Follow-up on ACTION-89, which test assertions do we really need?

----
Protocol-2013 - Verdict - mostly redundant

We have this paragraph: "if the charset parameter is specified it is
checked to ensure that it matches the encoding value from the supplied
XML. If there is a mismatch then a fault MUST be generated.†
[Definition: Use fault subcode contentTypeMismatch in the event that the
values do not match.†]

Protocol-2012 captures the first sentence, and Protocol-2013 apparently
covers the [Definition:].  Notice that the definition doesn't have any
RFC 2119 language.

My suggestion is to drop Protocol 2013, and if people want to refine
somewhat, we can either expand the scope of Protocol 2012 to encompass
the definition, or change it to read:

"if the charset parameter is specified it is checked to ensure that it
matches the encoding value from the supplied XML. [Definition: If there
is a mismatch then a fault with subcode *contentTypeMismatch* MUST be
generated.]†"

Thoughts?

----

Protocol-2020 - Verdict - mostly redundant with 2019.

I suggest rewording this:
"if using SOAP 1.2, and the contentType  property has an action
parameter, that parameter value MUST match this SOAPJMS_soapAction
value.† [Definition:  Use fault subcode mismatchedSoapAction if the SOAP
1.2 action does not match.†] "

to this:

"if using SOAP 1.2, and the contentType property has an action
parameter, and that [Definition: parameter value does not match this
SOAPJMS_soapAction value, a fault with subcode *mismatchedSoapAction*
MUST be generated.†]

----

Protocol-2023 - Verdict - mostly redundant with 2022

I suggest rewording:
" MUST appear in the JMS message in the JMS property named
SOAPJMS_requestURI.†  [Definition: Use fault subcode missingRequestURI
if the SOAPJMS_requestURI is missing from the message.†] "

to this:
"Appears in the JMS Message as the property SOAPMSG_requestURI.
Implementations MUST generate a fault with [Definition: subcode
missingRequestURI if the SOAPJMS_requestURI is missing from the message.]†

----

Protocol-2024 - Verdict - we should rewrite this so that the conformance
statement can appear in its entirety.  I'm still contemplating the best
way to do this.

----
Protocol-2035 - Verdict - Redundant.

We have individual conformance statements about each property.

----
Protocol-2039 - Verdict - Probably redundant - definitely weird because
it applies to an entire table.

JMSDeliveryMode - this is already written as a conformance statement but
we don't treat it that way.

JMSExpiration - vague text here doesn't need a conformance statement

JMSPriority - we could instead state that "this *SHOULD* be copied from
the JMSPriority of the request message.

JMSCorrelationID - covered by Protocol-2038

JMSDestination - covered by Protocol-2037

SOAPJMS_requestURI - hmmmm

SOAPJMS_bindingVersion - covered by Protocol-2010

SOAPJMS_contentType - covered by Protocol-2016

body - covered by RFC 2119 statements in section 2.4


----
Protocol-2041 - Verdict - Redundant.

Same as Protocol-2035

-----

This concludes my efforts for Action 89.  Obvious follow-up is for us to
discuss these items in the working group, and figure out the next action
items.

-Eric.

Received on Thursday, 25 June 2009 21:55:27 UTC