:

> The remainder of the discussion has been almost entirely about
> representation, though the what-to-store question has come up briefly
> once or twice.  I think we mostly have consensus that a new type
> should be about addresses, and not something more complete, but I
> could be misreading -- I have a fairly strong opinion on that
> particular question.

Well, I see nothing limiting us to just that one type.

I don't see a problem with having an address extensible union type (no
port), a "port" extensible union type (really a "transport-specific
data" type), and a struct that just has those two things (an "endpoint"
type). Then we have both a "just an address" type, and a convenient
struct for RPCs to use as an endpoint.  The decision for how many layers
to include in the address specification can just be done a per-RPC
basis; if something wants more than just the address and the port, just
make a new struct.

Actually, I see a slight problem in how I find it likely that all
addresses ever encoded with this will be very inefficient. But I can
live with it.

-- 
Andrew Deason
adeason@sinenomine.net