[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [business-transaction] Issue ex-4
Peter, I'm happy with the modifications you've proposed. Mark. ----- Original Message ----- From: "Furniss, Peter" <Peter.Furniss@choreology.com> To: <business-transaction@lists.oasis-open.org> Sent: Monday, March 01, 2004 1:55 PM Subject: [business-transaction] Issue ex-4 > As I mentioned on the last call, Mark had promptly supplied the text to > complete issue ex-4, and I had > neglected to pass it on. > > > Add the standard qualifier cancel-on-zero-participants > > <btpq:cancel-on-zero-participants> > <btpq:value> > true/false > </btpq:value> > </btpq:cancel-on-zero-participants> > > With associated text: > > The cancel-on-zero-participants qualifier causes a cohesion composer to > be automatically cancelled if its list of registered participants > becomes zero after either a PREPARE_INFERIORS or CANCEL_INFERIORS > message is received. This cancellation happens if the value for > cancel-on-zero-participants is set to true, and the result will be a > TRANSACTION_CANCELLED message returned to the terminator as a reply to > the PREPARE_INFERIORS or CANCEL_INFERIORS message, instead of > INFERIOR_STATUSES. > > Then in the text for BEGIN we need to add something to the section on > qualifiers: the standard qualifier cancel-on-zero-participants may be > present on BEGIN for a cohesion; if the value is true, then automatic > cancellation of the cohesion will occur as per the rules for > cancel-on-zero-participants. > > In the text for PREPARE_INFERIORS/CANCEL_INFERIORS for qualifers: the > cancel-on-zero-participants qualifer may be present; if it is, then it > overrides any previous value for this qualifier that may have been > provided when the cohesion was started via BEGIN. > ------------- > > Mark had said he didn't mind whether the default was true or false. I'm > not sure whether a default is needed. If the qualifier is absent, then > it is the same as false (existing behaviour). Do we even need a value ? > > I've added the bit on the first paragraph from "terminator .." to the > end, to clarify the sequence. (Terminator:decider relationship is always > request/reply so the TRANSACTION_CANCELLED has to be a reply to > something. > > > > > To unsubscribe from this mailing list (and be removed from the roster of the OASIS TC), go to http://www.oasis-open.org/apps/org/workgroup/business-transaction/members/leave_workgroup.php. > >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]