Tutorial Exercises
Questions from Week 1
 | From your own practical experience, describe the disadvantages of the conventional
approach to systems development. |
| From the point of view of management, what are the major drawbacks of the conventional
approach? |
| What major benefits can be attributed to the N.C.C. standards approach to systems
development? |
| In your opinion, are the criticisms given above fair? Explain. |
| To what extent are the criticisms of this method dependent upon its being a manual
documentation approach and, hence, would be ameliorated in a machine implementation? |
To help you answer these questions you should look over the notes you took in the first
lecture (25/01/1999). Also, you might light to look over my own potted history of system
design.
Back to top
Questions from Week 2
| What do you understand by the term top-down design? |
| Why does Jackson think the practice of top-down design is 'psychologically unrealistic'? |
| In systems design problems traditional top-down methods impose a hierarchical structure
on the environment in question. What is the major drawback of this approach? |
| How does top-down enforce the 'riskiest ordering of decisions'? |
Back to top
Questions from Week 3
| Why might we consider the software development process to be a feedback process? |
| Should we use the functional requirements as the focus for system design? |
| What role do assumptions play in system design? |
Back to top
Questions from Week 4
| Why is it claimed that JSD facilitates system maintenance? If the claim is true, what is
its significance? |
| How would you explain the statement 'a JSD model is a realisation of an abstract
description of the real world'? |
| Use simple (???) payroll system to show how the validity of the system's outputs is
crucially affected by the quality of the model or simulation of the real world. |
| If the system developer and the user have different understandings of the computer
system's behaviour, any information produced by the system is suspect. Is this statement
correct? Does JSD do anything to help overcome this problem? |
Using information from the Stitch in Time case
study:
| Draw up a candidate list of events |
| Draw up a candidate list of entities |
What entities will you reject from your model, and why?
Back to top
Questions from Week 5
Exercise 1 - Actions & Entities
From the following brief description, separate out phrases
which indicate the major nouns (to identify entities) and the
verbs (to identify actions). From these:
- Produce an action list to the entities you have
discovered;
- Allocate the actions to the entities you have discovered;
- Produce entity (process) structure diagrams for the
entities;
- Produce the equivalent structure text.
A certain machine is used to either make or finish
components. The machine is set up at the start of each day by an
operator. When the machine is used for making new components it
is loaded with raw materials; when used for finished components
that have already been made it must obviously be loaded with the
unfinished components. A product must be ordered before it is
made and it is put into a warehouse after it has been finished
and inspected before being delivered. If it is found faulty at
inspection, it may be re-finished or rejected. If re-finished,
it is assumed to be fit for delivery.
Exercise 2 - Backtracking
The life history of the entity HND-student contains the
following possible events:
- applies for a course;
- enrols for the course;
- progresses to the 2nd year;
- turns down an offer of a place;
- withdraws during the 1st year;
- withdraws during the 2nd year;
- passes exams into 2nd year;
- fails exams in 2nd year;
- accepted for course;
- rejected for course.
| Draw a process structure diagram for the student entity
without using backtracking. |
| Produce the equivalent structure text for the student
entity. |
| Redraw the diagram using backtracking. |
| Produce the equivalent structure text for the backtracking
version. |
Exercise 3 - Network model
Competitors for a motoring speed trial start from a single
grid position at different times and complete four laps which
vary in length as different parts of the circuit are brought
into use. As each car passes the checkpoint/finish a radio
signal relays car identification and time of day. Marshals
signal the starting time as each competitor sets off. Points are
scored as a function of time taken. On the assumption that each
record is sent to a single queue and that all competitors who
start also finish, produce an initial network model for a system
which will calculate each competitors final score for public
display.
Back to top
Questions from Week 6
Read through the Poly
Fleet Hire case study (Acrobat 3.01 format), then:
| Amend the specification so that a report can be generated which provides a list of all
those completed car hirings that lasted for more than eight days in the last nine months.
The report process is to be run at the manager's discretion (i.e., it is not a scheduled
process). |
| Make sure you draw/amend any structure diagrams & structure texts as necessary. |
Back to top
|
