Re: I-D Action: draft-smith-6man-in-flight-eh-insertion-harmful-00.txt

[Nick Hilliard <[e-mail]>]

Brian E Carpenter wrote on 13/10/2019 20:34:
> Of course. In the limited-domains draft we're advocating an explicit
> mechanism for defining the boundary; which of course wouldn't automatically
> make it watertight, but would offer some hope of making it watertight.
> But that being hand-waving, I'll stick to what I said: defining the boundary
> by means of adddress scope and router config is the state of the art.

you can define this as carefully you want, but people are going to 
ignore the scoping references and will expect that if the underlying 
protocol is notionally based on "ipv6", that it will operate correctly 
in general purpose ipv6 networks, e.g. the wider internet.

If your suggestion is to produce a protocol which is only expected to 
work within carefully defined walled gardens, that's fine and I think 
you should go ahead and do that.

But please don't say it's ipv6 or that it's compatible with ipv6 because 
it's not; it's something else.

Nick