Lecture of 4 January 1994

• Hand out syllabus
• Texts
  • Mandatory: Object Oriented Analysis and Design, 2nd Ed., Grady Booch, Benjamin Cummings, 1994
  • Recommended: The C++ Programming Language, 2nd. Ed., Bjarne Stroustrup, Addison Wesley, 1991
  • Free: Dylan: an object-oriented dynamic language, Apple Computer Eastern Research and Technology, 1992
• Other Materials
  • Usenet newsgroup comp.object
  • The comp.object Object-Orientation FAQ
• Programming Languages and Tools
  • C++ Programming Language
  • Dylan Programming Language
  • (Smalltalk Programming Language)
  • World-Wide-Web (WWW) and Mosaic hypermedia browser
• Grading criteria -- see syllabus
• Project style:
  • Project groups (5 members each) will be determined before the first project is started
  • Several projects will be assigned to each group
  • All products of analysis, design, and programming will be made available online to the instructor via the WWW.
  • Grades will be based upon the entire project including all elements of analysis and design, not just on working code
• Workload: (less that CDA 3101, more than COP 3530)
• Reading Assignment: Booch ppg. 1-40
• Practice vs. Science (techne)
  • Art of Computer Programming, D. Knuth
  • Science of Computer Programming, D. Gries
• To some of you, many of the concepts and techniques presented in the class will seem familiar. To others, they will be new. The class does not espouse the one true way of programming. It is dangerous to hold out too much belief in any method or activity. The goal of the class is to provide you with a set of tools for carrying out object-oriented design of software and implementation using object-oriented programming techniques.

Inherent Complexity of Software

• Industrial-Strength Software
  • Large, complex applications
  • "Rich set of behaviors"
  • Long lifetime with many users who come and go
  • Difficult for a single programmer to comprehend all the subtleties of design
  • "The world is only sparsely populated with geniuses. There is no reason to believe that the software engineering community has an inordinately large proportion of them." (Peters)
• Sources of Complexity of Software
  • Complex Problem Domain
    • Software systems are easily mutable and universally applicable to any domain requiring information processing or control. These include fantastically complex systems
    • Because of the mutability of these systems, customers tend to change their requirements on a regular basis. It is important to distinguish between software maintenance and evolution.
  • Difficulty of Managing the Development Process
    • Increasing sophistication of users leads to increasing demands of software products. This requires developers to implement large programs.
    • Demands for such large programs have led to the need on the part of software developers to reuse code effectively in order to avoid extra work.

      "Don't work, shirk. Don't create, imitate. Follow the path of least resistance." (Unknown-- Seen on a VDT at FSU, ca. 1975.)

    • The increasing size of applications requires the use of programming teams, and team development poses unique challenges.
  • Flexibility Offered by Software
    • There are innumerable components that can be crafted in software. The lead time for production of most of the basic components is relatively short (hours, days).
    • Physical manufacture of products has a long lead time (weeks, months).
    • Due to this difference (and others) software components are rarely standard, and are recrafted over and over by new programmers.
  • Difficulty of Characterizing Behavior of Systems
    • It is difficult to characterize and constrain the behavior of state machines when the number of potential states becomes high.
    • One must think of a computer as a Turing Machine (even though each computer is in fact a finite state machine).
    • Computation theory (Rice's Theorem in particular) tells us that any non-trivial property of a Turing Machine is undecidable. Thus, we cannot resort to any automated means of determining that our programs are correct. This is where the need for art in computer programming arises.