[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: RE: Some more diagrams for the model..
At 09:59 AM 4/23/01 +0100, Peter Furniss wrote: >Sazi, > >Re your simple protocol diagram. What happens to service2 ? Or is part of >the diagram missing (for me, at least). Service 2 also get a terminate message. (It is missing from the diagram) > >Is it possible for the work of (on) the two services to get different >results (confirm one, cancel the other) ? The original BEA proposal >inferred this was (I think - remembering, not checking in my haste) , but >referred to this as withdrawing. > The simple diagram just show the actors simply, not the details of the messages exchanged between. Yes different services may get different messages (terminate, cancel etc) >Are your actors (especially the sub-coordinator) representations of the >static capability to handle the cohesion protocol or the instance involved >in a particular one. Sub-coordinator is an actor who is capable of handling the protocol (note that in an implementation both service and sub-coordinater may be the same entity, but as far as the protocol concern there is a sub-coordinator and a service) > You have two registrations of sub-coordinator to >coordinator, but only one terminate reply - is this reply then structured to >say which of the registrations are cancelled, which confirmed ? Or would >there be a separate withdraw instruction for a cancelled registration ? The >former (at least) would make the coordinator : sub-coordinator protocol >similar to the cohesion outcome protocol that Alastair's last message talked >about, rather than the atom-level protocol in our main document. Again, the terminate message to the second service is missing in the simple diagram -- see the larger sequence diagram where each service gets message from the coordinator. >The neat >thing about the atom level protocol is that you can use it to do the >cohesion things, but the service/participant/sub-coordinator side need know >nothing about it (more precisely, does not need to implement anything >special to be involved in all sorts of different cohesion structures). > Yes, that is also my point that the service/participant/sub-coordinator do not know whether they are in an atom or not, they are just part of a transaction. >------------------------------------------------ >Peter Furniss >Choreology Ltd > >email: peter.furniss@choreology.com >phone: +44 20 7670 1679 >direct: +44 20 7670 1783 >mobile: +44 7951 536168 >13 Austin Friars, London EC2N 2JX> > > >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC