home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.object
- Path: sparky!uunet!projtech!sally
- From: sally@projtech.com (Sally Shlaer)
- Subject: Re: SHlaer-Mellor OOA Model, Problem with definition of Object
- Message-ID: <1992Jul21.175138.18096@projtech.com>
- Organization: Project Technology, Inc., Berkeley, CA
- Date: Tue, 21 Jul 1992 17:51:38 GMT
- Lines: 94
-
- >> [from sally@projtech.com (Sally Shlaer):]
- >> In summary, in OOA we are interested in the typical instance.
- Because we
- >> are dealing with analysis, we are not interested in particular
- instances.
-
- > [from johnson@m.cs.uiuc.edu (Ralph Johnson) :]
- > It seems to me that there are plenty of times in analysis where you
- are
- > concerned with individual instances. [Then a pretty example follows.]
-
- Yes, you are right. I was not clear. I should have said we are not
- interested in *specified* instances. We are obviously interested in
- individual unspecified instances that meet certain criteria. There are
- many examples of this in Object Lifecycles.
-
- ===================================
- On the terminology discussion :
- >> [from sally@projtech.com (Sally Shlaer):]
- >> In OOA, we use "object" to mean a single, typical but *unspecified*
- instance.
- >> So a Car object means a single car, I don't care which one. Most
- OODers
- >> today (I think) would find this a reasonable use of the word object
-
- > [from johnson@m.cs.uiuc.edu (Ralph Johnson) :]
- > If I did, I wouldn't have complained!
-
- Fair enough :-).
-
- >> [from Sally] Now let us switch over into OOD land. As I see it,
- >> the word "class" today combines two ideas:
- >> + the typical unspecified instance (I don't care which instance
- >> you are thinking about, this operation works the same way on any
- instance).
- >> + the collection of all existing instances (as in class data,
- iterators
- >> and similar class methods)
-
- > {from Ralph] I only use "class" for the first concept.
-
- OK. Then I think this is what we are saying: A typical unspecified
- instance is a class to Ralph (who I am taking to be a spokesman for
- OOD), and an object in S-M OOA. OOA does not have a word for the
- collection of existing instances.
-
- Questions: 1. Do you have a word for the collection of existing
- instances?
- 2. What do you mean when you say "object"? Can you answer in terms of
- "how many" and specified vs. unspecified -- or is there another
- dimension that we need to take into account?
- 3. When you said "Like you, I consider a collection of instances to be
- an object", you were using the OOA definition of object -- right?
- 4. When you said "Moreover, I don't consider a class to be an object
- during high level design", were you using the same or a different
- definition? (This one has me somewhat more puzzled.)
-
- ======================================
-
- Now, since you asked earlier, here is the historical truth on why we
- chose the word "object" to mean typical unspecified instance.
-
- In 1986-87, when we were writing the first book on S-M OOA, there were
- three terminologies current: the structured methods terminology, the
- database terminology, and oo terminology.
-
- The structured methods people (the Yourdon courses of that era and Matt
- Flavin's book) used "object" to mean typical unspecified instance. So
- "object" was closely associated with analysis.
-
- The database people were using "entity". These people were somewhat
- more closely associated with design. Based on this, we were slightly
- more inclined to use "object" for the concept in an analysis context.
-
- But as you may remember, the terminology in the OO community was fairly
- variable at this time, particularly when dealing with concepts that were
- not strictly tied to a particular OOPL. We read copious amounts of the
- literature, attended conferences, and talked to everyone we could locate
- to try to find the center of mass -- and "object" or "object-instance"
- seemed to be the center of mass for "typical unspecified instance", as
- well as we could tell.
-
- So the choice was made in 1986-87 taking into consideration all the
- factors we could think of that might be relevant. Now we are now stuck
- with this choice (can you imagine the outrage of our clients if we were
- to try to change it now?? Or the bizarre nature of this discussion if
- we weren't consistent from one book or paper to the next??). So the
- right thing for us to do is to be careful to define our terms and to
- write as clearly and consistently as we can. It is also right for us to
- understand exactly how our terms bother/confuse/irritate others -- hence
- my interest in this discussion.
-
- Sincerely
- Sally Shlaer sally@projtech.com
-