O.K that's why I asked about what the source side looked like. A
degenerate outbound wire seems right. One thing on this is can we
cache the degenerate wire so the wire only need to be constructed
once while the runtime is in operation?
We have the problem I outlined, e.g. if the target is expecting to be
in a transactional context and the source never "manually" sets one
up...
Jim
On Nov 30, 2006, at 9:10 AM, Jeremy Boynes wrote:
On 11/30/06, Jim Marino <[EMAIL PROTECTED]> wrote:
This I think highlights a problem with the spec around using
CurrentCompositeContext from unmamanged code in general...a locate
service only returns "half" of an invocation chain since the source
is not a component and is "unmamanged" code". Can you describe you
client code, is it a JSP or is it from a component?
That shouldn't matter should it? I think Greg's problem is that there
is a data incompatibility between the source side (the supplied Java
interface) and the target side (the WS reference). Normally the
connector fixes this by adding a transform interceptor to convert from
the Object on the source side to the OMElement on the target
side.However, locateService bypasses the connector so this does not
get done.
I think Greg's proposal is right - create a degenerate outbound wire
and then connect it to the target's inbound one. We know we needed to
do that for policy at some point but it looks like we'll have to do it
sooner due to this transform issue.
--
Jeremy
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
