John,

It should be SHOULD.  The M bit means "use" Tasteful.  The "O" bit means
use Stateful.  Two different contexts.  I was here when they were put in
ND and recall why.  One reason is that not everyone believed that just
stateless was acceptable and that was vision on those persons part.  The
reality is that most users will not use stateless but Statefull simply
out of habit first of all and second the trust model will not be
delegated to routers for IPv6 for some time until users are more
comfortable with the stateless model.  I also consider not doing a
SHOULD for these bits is irresponsible engineering of a standard on our
part.  If a user wants stateful which is clearly permitted in ND and in
the IPv6 architecture they should be able to use it. But reality is that
as any product engineer knows when you develop the code first time and
see MAYs you usually skip it. But that is just a side issue and logic.
The reason is as I said.  M/O bit says use stateful. For this to not be
a SHOULD is to denigrate a part of the IPv6 architecture and this is
simply wrong.

Putting words in that if you do this you will hurt yourself tells us it
should be a SHOULD.  If something can hurt users and this can it can
affect interoperability and deployment.  I could also argue if the "M"
bit is set and nodes don't go to a stateful server because they "cannot"
it is another form of DOS.

The case where I personally don't believe stateful is valid is for small
mobile devices for users like 3GPP, 802.11 hot-spots, military
operations, et al.  But in those cases.  Except in one scenario which I
won't get into.  Long term I see stateless used  more and more, but I
don't see Enterprises or any very large entity using stateless for a
very long time. But in cases where it might not make sense the market
will decide and not set the M bit.  But if they set it then that means
it's a SHOULD.

SHOULD means implement this unless you have a good reason to not
implement it.  For devices I spoke of above there is good reason to not
implement it IMO so don't.  

Small device only believers and die-hard stateless is the only model
should not use this forum to backdoor remove what is part of the IPv6
architecture.  I believe if we do that then it must be raised to a
higher level and question of Internet Architecture principles based on
the base specs for the architecture.

The other problem is that stateless brings with it many problems we will
have to deal with in deployment till we learn more. The users will see
this too and will be conservative and drop back to stateful.

This is a very serious issue.

Regards,
/jim

 

--------------------------------------------------------------------
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