Updated proposal for ISSUE-31, my ACTION-156

The current proposal for
http://www.w3.org/2002/ws/soapjms/tracker/issues/31

is in the issue.

The key question I was considering was whether or not I agreed with Mark
that we should make the property mandatory if there was a fault.  Since
the aim here is to be analogous to HTTP (which returns a 500 error for
faults), some indication in the Message properties is appropriate.  So I
concur with the overall proposal.

I suggest slightly altering the replacement text.  Key changes I have below:
 * Clarify that it is a "boolean JMS Message" property.
 * Add an example of setting it.

My update to the proposal follows:

In section 2.2.3:

 * This property indicates whether a SOAP/JMS message corresponds to a
SOAP fault. For senders, this property is set to true when responding
with a SOAP fault. When this property is true, the sending software MUST
set a boolean JMS Message property named SOAPJMS_isFault with a value of
true, as in: Message.setBooleanProperty("SOAPJMS_isFault", true).
 * For receivers, this property is derived from the boolean JMS Message
property named SOAPJMS_isFault — if present and containing a value of
true, the value of soapjms:isFault is true. If omitted, or present with
a value of false, the value of soapjms:isFault is false.

-----
In section 2.6.2.3, add a new row to the JMS Message properties section
of the table

SOAPJMS_isFault   -   Set to true if the response is a SOAP fault,
otherwise it can be absent.

------

This concludes my ACTION-156.

-Eric.

Received on Thursday, 18 March 2010 02:26:37 UTC