I added 2 footnotes into the corresponding areas saying that "The
concurrency management type CONTAINER may be specified for stateful
session beans, but doing so has no impact on the semantics of
concurrency management for such beans."
-marina
On 12/17/12 11:11 AM, Marina Vatkina wrote:
> Thank you Linda. Should the spec say something as well?
>
> -marina
>
> On 12/15/12 2:07 PM, Linda DeMichiel wrote:
>> This was discussed in the EJB 3.1 expert group, which went back and
>> forth over the
>> handling of the SFSB case. I don't know whether its presence at the
>> time of the
>> EJB 3.1 release was intentional or not. However, it is harmless, and
>> cannot be
>> removed.
>>
>> -Linda
>>
>>
>> On 12/14/2012 6:47 PM, Marina Vatkina wrote:
>>> Experts,
>>>
>>> ejb-jar xsd and the javadoc for the ConcurrencyManagement annotation
>>> say that the concurrency management type can be
>>> specified on a SFSB, but if specified, only Container is allowed
>>> ([1] and [2]).
>>>
>>> Is it intentional? What's the purpose of it if it is? There is
>>> nothing in the spec itself that describes this option.
>>>
>>> I think it should be removed until we add concurrency management
>>> options to the SFSBs.
>>>
>>> WDYT?
>>>
>>> -marina
>>>
>>> [1]<xsd:complexType name="concurrency-management-typeType">
>>> <xsd:annotation>
>>> <xsd:documentation>
>>>
>>> The concurrency-management-typeType specifies the way concurrency
>>> is managed for a singleton or stateful session bean.
>>>
>>> The concurrency management type must be one of the following:
>>>
>>> Bean
>>> Container
>>>
>>> Bean managed concurrency can only be specified for a singleton bean.
>>>
>>> </xsd:documentation>
>>>
>>> [2] " This annotation may be applied to stateful session beans, but
>>> doing so has no impact on the semantics of
>>> concurrency management for such beans. The concurrency management
>>> type for bean-managed concurrency (|BEAN|) does not
>>> apply to stateful session beans."
>>
>