Re: [MMUSIC] [rtcweb] Unified Plan for SDP Handling - Glare

[Adam Roach <[e-mail]>]

On 7/22/13 23:52, Christer Holmberg wrote:

> The draft also talks about modifying streams (section 3.4.2). Are there specific modification use cases where no glare is important, or is it about being able to modify streams in general and avoid glare?

I was a matter of realizing that our design allows, with very little 
additional effort, significant reduction in the probability of glare for 
stream modification in general. As this seemed like very low hanging 
fruit, it seems to be sensible to include it as part of the design.

So, to answer you question directly: it's not driven by specific use 
cases; it's just general glare probability reduction.


> So, moving ahead with unified plan basically means adopting the 
> requirement to solve glare, but we may also look at other mechanisms?

It is my understanding that making glare impossible or nearly impossible 
for the add/remove stream case is an important requirement for any plan 
(a, b, unified, or anything else). At the moment, the proposal in the 
unified plan document is the best that the authors can come up with. I 
think we're all open to the prospect of alternate designs that achieve 
the same goal, although I would hope we don't spend too much time 
exploring the solution space.

/a