Re: [dispatch] Use of BFCP or MSRP over datachannel or websocket

Partha,

It's not entirely clear to me that RTCWEB is the "compelling" use case
for MSRP over websockets.  I'd be interested to hear who might think
that is the case.  I can envision other Real Time applications using
web servers to make use of this, as well.

I don't see any requirements in the RTCWEB WG (charter or
requirements) that indicate this is a compelling use case:
http://www.ietf.org/id/draft-ietf-rtcweb-use-cases-and-requirements-10.txt
There are no use cases for IM, Chat or MSRP specifically that I could
find and I couldn't find anything in the RTCWEB archives:
http://tools.ietf.org/html/draft-ietf-rtcweb-use-cases-and-requirements-10
If there has been discussion, please provide a pointer.

The discussion of MSRP over websockets should continue on this mailing
list.  Per the dispatching of IETF-86 topics, the current proposal is
to AD sponsor this document unless there are very compelling reasons
not to publish.
http://tools.ietf.org/wg/dispatch/trac/wiki

The proposal had been for the general discussion of "foo" over
websockets to happen on the RAI list.    The last I looked there are
at least 200 people subscribed to DISPATCH that are not on the RAI
area mailing list.  I strongly encourage folks with an interest in the
topic to make sure they are subscribe to the RAI area mailing list:
https://www.ietf.org/mailman/listinfo/rai

Regards,
Mary.

On Sun, Mar 31, 2013 at 11:52 AM, Parthasarathi R
<partha@parthasarathi.co.in> wrote:
> The compelling reason for this draft is RTCWeb. In case of non-RTCWeb,
> TCP Shall be used which avoids the overhead introduced due to
> WebSocket. Please clarify in case WebSocket has some specific
> advantage over TCP in non-RTCWeb environment.
>
> In case dispatch or BFCPbis are not the right WG to make recommendation,
> then let us discuss in RTCWeb WG itself. Also, Please suggest if any
> other mailing list like (rai@ietf.org) is the right place to discuss.
>
> Thanks
> Partha
>
>> -----Original Message-----
>> From: dispatch-bounces@ietf.org [mailto:dispatch-bounces@ietf.org] On
>> Behalf Of Mary Barnes
>> Sent: Thursday, March 28, 2013 8:28 PM
>> To: DRAGE, Keith (Keith)
>> Cc: dispatch@ietf.org
>> Subject: Re: [dispatch] Use of BFCP or MSRP over datachannel or
>> websocket
>>
>> I don't think it's the job of dispatch or BFCPbis WG to define or even
>> make recommendations for RTCWEB.  That choice is up to the RTCWEB WG
>> and it shouldn't be influencing this work.  There are applications
>> other than RTCWEB that can and will make use of BFCPbis over
>> websockets.  I really don't understand why we are getting so hung up
>> on this - it's the BFCP messages themselves that are important - how
>> one chooses to transport is up to the application.
>>
>> Mary.
>>
>> On Thu, Mar 28, 2013 at 7:35 AM, DRAGE, Keith (Keith)
>> <keith.drage@alcatel-lucent.com> wrote:
>> > From what I have seen of the discussion so far, there appears to be a
>> body of support that there is a use case for both BFCP and MSRP over
>> websocket, for use in a non-rtcweb environment.
>> >
>> > If when we get to the rtcweb environment we allow this, and also
>> allow a usage over datachannel version, we essentially have two
>> solutions in the market place.
>> >
>> > If that occurs, what guidance will we give to a rtcweb developer as
>> to which to implement?
>> >
>> > Or do we envisage that only one will be allowed (and if so which) in
>> an rtcweb environment?
>> >
>> >
>> >
>> > Regards
>> >
>> > Keith
>> > _______________________________________________
>> > dispatch mailing list
>> > dispatch@ietf.org
>> > https://www.ietf.org/mailman/listinfo/dispatch
>> _______________________________________________
>> dispatch mailing list
>> dispatch@ietf.org
>> https://www.ietf.org/mailman/listinfo/dispatch
>