[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [wsbpel] Question on Profile Usage Re: [wsbpel] Issue 82 - resolution]
You are quite right that tools may treat external behavior descriptions as "templates" for implementation. So "external description" vs "template" is not the best nomenclature. The abstract class analogy you cited is interesting. We have very specific completion rules for each profile and I can imagine an abstract class variation that forbids the addition of new methods, say through an additional keyword modifier. That would be analogous to what we need. -----Original Message----- From: andrew.francis@mail.mcgill.ca [mailto:andrew.francis@mail.mcgill.ca] Sent: Wednesday, January 19, 2005 6:15 AM To: Satish Thatte Cc: rkhalaf@watson.ibm.com; wsbpel@lists.oasis-open.org; andrew.francis@mail.mcgill.ca Subject: RE: [wsbpel] Question on Profile Usage Re: [wsbpel] Issue 82 - resolution] Hello Satish: > I believe your example falls into the (intended) > "external behavior description" profile which forbids > executable completions that add interactions along the > partnerlinks in the abstract process since the intent of > such an abstract process is to fully specify externally > visible behavior. This is true. However I think of the aforementioned scenario as a process template. I envision the BPEL abstract process being translated into an abstract class in some language, the implementor overriding methods that customise behaviour (this is the template design pattern). And here lies the problem. Different names for similar activities are going to result in a proliferation of profiles and tool sets. Also how do I communicate, preferably in a machine processable form, the description you or I, just provided? I think it is just easier to agree on an abstract process's behaviour..... An analogy. Most object-oriented languages have some notion of an abstract class, its behaviour supported to different extents by the language. Most educated programmers understand abstract classes. And scores of design patterns (read use cases) use abstract classes. However I don't see Java, C#, C++, using profiles to describe how the abstract class should be used. Cheers, Andrew
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]