TO: 06f522 students

To: 06f522 students

From: R. Lechner

Subj: Assignment 1 - due Tues. Sept. 12, and more about 06f522:

RJLRef: $PH/06f522/06f522asgnt1.{doc,htm}

Greetings. This first assignment is to bring you on board re: the first phase of 06f522, which is information modeling. This 2-part assignment (asgnt1a and asgnt1b) is due on Sep 12 in class. Hand-drawn sketches are adequate for this assignment.

The problem requirements are inserted on Figures 9-17 and 9-22 of

$PH/06f522/LarmanApplUMLandPtrns/larman3EdDgmsCh01-14/09_domainModelsR1.ppt

Please let me know if you have problems accessing this URL. [Note that $PH = http://www.cs.uml.edu/~lechner/ in your browser or ~lechner/public_html if you are logged into any uml/cs server.]

(These 2 figures are extracted from the diagram file for Chapter 9 on Larman's website). My R1 revision includes figure captions and this assignment. (Each Figure is duplicated in html and rtf formats. I only changed the first one. If I ever do R2 of this file I will eliminate the duplicate slides.)

More about 06f522:

The subject of information (data and class) models is one of three in the course.

Next will come behavioral models (state machines or StateCharts) and finally Model-Driven Architecture MDA (or MDDevelopment MDD). This will introduce our own IDE (Integrated Development Environment) called COOL (for Collaborative O-O Laboratory).

In part 1, besides the material in Larman's text I hope we can appreciate how data models for analysis have evolved from the design constraints (single- inheritance, invariant, non-overlapping, complete decomposition) of C++/Java. Analysts must deal with multiple dynamic overlapping partial taxonomies as discussed in 'Generalization Semantics' UML-2 Issue v 0.34. (I searched the http://www.omg.org/issues/ tree but failed to find this (31pp) document.)

The following link is to an early (1999) view of UML status and evolution (to v1.4.).

Its slides #26-30 mention generalization semantics:

http://ias.uniklu.ac.at/projects/uml/OTG99/

Haim Kilov was an early advocate of such generalization. See 'Information Modeling' (his book with Ross) and more recent business requirements texts.

PS: Kilov recommended an entertaining short lecture by famous logician Wladyslaw Turski on his long career in computer HW and SW. I saved it at $PH/turski-ItWasFun.ppt. This online link failed:

http://www.mimuw.edu.pl/etaps03/Workshops/It%20was%20fun.ppt

This recent tutorial (>100 slides) is an extensive review of UML2 and MDA:

http://www.omg.org/news/meetings/workshops/MDA_2004_Manual/0-3_Tutorial1_Selic.pdf

We may get to discuss this in relation to COOL in Part 3.

By the way, I usually give one or two hour exams but replace the final exam with a group project.