Re: [Fwd: I-D Action: draft-carpenter-6man-why64-00.txt]

Le 09/01/2014 04:19, Lorenzo Colitti a écrit :
> On Thu, Jan 9, 2014 at 11:58 AM, Brian E Carpenter
> <brian.e.carpenter@gmail.com <mailto:brian.e.carpenter@gmail.com>> wrote:
>
>      > "For all unicast addresses, except those that start with the
>     binary value
>      > 000, Interface IDs are required to be 64 bits long".
>
>     Actually that (truncated) sentence is obsoleted by
>     draft-ietf-6man-ug which
>     is in the RFC queue. New version:
>
>         "For all unicast addresses, except those that start with the binary
>          value 000, Interface IDs are required to be 64 bits long. If
>     derived
>
>
> ... which is exactly the sentence I cited, right? I did look at
> draft-ietf-6man-ug before I blurted out, you know :-)

Let me try a clarification.  If it dares too far, excuses.

RFC4291 Addr Arch document says prefixes are of length n and IIDs of 
length 128-n.  Further it says that IIDs are required to be 64bit long.

A new document now in RFC Editor queue says that IIDs must be 64bit long 
and, additionally, when derived from a 48bit MAC address they must be 
constructed as 'Modified EUI-64'. (this is part of the u/g discussion).

This is not the end of the story, but could be a path to it.  Let me say 
why.

One agrees that addresses are n/128-n.  Then one makes them 64.  Then 
make them 64-but.  And further one may try to interpret them as along 
these hypothetical lines:
>    "For all unicast addresses, except those that start with the binary
>     value 000, Interface IDs are required to be 128-n bits long. If derived
>     from an IEEE MAC-layer address, they must be constructed in Modified
>     EUI-64 format."

This is just speculation on my part.

Alex

>
>
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> ipv6@ietf.org
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>