W3C home > Mailing lists > Public > public-soap-jms@w3.org > April 2009

clarification of precedence

From: Amelia A Lewis <alewis@tibco.com>
Date: Thu, 2 Apr 2009 10:11:07 -0400
To: SOAP/JMS (list) <public-soap-jms@w3.org>
Message-ID: <20090402101107299636.bf9efa33@tibco.com>
After review of sections 3.4.4 and 3.4.5 of the editor's copy of the 
specification, I find the following related to precedence:

If a property is specified at multiple levels, the most specific 
setting will take precedence (port first, then service, then binding).

(3.4.4, paragraph two; this is followed by an example of this 
overriding in action)

Section 3.4.5 says, summarizing: some of these properties may be set in 
the URI (para one); the URI is specified in address/@location (para 
two).  Para three:

Properties expressed in the URI [IETF RFC 3987] override any values set 
in the markup as described above.

(again, followed by an example)

It seems to me that this is complete, *with one exception*, which is 
indicating the precedence of information obtained from the 
"environment" versus that obtained from a WSDL.  That is, presuming 
that a URI is obtained from Somewhere Else, that URI is, in principle, 
according to our rules of precedence, "more specific" (applying to this 
particular run of this particular codebase on this particular machine 
in this particular interaction) than any information appearing in the 
WSDL.  That assumes, of course, that the URI can be associated with a 
particular port, if it differs from the address/@location.

Recommended: "Properties set in the environment (of the machine or of 
JNDI, for instance, either as components of a URI or set independently) 
take precedence over those specified in WSDL."  This would appear after 
para three of 3.4.5, above; it would either be the second sentence of 
the paragraph, or a separate paragraph.

I have a sad tendency toward parenthetical asides, however.  It might 
be improved by a blue pencil.

Amelia A. Lewis
Senior Architect
TIBCO/Extensibility, Inc.
Received on Thursday, 2 April 2009 14:11:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:17:20 UTC