On Thu, Jun 23, 2011 at 7:11 PM, Noel Chiappa <[email protected]> wrote:
>    > From: Alia Atlas <[email protected]>
>
>    > the question is whether the same bitpattern must belong to the same
>    > site and end-host.
>
> I would tend to say 'yes' - but I'm sure some bright spark will find a use
> for having that not be true!

Useful or not - that is not possible with what is currently specified.

>    > I'm not happy about separate namespaces not being separate (e.g.
>    > independently allocatable) without that being clearly written down.
>    > What is written down is that they are separate name-spaces.
>
> I think that's more of a goal, or perhaps more accurately a gross
> simplification made for the sake of a clear, simple explanation, than a
> strict statement of fact. (If nothing else, they aren't independently
> allocated.)
>
> So maybe the wording could use some clarification - as long as we don't
> make it harder to understand for the average network engineer out there
> (especially bearing in mind what Bob Dobbs said about average people...)

I strongly think it needs clarification; Joel's text is a good start.
Certainly, starting from the natural
assumption about two namespaces being completely independent, it was "clear"
that there were problems as specified.

Alia
_______________________________________________
lisp mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lisp

Reply via email to