[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: Re: [wsrp-wsia] Issues agenda for 9/26
Current plan: Vote: 1. (#6') Is groupID required? - Proposal is that with the initEnvironment() operation the need for groupID has been removed and therefore the field should be removed from the protocol. Discuss: 1. (#74) Should the Consumer MAY or MUST releaseHandles? - Draft v0.7 makes this change ... is it acceptable? 2. (#75) How to ensure consumerName is unique? - Draft v0.7 suggests that the URL of the Consumer could provide such a unique name. More needed? 3. (#78) Should we specify storage of consumerHandle? - Draft v0.7 incorporates the suggested changes. Are they acceptable? 4. (#79) Why return consumerHandle in modifyConsumer's consumerContext? - Draft v0.7 introduces a new structure to eliminate this potential source of error. 5. (#80) cloneEntity() and modification of properties - in same call? - Draft v0.7 eliminates this as the F2F decided to not merge properties into the other operations. 6. (#82) What is the need of previousMode (& previousWindowState)? - Draft v0.7 replaces these with a constant for requesting the Consumer return to the previous mode and a nomative statement that the Consumer MUST revert to VIEW_MODE ( VIEW_NORMAL in the case of windowState) if it does not have a previous value to revert to. 7. (#83) Is markupType a MIME type? - While the spec is silent on this currently, I think reusing MIME types makes more sense that developing another set. 8. (#92) Maximum size of a handle - Discussion on email list has made the following points: 1. For portals to operate efficiently, a handle should be usable as a database key (usually this means 255 bytes or less). 2. For security reasons, it would be good if the handle is not smaller than 256 bytes. 3. Producers may find URLs to be the most straightforward way of generating a handle ... this drives toward the length being unlimited. 9. (#93) Payload extensibility mechanism - Draft v0.7 explores changing this to the schema <any /> mechanism. This appears to require changing to style="document" and encoding="literal". Could people explore whether the revised WSDL works with various tools 10. (#94) Is getMarkup() able to modify/return state? - Current spec says the semantics are to return markup that reflects the current state. Propose we ratify that as the semantics. 11. (#95) destroyEntities() & refHandle - The F2F introduced destroyEntites(). One question that has arisen is whether a refHandle can be passed to this operation. Sense of the email discussion is NO. 12. Open items I have noted as resolved at the F2F. Is there agreement that these were the resolutions of the issues? - (#12) => split getDescription() into 2 operations - (#13) => split getDescription() into 2 operations - (#22) => split releaseHandles() into 2 operations, both of which return void or a fault message on failure - (#15) => withdrawn by Thomas - (#17) => initEnvironment() is in Markup factor with Producer metadata declaring need for usage - (#20) => entityState is required & is returnable by performInteraction() as well - (#32) => simplified property operations for entity configuration should be proposed - (#49) => split getDescription() into 2 operations - (#50) => No. Metadata is only defined in messages from get___Description () operations - (#54) => agreement with Gil's suggestion - (#55) => in-band operation is defined. Out-of-band process comments needed in spec. - (#56) => simplified property operations for entity configuration should be proposed - (#59) => wsdl field retained. If information was gathered prior to calling getServiceDescription(), it can be ignored. - (#60) => This was a suggestion for a presentation at the F2F - (#61) => F2F decided on 4 factors ... reflected in v0.7 - (#62) => process outlined for Consumer flag & entity fault message. Review how clear v0.7 lays out the scenario - (#64) => All permutations were approved. Persistent state only produced by entity. In the case of Consumer UI gathering property updates, the setEntityProperties() sets the entity's state regardless of where it is stored. - (#65) => This was a suggestion for a presentation at the F2F - (#68) => yes, initEnvironment() is needed. - (#76) => Removed this metadata - (#77) => Removed sendPublishedState flag - (#87) => Mode is a piece of state related to the Consumer's viewer, not the entity's state
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC