anonymous morphed into Pete W
Pete W morphed into Pete Walker
Martin C will be a few minutes late joining the call
Michael Rowley: 1. Preliminaries
- Roll Call
http://www.oasis-open.org/committees/membership.php?wg_abbrev=sca-j
- Appointment of scribe. List attached below
- Agenda bashing
- Approval of minutes from previous meeting(s) http://www.oasis-open.org/apps/org/workgroup/sca-j/download.php/25773/SCA%20Java%20minutes%202007-10-18.doc
2. Accepting new issues:
#TBD - A hole in the algorithm of introspecting property/reference from an unannotated impl class
Raiser: Raymond Feng
3. Accepting OSOA Errata to the input specs
http://lists.oasis-open.org/archives/sca-j/200710/msg00073.html
http://lists.oasis-open.org/archives/sca-j/200710/msg00074.html
http://lists.oasis-open.org/archives/sca-j/200710/msg00075.html
4. Issues discussion
Issue #3 - Local services expose implementation classes as their type http://lists.oasis-open.org/archives/sca-j/200710/msg00059.html
(to be prepared by Michael R.)
#TBD - A hole in the algorithm of introspecting property/reference from an unannotated impl class The proposal in the issue.
Issue 11 - Semantics of getCallbackID() are underspecified http://lists.oasis-open.org/archives/sca-j/200710/msg00028.html
Issue #4: Dependency Reinjection
http://lists.oasis-open.org/archives/sca-j/200710/msg00037.html
4. Adjourn
------
Rotating scribe list:
Jim Marino BEA Systems, Inc.
Bryan Aupperle IBM
Jean-Sebastien Delfino IBM
Simon Nash IBM
Martin Chapman Oracle Corporation
Uday Joshi Oracle Corporation
Michael Keith Oracle Corporation
Jason Kinner Oracle Corporation
Jeff Mischkinsky Oracle Corporation
Ron Barack SAP AG
Sanjay Patil SAP AG
Peter Peshev SAP AG
Roberto Chinnici Sun Microsystems
Peter Walker Sun Microsystems
Sriram Narasimhan TIBCO Software Inc.
Pradeep Simha TIBCO Software Inc.
Scott Vorthmann TIBCO Software Inc.
---- scribed once -----
Anish Karmarkar Oracle Corporation
Mike Edwards IBM
Michael Beisiegel IBM
Ashok Malhotra Oracle Corporation
David Booz IBM
Simon Nash: Simon Nash is scribe
Simon Nash: we don't have quorum... only 7 out of 19
Simon Nash: will wait for more people
Simon Nash: meanwhile have informal discussions
Simon Nash: Issue 11
Reza Shafii: could we have the jira link?
Simon Nash: in the nemail
Michael Rowley: http://www.osoa.org/jira/browse/JAVA-11
Simon Nash: http://lists.oasis-open.org/archives/sca-j/200710/msg00028.html
Reza Shafii: thanks
Simon Nash: use the one posted by Mike
Simon Nash: i am one step behind
Simon Nash: informal discussion: proposal is OK
Simon Nash: Action: Simon to write a proposal for voting
Simon Nash: it will be along the lines of: the value is null until (....xxx happens...)
Simon Nash: martin: could use the time to start electronic ballots
Simon Nash: mike: open the email to talk about dependency reinjection
Simon Nash: http://lists.oasis-open.org/archives/sca-j/200710/msg00037.html
Simon Nash: Martin has joined so we are only 2 short of quorum now
Simon Nash: suggestion is to add a MAY into the spec
Simon Nash: if using a setter the setter will be called a second time
Simon Nash: under what circumstance would callback or property be modified?
Simon Nash: The difference for stateless services is that the callback field would not be available if the component is
651 servicing a request for anything other than the original client. So, the technique used in the previous
652 section, where there was a response from the backendService which was forwarded as a callback from
653 MyService would not work because the callback field would be null when the message from the backend
654 system was received.
Simon Nash: much discussion about whether there is a valid case that does not have a non-null proxy injected when the instance is created
Simon Nash: Mike: if the proxy is non-null then it could be updated rather thann reinjected
Simon Nash: Simon: the spec should not force the implementation to work that way
Simon Nash: Simon: are there any cases where the proxy is null initially then needs to be made non-null later?
Simon Nash: we might need another issue if the spec is not clear on where the callback should go to
Simon Nash: multiple clients wired to one service on composite-scoped component
Simon Nash: where does the callback go to? is this clear currently? could the spec be fixed easily to make?
Simon Nash: this clear? if not, then perhaps this case should be disallowed?
Simon Nash: another potential reinjection scenario arises with the ability to redirect the callback
Simon Nash: if the callback can be redirected to a different binding then this may make it hard to update the proxy
Simon Nash: action: simon to look at the specific case of "multiple clients wired to one service on composite-scoped component" and make a proposal for where the callbacks should go in this case.
Simon Nash: what hooks would there be to get called after reinjection?
Simon Nash: there's @Init that gets called after the inital set of injections
Simon Nash: should there be a similar thing for reinjection?
Simon Nash: meeting adjourned
Simon Nash: (after Jason Kinner was added to the roll)