Firefox 47 WebRTC (and Web Audio) Release Notes

[Maire Reavy]

Firefox 47 is now our Developer Edition version, and it's scheduled to go to Beta on April 18th and to Release on June 7th.  Detailed WebRTC (plus Web Audio) release notes for Firefox 47 are here.

If anyone has any questions about the release, the notes, or possible bugs in the code, feel free to reach out to me directly.

Cheers,

-Maire

---------------------------------------------------

Noteworthy Changes:

• Recording:

• Periodic blinking is fixed (See bug 1234230)

• Fire error if mime type is unsupported (See bug 963238)

• Recorded videos of longer than a few seconds would not play (See bug 1213775)

• Clear screen and other privacy improvements to about:webrtc (See bug 1209252)

• Make RTCRtpSender.setParameters return a promise. (See bug 1247622)

• Add support for RTCRtpEncodingParameters.scaleResolutionDownBy (See bug 1244913)

• Increase buffer sizes on older Windows to avoid packet loss at very high (HD) bitrates (bug 1217677); allow Jumbo ethernet frames when in e10s mode (bug 1252755)

• MediaStream in video element doesn't guarantee video size on "loadedmetadata" (See bug 1240478)

• Firefox always sends full band Opus audio even though the negotiation specifies wideband audio (bug 1249098); support negotiation of mono Opus (bug 818618). Also fixed no audio on browser side when Opus payload types in offer and answer do not match (bug 1221473).

• High CPU usage on some systems when IPV6 interfaces are configured (See bug 1224845)

[Ray Wang]

May i ask a question here?

I know that Chrome version X corresponds to Webrtc version X.

but what's the relationship between Firefox version and Webrtc version?

for example, i'm running a firefox 45.0.1 version, does this version using the code in Webrtc M45 release?
is this relationship the same with Chrome?

在 2016年3月24日星期四 UTC+8下午10:08:40，Maire Reavy写道：

[Maire Reavy]

On 4/7/2016 9:44 PM, Ray Wang wrote:

May i ask a question here?

I know that Chrome version X corresponds to Webrtc version X.

but what's the relationship between Firefox version and Webrtc version?

We update the code we import from webrtc.org periodically.  Currently we are using branch 43 of webrtc.org and are in the process of upgrading to branch 49.

for example, i'm running a firefox 45.0.1 version, does this version using the code in Webrtc M45 release?

Firefox 45 was the first release to use webrtc.org branch 43.  (It was in the noteworthy changes for Fx 45: https://wiki.mozilla.org/Media/WebRTC/ReleaseNotes/45#Noteworthy_Changes)

When we do an upgrade, we call it out in the release notes.  Going forward I can include a regular reminder in the notes what version of the webrtc.org branch that particular release of Firefox is using.  There is no correlation between release number of Firefox and the branch number of webrtc.org.  You'll have to look at the Firefox release notes for that version to know. 

FYI: We plan to complete the upgrade to branch 49 for the Firefox 49 in June.  (It so happens the two version numbers will be the same, but that's just by chance this time around.)

Cheers,
-Maire

在 2016年3月24日星期四 UTC+8下午10:08:40，Maire Reavy写道：

Firefox 47 is now our Developer Edition version, and it's scheduled to go to Beta on April 18th and to Release on June 7th.  Detailed WebRTC (plus Web Audio) release notes for Firefox 47 are here.

If anyone has any questions about the release, the notes, or possible bugs in the code, feel free to reach out to me directly.

Cheers,

-Maire

---------------------------------------------------

Noteworthy Changes:

• Recording:

• Periodic blinking is fixed (See bug 1234230)

• Fire error if mime type is unsupported (See bug 963238)

• Recorded videos of longer than a few seconds would not play (See bug 1213775)

• Clear screen and other privacy improvements to about:webrtc (See bug 1209252)

• Make RTCRtpSender.setParameters return a promise. (See bug 1247622)

• Add support for RTCRtpEncodingParameters.scaleResolutionDownBy (See bug 1244913)

• Increase buffer sizes on older Windows to avoid packet loss at very high (HD) bitrates (bug 1217677); allow Jumbo ethernet frames when in e10s mode (bug 1252755)

• MediaStream in video element doesn't guarantee video size on "loadedmetadata" (See bug 1240478)

• Firefox always sends full band Opus audio even though the negotiation specifies wideband audio (bug 1249098); support negotiation of mono Opus (bug 818618). Also fixed no audio on browser side when Opus payload types in offer and answer do not match (bug 1221473).

• High CPU usage on some systems when IPV6 interfaces are configured (See bug 1224845)

--

---
You received this message because you are subscribed to the Google Groups "discuss-webrtc" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss-webrt...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/7eddccdb-d405-4218-ab9e-42879aa914a4%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Ray Wang]

Thanks Maire, that is clear to me now.

在 2016年4月8日星期五 UTC+8下午1:36:54，Maire Reavy写道：

[PhistucK]

By the way, I am surprised you are using it (for anything other than tests), since specifications usually require multiple independent implementations. Using the same implementation may kind of prevent it from advancing to a candidate recommendation and the like.

Are you using the WebRTC.org code for anything other than tests? If so, what features are implemented using WebRTC.org code?

Thank you!

☆PhistucK

To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/df0ad3ea-e3e0-4066-8801-ee9fec880bc9%40googlegroups.com.

[Philipp Hancke]

Firefox is using mostly the audio/video engines from webrtc.org, not the networking parts.

If you want something completely independent, go for Microsoft Edge. Video codecs are not interoperable yet but audio, ice and DTLS are which seems sufficient proof of interop to me.

To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/CABc02_KRGc9DH7haNs%3DGdtSMT89Fw%2BjM%3Dr0SZn_92tXsZvSQFg%40mail.gmail.com.

[PhistucK]

Thank you. :)

☆PhistucK

To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/CADxkKiJg25y8%3DD-KejV7%3DyG5aWFXS8LDqegibc6wULAqkLiHTA%40mail.gmail.com.

[Maire Reavy]

On 4/8/2016 3:10 AM, PhistucK wrote:

By the way, I am surprised you are using it (for anything other than tests), since specifications usually require multiple independent implementations. Using the same implementation may kind of prevent it from advancing to a candidate recommendation and the like.

Are you using the WebRTC.org code for anything other than tests? If so, what features are implemented using WebRTC.org code?

Firefox uses totally independent implementations of a large part of WebRTC except for some of the underlying media engines and RTP code from webrtc.org (much as we and Chrome (and others) share libjpeg-turbo, libvpx, libopus, libsrtp, sctp and others).  We do make modifications to the webrtc.org engine code, and we also contribute bugs and patches back to webrtc.org.

Firefox uses totally independent implementations of all of the DOM APIs,  ICE, TURN, Signaling (JSEP), SDP, DataChannels, encryption (outside of libsrtp itself), plus all code implementing MediaStreams/Tracks and GetUserMedia.  Implementing all these things ourselves has in fact shaken many bad assumptions out of Chrome's code -- and ours.  This sort of reliance on lower-level media/network libraries (or OS libraries) is quite common among browsers.

There are independent media engine implementations by Ericsson and (now) Microsoft.

Cheers,
-Maire

To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/CABc02_KRGc9DH7haNs%3DGdtSMT89Fw%2BjM%3Dr0SZn_92tXsZvSQFg%40mail.gmail.com.

[PhistucK]

Thank you for the detailed explanation!

☆PhistucK

To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/57082D0A.7060801%40reavy.org.

[Alexandre GOUAILLARD]

you can see some visual representations of the differences between the stand alone webrtc.org stack, chrome, firefox, and safari/webkit there:

http://www.slideshare.net/alexpiwi5/webrtc-browsers-n-stacks-implementation-differences

Note on slide 33, the Cico code has been rewritten to remove limitation to 1 video stream and others last year. AFAIK, the rest of the information still stand.

To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/CABc02_%2B8YfkxhxanvfMUwWwdgfjfO%3DapALKbaZfSm0pPvFMyvg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Alex. Gouaillard, PhD, PhD, MBA

------------------------------------------------------------------------------------

Principal Architect - Citrix, San Francisco

President - CoSMo Software Consulting, Singapore

------------------------------------------------------------------------------------

sg.linkedin.com/agouaillard

•