Re: [rtcweb] Data Channel Negotiation and reopening of decisions

[Michael Tuexen <[e-mail]>]

On Feb 15, 2013, at 6:14 PM, Martin Thomson wrote:

> On 14 February 2013 18:59, Michael Tuexen
> <[e-mail]> wrote:
>> You register at IANA PPID X for binary and PPID Y (Y != X) for text.
>> No signaling required.
> 
> Great!  That removes the need for an in-band protocol entirely. (I
> missed the IANA bit when I read through the description.)
> 
>> I expect each data channel consisting of a pair of streams. Each stream
>> is contained in at most one data channel.
> 
> Correct. The difference between what you have now and what I am
> proposing is simple: you match arbitrary stream X to arbitrary stream
> Y in the opposite direction through signaling or control messages.  I
> am proposing to match stream X to stream X always.  Thus, if you send
> on stream 7, this creates a data channel on the other end that also
> sends on stream 7.
I think I understand what you are proposing. But what happens, if
both sides at about the same time open want to open a data channel.
For both sides outgoing stream X is free, so they use this. So the
endpoints end up with one data channel instead of two.

For me, the endpoints trying to set up two data channels, and finally
having only a single one is a bad thing. You could argue, that this
comes up only in specific situations. That might be true, but for
an application writer this would mean that his code works most
of the time, but sometimes not. This is a bad thing for me.

Best regards
Michael
>