RE: Code examples: Existing API and JSEP

Sorry for the format of my reply (crappy web access outlook)
________________________________________
From: Cullen Jennings [fluffy@cisco.com]
Sent: Friday, January 27, 2012 5:19 PM
To: Justin Uberti
Cc: Adam Bergkvist; public-webrtc@w3.org
Subject: Re: Code examples: Existing API and JSEP

On Jan 26, 2012, at 9:24 PM, Justin Uberti wrote:

>>
>> - Special cases such as glare and multiple offers have to be handled in
>> JavaScript.
>> A simple example with the existing API (and ROAP) is quite powerful sice
>> special
>> cases are handled under the hood.
>>
>
> Right, this is what we acknowledged as a tradeoff of JSEP; by pushing more
> control into JS, more code is needed. But as you can see, the amount of
> code is not that large, and this could easily be placed into a JS library.

Uh, I think the coded needed is a lot more than this and I am still waiting for an answer on what it is that you can do with JSEP's "more control"

Adam:
Cullen is right about that the JSEP example needs more code. E.g., this example do handle the case where one side adds several streams in a row, but it doesn't include code to handle glare (or any other special case that I didn't think of). I think we need to identify where it gives more control otherwise it's just a lot of steps that you have to perform in JavaScript in a certain order.

>
>>
>> It's also a question if PeerConnection.createOffer() and
>> PeerConnection.createAnswer()
>> actually can have a return value? If the browser needs to reach down into
>> the platform
>> to gather information these methods may need to be async and use a
>> callback.
>>
>
> Good point. I'd be interested in knowing whether anyone would need this.

Yes - if you were implementing video on something like a iPhone, you would.