Introduction

Tom Kelliher, CS23

Feb. 5, 1996

Introduction

Syllabus

1. Contract between students and instructor

2. Keys: Responsibility, Commitment, Discipline

3. Optional C++ book (Lippmann)

4. Lippmann and Oualline on reserve

5. Late assignments accepted only with prior approval

6. Unannounced quizzes to encourage keeping up on reading

7. Use of WWW --- abuse of WWW

8. Perspective on collaboration --- useful aid, poor crutch

9. Working in Unix, Unix Lab, reference books, keys

10. Tentativeness of schedule

What I Expect

1. Demonstration of reasoning, problem solving and decomposition skills

2. Communication, correctness demonstration skills

3. Ability to detect, resolve ambiguity; exercise decision skills

4. ``Fluency'' in object oriented programming, comfort with Unix, grasp of data structures, algorithms:
   1. Class construction, design, mechanics
   2. Invariants, pre-conditions, post-conditions
   3. Induction, proof techniques
   4. Recursion, dynamic memory allocation/deallocation
   5. Unix, Unix tools: g++, gdb, emacs, etc.

Other information

Need to set-up Unix accounts for those who don't have them.

Problem Solving

Problem --- Solution Design --- Implemented Program

Two components to every program:

• Data structures (implies abstract)
• Algorithms: control structures (includes abstract)

The ``old'' paradigm: modular programming

• Top-Down, control flow

The ``new'' paradigm: object oriented programming

• Bottom-Up, message passing, model the problem

Software Life Cycle

Not really a linear process

1. Specification
2. Design: interfaces, modularity
3. Verification: correctness proof, invariants, etc.
4. Coding: relatively unimportant
5. Testing
6. Refinement
7. Production
8. Maintenance

What is a Good Solution?

One which minimizes total cost over all phases of the life cycle

Examples:

1. Calculating a tip
2. Double checking a grocery receipt
3. Automating LaTeX to HTML conversion

Modular Design

Advantages:

1. Models programming teams
2. Isolates programming errors given problem descriptions
3. Localizes ``like'' entitites
4. Lends abstraction, eases understanding
5. Code reusability
6. Eases modification
7. Error reduction through infomation hiding

Disadvantages:

1. Requires discipline, with emphasis on design
2. More work required on small, one-shot projects

What are objects?

Are they related to abstract data types?

What is an ADT?