The most important/practical reason this should probably be in CDI is that most JTA experts are probably not the right people to grok/implement this, whereas most CDI experts probably know enough about JTA to implement it (not to mention the fact that a JTA version is likely not forthcoming soon/easily).

The reason to standardize it would be that software that depend on this scope (such as the JMS context concept in JMS 2) can simply rely on the fact that it will be in the standard runtime instead of having to pack it themselves.

From: Lincoln Baxter [mailto:lbaxter@redhat.com]
Sent: Tuesday, May 22, 2012 8:05 PM
To: Reza Rahman
Cc: cdi-dev@lists.jboss.org
Subject: Re: [cdi-dev] Transaction Scope for CDI

Is CDI really the right place for this Annotation?

It sounds like this has behavior specific to JTA, which leads me to think that it should be added in the JTA spec, not the CDI spec. What needs to change in the core CDI behavior to make this possible, if anything? If this can be achieved with existing SPIs, then I don't see why CDI would need to change at all to support this.

~Lincoln

---

From: "Reza Rahman" <reza_rahman@lycos.com>
To: cdi-dev@lists.jboss.org
Sent: Tuesday, May 22, 2012 7:28:13 PM
Subject: [cdi-dev] Transaction Scope for CDI

CDI enthusiasts,

Pete, I and Nigel (JMS 2 spec lead) have been discussing the issue of the transaction scope behind the scenes the past few weeks. Attached is what we came up with and feel it meets the various related use-cases the most effectively. The downside is that it is quite involved (conceptually) and might take a bit of patience to absorb. Please give it a read and let me know your thoughts.

Cheers,

Reza

_______________________________________________
cdi-dev mailing list
cdi-dev@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/cdi-dev