PROGRAM DEVELOPMENT.

Stages involved in the program development cycle.

The process of program development can be broken down into the following stages:
1. Problem recognition (Identification of the problem).
2. Problem definition.
3. Program design.
4. Program coding.
5. Program testing & debugging.
6. Program Implementation and maintenance.
7. Program documentation.

Problem recognition.
Problem recognition refers to the understanding and interpretation of a particular problem.
The programmer must know what problem he/she is trying to solve. He/she must also understand
clearly the nature of the problem & the function of the program.
In order to understand a problem, look for the keywords such as compute, evaluate, compare, etc.
Usually, a programmer identifies problems in the environment and tries to solve them by writing
a computer program.
There are 3 situations that cause the programmer to identify a problem that is worth solving:
1. Problems or undesirable situations that prevent an individual or organizations from
achieving their purpose.
2. Opportunity to improve the current program.
3. A new directive given by the management requiring a change in the current system.
Sample problem: Develop a program that can be used to calculate/find the area of a circle. Use
the equation A = π * r2.

Problem definition (Problem Analysis).

In Problem definition, the programmer tries to define (determine) the:
(i). Output expected from the program.
(ii). Inputs needed to generate the output information.
(iii). Processing activities (requirements), and
(iv). Kind of files which may be needed.
 The programmer should write a narrative on what the program will do, and how it is meant to
achieve the intended purpose. Within this narrative, he/she is required to determine what data
is to be input & what information is to be output.
For example:
In calculating the area of any circle, the parameters needed to determine the area of any circle
are:
1. Input: (a) Pie (π) which is a constant.
(b) The radius of the circle.
2. Process: The formula for calculating area of a circle, which is π * radius * radius.
3. Output: The area of the circle (A).
At the end of the problem definition, the programmer is required to write a requirements
report/document for the new program. This document will enable the programmer to come up
with a program design that meets the needs at hand.

Note. Problem definition should be done thoroughly to ensure user satisfaction, and to facilitate
the subsequent stages in the program development cycle. A failure at this stage usually results in
a system that will not work as intended, or that may not work at all.

Program design
Program design is the actual development of the program’s process or problem solving logic
called the Algorithm.
It involves identifying the processing tasks required to be carried out in order to solve the
problem.
The design stage enables the programmer to come up with a model of the expected program (or a
general framework (outline) of how to solve the problem, and where possible, break it into a
sequence of small & simple steps.
The models show the flow of events throughout the entire program from the time data is input to
the time the program gives out the expected information.
√ The processing tasks must be in order & systematic. Therefore, the programmer
identifies the processing tasks required, and the exact order in which they are to be carried out.
√ The design process does not take account of the programming language to be used in the
final product, since it only defines program logic.
√ Program design provides for easy maintenance.
Note. It is important to design programs before entering them into the computer. The
programmer should only attempt to covert a design into a program code after ensuring that it is
logically correct. If possible, check the logical order on the desk.
Some programmers produce rough & ready solutions at a Keyboard, and continue to amend the
programs until eventually the program appears to do what was expected. This is not
recommended in programming because of the following reasons:
1. The final code may not be easy to follow, since it was just cobbled together.
2. Variable names & specific items of code may not be documented.
3. Programs produced by continuous amendments & changing of codes mostly lead to
unforeseen side effects.
E.g., there may not have been plan for testing the program or procedures, hence, the program
may easily fail.
4. A programmer may be asked to modify the code at a later date. Without sufficient
documentation, the programmer will be forced to trace through the program in order to gain an
insight into how the program functions.

Modular programming
Many programs are non-monolithic (i.e., they are not usually made up of one large block of
code). Instead, they are made up of several units called modules, that work together to form the
whole program with each module performing a specific task.
This approach makes a program flexible, easier to read, and carry out error correction.

Program coding
Program coding is the actual process of converting a design model into its equivalent program.
Coding requires the programmer to convert the design specification (algorithm) into actual
computer instructions using a particular programming language.
For example;
The programmer may be required to write the program code either in Pascal, C++, Visual Basic
or Java, and develop (invent) suitable identifiers, variable names, & their data types. However,
remember that, at this stage the coding is still a Pencil & paper exercise.
The end result of this stage is a source program that can be translated into machine readable form
for the computer to execute and solve the target problem.
Rules followed in coding a program.
1. Use the standard identifiers or reserved words.
2. Make the program more readable by using meaningful identifiers.
3. Don’t use similar variables.
4. Keep spellings as normal as possible.
5. Use comments to explain variables & procedures. This makes the program readable.
6. Avoid tricks – write the program using straightforward codes that people can readily
understand.
7. Modularize your program.

Program Testing and Debugging

After designing & coding, the program has to be tested to verify that it is correct, and any errors
detected removed (debugged).

TESTING:
Testing is the process of running computer software to detect/find any errors (or bugs) in the
program that might have gone unnoticed.
During program testing, the following details should be checked;
 The reports generated by the system.
 The files maintained in connection to the system’s information requirements.
 The input to the system.
 The processing tasks.
 The controls incorporated within the system.
Note. The testing process is a continuous process, and it ends only when the Programmer & the
other personnel involved are satisfied that when operational, the program will meet the objectives
and the growing demands of the organization.

Types of program errors

There are 5 main types of errors that can be encountered when testing a program. These are:
1. Syntax errors.
2. Run-time (Execution) errors.
3. Logical (arithmetic) errors.
4. Semantic errors.
5. Lexicon errors.
Syntax errors
Every programming language has a well-defined set of rules concerning formal spellings,
punctuations, naming of variables, etc. The instructions are accepted only in a specified form &
and must be obeyed by the programmer.
Syntax errors are therefore, programming errors/mistakes that occur if the grammatical rules of a
particular language are not used correctly.
Examples:
(i). Punctuation mistakes, i.e., if the programmer does not use the right punctuations & spaces
needed by the translator program, e.g., omitting a comma or a semicolon.
(ii). Improper naming of variables.
(iii). Wrong spellings of user defined and reserved words.
Reserved words are those words that have a special meaning to the programming language,
and should not be used by the programmer for anything else.
Syntax errors are committed by the programmer when developing, or transcribing the program,
and can be detected by the language translators, such as the Compiler as it attempts to translate a
program. Such errors must be corrected by the programmer before the program runs.

Logical (arithmetic) errors.

These are errors in the program logic.
Logical errors relate to the logic of processing followed in the program to get the desired results.
E.g., they may occur as a result of misuse of logical operators.
Logical errors cannot be detected by the translator. The programmer will detect them when the
program results are produced.
The program will run, but give the wrong output or stop during execution.

Run-time (Execution) errors.

These errors occur during program execution.
Run-time (execution) errors occur when the programmer introduces new features in the program,
which are not part of the translator’s standards.
For example; they may occur if:
(i). The computer is asked to divide a number by zero.
(ii). The number generated as a result of an instruction is too large to fit in a memory location.
(iii). When you raise a number to a very big power that cannot be accommodated in the
Register’s structure of the computer.
(iv). In case of a closed loop in the program, leading to a set of instructions being executed
repetitively for a long time.
Execution errors are not detected by the translator programs, but are detected by the computer
during execution. Sometimes, execution errors may lead to premature end of a program.
To detect and eliminate Execution errors, a test run should be performed on the program after it
has been translated.

Semantic errors.
These are meaning errors. They occur when the programmer develops statements, which are not
projecting towards the desired goal. Such statements will create deviations from the desired
objectives.
Semantic errors are not detected by the computer. The programmer detects them when the
program results are produced.
Example;
a). IF GP>=1500 OR 2200 THEN
TAX: = GP - (GP * 13%)
b). IF GP>=1500 AND GP<= 2200 THEN
TAX: = GP - (GP * 13%)
In the 1st statement, if the selection is between 1500 & 2200, the computer will pick only 1500 &
2200, and the other values will not be touched.
In the 2nd statement, the computer will be able to pick all the values between 1500 & 2200
because of the ‘AND’ operator.
Lexicon errors.
These are the errors, which occur as a result of misusing Reserved words (words reserved for a
particular language).

DEBUGGING:
The term Bug is used to refer to an error in a computer program.
Most programming errors often remain undetected until an attempt is made to translate a
program.
The most common errors include:-
 Improperly declared Constants and Variables.
 A reference to undeclared variable.
 Incorrect punctuation.
Debugging is therefore, the process of detecting, locating & correcting (removing, eliminating)
all errors (mistakes or bugs) that may exist in a computer program.

TYPES OF TESTING (Methods of error detection)

For the program to be assumed as correct, several testing needs to be conducted by the
programmer to ascertain/establish their validity.
There are several methods of testing a program for errors. These include:
1. Dry running (Desk checking).
2. Translator system checking.
3. Functional testing.
4. Use of Test data.
5. Use of debugging utilities.
6. Diagnostic procedures.
7. System test with actual data.

Dry Running (Desk checking):

Dry running is a method of checking a program for errors by making the corrections on a paper
before entering it in the program editor.
It involves going through the program while still on paper verifying & validating its possible
results. If the final results agree with the original test data used, the programmer can then type
the program into the computer and translate it.
√ Dry running helps the programmer to identify the program instructions, detect the most
obvious syntax and logical errors, & the possible output.
√ Dry running is much faster. This is because; it involves the use of human brain as the
processor, which has got a well inbuilt common sense.
Translator system checking:
This is a type of testing, which involves the computer & the translator programs.
After entering the program, it is checked using a translator to detect any syntax errors. The
translator can be a Compiler or an Interpreter, which goes through the set of instructions &
produces a list of errors, or a program/statement listing which is free from errors.
Functional testing (White-box testing):
This type of testing is based upon examining the internal structure of a program & selecting test
data, which give rise to the alternative cases of control flow.
Use of Test data.
The accuracy of a program can be tested by inputting a set of values referred to as Test data.
The test data is designed to produce predictable output.
There are 2 types of test data;
(a). Real data (live data): - test data obtained from the real problem environment (practical
applications).
(b). Dummy data: - assumed test data.
The programmer invents simple test data, which he/she uses to carry out trial runs of the new
program. At each run, the programmer enters various data variations including data with errors
to test how the system will behave. For example, if the input required is of numeric type, the
programmer may enter alphabetic characters. The programmer will then compare the output
produced with the predicted (actual) output.
Notes.
 Where possible, the program should be tested using the same test data that was used for
desk checking. More strict/rigid tests should be applied on the program in order to test the
program to its limits.
 Only Logical errors & Semantic errors can be corrected by the programmer using test
data.
 A good program should not crash due to incorrect data entry but should inform the user
about the irregularity and request for the correct data to be entered.
Use of debugging utilities.
After the program has been entered in the program editor, debugging utilities which are built in
the computer can be run during translation to detect any syntax errors in the program.
The errors are corrected and the debugging process is repeated again to find out more errors,
before the program is executed.

Diagnostic procedures.
For complex programs, diagnostic procedures, such as Trace routines, may be used to find
logical errors.
A Trace prints out the results at each processing step to enable errors to be detected quickly.

System Test with actual data.

This is whereby the new program is run in parallel with the existing system for a short time so
that results can be compared and adjustments made. In such cases, the system test is made using
actual data.
Implementation and Maintenance.

IMPLEMENTATION
Implementation refers to the actual delivery, installation and putting of the new program into
use.
The program is put into use after it is fully tested, well documented, and after training the staff
who will be involved in the running of the new program.
Structured Walk Through:
It is an organized style of evaluating/reviewing a program by a team of other programmers,
which then reports to the programming team.

REVIEW AND MAINTENANCE.

Once the program becomes operational, it should be maintained throughout its life, i.e., new
routines should be added, obsolete routines removed, & the existing routines adjusted so that the
program may adapt to enhanced functional environments.
The main objective of maintenance is to keep the system functioning at an acceptable level.
Program maintenance mainly involves: -
√ Correcting errors that may be encountered after the program has been implemented or
exposed to extensive use.
√ Changing procedures.
√ Hardware and software maintenance.
√ Changing parameters and algorithms used to develop the original programs.
√ Making any adjustments as new technology comes.
Note. Program maintenance runs parallel to the maintenance of the program documentation, i.e.,
any time maintenance is carried out on the program, the documentation should also be updated to
convey the right image of the system.

Program documentation.
After writing, testing, and debugging a program, it must be documented. In other words, the
programmer should describe all what he was doing during the program development stages.
Program documentation is the writing of supportive materials explaining how the program can
be used by users, installed by operators, or modified by other programmers.
Note. All the program development activities (i.e., from the initial stage up to the complete
program) should be documented/recorded in order to assist in the development of the program,
future modification of the program, general maintenance, machine & software conversion at a
later date, and program changeover.
Documentation can either be; Internal or External.
Internal documentation is the writing of non-executable lines (comments) in the source program
that help other programmers to understand the code statements.
External documentation refers to reference materials such as user manuals printed as booklets.

Types of program documentation.

There are 3 target groups for any type of documentation:
1. User-oriented documentation.
This enables the user to learn how to use the program as quickly as possible, and
with little help from the program developer.

2. Operator-oriented documentation:
This is meant for computer operators such as the technical staff. It is used to help
them install & maintain the program.
3. Programmer-oriented documentation:
This is a detailed documentation written for skilled programmers. It provides the
necessary technical information to help in future modification of the program.

Some documents used in program documentation.

(1).User guide/ manual.
This is a manual provided for an end-user to enable him/her use or operate the program
with minimal or no guidance.
A User guide is used in user-oriented documentation.
(2).Reference guide.
It is used by someone who already knows how to use the program but needs to be
reminded about a particular point or obtain more detailed information about a
particular feature.
(3).Quick Reference guide.
This could be a single sheet or card small enough to fit into a pocket. It is used by
the user to get help for the common tasks carried out within the program.
(4).Technical manuals.
They are intended for System analysts & Programmers. They assist in maintaining
& modifying the program design and code.

Contents in a program document.

Documentation includes:
1. Title of the program.
2. Function of the program.
3. Language used.
4. Hardware & Software required to support the processing of the system.
5. File specifications (details of the data structures used, & details of how data files
are to be organized, accessed, and kept secure).
6. Limitations of the program.
7. Format of the input & the output expected.
8. Design of the program using the design tools (i.e., detailed algorithms &
procedures used).
9. A listing of the Source program and the program flowcharts.
10. A carefully devised set of Test data, and a table of expected results.
11. Detailed instructions on how to run the program.