Comment on proposal to BINDINGS-88

[Simon Holdsworth <simon_holdsworth@uk.ibm.com>]

Comments on proposed resolution to BINDINGS-88
"JCA Missing Normative statements re: connection/@name and activationSpec/@name"

My main concern is that we should resolve
this issue with text that is consistent with what we have in the JMS binding
spec, where relevant.

Here's the text from the JCA binding
CD03 regarding connection factory create being "never":

•        /binding.jca/outboundConnection/connection/@name
– if the @create attribute is “never”, the name uniquely identifies
an existing instance of the managed connection factory. 
If the connection/@create attribute
is “always”, the @name value MUST be unique within the domain [BJC20007].


The definitions for JMS binding are
on the destination element's @jndiName attribute (note that for JMS, we
changed "name" to "jndiName" and collected together
the cases for @create in issue 31) these definitions describe the behaviour
of the @create attribute for all the resources in the JMS binding, destination
connection factory and activation spec.

The JMS binding deals with "never"
by saying that "the [...] attribute MUST specify the location of the
existing resource";  and with "always" by saying "if
the resource cannot be created at the specified location then the SCA runtime
MUST raise an error".

If we want the wording to be at least
consistent, we would update the JCA statements about connection and activationSpec
as:

•        /binding.jca/outboundConnection/connection/@name
– if the @create attribute is “never”, the @name attribute MUST specify
the location of the existing connection [BJC2000X].
If the connection/@create attribute
is “always”, and the resource identified by the @name attribute cannot
be created at the specified location then the SCA runtime MUST raise an
error [BJC20007]. 

and

•        /binding.jca/inboundConnection/activationSpec/@name
– if the @create attribute is “never”, the @name attribute MUST specify
the location of the existing activation spec [BJC2000X]. 
If the activationSpec/@create attribute
is “always”, and the resource identified by the @name attribute cannot
be created at the specified location then the SCA runtime MUST raise an
error [BJC20013].

It would be better I think to refactor
the JCA wording so that all the statements about the behaviour of the @name
and @create attributes are in one place as in the JMS binding spec, which
would effectively mean applying the resolution to issue BINDINGS-31 to
the JCA binding spec, but that of course requires someone take the time
to do so...

Simon Holdsworth
STSM, SCA Bindings Architect; Master Inventor; OASIS SCA Bindings TC Chair
MP 211, IBM UK Labs, Hursley Park, Winchester SO21 2JN, UK
Tel +44-1962-815059 (Internal 245059) Fax +44-1962-816898
Internet - Simon_Holdsworth@uk.ibm.com

---

Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU