PurifyPlus™

Runtime analysis is the practice of looking at a program's behavior based on what it does when it runs. This is different from "static analysis," where you analyze program behavior by looking at its source code or high-level structure; or "crash analysis," where you look at what a program leaves behind when it fails.

Debugging is one classic example of runtime analysis: You set breakpoints, step through the program, and look at argument values and variables, all to gain an understanding of what's actually happening as the program runs. When you're debugging, you're not interested in all the possibilities of what might happen, but, rather, what really does happen. That's what separates runtime analysis from static analysis.

Debugging is highly interactive and can be powerful, but it's limited in what it can show you. That is, it's hard to identify patterns and integrate higher-level behavior when your focus is narrowed down to the size of a source line or a variable.

Testing is another kind of runtime analysis: You feed in data or events and check to see that you got the right behavior from the program. This sort of testing is the cornerstone of software quality.

Still, traditional testing only shows you so much. A program can pass all its tests (give the right answers) and still have serious quality problems.

This is why the powerful runtime analysis tool - PurifyPlus - was created.