Once the users have accepted the design proposal, the system analyst must develop detailed design specifications or a detailed design. This is a time-consuming part of the project, but it is relatively straightforward.
In this phase, every fact of the system is considered in detail. Here is a list of some procedures for the detailed design of system activities.
Procedure for detail design of a system
- Designing output forms and screens,
- Planning input data forms and procedures,
- Drawing system flowcharts,
- Planning file access methods and record formats,
- Planning database interfaces,
- Planning data communications interfaces,
- Designing system Security controls, and
- Considering human factors.
This list is comprehensive, nor will all activities listed be used for al systems. Some analysts choose to plan overall logic at this stage, preparing program structure charts, pseudo-code, etc.
Normally, in the detail design phase, parts of the system are considered in this order:
- Output requirements
- Input requirements
- Files and databases
- Systems processing
- Systems controls and backup
First of all, analysts must know exactly the client’s demand that is the system’s output. The analyst must also consider the medium of the output-paper, computer screen, microfilm, and so on.
In addition, he/she must determine the type of reports needed (summary, exception, and so on) and the contents of the output what data are needed for the reports. The form of the printed outputs is also a consideration; they may be custom-printed if they go outside the organizations’ to customers or stockholders.
The analyst may wish to determine the report format by using a printer spacing chart, which shows the position of headings, the spacing between columns, and the location of data and page numbers. The analyst may also use screen reports, mock-ups on paper of how the screen will respond to user queries.
Once the desired output is determined, analysts must consider what kind of input is required to produce it. First, the input medium must be considered. Next, the analyst considers content again-what fields are needed, the order in which they come, and so on.
This in turn may involve designing forms that will organize data before it is entered. The analyst needs to plan some kind of input validation process, a check that data are reasonable as well as accurate-he/she would not expect a six-figure salary, for example, and for someone who works in the mailroom.
Finally, it is necessary to consider input volume at peak periods and whether the system can handle it. A mail-order house, for instance, may have to be ready for higher sales of expensive toys in the holiday season than at other times of the year.
Files and Database
It is necessary to consider how the files in a computer system will be organized: sequentially, directly, or by some other method. The analyst also needs to decide how the files should be accessed.
They might be organized as indexed files but be accessed directly or sequentially. Also, it is needed to determine the format of records making up the data files.
If the system has one or more databases. Or accesses databases used in other systems, then the analyst will have to coordinate with the database administrator, the person responsible for controlling and updating databases.
Now it is necessary to show the flow of data in the new system. One method is to use a standard ANSI flowchart to illustrate what will be used. Note that a systems flowchart is not the same as the logic flowchart used in programming.
The systems flowchart describes only the big picture. On the other hand, a logic flowchart represents the flow of logic within a single program.
Systems Controls and Backup
To make sure data are input, processed, and output correctly and to prevent fraud and tampering with the computer system, analysts will need to institute appropriate controls. In a batch system, in which data for the system are processed in groups, begin with the source documents, such as time cards or sales orders.
Each document should be serially numbered so that the system can keep track of it. Documents are time-stamped when received and then grouped in batches. Each batch. Each batch is labeled with the number of documents per batch; these counts are balanced against totals of the processed data.
The input is controlled to make sure data are accurately converted form force documents to the machine-processable form. Data input to online systems is backed up by system journals, files that record every transaction processed at each terminal, such as an account withdrawal through a bank teller. Processing controls include data validation procedures.
It is also important to plan for the backup of system files; copies of transaction and master files should be made on a regular basis. These file copies are stored temporarily in case the originals are inadvertently lost or damaged. Often the backup copies are stored off-site for security.
You may like also: