No problem, consider them removed in API.
Sorry, what's the WSDL model used in SCA again please?
On 8/30/06, Jim Marino <[EMAIL PROTECTED]> wrote:
On Aug 30, 2006, at 11:59 AM, Yang ZHONG wrote:
> Thank Ant so much for details, that's very helpful!
>
> I'll update the API:
> 2-1. remove INSTANCE
> 2-2. demonstrate the SCA way to get SymbolSpace.Registry
> implementation in
> JavaDoc
>
Why do we have the "inner" interface? I agree with Jeremy that it
doesn't seem very intuitive.
> SymbolSpace.Registry is the generic registry interface, it doesn't
> prevent
> us from having SCA specific interface(s).
> I can come up with a WSDL Registry interface proposal. Can someone
> tell me
> which model is used by SCA: WST, WSDL4J, EMF, home grown or
> anything else?
>
> On 8/30/06, Jeremy Boynes <[EMAIL PROTECTED]> wrote:
>>
>> I'm having trouble parsing this. Are you talking about an SCA
>> registry, or a WSDL registry, or the mechanism for locating it (via
>> IoC as Ant described or something else)?
>>
>> I was saying that in an IoC world there is /no/ provider part of the
>> API - a user never locates anything, they declare dependencies and
>> the container resolves them. The dependency would be on the actual
>> client interface (SymbolSpace.Registry I guess but I've already
>> admitted to not grokking the interfaces).
>>
>> --
>> Jeremy
>>
>> On Aug 30, 2006, at 11:28 AM, Yang ZHONG wrote:
>>
>> > Can someone propose a SCA specific API mentioned by Jeremy please?
>> > Something like
>> > interface ScaRegistry
>> > {
>> > Definition getDefinition (String nameSpace);
>> > Message getMessage (QName);
>> > }
>> > I don't know the SCA requirement much enough to make such proposal.
>> >
>> > At the same time, we can practise IoC in ScaRegistry service
>> locating.
>> > I hope I can learn from that practice and update the registry
>> > generic API
>> > accordingly.
>> >
>> > Thanks for pointing out a nice mechanism.
>> >
>> > On 8/30/06, Jim Marino <[EMAIL PROTECTED]> wrote:
>> >>
>> >>
>> >> On Aug 30, 2006, at 10:49 AM, Jeremy Boynes wrote:
>> >>
>> >> > As part of Tuscany there should not be any mechanism in a API
>> for
>> >> > "service location" - that is the responsibility of the IoC
>> >> container.
>> >> >
>> >> +1 (which nicely avoids external dependencies on some locator
>> >> implementation and evil Singletons)
>> >>
>> >> > --
>> >> > Jeremy
>> >> >
>> >>
>> >>
>> >>
>> ---------------------------------------------------------------------
>> >> To unsubscribe, e-mail: [EMAIL PROTECTED]
>> >> For additional commands, e-mail: [EMAIL PROTECTED]
>> >>
>> >>
>> >
>> >
>> > --
>> >
>> > Yang ZHONG
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>> For additional commands, e-mail: [EMAIL PROTECTED]
>>
>>
>
>
> --
>
> Yang ZHONG
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
--
Yang ZHONG
