[stir] Comments on draft-peterson-passport-divert-01

[Robert Sparks <[e-mail]>]

Are Identity Header field values ordered? (I don't think so - I think 
they form
an unordered set, and if some intermediary reordered them it wouldn't change
what they mean). I ask because this draft talks about "earlier PASSport"s. I
suggest that langauge be changed to say a "different PASSport" and where
applicable "already in the message".

(Aside: I wonder if we created unnecessary complexity by not allowing the
passports to appear comma separated on a single Identity: line so it 
could be
handled like a generic header field.)

The first paragraph of 4.1 could be misread to say you are replacing an
Identity header field value rather than adding a new one. You could 
avoid that
by saying "The resulting compact form Identity header field value to add 
to the
message" or better, expand the example to show before and after.

In the second paragraph of 4.2, I think it would be good to point back to
4474bis for what to do when there are invalid Identity header field values.

The Security Considerations section says "the only consequence will be that
downstream entities will have less confidence that the PASSport was mean 
to be
associated with this call". I don't think that's quite right. If the call is
retargetted and no PASSport is created for the retargetting, the 
recipient must
treat the original PASSport as invalid and thus must ignore it.

Nits:

Abstract: "originally destination"

Introduction : "numebr"

Section 3: 3rd paragraph. The second sentence (starting 'The "orig" 
value MUST
be copied' is very complex - can it be separated into several sentences?

The document needs to use "header field" and "header field value" a lot of
places it currently uses "header"