Hi, Good questions. Just came back from talking at IEEE APSCC 2009 about some of the topics you've mentioned. (Yes, I also do some research.:-) )
2009/12/11 herbst_andrew [email protected] > > I am not entirely sure if I can formulate a specific question. I am > unclear to what extent operations do things *other* than process messages > (I am aware this question might reveal a fundamental misunderstanding on my > part). > I think want you want to know is whether message is equivalent to the functional requirement fulfilled by a service. No, it's not. That's why I mentioned in my talk about doing identifying services and then doing semantic message modeling to design flexibility. What's missing from just looking at the message is the NFR. I happened to had an opportunity to listen to a talk by one of the famous UML guru and he mentioned importance of NFR but went on talk just about how FR relates to the architecture. There've also been several post in this forum about NFR, and came to the conclusion that what's been missing is the clear model of relationships between FR, NFR, and architecture so that customers will know how changes in one will effect the other. I'll write about this in my next paper because I've already developed a methodology to express it. > I am also inclined to ask a question of the form "Do messages determine > the content of operations or do operations determine the content of > messages?", if you can make sense of that way of framing things. > It's interrelated. Do both and make adjustments. > I am also interested in opinions as to how to make sense of this > statement by Erl: "A service groups a collection of related operations?". > Related in what sense? > This is a granuality question. I'm proposing related in FR, NFR, and semantic message model. H.Ozawa
