Re: [Tsvwg] Comments on draft-ietf-tsvwg-rsvp-user-error-spec-03.txt

["Adrian Farrel" <[e-mail]>]

Thanks Bruce!

I made the changes you suggested and added a note about ResvConf.

New version submitted.

Cheers,
Adrian
----- Original Message ----- 
From: "Bruce Davie" <[e-mail]>
To: "Adrian Farrel" <[e-mail]>
Cc: <[e-mail]>; "George Swallow" <[e-mail]>
Sent: Friday, March 28, 2008 4:52 PM
Subject: Re: Comments on draft-ietf-tsvwg-rsvp-user-error-spec-03.txt 


> 
> Adrian,
> [points of agreement clipped]
> 
> On Mar 28, 2008, at 10:58 AM, Adrian Farrel wrote:
> 
>>
>>>> 4.2
>>> [...]
>>>>    Implementations receiving a USER_ERROR_SPEC object on some  
>>>> message
>>>>    other than a PathErr, ResvErr, or Notify message MUST treat the
>>>>    error as a malformed message and process according to [RFC2205].
>>> RFC2205 allows ERROR_SPEC objects in ResvConf messages. I suggest  
>>> the  USER_ERROR_SPEC should also be allowed to appear in those  
>>> messages.
>>
>> In fact, the ERROR_SPEC is mandatory on the ResvConf, but RFC 2205  
>> says:
>>
>>         The ERROR_SPEC is
>>         used only to carry the IP address of the originating node, in
>>         the Error Node Address; the Error Code and Value are zero to
>>         indicate a confirmation.
>>
>> So there would never be cause to include USER_ERROR_SPEC. Or would  
>> there?
> 
> Your reasoning seems sound to me. Perhaps you want to put in a  
> sentence along these lines to avoid future readers scratching their  
> heads as I did.
> 
> So, with the few changes that I sent previously and that you agreed  
> to, I think this document would be ready for publication.
> 
> Bruce
> 
>