W3C home > Mailing lists > Public > xml-dist-app@w3.org > November 2001

S21 proposed sample implementation

From: Jacek Kopecky <jacek@systinet.com>
Date: Thu, 15 Nov 2001 23:36:01 +0100 (CET)
To: John Ibbotson <john_ibbotson@uk.ibm.com>
cc: <xml-dist-app@w3.org>, Stuart Williams <skw@hplb.hpl.hp.com>
Message-ID: <Pine.LNX.4.33.0111152309570.2958-100000@mail.idoox.com>
 Hello all. 8-)
 Here is a proposed sample implementation of the usage scenarion
S21 [1]. The scenario is basically about streaming potentially
huge messages from a sender to a receiver so that the entire
message need not be present at one time at any point in its
existence.

 Assumptions about the implementation:
 1) the content that is so huge will be inside the Body,
 2) receiver will be able to contain and process the full Header
element,
 3) the application will define a way of communicating a fault
occurring during the generation of the message without resorting
to the SOAP Fault mechanism (possible because of requirement 703b
[2]).

 The implementation:
 A message may contain some headers which will be handled
(possible because of assumption #2) by the receivers before
handling the body data or cached for handling after the body
data, as per the processing model.
 The data will be sent in chunks as separate children of the Body
element. This is possible because in streaming applications the
data is always built of small chunks each of which is separately
processible - packets of soundwaves, blocks of an image, frames
of video, whatever.
 A fault during generating will be communicated to the receiving
side in a special form of a chunk of data. (For applications
requiring out-of-band error handling streaming via a single
channel - a single SOAP message, for example - is not the way to
go.) The receiver will stop playing the audio or the video
stream, it will discard the image or it will do whatever is
suitable.

 I think I've got it covered. If my proposal is completely
incompatible with the other sample implementations in the current
US document, I hope it can be rewritten to fit. 8-)

 Comments, please?

                   Jacek Kopecky

                   Senior Architect, Systinet (formerly Idoox)
                   http://www.systinet.com/

[1] http://www.w3.org/2000/xp/Group/1/10/18-us/XMLProtocolUsageScenarios#_Toc527256181
[2] http://www.w3.org/TR/2001/WD-xmlp-reqs-20010319/#z703b
Received on Thursday, 15 November 2001 17:36:04 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:04 GMT