OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

ws-tx message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: WS-AT: Issues impact on interop, protocol correctness and specs


Dear Ian and Eric,

As the owners of virtually all of the open issues relating to WS-AT, Peter and I have worked on the following categorizations to help with the action item to sort out the impact of the remaining issues on interoperation testing for WS-C/WS-AT.

In addition, we have tried to distinguish where we think there are hard-core bugs against presumed design intent, and where the issues are simply proposing improvements, or purely editorial clarifications.

The issues are therefore sorted in two different ways:

A) Interop impact, assuming that the current interop scenarios are preserved.

Categories: INTEROP: affects existing interop test; WIRE/STATE: affects wire protocol and.or state tables, but does not affect existing interop tests; WORDING: purely editorial clarification.

B) Severity of change to protocol.

Categories: BUG: violates presumed design intent/correctness; OUTCOME: status quo is correct, but creates aggressive aborts that could be avoided; IMPROVEMENT: status quo works but feature could be implemented in a way which is arguably more elegant/ reduces implementation/composition difficulty; NO DESIGN CHANGE: does not alter working of protocol at all.

The executive summary is that there are only two issues which we believe affects protocol correctness (i.e. are bugs against presumed design intent): #41, #55. Neither of these affects the current interop tests.

Summary:

The number of issues in each category are as follows:

INTEROP: 4
WIRE/STATE: 4
WORDING: 11

===============

BUG: 2
OUTCOME: 1
IMPROVEMENT: 5
NO DESIGN CHANGE: 11:

Details

INTEROP

Affecting scenario #5.6:

i039   WS-AT: Coordinator should not distinguish protocol of orphaned participants
i043   WS-AT: Invalid state inappropriate response to orphaned volatile participants  
 
Affecting any/most of scenarios #5.n

i052   WS-AT: Replay message generates protocol errors
i053   WS-AT: Eliminate Replay message

WIRE/STATE:

i045   WS-AT: Meaning of wsp:Optional 
i055   WS-AT: User Commit and User Rollback should not return Aborted in None state
i056   WS-AT: User Commit and User Rollback row of CV state table contradicts resolution of 010
i041   WS-AT: Invalid events should not cause defined transitions
 
WORDING:

i035   WS_TX: term "coordinator" overloaded
i036   WS-AT: Coordinator state machine incomplete                 
i037   WS-AT: Register/Preparing in coordinator state table problematic     
i038   WS-AT: Incomplete actions on expiry                   
i046   WS-AT: Is logging mandatory?                      
i047   WS-AT: Is Completion protocol mandatory?                
i048   WS-AT: Internal events and actions undefined                 
i049   WS-AT: Definition of expiry inadequate/at odds with state table       
i050   WS-AT: Only root coordinators can accept Completion protocol registrations  
i054   WS-AT: Comms Times Out internal event should be removed from state tables 
i057   WS-AT: conceptual nature of protocol diagrams needs to be underlined   

===============

BUG:

i041   WS-AT: Invalid events should not cause defined transitions
i055   WS-AT: User Commit and User Rollback should not return Aborted in None state

OUTCOME:

i052   WS-AT: Replay message generates protocol errors

IMPROVEMENT:

i039   WS-AT: Coordinator should not distinguish protocol of orphaned participants
i043   WS-AT: Invalid state inappropriate response to orphaned volatile participants  

i045   WS-AT: Meaning of wsp:Optional 
i053   WS-AT: Eliminate Replay message
i056   WS-AT: User Commit and User Rollback row of CV state table contradicts resolution of 010

NO DESIGN CHANGE exactly equals WORDING

Yours,

Peter & Alastair



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]