Removing RemoteMediaStream. Adding one universal implementation of MediaStream that is used for both remote and local media streams.
Removed AudioDevice and VideoDevice since VideoCaptureModule and AudioDeviceModule now is reference counted.
Changes LocalAudioTrackImpl and LocalVideoTrackImpl to AudioTrackImpl and VideoTrackImpl so they can be used to repressent both remote and local tracks.
Renamed files to a better name.
Review URL: http://webrtc-codereview.appspot.com/151001
git-svn-id: http://webrtc.googlecode.com/svn/trunk@627 4adac7df-926f-26a2-2b94-8c16560cd09d
* Remove (most of) local libjingle mods. Only webrtcvideoengine and webrtcvoiceengine are left now, because the refcounted module has not yet been released to libjingle, so I can't submit the changes to libjingle at the moment.
* Update the peerconnection client sample app.
Review URL: http://webrtc-codereview.appspot.com/151004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@625 4adac7df-926f-26a2-2b94-8c16560cd09d
This is to avoid running into problems with jsoncpp and parts of libjingle
where we use STL and CRT routines that refer to the global locale for formatting.
If the current locale is e.g. Spanish, then numbers will be formatted as "12,34"
and not "12.34" as some parts (not all) of jsoncpp expect.
Code I noticed where we might run into this is here (but it's likely that there are
more places):
third_party\libjingle\source\talk\p2p\base\candidate.h (preference_str)
third_party_mods\libjingle\source\talk\app\webrtc\webrtc_json.cc
third_party\jsoncpp\src\lib_json\*writer*
third_party\jsoncpp\src\lib_json\*reader*
BUG=69
TEST=Verify that the signaling messages always contain numbers formatted as "1.23" and never "1,23" even though the regional settings on the machine specify otherwise (e.g. try setting it to Spanish).
Review URL: http://webrtc-codereview.appspot.com/140007
git-svn-id: http://webrtc.googlecode.com/svn/trunk@564 4adac7df-926f-26a2-2b94-8c16560cd09d
This CL modifies the ADM interface to ensure that an external ADM
can't call Create and Destroy any longer.
It also contains some minor style nits to conform better with
the Chromium style guide.
Review URL: http://webrtc-codereview.appspot.com/133014
git-svn-id: http://webrtc.googlecode.com/svn/trunk@552 4adac7df-926f-26a2-2b94-8c16560cd09d
Removed OnLocalStreamInitialized callback from the PeerConnection callback list. After adding OnAddStream trigger at the originator this callback was redundant. Also other modification is to provide same stream label in OnAddStream callback at the originator which provided in AddStream API.
Review URL: http://webrtc-codereview.appspot.com/138002
git-svn-id: http://webrtc.googlecode.com/svn/trunk@490 4adac7df-926f-26a2-2b94-8c16560cd09d
This CL will support negotiation of RTCP Mux feature. Earlier we were by default enabling and assuming remote end point will support this feature as well. This will also remove the maintaining of transport channels in WebRtcSession. Its left to cricket::Transport
Review URL: http://webrtc-codereview.appspot.com/131005
git-svn-id: http://webrtc.googlecode.com/svn/trunk@472 4adac7df-926f-26a2-2b94-8c16560cd09d
* I also merged the issue 113007.
This will be uploaded to the libjingle patch, so you may comment there if you want.
There's failure in the tests now, but I will let you review the threading change at the same time I will try to resolve the failure.
Review URL: http://webrtc-codereview.appspot.com/120002
git-svn-id: http://webrtc.googlecode.com/svn/trunk@426 4adac7df-926f-26a2-2b94-8c16560cd09d
1) Libjingle and webrtc use different jsoncpp, so we need to distinguish them.
2) Update the webrtcsession_unittest:
* Use the Thread::SleepMs
* Remove main, which has been defined in talk/base/unittest_main.
Review URL: http://webrtc-codereview.appspot.com/107003
git-svn-id: http://webrtc.googlecode.com/svn/trunk@356 4adac7df-926f-26a2-2b94-8c16560cd09d
