Re: [Stox] I-D Action: draft-ietf-stox-im-04.txt - forking MESSAGE

[Peter Saint-Andre <[e-mail]>]

On 10/03/2013 12:32 AM, Olle E. Johansson wrote:
> 
> 3 okt 2013 kl. 00:44 skrev Peter Saint-Andre <[e-mail]>:
> 
>>> - A SIP MESSAGE can fork. The draft doesn't mention that,
>>> something that may have impact on implementations. There's a JID on
>>> one side - but a forking address on the other side of this spec. -
>>> A SIP MESSAGE can be delivered inside of a dialog and outside of a
>>> dialog. Inside of a dialog is maybe out of scope for this draft,
>>> but in that case the remote target will be a contact address, maybe
>>> even a GRUU.
>>
>> Hmm, and here I thought we might be able to discuss forking only in
>> the media I-D. But I think you're right that we need to discuss it
>> here, too. What specification has the best coverage of forking? Or is
>> this part of the "oral tradition" in the SIP community?
> 
> Forking is covered in RFC 3261. Luckily there's no "early text" or that kind of issues
> with forking a message - you will only get one 200 OK back (as opposed to INVITE),
> which also means you will not see how many destinations that got the MESSAGE.
> You can not CANCEL a MESSAGE either. You may get authentication requests as always.
> 
> RFC 4328 states:
> 
> "When one user wishes to send an instant message to another, the
>    sender formulates and issues a SIP request using the new MESSAGE
>    method defined by this document.  The Request-URI of this request
>    will normally be the "address of record" for the recipient of the
>    instant message, but it may be a device address in situations where
>    the client has current information about the recipient's location."
> 
> So if you have a subscription for registration information on the AOR,
> you might select a contact/GRUU to send a message too. 
> 
> It also states:
> 
> "Note that a downstream proxy could fork a MESSAGE request.  If this
>    occurs, the forking proxy will forward one final response upstream,
>    even though it may receive multiple final responses.  The UAC will
>    have no way to detect whether or not a fork occurs.  Therefore the
>    UAC MUST NOT assume that a given final response represents the only
>    UAS that receives the request.  For example, multiple branches of a
>    fork could have resulted in 2xx responses.  Even though the UAC only
>    sees one of those responses, the request has in fact been received by
>    the second device as well.
> "

Thanks for the clarifications! I'll work to incorporate some of those
insights into the -im document.

> All of this makes MESSAGE quite a poor solution for messages you want kept private.
> You don't know where it's sent and even if there's authentication on the receiving side,
> anyone that gets the unauthenticated request still can access the message unless
> of course there's S/MIME involved. 

Indeed.

> Well, there was a reason MSRP was developed later. ;-)

"If at first you don't succeed..."

Mind you, we've had our share of failed extensions in the XMPP commuity,
too.

Peter