What Is Rapid Application Development?
Just like Agile, Rapid Application Development (RAD) grew out of a response to Waterfall methods, and a desire to build better software faster.
Join the DZone community and get the full member experience.
Join For FreeConceived in the 1980s, rapid application development, or RAD, was the first development methodology to challenge traditional Waterfall development practices. Though often mistaken for a specific model, rapid application development is the idea that we benefit by treating our software projects like clay, rather than steel.
Software is a unique engineering structure because it is transient. With traditional engineering projects like bridge construction, engineers cannot begin to build a bridge and then change their minds halfway through the process—that’s pure chaos. But a bridge built in software? Engineers can change that every day. RAD takes advantage of this by emphasizing rapid prototyping over costly planning.
Brief History of Rapid Application Development
Three quarters into the previous century, humanity began to want software. Fulfilling this desire required developers. These software projects consumed months of laborious planning and even more in development—just like traditional engineering projects. Software architects worked with end-users to draft functional requirements, then spent countless hours defining them in spec sheets.
With specifications prepared, development began. Anywhere from months to years later, users got their first glimpse of the product they themselves requested. And if it failed to meet their expectations, the engineers would refactor—the costs of which were extraordinary.
This process, which began with the blackboard, moved to the spec sheet, then to software, and terminated at the user, is known colloquially as the “Waterfall” approach. It mirrored traditional engineering assignments that worked with immutable materials like wood, cement, and iron — once set and paid for, these resources were costly to alter.
In the 1980s, Barry Boehm, James Martin, and others recognized the obvious: software was not a raw mineral resource. They saw software for what it was: infinitely malleable. Boehm and Martin took advantage of software’s inherent pliability when designing their development models: the Spiral Model and the James Martin RAD model, respectively. Since then, RAD has evolved to take on other forms and acted as a precursor to Agile.
Rapid Application Development Methodology
Though exact practices and tools vary between specific RAD methodologies, their underlying phases remain the same:
1. Define Requirements
Rather than requiring that you spend months developing specifications with users, RAD begins by defining a loose set of requirements. We say loose because among the key principles of rapid application development is the permission to change requirements at any point in the cycle.
Instead of committing to hard and fast specifications, developers gather the product’s “gist.” The client provides their vision for the product and comes to an agreement with developers on the requirements that satisfy that vision.
2. Prototype
In this rapid application development phase, the developer’s goal is to build something that they can demonstrate to the client. This can be a prototype that satisfies all or only a portion of requirements (as in early stage prototyping).
This prototype may cut corners to reach a working state, and that’s acceptable. Most RAD approaches have a finalization stage at which developers pay down technical debts accrued by early prototypes.
3. Absorb Feedback
With a recent prototype prepared, RAD developers present their work to the client or end-users. They collect feedback on everything from the interface to functionality—it is here where product requirements may come under scrutiny.
Clients may change their minds or discover that something that seemed right on paper makes no sense in practice. Clients are only human, after all. With feedback in hand, developers return to some form of step 2: they continue to prototype. If feedback is strictly positive and the client is satisfied with the prototype, developers can move to step 4.
4. Finalize the Product
During this stage, developers may optimize or even re-engineer their implementation to improve stability, maintainability, and a third word ending in ‘-ility.’ They may also spend this phase connecting the backend to production data, writing thorough documentation, and doing any other maintenance tasks required before handing the product over with confidence.
Both Boehm’s Spiral Model and James Martin’s RAD Model make use of these four steps to help development teams reduce risk and build excellent products. However, RAD has its drawbacks as well.
RAD: Advantages and Disadvantages
We’ve covered some advantages of RAD already, but let’s restate them and expand.
Advantage |
Description |
Speed | In the traditional Waterfall approach, developers were unlikely to go on vacation after delivering the product. Clients would invariably request changes ranging from the interface to functionality after first delivery. With RAD, projects are more likely to finish on time and to the client’s satisfaction upon delivery. |
Cost | In rapid application development, developers build the exact systems the client requires, and nothing more. In Waterfall, IT risks building and fleshing out complex feature sets that the client may choose to gut from the final product. The time spent building zombie features can never be recovered, and that means the budget spent on them is lost. RAD reduces this risk and therefore reduces the cost. |
Developer Satisfaction | In the traditional Waterfall approach, developers work in silos devoid of feedback and positive affirmation for a product well-made. And when they finally get the opportunity to present their work to the client, the client may not roll out the red carpet for them. Regardless of how proud developers are of their work, if the client isn’t satisfied, the end result is not a good one. In RAD, the client is there every step of the way and the developer has the opportunity to present their work frequently. This gives them the confidence that when the final product is delivered, their work receives appreciation. |
Those advantages sound pretty rosy, so let’s douse this warm positivity with a cold splash of reality.
Disadvantage | Description |
Scale | A close-knit team of developers, designers, and product managers can easily incorporate RAD practices because they have direct access to one another. When a project expands beyond a single team or requires inter-team communication, the development cycle invariably slows down and muddles the direction of the project. Simply put, it’s difficult to keep a large group of people on the same page when your story is constantly changing. |
Commitment | In Waterfall, the client spent most of their time apart from the development team after completing specifications. This allowed clients to focus on their primary tasks and developers to focus on building. In RAD, the frequent cycle of prototypes requires developers and clients to commit to frequent meetings that, on the outset, may appear to consume unnecessary cycles. |
Interface-Focus | The RAD methodology motivates developers to find the perfect solution for the client. The client judges the quality of the solution by what they can interact with. As a consequence, some developers forego best practices on the backend to accelerate development of the front-end-focused prototype. When it’s time to deliver a working product, they patch up the server code to avoid a refactor. |
With the pros and cons of rapid application development laid out, we can determine which types of projects benefit most from RAD, and which do not. If you need to build an internal business tool or even a customer-facing portal, like an app or website, RAD techniques will help your team deliver a better experience to your end-user.
However, if you are tasked with building mission-critical software (flight controls, implant firmware, etc.), a RAD approach is not only inappropriate, it may also be irresponsible.
Published at DZone with permission of Stanley Idesis, DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.
Trending
-
Creating Scalable OpenAI GPT Applications in Java
-
How to Create a Microservice Architecture With Java
-
Ways To Stop and Resume Your Kafka Producer/Consumer at Run-Time
-
Does the OCP Exam Still Make Sense?
Comments