[cdi-dev] Minutes from EG meeting 20/08/2012

Pete Muir pmuir at redhat.com
Tue Aug 21 09:45:31 EDT 2012


Pete Muir
Mark Struberg
Joe Bergmark
Mike Brock
Jozef Hartinger
Antoine Sabot Durand
Fabien Marsaud

Recap from last week

* Multiple Annotated Types - we need to schedule a meeting which Stuart can make, to discuss. Pete to arrange.
* Should we use private methods as sources of observers, producers and initializers. Mike checked JLS, it is possible. EJB spec set the precedent to use them. We will use them. Pete to write up.

Scopes for non Http Request

Joe had dug into this. EG conclusion was that we should require a remote invocation to *always* be in a new request scope, and that it must not be shared or interfere with the lifecycle of an existing request scope. Joe to write up spec changes with help from Pete.
We also discussed that we should explicitly require the application scope to be available during @Singleton @Startup @PostConstruct methods. Joe to write up CDI spec changes. Pete to discuss integration with Java EE spec leads.

Disposers on specialized beans

Mark outlined a potential issue here. Jozef and Pete think this is an implementation issue, not a spec issue. Mark to work with Martin to produce TCK test and if a spec issue is discovered, we will fix it then.

Short lived producers on longer scoped beans

AKA the manager pattern. CDI-230 was discussed and we decided this is better addressed by a FAQ. Issue has been rejected. No further action.


Pete proposed dropping @Veto, and instead being able to declaratively and explicitly control whether a bean is enabled or disabled. Pete to write up full proposal.


This is issue CDI-43. Full notes on https://issues.jboss.org/browse/CDI-43. Pete to write up spec changes based on EG discussion.

See you all the same time next week :-)

More information about the cdi-dev mailing list