W3C home > Mailing lists > Public > www-voice@w3.org > January to March 2013

RE: Implement TCP with scxml compliant state machine

From: Michael Bodell <bodell@247-inc.com>
Date: Mon, 11 Feb 2013 19:59:34 +0000
To: chris nuernberger <cnuernber@gmail.com>
CC: "VBWG Public (www-voice@w3.org)" <www-voice@w3.org>
Message-ID: <99155A875ACA0F41BDE9B4DA3E307426119B018D@HKNPRD0310MB350.apcprd03.prod.outlook.com>
We integrated with JS (our context is a voice browser that already has JS integration needs for VXML, CCXML, and SISR, so that integration was inexpensive).  But to get the data from JS in the state machine to the C++ code you needed to send events with data payloads, so it correctly enforced that data is only changed by the state machine - no directly shared data between state machine and external system.

From: chris nuernberger [mailto:cnuernber@gmail.com]
Sent: Monday, February 11, 2013 11:07 AM
To: Michael Bodell
Cc: VBWG Public (www-voice@w3.org)
Subject: Re: Implement TCP with scxml compliant state machine

Did you integrate with a scripting language for your SCXML model or did you use code generation or something like that to keep everything in c++?

Chris

On Mon, Feb 11, 2013 at 11:58 AM, Michael Bodell <bodell@247-inc.com<mailto:bodell@247-inc.com>> wrote:
FWIW we implement some of our SIP stack using SCXML state machines to capture the SIP flow (the SCXML implementation is a C++ one where the external events are actually C++ "event" objects that get delivered to other threads).  Because some of the SIP signaling specifications are complex with lots of contingencies the SCXML is easier to develop and test than the "native" code driven state machines we'd used before.  So I think there is benefit there for handling lots of different protocols (assuming your SCXML interpreter is robust).

From: chris nuernberger [mailto:cnuernber@gmail.com<mailto:cnuernber@gmail.com>]
Sent: Monday, February 11, 2013 10:49 AM
To: VBWG Public (www-voice@w3.org<mailto:www-voice@w3.org>)
Subject: Implement TCP with scxml compliant state machine

It would be extremely interesting to implement the TCP stack with an SCXML state machine.

You could then compare its characteristics (performance, security, correctness, flexibility) with a native C implementation and see where the benefits lie.

I wonder if any OS grad students would be interested in that for a summer of code project.

Chris

--
A foolish consistency is the hobgoblin of little minds - Emerson



--
A foolish consistency is the hobgoblin of little minds - Emerson
Received on Monday, 11 February 2013 20:00:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 11 February 2013 20:00:15 GMT