The official way to do this is to submit an "interpretation request" to
the IEEE. The instructions can be found here:
http://standards.ieee.org/reading/ieee/interp/. The request will then
be referred to the appropriate working group for response and approval.
The earliest that this could be completed by 802.11 is during their
meeting during the week of May 14-18.
This would result in an official response from the 802.11 working group
and the IEEE. Generally, the response will include referral to the next
maintenance task group for inclusion in a corrigendum or the next
revision of the standard.
-Bob
-----Original Message-----
From: Alper Yegin [mailto:alper(_dot_)yegin(_at_)yegin(_dot_)org]
Sent: Thursday, April 06, 2006 4:22 PM
To: Bob O'Hara (boohara); 'Yoshihiro Ohba'; 'Sam Hartman'
Cc: ietf(_at_)ietf(_dot_)org
Subject: RE: Last Call: draft-ietf-pana-framework-06
Hi Bob,
I think the issue at hand is not the interpretation of text, but the
conflicting text in the 802.11i standard. David Nelson has already shed
some
light to this.
Text in Clause 5 opens the door for non-802.1X protocols to utilize the
uncontrolled port, and we didn't find any limitations on what those
protocols may be -- hence concluded that PANA could be one of those.
This is
where PANA WG came from. I'd not say this is an "interpretation", but
rather
spec-reading.
How are we (in fact, IEEE) going to resolve this?
Alper
-----Original Message-----
From: Bob O'Hara (boohara) [mailto:boohara(_at_)cisco(_dot_)com]
Sent: Thursday, April 06, 2006 4:49 PM
To: Alper Yegin; Yoshihiro Ohba; Sam Hartman
Cc: ietf(_at_)ietf(_dot_)org
Subject: RE: Last Call: draft-ietf-pana-framework-06
Alper,
In my reading of 802.11i, such an AP will not be compliant. In
addition, PANA needs to have the 802.11 mobile client also support
PANA.
Currently, PANA packets would also be dropped at the sending client,
prior to opening the 802.1X controlled port.
In Yoshi's email, referenced at the URL below, he says that it is
possible to interpret the text in the current 802.11ma revision draft
to
allow the PANA operations described for bootstrapping an 802.11i AP.
The IEEE is the only body that has the authority to interpret their
standards. For 802.11, IEEE delegates that authority to the working
group that I chair. I can't forecast what an official interpretation
response might say, but the text in clause 5 is taken from the
overview
section of the standard. The text prohibiting non-802.1X frames from
passing on the uncontrolled port is taken from clause 6, the
description
of the Data SAP. My opinion would be that the text in clause 6 would
be
given precedence, both because of its placement and because of its
specificity.
-Bob O'Hara, Chair, 802.11 Task Group m
-----Original Message-----
From: Alper Yegin [mailto:alper(_dot_)yegin(_at_)yegin(_dot_)org]
Sent: Wednesday, March 22, 2006 11:55 AM
To: Bob O'Hara (boohara); 'Yoshihiro Ohba'; 'Sam Hartman'
Cc: ietf(_at_)ietf(_dot_)org
Subject: RE: Last Call: draft-ietf-pana-framework-06
Hi Bob,
Last two e-mails were in response to Sam's "implementability" concern.
Note that, when an IEEE 802.11 AP is used as a PANA EP and
bootstrapped
for
IEEE 802.11i security, it is not an "off-the-shelf" AP. Some amount of
coding needs to be done on the AP. The question is, would this
modified
AP
still be compliant with IEEE 802.11i or not (no, PANA WG does not mean
to
propose a change to IEEE standards.)
Please see Yoshi's e-mail and let us know what you think.
http://www1.ietf.org/mail-archive/web/ietf/current/msg41011.html
Thanks.
Alper
-----Original Message-----
From: Bob O'Hara (boohara) [mailto:boohara(_at_)cisco(_dot_)com]
Sent: Wednesday, March 22, 2006 8:34 PM
To: Alper Yegin; Yoshihiro Ohba; Sam Hartman
Cc: ietf(_at_)ietf(_dot_)org
Subject: RE: Last Call: draft-ietf-pana-framework-06
I have no doubt that an implementation can be made to work, when one
has
control of all the layers. The question is whether PANA bootstrap
will
work when all that is supplied is a PANA layer that must operate
above
an existing, presumably standards compliant, 802.1X/802.11i
implementation. When one can bypass a restriction of 802.11i (which
says to drop non-802.1X frames on the uncontrolled port), then PANA
bootstrap is possible.
However, what authority has PANA to change a standard developed in
an
entirely different standards organization?
-Bob
-----Original Message-----
From: Alper Yegin [mailto:alper(_dot_)yegin(_at_)yegin(_dot_)org]
Sent: Wednesday, March 22, 2006 7:19 AM
To: 'Yoshihiro Ohba'; 'Sam Hartman'
Cc: ietf(_at_)ietf(_dot_)org
Subject: RE: Last Call: draft-ietf-pana-framework-06
We (Samsung) have an implementation as well.
Alper
-----Original Message-----
From: Yoshihiro Ohba [mailto:yohba(_at_)tari(_dot_)toshiba(_dot_)com]
Sent: Wednesday, March 22, 2006 12:02 AM
To: Sam Hartman
Cc: ietf(_at_)ietf(_dot_)org
Subject: Re: Last Call: draft-ietf-pana-framework-06
If implementability of the specification is an issue, my company
has
an implementation of bootstrapping 802.11i PSK mode based on
running
PANA over Uncontrolled Port. The implementation works without
modifying a WiFi hardware or its firmware. We have a plan to
publish
the source code of the implementation in Open Diameter project.
Regards,
Yoshihiro Ohba
On Tue, Mar 21, 2006 at 11:45:25AM -0500, Sam Hartman wrote:
"Yoshihiro" == Yoshihiro Ohba
<yohba(_at_)tari(_dot_)toshiba(_dot_)com>
writes:
e email discussion over
Yoshihiro> the EAP mailing list quoted below, I had a short
Yoshihiro> conversation on this issue with Jesse Walker
during
Yoshihiro> IEEE 802 interim meeting in January in order to
Yoshihiro> follow-up the email discussion and understand the
input
Yoshihiro> from Jesse more. As far as I understand, he
seemed
to
Yoshihiro> agree on this possible interpretation while he
Yoshihiro> mentioned that there is no existing 802.11i
Yoshihiro> implementation that uses 802.1X Uncontrolled Port
for
Yoshihiro> non-802.1X frame exchange, but I may be still
Yoshihiro> misunderstanding something. Also, for the sake
of
Yoshihiro> completeness of the email discussion over the EAP
Yoshihiro> mailing list, the following email that I sent in
Yoshihiro> response to msg03872 should be quoted as well:
Yoshihiro>
http://lists.frascone.com/pipermail/eap/msg03879.html.]
So, the implementability of our specifications is a significant
concern. If we do not expect there to be environments in which
a
feature of our spec will be implementable, then we should remove
that
feature.
Implementability is sufficiently important that RFC 2026
explicitly
gives the IESG the ability to request an implementation report
even
for publication at proposed standard when it has questions about
whether a particular feature can be implemented interoperably.
--Sam
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf