1 cs 501 spring 2005 cs 501: software engineering lecture 9 requirements 3

33
1 CS 501 Spring 2005 CS 501: Software Engineering Lecture 9 Requirements 3

Post on 21-Dec-2015

214 views

Category:

Documents


0 download

TRANSCRIPT

1 CS 501 Spring 2005

CS 501: Software Engineering

Lecture 9

Requirements 3

2 CS 501 Spring 2005

Course Administration

Feasibility Study

Everybody should received comments shortly.

Teaching Assistants have been assigned.

Quiz 1

Uncollected answer books are at the reception at 301 College Avenue.

3 CS 501 Spring 2005

Discussion of Feasibility Study

General

Who will read the feasibility study?What is needed to decide to go ahead?

Scope and understanding of project

Statement of the task Preliminary requirements and technical analysisDeliverables and business considerations

Confidence

Outline plan. Does it inspire confidence?Visibility plan.  How will progress be reported?  Risk analysis. How are the risks to be minimized?

4 CS 501 Spring 2005

Quiz 1 (c)

A U.S. government agency, which manages huge numbers of documents, has received funds to build a large-scale computer system.

The agency proposes to use a waterfall model of software development, but an independent study recommends a phased development using iterative refinement in the first phase and a waterfall model in the second.

[On the next three slides, a short answer is given in Roman type. Material in italic extends the short answers.]

5 CS 501 Spring 2005

Quiz 1 (c)

i Before this system is implemented some difficult technical choices have to be made. How much should the choice of technology be considered during the feasibility study?

During the feasibility study, it is necessary to know that there is at least one technical approach that would allow the system to be built. It should be studied in sufficient detail to allow estimates of costs and time table.

If there appear to be several very different technical approaches possible, more than one might be analyzed in sufficient detail to understand the implications for the costs and timetable.

In this project, the feasibility study should also plan the process by which the technical decisions will be made.

6 CS 501 Spring 2005

Quiz 1 (c)

ii These technical choices are one of the reasons that iterative refinement is recommended for the first phase of development. Why?

The first iteration will use the team's initial technical knowledge, but will likely show weaknesses in the approach chosen. Subsequent iterations will replace or improve aspects of the technical choices until a version is established that meets the requirements of the project.

A waterfall model is unsuitable because of the large-scale changes that may need to be made. Because of uncertainty, it is difficulty to judge the cost and timetable of the project until this stage is complete.

7 CS 501 Spring 2005

Quiz 1 (c)

iii What is the advantage of using the waterfall model in the second phase?

The waterfall model with feedback is an effective way to build a system that is well understood, so that few changes are likely to be made in either the requirements or the design.

At its best, a waterfall process results in each stage being done once and being done well. For this system, the iterations made during the first phase should have created a high level of understanding and agreement about the entire system.

8 CS 501 Spring 2005

Software Development Principles and Practices

Software development

Software development is a craft. Software developers have a variety of tools that can be applied in different situations.

Part of the art of software development is to select the appropriate tool for a given implementation

9 CS 501 Spring 2005

Modeling Techniques for Requirements Analysis and Definition

The craft of requirements analysis and specification is to select the appropriate tool for the particular task.

• A variety of tools and techniques.

• Many are familiar from other courses.

• There is no correct technique that fits all situations.

As you build understanding of the requirements through scenarios and use cases, use modeling tools to specify requirements. The models provide a bridge between the client's understanding and the developers.

10 CS 501 Spring 2005

Decision Table: Bad Example

University Admission Decision

Note that the rules are too vague.

High SAT T F F F F F

High grades - T F F F F

Sports - - T T F F

Recommendations - - T F T F

Reject X X X

Accept X X X

11 CS 501 Spring 2005

Decision Table: Good Example

University Admission Decision

Note that the rules are now specific and testable.

SAT > S T F F F F F

GPA > G - T F F F F

Athletics code = - - T T F F

Recommendations > A- - - T F T F

Send rejection letter X X X

Send acceptance letter X X X

12 CS 501 Spring 2005

Data-Flow Models

External entities

Processing steps

Data stores or sources

Data flows

An informal modeling technique to show the flow of data through a system.

13 CS 501 Spring 2005

Data-Flow Example: University Admissions

Applicant

Applicationform Receive

application

Completedapplication

Evaluate

Rejection

Acceptance

Shows the flow, but where is the data stored?

14 CS 501 Spring 2005

Data-Flow Example: Assemble Application Stage

Applicant

Applicationform

Receive

Completedapplication

Supportinginformation

Pendingdatabase

Acknowledgment

Initiateevaluation

Applicantdatabase

Evaluationrequest

AND

AND

Acknowledgment

Does this model cover all applications? Are there special cases?

15 CS 501 Spring 2005

Data-Flow Example:Process Completed Application Stage

Rejection

Evaluation

Applicantdatabase

Evaluationrequest Acceptance Financial

aid

Offer

Specialrequest

The data-flow diagram will need specification of the decision-making process.

16 CS 501 Spring 2005

Procedural Models: Flowchart

Operation

Decision

Manual operation

Report

An informal modeling technique to show the decisions and paths that data takes through a system.

17 CS 501 Spring 2005

Flowchart: University Admissions

Form received New?

Database record

T

Notify student

F Update database

Complete?

Notify student

T

FEvaluate

18 CS 501 Spring 2005

Procedural Models: Pseudo-code

Example: Check project plan

check_plan (report)

if report (date_time) > due_date_time then error (too_late) if report (client) = none then error (no_client) if report (team) < min_team or > max_team then error (bad_team) if error() = none then comments = read_report (report) return (comments (text), comments (grade)) else return error()

An informal modeling technique to show the logic behind part of a system.

19 CS 501 Spring 2005

Entity-Relation Model

A requirements and design methodology for relational databases

• A database of entities and relations

• Tools for displaying and manipulating entity-relation diagrams

• Tools for manipulating the database (e.g., as input to database design)

Warning: There is much confusion about definitions and notation

20 CS 501 Spring 2005

Entity-Relation Diagram

An entity

A relation between entities

An entity or relation attribute

An inheritance relation

21 CS 501 Spring 2005

Example: CS 501 Project

Student

CS501 Student

Major

Project

5 to 7

1

Member of

Client team member

Client1:n

Tech contact

0:n0:1

0:1

22 CS 501 Spring 2005

Example: Schema for Web Data

CS 501 Spring 2005

23 CS 501 Spring 2005

MARC Format for Books

001 Control number245 Short title260 Publisher650 Subject heading650 Subject heading700 Author

The MARC format defines a set of tagged fields that are used to catalog books in libraries. Here are some examples.

Suppose you wish to use a relational database to store MARC records.

24 CS 501 Spring 2005

Part of an Entity-Relation Diagram for MARC

Book

Short title

Catalog record

Describes

Control numb

Subject heading

Is about

CreatorEditor of

Author of

1:n

1

0:n

0:n

0:n

0:n

0:n

1

25 CS 501 Spring 2005

Data Dictionaries

A data dictionary is a list of names used by the system

• Brief definition (e.g., what is "date")

• What is it (e.g., number, relation)

• Where is it used (e.g., source, used by, etc.)

• May be combined with a glossary

As the system is implemented, the data dictionary in the requirements is input to the system data dictionary, which is a formal part of the system specification.

26 CS 501 Spring 2005

Petri Nets

A Petri Net models parallelism

A

S1

Sm

S

S

A

Event 1

Event n

Event

A

Event 1

Event n

..

f(A; E) S

f(A; E1,..,En) S

f(A; E1,..,En) S1,..,Sm

27 CS 501 Spring 2005

Transition Diagrams

A system is modeled as a set of states, Si

A transition is a change from one state to another.

The occurrence of a condition, Ci, causes the transition from one state to another

Transition function:

f (Si, Cj) = Sk

Example S1 S2

S30

0

0

1

11

28 CS 501 Spring 2005

Therapy Control Consol: Scenario

"The set up is carried out before the patient is made ready. The operator selects the patient information from a database. This provides a list of radiation fields that are approved for this patient. The operator selects the first field. This completes the set up.

The patient is now made ready. The lock is taken off the machine and the doses with this field are applied. The operator then returns to the field selection and choose another field."

29 CS 501 Spring 2005

Therapy Control Consol: Finite State Machine

Example: Radiation Therapy Control Console

You are developing requirements for the operator's control console. In a client interview, the client describes the dangers inherent in using the machine.

The operator must follow a strict procedure before the machine is ready to run.

You use a finite state machine to specify the procedures. This shows the client that you understand the requirements and specifies the procedures for the developers.

30 CS 501 Spring 2005

State Transition Diagram

Patients Fields Setup ReadyBeam

on

Enter Enter Start

Stop

Select field

Select patient(lock on)

(lock off)

Discuss each state and transition with the client.

31 CS 501 Spring 2005

Therapy Control Consol: State Transition Table

SelectPatient

SelectField

Enter lock off Start Stop lock on

Patients

Fields

Setup

Ready

Beamon

Fields

Fields

Fields

Patients

Patients

Patients

Setup

Setup

Setup

Ready

Beamon

Ready

32 CS 501 Spring 2005

Prototyping Requirements

Rapid prototyping is the most comprehensive of all modeling methods

A method for specifying requirements by building a system that demonstrates the functionality of key parts of the required system

Particularly valuable for user interfaces

33 CS 501 Spring 2005

A Note on Object Models

This course teaches object models as a tool for design.

Some people recommend object models for requirements definition, but it is difficult to use them without constraining the system design.

Flow charts and finite state machines are supported by UML as design models, but can also be used for requirements.

*