W3C home > Mailing lists > Public > public-webrtc@w3.org > June 2018

Re: default components versus custom components

From: Peter Thatcher <pthatcher@google.com>
Date: Mon, 18 Jun 2018 02:12:07 -0700
Message-ID: <CAJrXDUH62uPpvoUUPUvSkBukycigbBd6zYBdi=vG+vcL4Lx4KA@mail.gmail.com>
To: "Zhu, Jianjun" <jianjun.zhu@intel.com>
Cc: Göran Eriksson AP <goran.ap.eriksson@ericsson.com>, Bernard Aboba <Bernard.Aboba@microsoft.com>, "public-webrtc@w3.org" <public-webrtc@w3.org>
It's probably better to think of "browser components" and "app components"
rather than "default components" and custom components".  I don't think we
should have implicit components, but rather explicit components.  Then the
app is always in control of what is used, and no implicit funny business.

On Sun, Jun 17, 2018 at 7:05 AM Zhu, Jianjun <jianjun.zhu@intel.com> wrote:

> I think the conflicts between a default component and a custom component
> should also be addressed. Let's say a website implements AV1, adds it as a
> supported payload type. There is no problem at this time because AV1 could
> only be a custom component. But someday, when AV1 is implemented in
> browsers, we need a clear strategy or API to determine whether the default
> AV1 component or the custom AV1 component will be used.
>
>
>
> Best Regards,
> Jianjun
>
> On 2018/6/16, 6:00 AM, "Göran Eriksson AP" <
> goran.ap.eriksson@ericsson.com> wrote:
>
>
>
>
>
>     Den 2018-06-15 23:54 skrev "Bernard Aboba" <
> Bernard.Aboba@microsoft.com> följande:
>
>
>
>         Peter said:
>
>
>
>         "I'm persuaded by "default components" for the sake of making
> simple things simple, but I'm not at all persuaded by the compatibility
> argument.  If someone wants to make custom RTP payloads/packetizations or
> use SCTP instead of RTP, let them."
>
>
>
>         [BA] As long as you have "default components" then custom
> components need only be developed when they are required.
>
>
>
>         In those cases where custom components are needed, presumably the
> developers understand their interoperability requirements - they know
> whether their custom RTP payload developed in the browser application needs
> to be able to talk to an equivalent custom module developed for their
> mobile application.
>
>
>
>     [GE] +1. Same if developer use QUIC based transport solutions- the Web
> site developer ought to know the requirements. What we may need to consider
> is the ability for a web site to lock down the ability to inject custom
> components.
>
>
>
>
>
Received on Monday, 18 June 2018 09:12:43 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:18:42 UTC