users@javaee-spec.java.net

[javaee-spec users] [jsr342-experts] Re: [jsr345-experts] Bean Validation aligment in EE 7 (was [jsr345-experts] Bean Validation support for EJBs?)

From: Jeff Genender <jgenender_at_savoirtech.com>
Date: Mon, 29 Oct 2012 17:03:39 -0600

Bill, can you please clarify "is definitely worth considering at some point". You mean dropping JAX-WS from EE in the future? I certainly hope that is not the case…

Jeff

On Oct 29, 2012, at 4:50 PM, Bill Shannon wrote:

> JAX-WS is not being replaced by JAX-RS. We have no current plans to prune JAX-WS.
>
> That said, we're not planning big updates to JAX-WS, mostly because we're busy doing other things and resources are finite. Having JAX-WS endpoints be CDI managed beans, with all the associated advantages, is definitely worth considering at some point.
>
> Antonio Goncalves wrote on 10/29/12 12:55:
>>
>> Hum... Werner, what you are saying is very important : is the EE EG clearly saying "we will not update JAX-WS because it's being replaced by JAX-RS" ? Prunning a technologie has a strong meaning but what do we want to say about JAX-WS (and JSP) ? "it's not prunned but we are not updating it" ?
>>
>> I think we should agree on the future of JAX-WS (and why not JSP) and express it loudly to the community. But it will be strange to have method level validation, injection, interception work in every spec, except JAX-WS
>>
>> On Mon, Oct 29, 2012 at 8:08 PM, Werner Keil <werner.keil_at_gmail.com> wrote:
>> Emmanuel/All,
>>
>> Thanks for that, and to Antonio for sharing.
>> As JAX-WS 2.0 is only in MR and there is no "real" update scheduled any time soon, that seems like a problem not so easy to tackle.
>> JAX-WS unlike RS has grown out of fashion to many, though I know first hand, large companies still use it and SOAP a lot, especially in projects already in production.
>>
>> Unless there was a demand and consensus to make such "soon legacy" JSRs optional at the very least with EE 8, maybe a "2.1" release could use CDI after all, but of course that depends on demand.
>>
>> Werner
>>
>> Am 29.10.2012 19:51 schrieb "Antonio Goncalves" <antonio.goncalves_at_gmail.com>:
>> Emmanuel's answer (he has no write access to the ML) :
>>
>> "The Java EE interception technology does not allow to intercept methods
>> calls when calling a method from within the bean. That's a general
>> problem that applies for all Java EE so at least that is consistent.
>>
>> I think it makes a lot of sense to support JAX-WS. But JAX-WS has no
>> integration with CDI so we could not ride along the CDI integration to
>> validate method calls.
>>
>> About servlets, I don't see the validation of methods called by the
>> servlet container as very useful. In your example, when is a request
>> object null?
>>
>> I discussed the JAXB validation integration possibilities with the spec
>> lead a long time ago but unfortunately, I don't think either of us
>> worked further on the subject. That is something I've seen asked several
>> times by the community though.
>>
>> Emmanuel"
>>
>> On Sat, Oct 27, 2012 at 5:49 PM, Antonio Goncalves <antonio.goncalves_at_gmail.com> wrote:
>> Hi all,
>>
>> A few months ago in the EJB 3.2 EG ML we talked about aligning BV in EJBs. We've also exchanged a few emails with Emmanuel Bernard (CCed) about BV aligment in other specification. As a developer I expect BV method level validation to work in every EE specification. What do you think ? If it's the case, I would expect this to work :
>>
>>
>> @WebServlet
>> public class MyServlet {
>>
>> public void doGet(@NotNull HttpServletRequest req, @NotNull HttpServletResponse resp) {}
>> }
>>
>> or
>>
>> @WebService
>> public class MyWS {
>>
>> @WebMethod
>> @NotNull
>> public String methodA(@WebParam(name = "myParam") @NotNull String s )
>> }
>>
>>
>> So that means a JAX-WS update will have to be made (JAX-WS 2.3 ?). With WebServices that would also mean that BV will be integrated with JAXB (but I think it's planned, Emmanuel correct me if I'm wrong).
>>
>>
>> What do you think ?
>> Antonio
>>
>>
>> On Wed, Aug 29, 2012 at 10:05 AM, Antonio Goncalves <antonio.goncalves_at_gmail.com> wrote:
>> Hi all,
>>
>> With the new Bean Validation 1.1 method-level validation I expect something to work out of the box (with or without CDI) :
>>
>> @Stateless
>> public class AccountService {
>>
>> public Account get(@NotNull Account model) { }
>> }
>>
>>
>> As for Stateful bean the question is "is there a possible integration between BV and EJB like JPA or JSF" ? For example, in the following code, shall we automatically validate the bean on @Remove (like @PreRemove on JPA) :
>>
>> @Stateful
>> public class ShoppingCart {
>>
>> @NotNull
>> private List<CartItem> cartItems;
>>
>> @Remove
>> public voide checkout { // validate cartItems ? }
>> }
>>
>> I don't really see the point in something like that, but I might miss something. Could there be another possible integration ?
>>
>> Antonio
>>
>> On Wed, Aug 29, 2012 at 12:41 AM, Marina Vatkina <marina.vatkina_at_oracle.com> wrote:
>> Experts,
>>
>> As you know, EJB spec currently doesn't support Bean Validation. Do we want to change that?
>>
>> Let me know if you think that:
>>
>> a) BV should be supported for EJBs without CDI being enabled (probably with an opt-in flag to be backward compatible)
>> b) BV should be supported for EJBs only if CDI is enabled
>> c) there is no need for standardizing BV with respect to EJBs
>> d) you don't care either way :(
>>
>>
>> thanks,
>> -marina
>>
>>
>> --
>> Antonio Goncalves
>> Software architect and Java Champion
>>
>> Web site | Twitter | LinkedIn | Paris JUG | Devoxx France
>>
>>
>>
>> --
>> Antonio Goncalves
>> Software architect and Java Champion
>>
>> Web site | Twitter | LinkedIn | Paris JUG | Devoxx France
>>
>>
>>
>> --
>> Antonio Goncalves
>> Software architect and Java Champion
>>
>> Web site | Twitter | LinkedIn | Paris JUG | Devoxx France
>