On 12/03/2013 13:12, John D. Ament wrote:
> Nigel,
>
>
> On Tue, Mar 12, 2013 at 8:27 AM, Nigel Deakin <nigel.deakin_at_oracle.com <mailto:nigel.deakin_at_oracle.com>> wrote:
>
> Community member "genomeprjct" logged this JIRA issue. This is a proposal that JMS takes advantages of the JCA and
> EJB changes described in the recent email by Bill Shannon that I forwarded recently.
>
>
> I thought we had previously noted that this user was in fact a member of the E.G.? A member who likely laments that the
> one piece of information that absolutely cannot be changed is his old Project Kenai username which has been imported to
> the java.net <http://java.net> JIRA?
Yes I think I did work that out at some point in the past, but I had forgotten :-)
> I personally believe we should be taking cues from the EJB and the Platform specs and at least have an answer for it in
> JMS 2.0. Based on the feedback I've seen from David Blevins, this is more or less a compromise based on size to be able
> to fit in to the current release.
We haven't even started discussing that the JMS spec should say (I didn't know this was coming). And in addition to
working out the annotations we want, we also need to develop TCK tests and a RI. David Blevins has already developed a
prototype RI; we haven't even started. I really don't think we should be rushing this important feature. But I'd welcome
views on this.
> I don't believe anything that was proposed was too substantial, especially since we don't require a JMS provider to even
> provide a Resource Adapter, it is only recommended at this point (pg 108, PFD).
I think the relevant spec to extend is the part of the EJB 3.2 spec which defines JMS message-driven beans. This defines
the interface it must implement, and the activation properties that must be supported. None of that is optional.
Nigel