W3C home > Mailing lists > Public > public-webrtc@w3.org > December 2017

Re: WPT test dependencies

From: Soares Chen <soares.chen@gmail.com>
Date: Fri, 1 Dec 2017 15:41:31 +0800
Message-ID: <CACUDQOGD_adw=Zvwx9yLNaiyuxJF-w0pgWKdo4XuNLybzv=dLQ@mail.gmail.com>
To: Bernard Aboba <Bernard.Aboba@microsoft.com>
Cc: "public-webrtc@w3.org" <public-webrtc@w3.org>
​​
On Fri, Dec 1, 2017 at 1:43 AM, Bernard Aboba <Bernard.Aboba@microsoft.com>
wrote:
> Why does the RTCIceTransport test depend on the RTCSctpTransport being
defined?
>
> For example, if a browser supports RtpSender/Receiver, DtlsTransport and
IceTransport but not SctpTransport couldn't an RTCIceTransport object be
obtained from sender.transport.transport ?

​There was some discussion about this in GitHub issue #1306. [1] In summary
there were issues of getting an instance of RTCIceTransport from an
RTCRtpsender/Receiver, because the transport attribute there is nullable.
At the time of writing the tests, RTCSctpTransport was used for getting
instance of RTCIceTransport, because there is a deterministic way that we
can be sure that peerconnection.sctp become non-null when setting session
description. That wasn't the case for RTCRtpsender/Receiver.

It looks like the newer editor's draft now has normative steps in setting
session description for setting the transport attributes to set the
transport attribute of RTCRtpsender/Receiver. This should make testing it
much easier and we can update the tests to retrieve RTCIceTransport
instances that way.

[1] https://github.com/w3c/webrtc-pc/issues/1306#issuecomment-305681038​
Received on Friday, 1 December 2017 07:42:14 UTC

This archive was generated by hypermail 2.3.1 : Friday, 1 December 2017 07:42:15 UTC