- From: Randell Jesup <randell-ietf@jesup.org>
- Date: Sat, 28 Apr 2012 11:25:38 -0400
- To: public-media-capture@w3.org
- Message-ID: <4F9C0BF2.5090508@jesup.org>
On 4/28/2012 4:29 AM, William Lin wrote: > look forward to this recorder proposal. > one question, will recorder be a html5 tag similar to <video>,<audio> > and can be used independent of Media Caputure or just part of Media > Capature to work for record MediaStream? thanks There's some interesting/useful symmetry to having record be a 'sink' for a MediaStream, as opposed to a method on it (record.src = mediastream; record.path = blah; record.max_length = blah; etc). It's certainly more future-proof. <video>/<audio> elements could be fed into it through a mediastream (video.captureStreamUntilEnded() from roc's Processing proposal, etc). > > On Sat, Apr 28, 2012 at 3:21 PM, Anant Narayanan <anant@mozilla.com > <mailto:anant@mozilla.com>> wrote: > > On 4/27/12 7:09 PM, William Lin wrote: > > Seems MediaStreamRecorder is removed. And all the function > record() > related is removed too. > No clear the story of how was this decision made. > Or in next step, recorder will be added? > > > I apologize, I didn't mention in my earlier email that the other > change was the removal of MediaStreamRecorder. We've been talking > on the list about removing it for a while now - once we have a new > proposal we'll be adding something similar to the spec. > > The main problem with the old MediaStreamRecorder was that it was > defined in terms of a MediaStream, but it seems more desirable to > have such a recorder work directly with the outputs of a > MediaStream, such as <video>, <audio>, or even <canvas>. One of my > TODO items is to come up with such a proposal, which I will do as > soon as possible! > > Thanks, > -Anant > -- Randell Jesup randell-ietf@jesup.org
Received on Saturday, 28 April 2012 15:27:03 UTC