Re: [Autoconf] Updated ad hoc addressing model document

[Carlos Jesús Bernardos Cano <[e-mail]>]

Hello Charlie,

On Tue, 2010-02-09 at 09:51 -0800, Charles E. Perkins wrote:
> Hello Carlos,
> 
> Questions/comments inline:
> 
> 
> On 2/9/2010 9:17 AM, Carlos Jesús Bernardos Cano wrote:
> 
> >> Agreed that standard "not that clever" behavior of the IP stack is
> >> putting prefixes, configured on interfaces, in the routing table?
> 
> What "not that clever" agent is putting them there?
> 
> >> Agreed that putting other prefixes is 'something clever'?

If an prefix::iid/lprefix on-link address is configured in one
interface, then a route to that prefix prefix:://lprefix is configured
and I can communicate with other addresses of the same prefix on the
same L2 link. If the address is off-link, something else is needed to
communicate.

A good example of the above are link-locals.

With current document, we always need something else to communicate.
This is not something I'm necessarily against, I just said that I agree
that something else is needed with current addressing model.

Carlos

> 
> "other prefixes" == ???
> 
> >> Agreed that with the proposed addressing model, under conditions that
> >> the 'something clever' is not functioning, L3 communication fails for
> >> links between two nodes that have L2 communication?
> >
> > I agree with this.
> 
> But the document says:
> 
> =>    If L2 communication is enabled between a pair of interfaces, IP
> =>    packet exchange is enabled regardless of the IP subnet configuration
> =>    on each of these interfaces.
> 
> Thus, Teco's assertion is wrong.
> 
> It seems to me that each of the three parts quoted above
> are either highly questionable, mysterious, or wrong.
> 
> What's to agree with?
> 
> Regards,
> Charlie P.

-- 
Carlos Jesús Bernardos Cano     http://www.netcoms.net
GPG FP: D29B 0A6A 639A A561 93CA  4D55 35DC BA4D D170 4F67