home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!dtix!darwin.sura.net!uvaarpa!gossip.urich.edu!urvax.urich.edu!MARTINHER
- From: martinher@urvax.urich.edu
- Newsgroups: comp.object
- Subject: Re: Summary on OO training
- Message-ID: <1993Jan1.224012.14594@gossip.urich.edu>
- Date: 1 Jan 93 22:37:50 GMT
- References: <Bzo45H.3pG@unx.sas.com>,<1992Dec23.203113.9755@actrix.gen.nz>
- Sender: news@gossip.urich.edu (USENET News System Admin)
- Reply-To: martinher@urvax.urich.edu
- Organization: University of Richmond
- Lines: 73
-
- I just spent a week each with an instructor from the GE Advanced
- Concepts Center and Desmond D'Souza of Icon Computing, Inc. They
- both teach OMT and the courses are excellent. I have no other
- connection with either organization except for the fact that I
- recommended that our organization adopt OMT (Rumbaugh, et al)
- as our analysis and design methodology.
-
- I hesitate to tell you how good Desmond really is, as I hope to
- learn much more from him in the future and if everyone else
- finds out how wonderful he is then he will become rich and
- famous and he may not have time for me. But, because he deserves
- credit for his skill and the contribution that he may to our
- organization I will take that chance.
-
- Desmond D'Souza is the founder of Icon Computing, Inc. of Austin,
- TX. They specialize in object-oriented technology and education.
- He writes a regular article for JOOP which I find invaluable and
- was one of the main reasons that I sugsted bringing him in to
- teach our staff. We first met here on the Net and then conversed
- by Email and phone last month. This week (Dec 28-31) Desmond
- himself taught us Object-Oriented Design Idioms and Architectures.
- His design concepts are based upon the analysis, modeling, and
- design discussed in "Object-Oriented Modeling and Design" by James
- Rumbaugh, et al. (OMT). He has suplemented this with the best
- work of Booch and other respected experts in OO Design, and his
- own outstanding contributions to our field. This he retains
- the excellent benefits of OMT and almost completely eliminates
- any of the correct criticisms of OMT's weak integration of the
- three models (Object, Functional, and Dynamic) and the weakness
- of the OMT design treatment. He both provides the missing
- design methods and explains how to explicitly relate and strengthen
- the three models by reference to each other. This is no small
- accomplishment and I applaud this contribution.
-
- He did all this while maintaining the interest of a diverse
- audience with widely disparate skills. He was at all times clear,
- concise, entertaining, and informative. He has a rich sense of
- humor and does not take himself too seriously. We had everyone
- from supervisors with little OO experience and mainframe Cobol
- wizards with no OO or object language experience to workstation
- C++ technology trailblazers in the classroom this week.
-
- (I myself am not ready to claim OO master status but within our
- organization I am the technology leader on MACs, PC, OO Design
- & Analysis, and I teach C++ internally.)
-
- He was consistently able to challenge me and the other leaders
- as he offered us new insights and much valuable information. At
- the same time he kept the interest and opened up the possibilities
- of objects to our more traditional staff. ALL of their COMMENTS were
- POSITIVE, both in front of Desmond and those I heard or overheard
- in private.
-
- Desmond claims that he has several more instructors of his caliber.
- If this is accurate then he has a very strong organization which
- should be considered by anyone looking for education and mentoring
- in OO concepts.
-
- There, I have gone and done it. Told the truth and maybe cost
- myself something valuable in the long run. But, my better nature
- insists that helping people to find Desmond D'Souza and Icon
- can only benefit all of us. Thanks Desmond.
-
- And please do not take anything I said here as a criticism of OMT
- in a generaly sense. OMT is the best framework that I have yet
- found for object oriented life cycles, the fact that it is not
- perfect is completely expected. The fact that it can be easily
- improved - not by contradiction but by expansion - is to OMT's
- credit and the credit of Jim Rumbaugh and his associates.
-
- Herb Martin "I refuse to misunderstand, simply because I did not ask."
- martinher@urvax.urich.edu
- (804) 274-4325
-