rdthomps@vela.acs.oakland.edu (Robert D. Thompson) (11/06/90)
Ian Kluft amazes us with, >First, they are different mostly because Coad/Yourdon's book is about object- >oriented *analysis*. Booch's book is about object-oriented *design*. OOA is Ok, Let's make sure that anybody who cannot read titles gets this right, "Object-Oriented Analysis" is about object-oriented *analysis* and "Object-Oriented Design" is about object-oriented *design* Whew, glad we got that one out of the way. Both books discuss methods for identifying objects, attributes, and relationships - not really sophisticated. However, Booch provides a more comprehensive look at the fundamental concepts of object-oriented methods (classification, decomposition, etc...) **My own** opinion is that neither book provides a discrete method for deriving class-specs by a step-by-step process. Booch, however, does go a long way to discuss the surrounding nature of such a process. I WOULD APPRECIATE ANY REFERENCES ON METHODS FOR DERIVING CLASS SPECIFICATIONS FROM REQUIREMENTS (IDENTIFICATION). I am currently working on a method. <Robert>
bobatk@microsoft.UUCP (Bob ATKINSON) (11/08/90)
Robert D. Thompson writes: >Coad/Yourdon's book >Booch's book >**My own** opinion is >that neither book provides a discrete method for deriving class-specs >by a step-by-step process. >I WOULD APPRECIATE ANY REFERENCES ON METHODS FOR DERIVING CLASS >SPECIFICATIONS FROM REQUIREMENTS (IDENTIFICATION). "Designing Object Oriented Software", by Wirfs-Brock, Wilkerson, and Weiner (Prentice Hall, 1990) is focussed on this sort of cookbook approach. They give a follow-your-nose way for deriving class specifications given requirements which is basically sound. They don't claim to have discovered the essence of all design problems, but instead present an approach that novices can follow regimentally and learn the exceptions to later. Bob Atkinson
marc@dumbcat.sf.ca.us (Marco S Hyman) (11/08/90)
In article <3697@vela.acs.oakland.edu> rdthomps@vela.acs.oakland.edu (Robert D. Thompson) writes:
**My own** opinion is
that neither book provides a discrete method for deriving class-specs
by a step-by-step process.
One of the questions I hear often, especially from newcomers into things OO,
is "how do I determine what should be an object/class?" This is not far
from rdthomps quest for a step-by-step process to derive class
sepcifications. Perhaps the advice I give to that question is appropriate
here -- read Robert Pirsig's "Zen and the Art of Motorcycle Maintenance,"
especially the chautauquas on QUALITY. Most know a quality design (or class
specification) when they see it. The steps when you don't see it is to
change it. Iteration is probably the key.
This is not the answer most want to hear. It's the best I can give.
// marc
--
// marc@dumbcat.sf.ca.us
// {ames,decwrl,sun}!pacbell!dumbcat!marcrdthomps@vela.acs.oakland.edu (Robert D. Thompson) (11/08/90)
Bob Atkinson writes, >"Designing Object Oriented Software", by Wirfs-Brock, Wilkerson, and Weiner >(Prentice Hall, 1990) is focussed on this sort of cookbook approach. They >give a follow-your-nose way for deriving class specifications given >requirements which is basically sound. They don't claim to have discovered >the essence of all design problems, but instead present an approach that >novices can follow regimentally and learn the exceptions to later. > > Bob Atkinson Thanks for the response Bob. However, I must point out that I (and I assume some others) are not looking for a "sort of cookbook approach" nor are we "Novices" in need of being lead "by-the-nose" I am developing a methodology for class-specification and documentation. I am looking for other insights. I am attempting to establish some consistency for my own group development (and others - if things go well). Anyway, I just thought I would make that clear to you. <Robert Thompson>