Melinda Shore wrote:
Keith Moore wrote:
I think it's interesting that you appear to consider "multiple addresses
per host" a narrower problem than "multiple network connections per
host", whereas I consider the latter to be a subset of the former.
Yeah - I do, too.
To expand on this a little, I think it's useful to
frame this in terms of policy (which seems to be the
direction it's already headed) and identify which
policies are attributes of which entities. There are
policies that apply to the network device (routing,
for example), policies which apply to an interface
(filtering and maybe QoS), and policies which apply to
a network. I don't think the distinctions are always
all that clean but I do think it's a useful exercise
to try to get some clarity around what belongs to whom.
If nothing else it might help sort out what's a tractable
problem and what isn't.
Policies are certainly an interesting set of questions raised by these
circumstances. But they do not stand on their own, they interact with
other concerns. I don't see any benefit in fragmenting the discussion
so that policies are discussed separately from, say, application level
issues, or network administration issues, or host administration issues
(note that the latter two are not the same), etc. To the contrary, I
don't think a policy discussion can be realistic unless it's held in
light of those other concerns.
And I don't know why you think that the discussion is "already headed"
toward policy when the group isn't even chartered yet. Certainly the
discussion on the IETF list isn't "already headed" that way.
Ietf mailing list