It seems that overriding MediaPipeline which will work in conjunction with custom external transport class (which conforms to TransportInterface interface) can help me to solve this. What do you think?
Moreover, is it possible to create these custom classes in C++ Firefox add-on and then, substitute traditional MediaPipeline and Transport classes in Firefox, so WebRTC will work as it used to but those classes will be my custom ones? Also, is it possible to give an access (javascript interface) to js-apps running in browser? Any thoughts/ideas helpful. On Thursday, July 11, 2013 11:46:10 AM UTC-7, Ethan Hugg wrote: > Agreed. I assumed this was about hacking up a proof-of-concept. > > Contributing another transport back to FF would need lots of discussion > > I'm sure. > > > > -EH > > > > > > > > On Thu, Jul 11, 2013 at 11:37 AM, Adam Roach <[email protected]> wrote: > > > > > Well, if he's moving off UDP, he'll need to change the SDP signaling, > > > which does impact the code inside media/webrtc/signaling. > > > > > > If this new transport is important, we probably need to take on work in > > > the IETF MMUSIC working group to specify tokens for indicating it. The > > > transport token isn't really an extension point that is commonly exercised > > > for nonstandard purposes, but it would be far worse to call something > > > "RTP/SAVPF" when it's not... > > > > > > /a > > > > > > > > > On 7/11/13 13:31, Ethan Hugg wrote: > > > > > > I suggest you take a look instead at the code in media/mtransport. I don't > > > think you would need to change media/webrtc/signaling to do what you want. > > > > > > -EH > > > > > > > > > > > > On Thu, Jul 11, 2013 at 11:04 AM, peetonn <[email protected]> > > <[email protected]> wrote: > > > > > > > > > Hi there, > > > > > > I'm curious of how does acquiring of video frames and sending them over > > > the network actually work? > > > I've looked at media/webrtc/siganling sources as well as read lots of > > > documents like these > > (http://tools.ietf.org/html/rfc5245#section-1,http://datatracker.ietf.org/doc/draft-ietf-rtcweb-jsep/?include_text=1,http://dev.w3.org/2011/webrtc/editor/webrtc.html#intro > > and others) but > > > still can't find a point in a code where frames are acquired and sent to > > > the udp socket. There are MediaStream, MediaConduit, MediaPipeline, > > > CC_SIPCCCall, CC_SIPCCDevice... and all of this is highly interconnected. > > > > > > What I'm trying to do in the end - is to substitute default UDP-transport > > > with another one (specifically this one http://www.named-data.net/). So I > > > need to understand in which way it is possible to do. > > > Will appreciate any help and ideas where to look at... > > > _______________________________________________ > > > dev-media mailing > > [email protected]https://lists.mozilla.org/listinfo/dev-media > > > > > > _______________________________________________ > > > dev-media mailing > > [email protected]https://lists.mozilla.org/listinfo/dev-media > > > > > > > > > > > > -- > > > Adam Roach > > > Principal Platform Engineer > > > [email protected] > > > +1 650 903 0800 x863 > > > _______________________________________________ dev-media mailing list [email protected] https://lists.mozilla.org/listinfo/dev-media
