In this article, you are going to look at the code you write from a very high level. Here we discuss how you decide to write your code before you begin writing it. This is probably the single most important decision you will make because what you decide will become the foundation upon which all future upgrades and additions to your application will be built upon.
There are two general approaches you can take when developing applications. Within each approach, you can establish your own specific methods for dealing with programming issues. I refer to these two approaches simply as the Page-centric and Program-centric architectures.
Page-centric architecture is probably the most common methodology for developing dynamic Web applications. This architecture focuses primarily on creating JSP templates. These templates are generally divided up by a logical directory structure, with multiple JSP pages in each directory. Each JSP page performs a specific functionviewing database information, processing email, storing shopping cart information, and so on. Figure 1 shows how this model works.
This is a fairly common approach, and is the one used most by ColdFusion, PHP and Active Server Pages (ASP) developers.
There are a number of advantages to this approach:
SimplicityCreating an application within this environment can be done with ease. A basic understanding of JSP is all that is needed to add or edit the application.
HomogeneityEverything in this application will be a JSP page of some kind. There may be some instances where a JavaBeans component or custom action is utilized, but the majority of changes can be made by directly editing a JSP template.
Time-to-MarketJSP-only applications can generally be built faster in the beginning because there are fewer elements involved.
AccessibilityBecause JSP is fairly easy to develop, especially with tools such as UltraDev and JRun Studio, even designers can participate in dynamic development. This means that it is not always necessary to have a Java programmer or team of programmers to make functional additions or changes to an application.
The overall compelling reason for using this approach is that it is fairly simple, and doesn't require a deep knowledge of programming or a team of developers. While this is a truly viable coding approach for many projects, this design has somewhat of a bad reputation among hardcore Java programmers due to its weaknesses.
Using the Page-centric approach is also more intuitive to understand from a browser's perspective. Each page has a function, and there is a clear connection between a page and the type of information it contains or the functionality it will deliver. For example, a page called EnterMovie.jsp will serve the function of entering new moviesthis is easy to understand.
However, taking this approach has its disadvantages. The main problem is that as a Page-centric application evolves over time, the code can become convoluted as different programmers make their logic changes directly in the JSP page. These pages can become overwhelmed with scriptlets as core business logic becomes embedded in every page. Ultimately, a Page-centric application must be managed carefully, or it can become a behemoth requiring a complete overhaul to add any new functionality.
Figure 1 illustrates graphically how the Page-centric structure works.
Figure 1 The Page-centric approach is easy to use and intuitive for the programmer.