W3C home > Mailing lists > Public > public-webrtc@w3.org > May 2013

Re: Fwd: Re: Asymetric API on DataChannels creation

From: <piranna@gmail.com>
Date: Wed, 29 May 2013 21:45:43 +0200
Message-ID: <CAKfGGh2W9DS_+nFxJTzzUnP7mEsK4Rfn=5ft966h7wTy9mPZBQ@mail.gmail.com>
To: Randell Jesup <randell-ietf@jesup.org>
Cc: public-webrtc <public-webrtc@w3.org>
Yes, the spec say it. The datachannel event is new, so I think it should
work the way I said.
El 29/05/2013 20:53, "Randell Jesup" <randell-ietf@jesup.org> escribió:

>
> On 5/29/2013 3:53 AM, piranna@gmail.com wrote:
>
>>
>> When you call to createDataChannel() you get a "connecting"
>> DataChannel, and when connection gets stablished you get an "open"
>> event over it. On the other hand, on the receiver side you only get a
>> "datachannel" event with the new DataChannel, thats already open.
>> Doesn't it makes more sense that the "datachannel" event raise a
>> "connecting" DataChannel to be able to do pre-open initializations and
>> later raise the "open" event instead of do both things at the same
>> time? This would also easy to reuse code on applications when both end
>> are symetric.
>>
>> If it's done this way just to ear a data flight, it would be mimic
>> just setting on the specification the "datachannel" event before
>> setting/changing the status to "open" and adding afterthat an "open"
>> event.
>>
>>
> The spec is incorrect if it states this; we agreed that both sides
> should get an onopen event to parallel WebSockets
>
> I believe adam is updating the datachannel spec this week
>
>    Randell
>
>
>
>
>
Received on Wednesday, 29 May 2013 19:46:11 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 15:19:33 UTC