-----Original Message-----
From: Christian Vogt [mailto:chvogt(_at_)tm(_dot_)uka(_dot_)de]
Sent: Monday, February 05, 2007 12:53 PM
To: Henderson, Thomas R
Cc: Jeffrey Hutzelman; ietf(_at_)ietf(_dot_)org; secdir(_at_)mit(_dot_)edu;
David
Ward; Gonzalo Camarillo
Subject: Re: secdir review of draft-ietf-hip-mm-04.txt
The following rules seem appropriate:
> - a HIP host SHOULD send a NOTIFY error if an unsupported Locator
> Type is received in a LOCATOR parameter, when such Locator
is also declared to be the Preferred locator for the peer
> - otherwise, a HIP host MAY send a NOTIFY error if
an unsupported Locator Type is received in a LOCATOR parameter
Shouldn't the transmission of the NOTIFY be a "MUST" in the
special case
where the LOCATOR parameter contains /only/ locators of
unsupported type?
The preferred locator would in this case remain the same as before,
meaning that it would be in DEPRECATED status.
The main problem in using NOTIFY as a MUST is that the NOTIFY parameter
is defined as being optional in the base spec. I think that if we agree
that we want a MUST, we will have to define a more explicit and reliable
way to reject the locator, such as a LOCATOR_FAILURE parameter in the
UPDATE.
Tom
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf