[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Speechsc] What is the scope of MRCPv2 session



I'm fine with this clarification (i.e. that the text in 6.2.1 is indeed correct). Are we all agreed, however, that as a consequence, the client cannot allocate more than one instance of the same resource type within a given SIP dialog? (If this was allowed and the text in 6.2.1 applied, the channel identifier would be the same for both resources and it would be impossible to disambiguate messages from the two resources).

To completely address this thread, there are two more adjustments needed:

1. The use-case of using two speechrecog resources (one in normal mode and the other in hotword) can only be realised using separate SIP dialogs. In section 9, change "same" to "separate":

  ... "the two modes can be invoked through separate resources
  allocated to the same SIP dialog"

2. What happens if the client attempts to establish two or more resources of the same type within a single SIP dialog? Presumably, 488 Not Acceptable Here? This error and the the restriction as a consequence of 6.2.1 should be called out.

Dave

----- Original Message ----- From: "Dan Burnett" <dan_burnett2000 at yahoo.com>
To: <santhana at huawei.com>; "'IETF SPEECHSC (E-mail)'" <speechsc at ietf.org>
Sent: Friday, September 08, 2006 5:08 PM
Subject: RE: [Speechsc] What is the scope of MRCPv2 session



I also updated the examples in 14.1 as well (both the
a=channel lines and the Channel-Identifier values).

-- dan

--- Dan Burnett <dan_burnett2000 at yahoo.com> wrote:

After carefully review your comments, Dave Oran's,
and
the text of draft -10, I think the key relevant text
in 6.2.1 here is the following:

"The unambiguous string (first part) MUST BE unique
among the resource instances managed by the server
and
is common to all resource channels with that server
established through a single SIP dialog."

Since the example in section 4.2 uses a single SIP
dialog to set up two resource channels with the same
MRCP server, the prefix values must be the same.  I
have corrected the example to use the same prefix in
the a=channel line for both resources.  This
correction will appear in draft -11.

-- dan


--- santhanakrishnan <santhana at huawei.com> wrote:

> Hi Oran,
>
>       Thanks for your detailed explanation. It
> clarifies enough of our
> queries.
>
> But I would like to highlight the main source of
> confusion again to you.
>
>
>
> Please refer the following section from the draft
>
> 6.2.1
>
>    All MRCPv2 requests, responses and events MUST
> contain the Channel-
>
>    Identifier header.  The value is allocated by
the
> server when a
>
>    control channel is added to the session and
> communicated to the
>
>    client by the "a=channel" attribute in the SDP
> answer from the
>
>    server.  The header value consists of 2 parts
> separated by the '@'
>
>    symbol.  The first part is an unambiguous
string
> identifying the
>
>    MRCPv2 session.  The second part is a string
> token which specifies
>
>    one of the media processing resource types
listed
> in Section 3.1.
>
>    The unambiguous string (first part) MUST BE
> unique among the resource
>
>    instances managed by the server and is common
to
> all resource
>
>    channels with that server established through a
> single SIP dialog.
>
>
>
> Here it says the prefix (First part) identifies an
> MRCPv2 session.
>
>
>
> But in the following example, the channel Id
> returned by the Server
>
> for Synthesizer and Recognizer belonging to the
same
> MRCPv2 Session
>
> is shown with different prefix(First Part).
>
>
>
> S->C:  SIP/2.0 200 OK
>
>           Via:SIP/2.0/TCP
> client.atlanta.example.com:5060;
>
>           branch=z9hG4bK74bf9
>
>           To:MediaServer
> <sip:mresources at server.example.com>
>
> ........................................
>
>           v=0
>
>           o=sarvi 2890844526 2890842809 IN IP4
> 126.16.64.4
>
>           s=-
>
>           c=IN IP4 224.2.17.12
>
>           m=application 32416 TCP/MRCPv2
>
>           a=setup:passive
>
>           a=connection:existing
>
>           a=channel:32AECB234338 at speechrecog
>
>           a=cmid:1
>
>           m=application 32416 TCP/MRCPv2
>
>           a=setup:passive
>
>           a=connection:existing
>
>           a=channel:32AECB234339 at speechsynth
>
>           a=cmid:1
>
>           m=audio 48260 RTP/AVP 0 96
>
>           a=rtpmap:0 pcmu/8000
>
>           a=rtpmap:96 telephone-event/8000
>
>           a=fmtp:96 0-15
>
>           a=sendrecv
>
>           a=mid:1
>
>
>
> Here I wonder either the example or the draft
> statement seems to be
> incorrect.
>
> Please clarify on this.
>
>
>
> Regards,
>
> Santhanakrishnan;
>
>
>
>
>
>
>
> -----Original Message-----
> From: David R Oran [mailto:oran at cisco.com]
> Sent: Friday, June 16, 2006 2:37 AM
> To: IETF SPEECHSC (E-mail)
> Subject: Re: [Speechsc] What is the scope of
MRCPv2
> session
>
>
>
>
>
> On Jun 13, 2006, at 5:01 AM, Burger, Eric wrote:
>
>
>
> > Do we need to tighten the language? Maybe
> introduce a new term?
>
> > Pick two from the set {MRCPv2, SIP, control,
media
> control,
>
> > resource control, or something else descriptive
of
> your choice}.
>
>
>
> >
>
> >
>
> > Please follow-up to the list with your thoughts.
>

=== message truncated ===


__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

_______________________________________________
Speechsc mailing list
Speechsc at ietf.org
https://www1.ietf.org/mailman/listinfo/speechsc



_______________________________________________
Speechsc mailing list
Speechsc at ietf.org
https://www1.ietf.org/mailman/listinfo/speechsc