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


Help: OASIS Mailing Lists Help | MarkMail Help

soa-rm-ra message

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

Subject: Re: [soa-rm-ra] UML Model Check

Following the action fro the last telcom, I've listed a few questions and comments to the left slides presented by Chris.

Slide 12. A viewpoint is a tricky thing: a viewpoint on differed from a viewpoint of the same thing. An external stakeholder with regard to SOA ecosystem sees different things that an internal stakeholder when trying to address the same service, for example. Probably, we have to distinguish between these viewpoints where possible.

Slide 13. It is not necessary true that Actors: "work with/to requirements and objectives, which are specific, measurable". Some actors a re people (have (enterprise) goals, (individual) needs, which are generic, immeasurable/difficult to measure) and some are non-human but looking for relatively uncertain semantical business functions and RWE.

Slide 13. "If all action implies intent, can a non-human actor have intent? or just following orders?" I think that the order is an expression of intent, i.e. an intent presents in all cases.

Slide 13. "Ecosystem is generally a peer network of stakeholders" - disagree. Ecosystem includes Actors and Stakeholders, i.e. there may be non-human entities in the ecosystem.

Slide 13. "System is a network of actors. An actor qua actor has no ‘skin in the game’ – no stake to hold. Not problem for delegate (works according to a ‘script’)" - Well, system includes Actors but I would not say "System is a network of actors" - this narrows the scope too much. An actor has certain behavioral pattern and it may be associated with quite complex business logic that may have 'a stake to hold'.

Slide 15. "Move explanation of the UML to an appendix" - do we need to repeat the context of the UML over there?

Slide 15. "Indicate limitations of the UML in fully describing RAF principles" - indication is not enough, IMO. If we do so, we have to have particular case to demonstrate. Actually, I am and always was surprised - why we use only Class Diagrams? I find it is difficult or inefficient to express all cases in Classes, sometimes we need Use-Cases, sometimes - Component diagrams.
Slide 15. "Rename Principle 3 to ‘Distinction of Concerns’ (not Separation of concerns’) – the concerns are not completely separated between the different viewpoints" - commonly known name is "Separation of concerns" while all things are somehow connected. I do not think that concerns must be totally separated between viewpoints - they are about connected things, aren't they?

Slide 15."Formalise much fewer terms" - it is an open question in the light of the SOA Ontology Technical Standard

- Michael

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