Create reusable and redistributable components with dojo and Ajax

Brief introduction

Adding functionality to existing applications has proven to be one of the most challenging tasks in software development. In addition to not changing existing code, make sure that the new code meets all the ease of use and performance requirements. In this article, learn how to introduce new features in the late stages of a Web project, and how to seamlessly integrate modifications into existing projects.

The methods described in this article are particularly relevant in the following stages of the software development lifecycle:

Development phase. The customer asks the development team to integrate new features into the product to improve ease of use (this is increasingly common as the project uses an iterative agile development approach.) This requires rapid analysis and integration of new functionality over a short period of time, and does not compromise application robustness and standards.

Maintenance phase. You often need to add and remove features. The development team may also need to access or cancel certain features for a particular customer. Other customization work is often required.

After analyzing this problem, the team should develop a flexible architecture that supports seamless additions and deletions without the need to modify code for product or application core functionality. Ideally, code that is part of the core functionality should not be modified at all, because a lot of time has been invested in the testing and stability processing of this part of the code. The use cases in this article demonstrate how to develop such an application and point out some of the challenges.

Case

A blog application has been developed and put into use. However, the customer has put forward a new request: In the blog system to add mail features. The customer already has a mail server that wants to integrate the two applications in some way. Customers want to:

Each time the user clicks a valid URL, a window is provided to them.

This window contains all the options provided by the mail component.

The mail window can send mail to the mail server.

Challenge

The development team is unwilling to modify the core functionality of the blogging system in order to add mail functionality, because the risk is too high. If the development team is using third-party software to develop core functionality that does not allow them to modify the source code, the situation becomes even more cumbersome. In addition, managers are also opposed to modifying existing presentation and business tiers.

Solution

The development team can choose between two methods. Figure 1 shows a method where core functionality and new functionality are not tightly integrated, and the architecture makes the new functionality loosely coupled with core functionality:

Figure 1. Core features and new features are not tightly integrated

Figure 2 shows a tightly integrated approach to core functionality and new functionality, in which the new functionality becomes an intrinsic part of the core functionality.

Figure 2. Tight integration of core functions and new features