[jbossws-dev] EJB3 endpoints into WAR deployments

[Jim Ma]

Okay. Alessio. Without the @WebContext support in nested EJB3Endpoint 
will be easier to
fix .  I'll get a quick fix for this issue and let you look at this 
patch later.

Cheers,
Jim

On 08/31/2011 02:01 PM, Alessio Soldano wrote:
> Hi Jim,
>
> On 08/31/2011 04:46 AM, Jim Ma wrote:
>>>> I'd say we won't have DeploymentType abstraction anymore.
>>>> We will just have EndpointType abstraction with the following values:
>>>>     * JAXWS_POJO
>>>>     * JAXWS_EJB3
>>>>     * JAXRPC_POJO
>>>>     * JAXRPC_EJB21
>>> The jaxrpc side of the issue is still to analysed, anyway, I see your
>>> point, we can simply derive the DeploymentType info (JAR or WAR) from
>>> the new EndpointType, so the former can simply be removed.
>> But we still need let each endpoint easily know its
>> deploymentType(war,ear or jar) information and other
>> deployment context information(WarMetaData, EARMetaData or we wrapped
>> JSEArchiveMeteData, EJB EJBArchiveMetaData).
>> Then we can decide the address and context root when the ws endpoint is
>> actually published.
> I believe that the jar/war info can be derived from the EndpointType
> once that's set, however if I'm proven otherwise, you can keep the
> DeploymentType for this.
>
>
>>> In the next future we'll work on allowing different context-root for
>>> endpoints in the same webapp: that also implies deep changes in the
>>> container integration, as most probably multiple webapps need to be
>>> started to be able to publish endpoints in different context-roots and I
>>> can foresee multiple issues with that (classloading just to say one).
>> The @WebContext annotated with EJB3Endpoint is supported outside war
>> file . We won't consider to support this case this time to keep thing
>> simple ?
> Right, that can go in the future effort.
>
>
>> JAXWS_POJO WS endpoints coexists with EJB3Endpoint in a war file and
>> there is jboss-web.xml to define the Context root with specific value.
>> and EJB3Endpoint annotated with @WebContext.
>>
>> Alessio, can your new created  dynamically publish endpoint function
>> help on this case ?
> I'm not sure about that; we'll better evaluate when working on this
> additional effort; theoretically, the new publish endpoint function does
> not add anything really relevant here, but might be convenient. We'll see.
>
> Cheers
> Alessio
>
>