>>>>> "Joe" == Joe Salowey <[email protected]> writes:
Joe> (http://tools.ietf.org/html/draft-aboba-radext-wlan-13) - this
Joe> is an draft attribute designed to indicate the EAP lower layer.
Joe> It contains several values to cover many of the EAP lower
Joe> layers. IT does cover IKEv2 and PANA. Its likely that list
Joe> values need to be augmented and cleaned up.
Joe> NAS-Port-Type and Tunnel-Type seem to have limitations. My
Joe> initial suggestion is the following:
Joe> - Define the EAP-Lower-Layer attribute in the document
Since there are are already approved normative references waiting on
eap-chbind, I'd prefer to define the attribute here.
However if Bernard's draft is ready to last call I'd be fine with that.
Joe> - If the lower layer protocol does not define a specific
Joe> attribute indicate the lower layer type then EAP-Lower-Layer
Joe> attributes MUST be included. Would it also be useful to
Joe> include NAS-Port-Type?
I think nas-port-type should be a SHOULD here.
For things like abfab it makes no sense at all.
for things like PANA it may or may not make sense.
Joe> - If the lower layer protocol does define a specific attribute
Joe> then that attribute MUST be included and EAP-Lower-Layer MAY be
Joe> included. Would it be easier just to always use
Joe> EAP-Lower-Layer?
I'd have a hard time justifying such a MUST for something like abfab on
the grounds RFC 2119 gives for when a MUST is appropriate. I think
SHOULD rather than MAY include eap-lower-layer better captures what's
going on.
Joe> Thoughts?
I'd be happy to implement this within the next couple of weeks.
Obviously if Bernard's draft is ready and we can just send it on and use
eap-lower-layer that would be less work for me.
_______________________________________________
Emu mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/emu
