|
-
July 14th, 1999, 05:02 AM
#1
Meyers' opinion on Use Cases
Mention the name Bertrand Meyers and someone usually rises to the occasion but I don't want to start off some thread about him personally or Eiffel for that matter. But..last night I was reading the small section on Use Cases in BM's OOSC2 book and wondered what other (more experienced?) engineers made of his comments. If I remember correctly his basic points were:-
1. UC's focus on the ordering of functions rather than the behaviour of an abstraction,
2. The user perspective is not always the best way to capture the systems requirements, and
3. UC's focus on the functions not the data in the system.
4. Working with UC's can be tantamount to doing top-down functional decomposition.
I can appreciate something about what he says although my first reaction was that point [3] slightly contradicts point [1] as I thought that the general OO position seems to put more emphasis now on the behaviour-oriented rather than the simple data-oriented view of abstraction.
Does anyone else want to offer an opinion?
Neil
// Just keep banging the rocks together..
-
July 14th, 1999, 08:34 AM
#2
Re: Meyers' opinion on Use Cases
For some time before using UML I've been using DFDs for analysis which I modified so that with numbering processes and data flows with majors and minors I was able to give some additional information about the sequence of events and by using Gane/Sarson location field at the bottom of the process I was able to tip 'actors' (in UML terminology) involved...
In my latest three projects I used UC. With only one app being business app (other two being a network monitoring tool, and a replication system) use cases were sufficient.
Several days ago while surfing the web I surfed on to this page:
http://home.earthlink.net/%Emadyoshi/igs/igs.html
These guys used an interesting approach. Simpy using DFD and UML together. After all the objective of the analysis is to give a clearer picture and our job is about meeting the objectives using whatever tools necessary, right? :-)
Hm. And a thought. If using UC technique and expanding each high level Use Case to many low level Use Cases - this is legal, keeping actors the same so the model is consistent, covering all branches in typical course of events and describing data exchanged during each interaction of actor-use case or use case-use case you get - well - the extended DFD...
---------------------------------------------
Tomaz Stih, B.Sc.CS [email protected]
Ob sotoccju 10 Nameco Group
SI-1000 Ljubljana http://www.nameco.com
Europe
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
Click Here to Expand Forum to Full Width
|