[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [wsrp-interop] Groups - WSRP Interop SC call modified
Hi Andre, I disagree here, this is yet another usage pattern for MarkupContext. I think that MarkupContext should be handled in the same way, no matter whether it was returned from getMarkup() or performBlockingInteraction(). Furthermore we have added MarkupContext to the UpdateResponse as an optimization to save the additional getMarkup() call if markup can be returned from pba() - that's one reason why MarkupContext should be handled in the same way for both operations. Why do you want to return a title in pba() and the let the Consumer call getMarkup() right after this to obtain the markup? Why couldn't in this case the title be returned in the getMarkup() response? If you just want to change the title in pba() but remain the markup in the portlet fragment this would implicitly mean that the consumer has to cache the markup always - even if we don't call it caching here? Couldn't for that purpose the usedCachedMarkup flag be set to true? (Consumer supporting caching would adher to the pattern you want). So again even if I stress it, here is my interpretation/usage pattern of MarkupContext. 1. if pba() doesn't return MarkupContext in the UpdateResponse, i.e. it doesn't occur on the wire (btw. it isn't nillable), then the Consumer subsequently calls getMarkup() to obtain the markup. 2. if MarkupContext is returned within UpdateResponse, then the Consumer handles it in the same way as if would have returned it from getMarkup() and doesn't call getMarkup() any more. 3. if markupString and markupBinary is missing, then it is interpreted as "there is no content to display in the portlet's fragment", i.e. the consumer can choose to display "" for a better end user experience instead of throwing faults. In this case also the mimeType is missing in MarkupContext. 4. if either markupString or markupBinary is there, then it is interpreted as "display exactly this markup". In this case the mimeType must be passed with MarkupContext. Mit freundlichen Gruessen / best regards, Richard Jacob ______________________________________________________ IBM Lab Boeblingen, Germany Dept.8288, WebSphere Portal Server Development Phone: ++49 7031 16-3469 - Fax: ++49 7031 16-4888 Email: mailto:richard.jacob@de.ibm.com |---------+----------------------------> | | Andre Kramer | | | <andre.kramer@eu.| | | citrix.com> | | | | | | 07/18/2003 09:21 | | | AM | |---------+----------------------------> >--------------------------------------------------------------------------------------------------------------------------------------------------| | | | To: wsrp-interop@lists.oasis-open.org | | cc: | | Subject: RE: [wsrp-interop] Groups - WSRP Interop SC call modified | >--------------------------------------------------------------------------------------------------------------------------------------------------| Rich, thanks for the clarifications. However, I still have a problem with 3b. How can performBlockingInteraction return a new peferredTitle and no markup? In this case markupContext is not nill or missing and has: (a)markupContext.peferredTitle = "some new title" (b)markupContext.mimeType, markupContext.markupString and markupBinary are all null. Should (b) not be interpreted as "set title and call getMarkup", rather than skipping getMarkup and displaying an "" to the user (the latter is what I understood from the minutes)? I don't want to be forced to use the early markup optimzation just to set a title. And, in future CacheControl may be returned without markup from pbia, so we should not be defining rules like "if these fields are null then this means skip getMarkup" (or checking for (a and b) as a special case) as this can not be expressed in wsdl. I would prefer to make (b) mean call getMarkup for pbia. For getMarkup (b) can mean show a "" to the user just to make consumers more robust. regards, Andre -----Original Message----- From: Rich Thompson [mailto:richt2@us.ibm.com] Sent: 17 July 2003 18:36 To: wsrp-interop@lists.oasis-open.org Subject: RE: [wsrp-interop] Groups - WSRP Interop SC call modified Sorry my notes weren't clearer. 1a) We thought it would be useful to have a convention for a value to use in this common case. We preferred to take it from the wsrp namespace, but a convention can not define new values for a namespace. As a result we have an odd, but reasonable value for the convention. 3b) Apparently I wasn't clear enough. As an optional element, the MarkupContext can be missing. Some implementation may prefer to actually send it as a nil element. In both cases, this will stimulate the Consumer to invoke getMarkup. Returning a non-nil MarkupContext will be taken as having provided the markup for this user interaction. 4bi3) Yes, the Producer may throw an OperatioFailed fault. In no case is it sensible for a Consumer to presume that the persistent state of a POP may be written. Specifying this will likely return normally when no persistent state change was attempted and return a fault when a state change is attempted. Rich Thompson Andre Kramer <andre.kramer@eu.citrix.com> To: wsrp-interop@lists.oasis-open.org cc: 07/17/2003 01:15 PM Subject: RE: [wsrp-interop] Groups - WSRP Interop SC call modified Apologies for joining late on the the call. 1a - the convention of userContextKey="wsrp:minimal" is weird, but it's just a convention, right? Some systems have a guest system group that can be used as the basis for a real userContextKey. I don't understand 3.b: "The signal for getting getMarkup called is a nil MarkupContext or missing." - or missing what? Could it be missing mimetype? I would like to be able to return a new preferredTitle with and without markup from pbia. Also, in future we may be passing back a CacheControl without markup. 4 b(producer side) i(POP being accessed) 3(PortletStateChange=readWrite -> make no sense, throw OperationFailed) - in the general case this is just: "the producer may throw an Operation failed". i.e. only makes "no sense" for JSR portlet? My question is: how does a non-JSR168 portlet know that POP readWrite is not allowed? We don't expect it to catch OperationFailed and retry with readOnly do we? thanks, Andre -----Original Message----- From: richt2@us.ibm.com [mailto:richt2@us.ibm.com] Sent: 17 July 2003 17:35 To: wsrp-interop@lists.oasis-open.org Subject: [wsrp-interop] Groups - WSRP Interop SC call modified WSRP Interop SC call has been modified by Rich Thompson (richt2@us.ibm.com). Event description: Shared timeslot with Conformance SC and biweekly TC call. USA Toll Free Number: 877-718-0936 USA Toll Number: +1-712-923-6878 PARTICIPANT PASSCODE: 402957 Agenda: Date: Thursday, 17 July 2003 Time: 11:00am - 01:00pm Eastern Time This event is one in a list of recurring events. Other upcoming dates include: Thursday, 24 July 2003, 11:00am to 01:00pm Eastern Time Thursday, 31 July 2003, 11:00am to 01:00pm Eastern Time Thursday, 07 August 2003, 11:00am to 01:00pm Eastern Time Thursday, 14 August 2003, 11:00am to 01:00pm Eastern Time Thursday, 21 August 2003, 11:00am to 01:00pm Eastern Time Thursday, 28 August 2003, 11:00am to 01:00pm Eastern Time View event details: http://www.oasis-open.org/apps/org/workgroup/wsrp/wsrp-interop/event.php?event_id=2503 PLEASE NOTE: If the above link does not work for you, your email application may be breaking the link into two pieces. You may be able to copy and paste the entire link address into the address field of your web browser. Referenced Items Date Name Type ---- ---- ---- 17 Jul 2003 20030717 WSRP Interoperabil... Minutes You may leave a Technical Committee at any time by visiting http://www.oasis-open.org/apps/org/workgroup/wsrp-interop/members/leave_workgroup.php
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]