Spec issues 20050516 (attached).

[Gary P Cole <Gary.P.Cole@Sun.COM>]

|3. What is the 'namespaceURI' value for a SPML1.0 (DSMLv2) attribute expression?
|20050325 - Overlap?
|20050329 - Still may want to mention this as an example in the main spec.
|20050331 - SPMLv2 DSML Profile doc should specify this.
|20050510 - CLOSED.  If relevant, this belongs in DSML Profile doc.

4. Search on references.
		A requestor can query for Person instances based on “owner” 
      	only if the provider defines “owner” as an element or attribute of Person 
              in the schema for the target.  In such a case, what would the value of “owner” look like? 
              Would “owner=’joebob’ work?
20050329 - Discussed in con-call.  
           - Jeff said to say only what we do (not what we don't do).
           - Hal said to say we don't if people will assume that we do.
           - Compromise: ability to search on references is unspecified.
20050329 - Jeff Bohren suggested that maybe we *could* search on references.  I said I'd help.
20050412 - Jeff Bohren will enhance XSD 16 to support search on capability-specific data.
20050510 - XSD 17 adds QueryClauseType plus <and>, <or>, and <not> elements.
20050516 - (d07) Discussed search in Reference Capability, references in Search Capability.


|5. Should Custom Capabilities point to a "registry" of custom capabilities.
|20050222 - Darran suggested that the PSTC should maintain such a registry.
|20050510 - Agreed during conference call to remove from spec.
|20050512 - (d07) Fixed in Draft 7.
|

6. Update "Security and privacy considerations".
20050215 - Darran says that we should update this.  Wants a volunteer.
20050222 - Hal Lockhart volunteered to update this. Keep this in the spec.
20050329 - Hal thinks he can do this during this week.

7. Appendix I (throughout): Document references need validation and update.
20050329 - Raj volunteered during the con-call to do this.
20050510 - Hal: Beware normative references to non-std docs.

8. Appendix L: Glossary.  Definitions need review and update.
20050222 - Point to a separate document for Glossary.
20050329 - Raj volunteered during the con-call to do this.

|9. Conversational flow. 
|20050405 - Jeff sent email: "Suggested text for version 3.1.1".
|20050405 - I'd already removed the word "blocking".
|           Overlooks the case where provider chooses asynchronous execution.
|20050419 - (Jeff Bohren) wrote a followup (ellided for brevity):
|20050421 - Sent email to list: "Conversational flow" asking for help.
|20050510 - (d07) Removed diagrams; don't need requestID with sync transport.
|

10. What if reference capability declaration contains no reference definition?
	[Ed. If the provider did not declare any typeOfReference within target2’s declaration 
	of the Reference Capability for Person, would this mean that an instance of Person on target2 
	may use any type of reference refer to an instance of any schema entity on any target?]
20050510 - Agreed during con-call to discuss this on the list.
20050512 - Sent email: "Spec issue #10: Reference capability declaration without..."

11. What if a reference definition does not specify a targetID on the canReferTo schema entity?
	[Ed. If the provider did not specify a targetID within “Person-owns-Account”, 
	would this be an error (since target2 does not support Account), or would this mean that 
	an instance of Person on target2 may use the “owns” type of reference refer to 
	an instance of Account on any target?]
20050510 - Agreed during con-call to discuss this on the list.
20050512 - Sent email: "Spec issue #11: Reference definition CanReferTo without targetID..."
20050516 - AGREED: Specify targetID is required if multiple targets.
20050516 - Gary to fix in Spec Draft 8.

|12. Rami wants each request element to have a corresponding response element for clarity.
|20050404 - We can do this without defining another type.  Sent email: "Request/response pairs."
|20050405 - (d06) Illustrate the addition of a new top-level element bulkModifyResponse of type spml:ResponseType.
|20050419 - Agreed during con-call to define a top-level element for each request and response.
|20050510 - (d07) XSD 17 addresses this.
|
|13. Spec should describe Complex References (per M.Raeppler's Issues re: Suspend Capability)
|20050417 - (d07) Discussed Complex References in Reference Capability section.
|20050417 - Posted draft_pstc_spmlv2_reference_capability.doc to the list.
|20050419 - Martin Raeppler liked the draft, and wanted it in Draft 7.
|

14.  How does a provider declare the structure of reference data for complex references?
	Reference Definition contain xml schema (or refer to schema entity)?
20050510 - Agreed during con-call to discuss this on the list.
200500512 - Sent email: "Spec issue #14: Declaring the Structure of Complex Reference Data"


|15. Company affiliations for Gerry Woods and Darran Rolls are incorrect.
|20050421 - QUESTION. Sent email to list: "Company affiliations".
|	- Both were affiliated with OASIS members when they contributed.
|	- Both now belong to companies that are not OASIS members.
|	- Neither company wishes to disavow the contribution of its former employee.
|20050421 - SOA Software is now a member; list Gerry Woods as SOA.
|20050421 - Darran Rolls says to keep him listed as Sun.
|20050421 - (d07) Updated company affiliations for Bohren, Woods, and Rolls.
|
 
16. Line 559: the IdentifierType is used to identify targets, but is not limited
to targets. To date only targets are identified using an element of type
IdentifierType, but that may well change in the future. The type
IdentifierType should be considered a generic type for which
PSOIdentifierType is a specialization.
20050421 - DEFER. Sent email to list: "IdentifierType...".  Part of XSD issue #67 (targetID).
20050510 - Part of the "targetID" debate. (JeffB will publish side-by-side examples.)
Draft 18 - TargetID is now an attribute.  IdentifierType is used only by PSOIdentifierType.

 
|17. Draft 6, Lines 574 to 576. Strike these lines. This is not appropriate for normative text
|and is too basic to put into a spec document.
|20050421 - (d07) Fixed in Draft 7.
| 
|18. Draft 6, Line 697: String "The PSTC believes that". If we believe something we should
|just say it.
|20050421 - (d07) Fixed in Draft 7.
|
|19. Draft 6, Line 722: use "discovery" in place of "bootstrap"
|20050421 - (d07) Fixed in Draft 7.
| 
|20. Draft 6, Line 804: change "The PSTC recomends that the provider explicitly declare"
|to "The provider should explicitly declare". Everything in this documents
|should be what the TC believes by definition.
|20050421 - (d07) Fixed in Draft 7.
| 
|21. Draft 6. Globally change psoId->psoID, targetId->targetID, containerId->containerID.
|20050421 - AGREE. Need to do this in the XSD (issue #75).
|20050510 - (d07) Fixed in Draft 7.
 
22. Draft 6, Line 962: An add request MAY omit both the target ID and container ID if the
PSP has only one target and containment is not being specified. 
20050421 - Sent email to list: "AddRequest targetId, containerID and psoID."
20050512 - Re-sent email: "Spec issues #22 and #23: AddRequest targetId...."
20050516 - AGREE.  Gary to fix in spec Draft 8.

23. Draft 6, Line 962: An add request MAY contain a PSO ID. This section implies otherwise.
20050421 - Sent email to list: "AddRequest targetId, containerID and psoID."
20050512 - Re-sent email: "Spec issues #22 and #23: AddRequest targetId...."
20050516 - AGREE.  Gary to fix in spec Draft 8.

|24. Draft 6, Line 1128: Pso should be PSO
|20050421 - (d07) AGREE.  Fixed in Draft 7.
|
|25. Draft 6, Line 1158 and 1161: pso should not be in red.
|20050421 - (d07) AGREE.  Fixed in Draft 7.
|
|26. Draft 6, Line: 1189 In the example change toPsoId to psoID. Change XSDs also.
|20050421 - DISAGREE. Sent email to list: "ReferenceType#toPsoId".  Was XSD issue#57.
|20050510 - CLOSED. Jeff Bohren agreed during conference call that XSD#57 should hold.

27. Effective date must be UTC format date/time string (with no offsets).
20050510 - (d07) AGREE.  Fixed in Draft 7. 
20050510 - OPEN. Jeff Bohren will send better sample text from SAML spec.