Re: [MMUSIC] BUNDLE TEXT: Offerer procedures (June 12th)

[Paul Kyzivat <[e-mail]>]

Sure, this is all good.

On 6/14/13 9:30 AM, Christer Holmberg wrote:
>
> Hi Paul,
>
> I will try to submit a new version of the BUNDLE draft today, or at latest during the weekend.
>
> I have done quite much re-structuring/re-wording, and I also try to implement most of your suggestions.
>
> Allowing the Offerer to assign a new BUNDLE address to multiple "m=" lines will require some editorial work, though, so I would like to have some more discussions about it before I change the text. Just to make sure you don't think that I don't care about your opinion :)
>
> Regards,
>
> Christer
>
>
>
> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On Behalf Of Christer Holmberg
> Sent: 14. kesäkuuta 2013 14:36
> To: Paul Kyzivat; [e-mail]
> Subject: Re: [MMUSIC] BUNDLE TEXT: Offerer procedures (June 12th)
>
> Hi,
>
>>> 6.1.  General
>>>
>>>      This section describes the usage of the SDP Offer/Answer mechanism
>>>      [RFC3264] to negotiate the usage of the BUNDLE mechanism, to
>>>      negotiate the bundle address, and to add, remove and reject SDP
>>>      Media Descriptions ("m=" lines) associated with a BUNDLE group.
>>>
>>>      The generic rules and procedures defined in [RFC3264] and
>>>      [RFC5888]apply when the SDP Offer/Answer mechanism is used for BUNDLE.  For
>>>      example, if an SDP Offer is rejected, the previously negotiated  SDP
>>>      parameters and characteristics (including those associated with BUNDLE groups) apply.
>>>
>>>      When an endpoint generates an SDP Offer, or an SDP Answer, which
>>>      contains a BUNDLE group, the endpoint MUST assign SDP media-level
>>>      mid value for each "m=" line in the BUNDLE group.  In addition, the
>>>      endpoint MUST insert an SDP session-level group:BUNDLE attribute,
>>>      and place each mid value associated with the BUNDLE group in the
>>>      attribute mid value list.
>>>
>>>      Until the Offerer knows whether the Answerer supports the BUNDLE
>>>      mechanism, the Offerer MUST, for each "m=" line in a BUNDLE group:
>>
>> Under what circumstances may the offerer presume to *know* whether the answerer supports BUNDLE?
>>
>> The minimal answer is that if an offer included a bundle group, and
>> the answer also contains a corresponding bundle group, then the answerer supported *that* bundle at the time of *that* answer, and continues to do so at least until there is another offer.
>
> Yes.
>
> I was actually thinking about removing the "until the Offerer knows" text, and simply generally say that the bundle address is negotiated by assigning a unique address to each "m=" line.
>
> But, as you want to be able to assign a bundle address to multiple "m=" lines before it has been negotiated, it would not work.
>
>> But we want to assume more - at least that if *a* bundle was supported
>> by both sides in one O/A exchange, then both sides can be presumed to
>> *understand* bundle, including the use of the same address info for multiple m-lines in a bundle offer, in future offers within the same signaling session.
>
> I would rather say SDP session, to keep it independent from the signaling protocol. That is the same as for any other parameter you negotiate using SDP O/A - it's valid until re-negotiated.
>
>> Is, or should there be, a way to go back on that presumption - to stop
>> supporting bundle? I guess if an end wanted to stop supporting bundle
>> while a bundle was in effect it would need to send an offer that didn't include any bundles. But it might want to send such an offer just because it doesn't currently need a bundle though it still supports them.
>
> Sure, but if anyone then wants to start using BUNDLE again, it will send an Offer and suggest the usage.
>
>> Perhaps we should say that an offerer or answerer who supports bundle
>> should *always* include an a=group:bundle in its offers and answers. If it doesn't currently want a bundle it can simply include "a=group:bundle" without any mid  values.
>
> Or, we could leave it up to the signaling protocol. For SIP, one could define a media feature tag. That would also allow SIP entities to request that INVITEs are sent to entities to support BUNDLE (and have indicated so during registration).
>
>
> --------------------------------------
>
>
>>> 6.2.1.  SDP Offerer Bundle Address Request and Usage
>>>
>>>      An Offerer can assign an address to multiple "m=" lines in a
>>>      BUNDLE group, once the address has been selected as the Offerer's local
>>>      bundle address.  An Offerer MUST NOT assign an address to multiple
>>>      "m=" lines until the address has been selected as a bundle address
>>>      for that BUNDLE group.
>>>
>>>      In order to negotiate (or, to re-negotiate) the bundle address
>>>      associated with a BUNDLE group, the Offerer, in the SDP Offer,
>>>      assigns a unique address to each "m=" line in the BUNDLE group.
>>>      In addition, the Offerer indicates which unique address it wishes to
>>>      use as its local bundle address.  There Offerer places the mid value,
>>>      associated with the "m=" line that contains the preferred address
>>>      first in the SDP group:BUNDLE attribute mid list.  The Answerer
>>>      then selects the local bundle address for the Offerer ([ref-to-be-added]).
>>
>> I don't understand what the last sentence above is trying to tell me.
>
> The intention is to say that the Answerer selects the Offerer's bundle address.
>
>
> --------------------------------------
>
>
>>>      If the Offerer, in a subsequent SDP Offer, wants to re-negotiate
>>>      the bundle address associated with a BUNDLE group, it MAY assign the
>>>      previously negotiated bundle address as a unique address to one of
>>>      the "m=" lines in the BUNDLE group.
>>>
>>>      If the Offerer assigns the bundle address to more than one "m="
>>>      line in a BUNDLE group, the first mid value in the SDP group:BUNDLE
>>>      attribute mid value list MUST represent an "m=" line to which the
>>>      bundle address is assigned.  Hence, in order to re-negotiate the
>>>      bundle address, the Offerer needs to assign a unique address to
>>>      each "m=" line in the BUNDLE group, as described above.
>>
>> I still object to this. It is a significant cost, and I don't understand how it adds any value over simply assigning a new address to all the bundled m-lines.
>
> The problem is not assigning a new address to all m- lines.
>
>>From a specification perspective, it becomes a little more tricky when you add a new bundle address to some m- lines, and unique addresses to others.
>
> Also, the Answerer will have check and figure out that the suggested new BUNDLE address has been assigned to multiple "m=" lines, and that it therefore cannot move it out of the BUNDLE group without rejecting it, etc.
>
> I guess one alternative could be to say that a suggested new BUNDLE address would have to be assigned to each "m=" line in the BUNDLE group. I.e. the Offerer would not be allowed to assign both a new BUNDLE address and unique addresses (e.g. for "m=" lines being added to the BUNDLE group).
>
> --------------------------------------
>
>
>>> 6.2.2.  Bundle Address Synchronization (BAS)
>>>
>>>      When an Offerer has sent an SDP Offer, a bundle address has been
>>>      selected by the Answerer, and the bundle address was in the SDP
>>>      Offer not assigned to each "m=" line in the BUNDLE group, the Offerer
>>>      MUST send a new SDP Offer, in which it assigns the negotiated bundle
>>>      address to each "m=" line in the BUNDLE group.  This procedure is
>>>      referred to as Bundle Address Synchronization (BAS).
>>>
>>>      The Offerer MUST NOT modify any parameters associated with the
>>>      BUNDLE group when it performs a BAS.
>>
>> I still don't understand the point of this restriction.
>> What value does it add?
>>
>> Suppose that a need to make another change arises after the offer that
>> proposed the bundle, but before the answer has been received. Then, when the answer is received it will be necessary to send a BAS before sending another offer to make the new change. Why wait?
>>
>> ISTM that it would be better to say that after receiving an answer, if
>> there is at that moment no need to send another offer, if the corresponding offer had bundled m-lines with different addresses, then a BAS must be sent.
>
> The main reason for not allowing any parameter modifications in the BAS was to decrease the likelihood that the Answerer would reject the BAS (because it doesn't accept the suggested parameter changes).
>
> So, we can for sure allow parameter modifications in a BAS, but I think we need some text saying that an Offerer must perform a BAS until it is accepted.
>
>
> --------------------------------------
>
>
>>> 6.2.4.  Removing A Media Description From A BUNDLE Group
>>>
>>>      When an Offerer removes an "m=" line from a BUNDLE group, the
>>>      Offerer MUST assign a unique address to the "m=" line that is removed.  In
>>>      addition, the Offerer MUST NOT place the mid value that was
>>>      previously defined for the "m=" line in the SDP group:BUNDLE
>>>      attribute mid value list associated with the BUNDLE group.
>>
>> nit: I don't think this should say anything about the mid value that
>> was previously defined for the m-line. That implies that mid values must be preserved from one o/a to another. While they probably will in most cases I don't think that should be required.
>> If I want, I should be able to "relabel" the m-lines with new mid
>> values. This reduces the amount of state that is required to be preserved from one o/a to the next.
>>
>> The statement "remove an m-line from a bundle group" already implies
>> that the group does not contain a mid-value identifying that m-line
>> (in
>> *this* offer or answer).
>
> Maybe I could say that the mid value that was previously defined is no more valid, and must not be present in any Offer or Answer.
>
>
> --------------------------------------
>
>
>>> 6.2.5.  Rejecting A Media Description In A BUNDLE Group
>>>
>>> I agree with prior comment that using "Rejecting" here is weird and confusing. I agree that "disabling" is probably more appropriate.
>>> But then "disabling a media description in a bundle group" is still a little confusing, because it is then no longer "in" the bundle group.
>>>
>>> It might be clearer to just incorporate this into the prior section on removing a media description from a bundle group. E.g., in 6.2.4:
>>>
>>>     When an Offerer removes an "m=" line from a BUNDLE group, the Offerer
>>>     MUST assign either a unique address *or a zero port* to the "m=" line
>>>     that is removed.
>
> I at least intend to put back "disabling". I was also thinking about combining the sections, as you suggest, as the only difference is that the port value is set to zero in the reject/disable case.
>
>>>      When an Offerer rejects an "m=" line in a BUNDLE group, the
>>>      Offerer MUST assign a unique address, with a zero port value
>>> [RFC3264], to
>>
>> as long as the port is zero the address need not be otherwise unique.
>
> True. I guess your suggested text above ("unique address or a zero port") will fix that.
>
> As a side note, I intend to replace "unique address" and "bundle address" with "unique transport parameters" and "bundle transport parameters", as suggested by Martin.
>
> ...except in cases where the text actually talks about the IP address, that is :)
>
> Regards,
>
> Christer
> _______________________________________________
> mmusic mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mmusic
> _______________________________________________
> mmusic mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mmusic
>