>
>   | I was wondering if this has been mentioned before.
>
> It has been discussed before.   It is a truly wonderful idea...
>
> If it could be done, there'd be no adverse effects, the problem
> is how the extra identifier gets distributed to the nodes.   Your
> average system has no idea what its vlan-id might happen to be (not
> that there's any reason to restrict the usage that way).

it doesn't matter i reckon. I mean from the other nodes point of view it's
just another link-local address ( well, as long as
when it determines what the address type is, it matches for fe80::/10 and
not
fe80::/64

>
> It cannot use the net to discover the id, as at the stage it is
> configuring its link local address, it has no way to use the net.
>
> The earlier suggestion kept the 0's in the format on the wire, and
> just used the id for local internal identification (within the node).
> That means that each node cinvent its own numbers, they don't need to
> be consistent with those chosen by anyone else.   But having the
> address the node uses different from the one passed to other nodes
> complicates any protocol that actually uses the value of the address
> (the node has to know to use the address with the ID in it for addressing,
> but the address with 0's replacing the ID for all other purposes, which is
> messy).
>
> In any case, this suggestion went nowhere.

as long as the node continues to use,e.g. fe80::1:2c0:dfff:fe07:6d62 for the
specified
interface i don't see why this would be a problem to other nodes. Only when
implementations start matching
to the /64 boundary instead of the /10 boundary this may be a problem.

using the format fe80::<id>:<interface-id>, would allow a single link-local
addresses
to be shared among various interfaces on the node and will also enable the
node to determine
which interface a destination link-local address is going out of. Also, why
not fe80:<zone-id>::<interface-id>
instead of address%<zone-id>? Then you wouldn't have problems described in
draft-ietf-ipngwg-scoping-arch-03.txt for URLs

but then again it would be difficult to say fe80:eth0::<interface-id> or
fe80:vlan0::<interface-id>
unless you can map eth0 and vlan0 or vlan1 to a standardised numerical index
or value.

Regards,
Sean



--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to