What about two or more instances of the SAME Info package? Do we need to say anything about that, and if so what? If we limit it to one Info package, this issue doesn't arise.
John > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Dean Willis > Sent: 17 November 2008 20:00 > To: Jonathan Rosenberg > Cc: IETF SIP List > Subject: Re: [Sip] comments on draft-ietf-sip-info-events-01 > > > On Nov 17, 2008, at 11:14 AM, Jonathan Rosenberg wrote: > > > > We do not *NEED* to define what happens when INFO has > multiple info > > packages in it. If the syntax allows just one, and the spec allows > > just one, there can only be one. > > > > And I will reiterate that, this is a separate question from > whether > > an INFO can contain multipart, where one body is an INFO package > > object and the other is for a different reason (AIB). We are > > debating whether the INFO framework itself allows for multiple > > packages per INFO. > > > But if there are multiple body parts (only one of which is an INFO > body) , INFO requires a mechanism for finding the right body part. > > We can't delimit this based on the MIME type for the same reason we > can't just do a one-to-one map between INFO and MIME -- sometimes a > single MIME type is used in different ways for different applications. > > So far, the only reasonable way proposed to identify the MIME body > associated with an INFO package is a MIME CID identifier that points > from the INFO package header to the body. According to this model, > INFO compliant UAs must understand multipart MIME (as the body > handling draft requires), and they must understand Content-ID. > > If we have this mechanism, it works just as effectively and > automatically for multiple INFO bodies per message as it does for a > singular body per message. > > > > We could probably say nothing about singular or multiple INFO > package > bodies per INFO message in the draft, and nothing would > change in the > mechanism at all: > > 1) Parse header > 2) If current header is an Info package identifier, parse out the CID > 3) Search through the bodies until you find the right CID > 4) If there are more headers, go to step 1 > > > One could probably complicate this by adding a state about "But if > you've already found one Info package header, ignore all subsequent > ones and their associated body parts", but that actually makes the > logic MORE complicated, not less. > > -- > Dean > > _______________________________________________ > Sip mailing list https://www.ietf.org/mailman/listinfo/sip > This list is for NEW development of the core SIP Protocol > Use [EMAIL PROTECTED] for questions on current sip > Use [EMAIL PROTECTED] for new developments on the application of sip > _______________________________________________ Sip mailing list https://www.ietf.org/mailman/listinfo/sip This list is for NEW development of the core SIP Protocol Use [EMAIL PROTECTED] for questions on current sip Use [EMAIL PROTECTED] for new developments on the application of sip
