iRise Application Simulator 3.0
Far more successful than any other attempt weve seen to enable application design by nonprogrammers, the 3.0 release of iRise Application Simulator gives business unit stakeholders a clear idea of Web application appearance and behavior while communicating their requirements to downstream development staff. Application Simulator 3.0 costs $250,000.
EVALUATION SHORT LIST
Unlike the other software quality assurance tools reviewed in this package, Application Simulator from iRise is intended for use by nonprogrammers—not the kind whose first to-do item is "learn Visual Basic," but rather the business analysts and other stakeholders who need to work with developers and share a common understanding of desired application appearance and behavior.
In more than a decade of reviews, eWEEK Labs has yet to see a more successful approach to that goal. In fact, we have yet to see anything close to what iRise has achieved, and its therefore a pleasure both to give this product its first-ever review and to award it the honor of Analysts Choice.
This is an enterprise product at an enterprise price: The project license package that combines server-based facilities, five full-function analyst seats and as many as 100 simultaneous users of the lighter-weight Web client tool is stickered at the nice round figure of $250,000. That may sound less intimidating if expressed as four equivalent full-time staff member years, which may be quickly saved by improved communication of business unit needs (not to mention potential gains from reduced time-to-market for a customer-facing application).
When we first learned of this product, we were, frankly, skeptical. Weve seen tools that let users lay out screens and draw diagrams of how one screen leads to another. But these tools have fallen short when it comes to identifying gaps between a data model and a desired application function. Weve also seen tools that let programmers lay out screens and populate their controls with code that defines behavior and interacts with data. But these tools dont enable overall application design without unreasonable investment in writing code.
We have not previously seen anything that succeeded in hitting the strategic target of laying out screens, describing their connections, and testing their functions using actual data without ever writing code or even anything that looks like code—quickly enough and clearly enough that different ideas can be tested and oversights rapidly identified at minimal cost. The iRise product, in contrast, detects logic errors to prevent impossible specifications from reaching a development team: "You cant design a simulation that wont run," asserted one iRise engineer, and we were unable to refute that.
During tests, we easily flipped between the design tools, with their PowerPoint-like ease of manipulating objects, and the simulation and analysis tools. The latter modes gave us a choice between viewing an application as it might appear to a user or viewing highly annotated screens that included references to statements of requirements and other information.
Only once did we think wed found a desired control flow that the iRise visual tools could not express. However, on further thought, we realized that the problem lay in our database model, not in the iRise semantics.
The products requirements-tracking aids and other collaborative tools are clearly designed in rather than bolted on. Features such as threaded discussions or e-mails to affected parties with notice of requirements changes are integrated well with the products standards-based framework of XML files and Web-based tools.
Discuss this in the eWEEK forum.