Re: [MMUSIC] Mismatch between sdp-bundle-negotiation and RFC5888

[Roman Shpount <[e-mail]>]

On Sun, Dec 2, 2018 at 1:22 PM Christer Holmberg <
[e-mail]> wrote:

> >>>>> I know this is a really late notice, but I think there is a slight
> mismatch between RFC5888 and sdp-bundle.
> >>>>>
> >>>>> RFC 5888 Section 9.2 (
> https://tools.ietf.org/html/rfc5888#section-9.2) says:
> >>>>>
> >>>>>   SIP entities refuse media streams by setting the port to zero in
> the corresponding "m" line. "a=group" lines MUST NOT
> >>>>>   contain identification-tags that correspond to "m" lines with the
> port set to zero.
> >>>>>
> >>>>> Considering that sdp-bundle redefines the meaning of port zero when
> 'bundle-only' attribute is present, RFC5888 will need
> >>>>> to be updated to allow including bundle-only media streams in the
> "a=group" lines.
> >>>>
> >>>>…or, one could argue that the text in RFC5888 only applies to the case
> when port zero is used to refuse a media stream.
> >>>
> >>> One could argue this, but the language in RFC 5888 is quite blunt.
> >>
> >> So, you really want to bring BUNDLE back to the WG again?
> >
> > No, I really do not want to bring it back. Somebody implementing BUNDLE
> asked me, so I wanted to ask on the list. I would suggest we argue that
> > BUNDLE overwrites all uses of port 0 to disable the media stream,
> including the one in RFC5888.
>
> I don't think BUNDLE overwrites ALL uses of port 0 to disable the media
> stream.
>
> In any case, my understanding was that you want to add something to BUNDLE?
>
> Ideally I wanted to add a section to sdp-bundle that updates RFC5888 and
specifies that mid can be added to the group even if m= line is disabled
with port 0, if this m line is enabled using other means, such as SDP
attribute. In other words something similar to section of sdp-bundle that
updates RFC 3264.

Regards,
_____________
Roman Shpount