[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [ws-rx] NEW ISSUE: Inefficient close/terminate of offered sequences
Paul, On 17/05/06 10:16, Paul Fremantle wrote: > Finally, Your statement that the originator should not be able to > terminate contradicts the spec. Either side can terminate the spec at > any time. Section 3.3 of WD12 (and on) seems very clear to me. Only the RM Source may terminate a sequence, at least when using the terminate sequence request. I believe the implicit assumption here is the RM Source shouldn't have any resources it needs to reclaim associated with a closed sequence. Therefore, the RM Destination need only inform the RM Source of closure (using a Final sequence acknowledgement) in regular processing. The Sequence Terminated Fault covers other cases. Again, I don't see the issue you have stated. Mechanisms which are needed for clean closure and termination of sequences in either direction already exist. thanx, doug > Paul
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]