lwri...@frogooa.com
未讀,2011年3月8日 下午1:47:552011/3/8登入以回覆作者
登入以轉寄訊息
你的權限不足,無法在這個群組刪除訊息
該群組的電子郵件地址為匿名,或你需要檢視成員電子郵件地址的權限才能查看原始貼文
收件者:miUML
I don't know if there are newer versions of the meta-model laying
around, but I got to looking at the latest posted versions and managed
to resolve some of my discomfort. The key was the relationship, R23,
between Identifier and Relationship, and my continued discomfort with
R27 between Class and Identifier. I came to the following epiphany:
Within the context of OOA, Classes don't participate in Relationships!
Instances participate in Relationships. The one exception is the
Associative Class, and that one requires some more thought.
The visual representation of the class model certainly makes it appear
like Classes are participating in Relationships, but if you draw those
relationships using a potato diagram, you don't draw lines between the
potatoes (classes), but rather between the eyes (instances). The verb
phrase, "participates in", also makes this clear. I have had a
longtime, nagging suspicion that all the meta-models I've seen to date
have been a little polluted by drawing tool considerations.
This clarification allows one to remove objects like Link and
Population, which were also striking me as somewhat redundant, but I
couldn't quite figure out why, until now.
This needs some more work, but I thought I'd throw it out for
discussion.