Re: [rtcweb] Unified plan IPR

On 7/23/13 11:58 PM, cowwoc wrote:
> On 23/07/2013 12:51 PM, Adam Roach wrote:
>> On 7/23/13 11:22, Roman Shpount wrote:
>>> The situation would be a bit clearer if patent holders were to
>>> provide the licensing policy regarding this IPR release. Given that
>>> Ericsson is actively involved in this working group, I think it would
>>> be reasonable to ask them for this.
>>
>> If process has been properly followed, the IPR holder has already been
>> notified by the IETF executive director. See
>> http://www.ietf.org/rfc/rfc4879.txt (section 1 paragraph 1)
>>
>> I doubt agitating for action on these mailing lists is likely to
>> produce useful results.
>>
>> /a
>
> Hi Adam,
>
>      I'm a bit concerned about the optics of what just happened.
>
>   * The Working Group has been pushing for the use of SDP since 2011
>     (see http://www.ietf.org/mail-archive/web/rtcweb/current/mail15.html).
>   * The first post related to the use of SDP in WebRTC came from
>     Christer Holmberg of Ericsson on September 14th, 2011.
>   * One of the Chairs of the Working Group and one of the Specification
>     editors are from Ericsson.
>   * There has been a substantial push against the use of SDP by some
>     mailing list participants, but this was rejected by the Working Group.
>   * Suddenly we find out that Ericsson has filed two patents related to
>     the use of SDP in WebRTC and these were filed *after* Ericsson
>     actively pushed for the use of SDP.
>
>      Isn't there a conflict of interest here?

This was quite a theory, and I was hesitating a long time before 
deciding if I should respond or not. But what I can say (and some parts 
are verifiable, some are not) is:

* I had no idea about these patents before these emails

* I think all document editor's (including the one from Ericsson) has 
been very careful in not pushing their way, but rather document the 
consensus of the WG (and bring open issues to the attention of the WG 
for discussion)

* The person from Ericsson currently editing, as well as myself, have 
been arguing all the time that SDP is not an API surface. We have also 
contributed proposals on API surface to use (to avoid having to touch 
the SDP). A reason for this has been to allow the removal of SDP and O/A 
without breaking apps.

* A long time ago, there were no createOffer/Answer or setLocal/Remote 
methods on PeerConnection. Instead the PeerConnection would emit a blob 
of data that should be applied at the remote PeerConnection to make 
things work whenever needed. There were no need to use SDP or O/A in the 
design. In the discussion (and show of hands) I think everyone working 
for Ericsson preferred to stay at that model, and a reason was (again) 
that it enabled removing SDP and O/A without breaking applications. The 
consensus however lay elsewhere, and of course we accepted this.

* I have repeatedly argued for moving functionality from SDP signaling 
to RTP (RTCP) signaling.

I have not read those patents so I don't know if the above helps really, 
but going through all archives and notes (and you'd have to dig into 
IETF archives) I think you will find that we've not pushed particularly 
hard for SDP. We have seen this technology (WebRTC) as a really good 
thing for the community at large, and have put up resources to help its 
development (and have accepted consensus aslo when it is not what we 
prefer). I must admit your accusation hurts a bit.

Received on Thursday, 25 July 2013 05:45:05 UTC