Perhaps foolishly stepping in here when I am behind in other work :) ...

I agree that both the issue of being able to distinguish an unknown EH from
an unknown upper layer header and having a uniform start to extension
headers are important and both must be in place in order to get the full
value.  However, I am fine with these being resolved separately.

Since the distinction between an unknown extension header or an unknown
next layer protocol is important, let me informally propose a solution...

--
My suggestion for how to solve this problem lies in setting aside a next
header range for future extension headers.  For example set aside the range
240-247.

Within that, I'd suggest reserving 247 as a special case, to allow for
expansion.  This following probably wouldn't be officially defined now, but
my idea there is that if "next header=247", then the 3rd octet (or 3rd and
4th octet) of the extension header would further define which extension
header this is.

Also, we might want to make 246 (or 240-243) a special value to indicate
that this extension header "transforms" the rest of the packet in a way that
will make it impossible to meaningfully parse the header and what follows.
If it didn't already have a next header value, ESP would be an example of
this.  A mechanism like I suggested for 247 could make up for the fact that
this is a limited range.

Of course, the 3 ranges/subranges I mention could be different sizes than I
suggested and could be distinct ranges.

(Eventually you may have to deal with running out of values for upper layer
protocols, but I'm not going into that.)
--

Feel free to pick away...

-- Jim

On 3/20/08 1:21 PM, "Markku Savela" <[EMAIL PROTECTED]> wrote:
> This seems moot all text in the draft...

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to