- From: Amelia A Lewis <alewis@tibco.com>
- Date: Thu, 22 May 2008 11:55:50 -0400
- To: Phil Adams <phil_adams@us.ibm.com>
- Cc: SOAP/JMS (list) <public-soap-jms@w3.org>, Roland Merrick <roland_merrick@uk.ibm.com>
Heyo, On 2008-05-22 11:44:21 -0400 Phil Adams <phil_adams@us.ibm.com> wrote: > Of course, my thinking here is restricted to the application server > environment, since that's what my focus is. There might be other > "runtimes" > out there that want to play in the SOAP/JMS sandbox as well that > would > operate differently and might have different testing characteristics. I suspect that this is what triggered my response. We deliver JMS as a standalone messaging application; I don't know that we deliver it within a web application server environment (but I don't know the entire TIBCO software line, mind). Doing the least necessary to verify conformance seems to me to be the key. We *are* defining at the API level. That's the only level we *can* define at, interoperably. Vendors may be supplying other APIs that make it easier, but ... fundamentally, we're defining which APIs are called in order to generate a SOAP message, and which are or should be called to consume it. We can't test wire-level conformance, because JMS ain't got it. We *can* define a serialization, using the same sets of APIs that we are effectively using to define the protocol, and verify that the output is conformant/consistent. Do we need more for bootstrapping than the JNDI environment? Amy! (could folks please stop copying the -request list address in reply-alls?) -- Amelia A. Lewis Senior Architect TIBCO/Extensibility, Inc. alewis@tibco.com
Received on Thursday, 22 May 2008 15:57:04 UTC