Development Philosophy

In the past, most application development centered around a lengthy, linear process: analysis, design, implementation, testing and release. This cycle might stretch out over a year or more, and often times the requirements gathered at the beginning of the project were not even close to the real needs of the business by the time the software was actually released. Once the business strategists had given thier input in the analysis phase, all that they could do was sit back and hope that they had not missed anything. Changes, especially later in the cycle. were extremely costly.

At Acolyte, we have adopted a different strategy for developing applications. We target shorter, more focused development cycles that begin to deliver business value quickly. Each development cycle builds on the previous one, allowing business strategists to make course corrections and adjustments throughout the project. This approach has demonstated much higher success rates, fewer missed deadlines, and fostered cooperation betweeen business people and technical people.

Object Oriented Analysis & Design

Another key aspect of Acolyte's approach is object oriented analysis and design. When applied correctly, object oriented techniques can greatly reduce software maintainence costs and provide a natural path for upgrades and enhancements. By insulating various behaviors and responsibilities within the object model, integration with legacy systems can be accomplished seemlessly.

Relational Database Design

One of the primary roles of applications is to collect and manage data. A relational database managements system (RDBMS) is often used to store and retrieve this data. Acolyte has many years of industry experience designing both small and large data models for a number of RDBMS engines including Oracle, SQL Server, MS Access and mySQL. Certain nuances in the design are tailored to the specific requirements of the customer and the RDBMS being used.

In general, Acolyte's philosophy is to begin by designing a normalized data model. This means that there will not be redundant data stored in different tables, and that data can be updated in one place. A normalized data model is the most reliable, consistent and flexible. In certain situations, a highly normalized data model may suffer performance problems. Where necessary, Acolyte will de-normalize data and tune the design to provide optimal efficiency.