[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: mech-v2-05pre



Ok, I willing to drop my objection.  It seem like
a slight overspecifiction, but that's better then
having ambiguous text.

It just seemed to me in the begining that this was
an implementation detail and how and where it's
handled was up to the implementation.

The way text reads now, it seems like it's the job
of the tunnel driver to drop the packet, not the l3 (which
in my opinion is kind of backwards, i.e. ethernet doesn't
look at the IP header).

-vlad

Pekka Savola wrote:
On Thu, 26 Aug 2004, Fred Templin wrote:

--- Pekka Savola <pekkas@netcore.fi> wrote:

Of course, if there's another spec down the road which modifies proto-41 configured tunnels behaviour somehow so that other IP protocols than just "6" are OK, that's fine as well. The question is just about whether unmodified configured tunnels could support protocols other than "6", which doesn't seem to be the case..

As far as I can tell, no one is suggesting *support* for protocols other than "6".

The suggestion is for the specification to declare as out-of-scope
the behavior for packets discovered to have version other than "6"
after the encapsulating IPv4 header is discarded.


Right, and in such a case there'a very little difference to just
specifying drop here (and specifying the other protocols elsewhere),
or declaring it out of scope (and implying a drop and specification
elsewhere).

Just because this spec says they MUST be dropped doesn't mean some other spec cannot say that instead of dropping, something else needs to be done about them.

(Example: options which are "must be zero and ignored", which later
specifications define and specify meanings for.  In this case, there's
just no need for "ignored" because we don't expect interoperability.)


-- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Vladislav Yasevich Linux and Open Source Lab Hewlett Packard Tel: (603) 884-1079 Nashua, NH 03062 ZKO3-3/T07