[ http://issues.apache.org/jira/browse/WODEN-68?page=comments#action_12446095 ] Arthur Ryman commented on WODEN-68: -----------------------------------
Features at Risk <Marsh> http://lists.w3.org/Archives/Public/www-ws-desc/2006Oct/0038.html <Marsh> http://lists.w3.org/Archives/Public/www-ws-desc/2006Oct/0035.html Jonathan: Youenn is going to be late but he sent a proposal on MTOM <plh> http://lists.w3.org/Archives/Public/www-ws-desc/2006Oct/att-0047/wsdl20-mtom-extension.html Jonathan: Arthur proposed using content negotiation Arthur: why can't the client using the MTOM mime type Tom: not convinvce HTTP content negotiation is a good way to workaround the description, the WSDL should describe it Philippe: agree with Tom, we have not relied on content negotiation Asir: I understand this is just a temporary measure until policy assertion is defined Jonathan: policy will become the preferred way to advertise extensions Asir: +1 to Jonathan Jonathan: there are already policy assertion in use for WSDL 1.1 so this is a clean migration path to WSDL 2.0 Philippe: this will not satisfy the Canon requirement Jonathan: Youenn's proposal does use content negotiation in the case where MTOM is optional <Marsh> Arthur: Like to revisit the assumption of tying f&p removal with this extension. <Marsh> Arthur: Canon wanted to use f&p to describe MTOM. But now they're using WSDL extensions. <Marsh> ... They could have used extensions initially. <Marsh> ... We have replaced one extensibility mechanism with another, not functionality. <Marsh> ... Other group should take this up <Marsh> Philippe: They had a standard mechanism before, now they've lost it. <Marsh> Asir: XMLP WG is interested in this, +1 to Arthur's statement about domain experts. <Marsh> Arthur: Canon's use of MTOM feature was not well-defined, it relied on private stuff. <Marsh> ... No prior submission from Canon to define a standard MTOM f&p behavior. <Marsh> Philippe: Trying to represent Canon's position. <Marsh> ... thought using the feature URI was sufficient, maybe that wasn't the case... <Marsh> Asir: Disagree's with Philippe. Section 2.1 of Canon's proposal has a lot of stuff that would have to be spelled out in F&P. <JacekK> zaim, aaaa is me <Marsh> Arthur: If they wanted to use this feature as an interoperable mechanism, they need to have a spec. <TomJ> The spec they had was WSDL, now they are specifying it fully <Marsh> Charlton: Agree with Arthur, would need a spec. <asir> XMLP minutes are at http://www.w3.org/2006/10/18-xmlprotocol-minutes.html Philippe: Agree that the MTOM spec was not sufficient and they would need to specify the operation semantics ... XMLP is considering defining an MTOM policy assertion Tom: If XMPL is defining a spec will they answer the questions we raised? Asir: See the link I pasted. I expect the semantics to be specified. Jonathan: need to coordinate XMLP, Policy, and us ... there is a clear need for one or more description mechanisms for MTOM, we need to decide who defines Asir: there is an existing MTOM policy assertion and this will be a good starting point Philippe: no spec for it Asir: MSFT is planning to publish it <Roberto> it will all happen "one day" Jonathan: it is in the test suite Philippe: is it published somewhere Asir: on the Web, google it <plh> I don't find the test suite using google. do you have a pointer Asir? Tom: so the feeling is you don't want it described here since it will interfere with planned work on a policy assertion Jonathan: this group alone will not be able to close the issue since other groups are involved <TomJ> I am just not convinced that WSDL 2.0 readers/users will be happy that we just walk away from specifying how the MTOM is engaged in WSDL Jonathan: we could adopt the proposal as a working draft because its in our scope, but then we'd have to manage all the comments and prove there are implementations ... what can we do to move forward now? there was a call to resolve the F&P issue ... can be adopt Youenn's proposal as a draft? Asir: we can send Youenn's proposal to XMLP Jonathan: would we request a way to describe MTOM or specifically a WSDL extension? Asir: should let them decide Charlton: we could give them a list of requirements Jonathan: perhaps Canon should transfer the requirements to XMLP since they are a member of that <Roberto> our charter talks about MTOM: "support for this mechanism will be provided" Philippe: do we have to wait for the policy assertion? <plh> Philippe: looks like XMLP will do the policy assertion. Do we have to wait for them to know if they are going to do a WSDL extension? Jonathan: there are potentially 3 ways for MTOM 1) youenn's extension, 2) the XMLP assertion, 3) the MSFT assertion - these could all be different Arthur: it would be desirable to have one semantics that could be used in policy and elsewhere, e.g. an extension Philippe: we need to understand what XMLP will do Jonathan: what can we to make progress Tony: drop F&P Jonathan: we could also adopt some requirements, e.g. to have a way to describe MTOM Asir: we could continue to consider the Canon proposal and drop F&P Tom: I feel this group should define MTOM ... wasn't MTOM a normative part of the spec? Arthur: no, just a primer example Philippe: XMLP is the right place to define MTOM <asir> +1 to Philippe Philippe: Canon is a member of F&P Jonathan: the reason we have a separate document for MTOM is that is was not a part of the Adjuncts Tom: I assumed we had spec'ed a uri for MTOM Jonathan: the MTOM spec defines a feature URI, but it is insufficient for defining the behavior Tom: I am concerned that we are losing a capability, e.g. the ability to use a SOAP feature in WSDL 2.0 Jonathan: I believe WSO2 supports WS-Policy so it interops with MSFT Charlton: is WS-Policy in Axis2 Tom: not is Axis2 but is at Apache <asir> Apache Neethi Tom: I propose we drop F&P and move MTOM to XMLP Philippe: last week I doubted XMLP would take on this MTOM work, but now they have indicated they want to Roberto: I agree that MTOM should be handled by XMLP Arthur: IBM is interested in discussing MTOM in XMLP Jonathan: how do we transition Youenn's spec - Coordination? ... the proposal is to drop F&P and move MTOM to XMLP Philippe: should we discuss policy versus extension? Asir: let them review the spec with us Jonathan: it's already in the current spec ... Is there any objection to remove F&P and move MTOM to XMLP? ... no objections raised, so this is RESOLVED. <scribe> ACTION: Remove F&P - editors for each Part should make necessary edits [recorded in http://www.w3.org/2006/10/19-ws-desc-minutes.html#action04] <scribe> ACTION: contact Coordination Group and move Youenn's proposal to XMLP [recorded in http://www.w3.org/2006/10/19-ws-desc-minutes.html#action05] > Remove Feature and Property Components > -------------------------------------- > > Key: WODEN-68 > URL: http://issues.apache.org/jira/browse/WODEN-68 > Project: Woden > Issue Type: Bug > Components: Parser > Reporter: Arthur Ryman > Assigned To: John Kaputin > > The W3C WSDL 2.0 WG decided to drop Feature and Property components from the > spec. We need to remove these from Woden. [1] > [1] http://www.w3.org/2006/10/19-ws-desc-minutes.html#action04 -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
