Re: [MMUSIC] BUNDLE DECISSION: Full bundle or no bundle - there aint anything in between

[Paul Kyzivat <[e-mail]>]

I agree with the sense of what you state below.
I have some nits with the words you use.

On 5/24/13 6:28 AM, Christer Holmberg wrote:
> Hi,
>
> At the interim yesterday, we agreed that an answerer MUST always include
> identical ports* (usage of port zero within a bundle is a separate
> topic) for all m- lines within an bundle group.
>
> If the answerer wants to use different ports* for some/all m- lines, it
> MUST NOT include those m- lines in a bundle group

OK so far. But:

> (if they already are
> in a bundle group, a separate offer is needed to remove them).

This depends on what you mean by "already in a bundle group".
Certainly they are already in a bundle group in the offer, so presumably 
that is not what you mean.

I presume you mean that they were already in a bundle group in the prior 
O/A.

I think this can be stated more clearly. How about:

The answer to an offer containing a bundle group may remove an m-line 
from the bundle group if (and only if) the address/port of the m-line in 
the offer is distinct from the address/port of all the other m-lines in 
the offered bundle group. When doing so it must supply a unique 
address/port for that m-line in the answer.

	Thanks,
	Paul

> WHAT IT MEANS: There is no way for an answerer to tell that the offerer,
> for certain m- lines, can use the same port*, eventhough the answer will
> use separate ports*.
>
> I will work on some actual text proposal later, but please indicate if
> you OBJECT to this approach in general.
>
> Regards,
>
> Christer
>
> NOTE: “port” = port:address combination – we are working on coming up
> with better terminology J
>
>
>
> _______________________________________________
> mmusic mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mmusic
>