. xDIVA

Welcome to the Home of xDIVA

xDIVA stands for (extreme Debugging Information Visualization Assistant).

Disclaimer

Currently, xDIVA is a research prototype which welcome your opinions. It is already usable to test out the ideas, however, we apology if you run into assertion fails, unhandled exceptions, or bugs. xDIVA is not implemented by full-time professional programmers. The xDIVA developers are the graduate students of Software Engineering Lab @ Computer Science and Information at National Taiwan Normal University . In addition, as the progress is made, some newly added features may create bugs for other components. This issue will be addressed by employing unit testing and regression testing in the near future.

Please add new ticket for your problems and opinions. You can also write email to ypc@… who is the project creator of xDIVA.

Starting Points

Tutorials
Gallery
Publications
Downloads
Installation Guide
Documentation and Publications?
The xDIVA install Q&A
Thins to check before run your xDIVA
Known bugs or Things that have not been taken care of?

About

The problem that xDIVA deals with

    In general, the visualization problem remains a difficult one if the domain of data is not constrained or limited. A simple example can explain the problem. Consider a class X with three attributes which     have poor naming.

   class X {
    int a;
    int b;
    int c;
   };

    The meaning of class X can be interpreted in several ways depending on the application domain. For example, X can be a vector in a 3-dimensional space. A visualization with an arrow pointing to the     direction (a,b,c) is most suitable for the interpretation. Next, in a commercial scenario a,b and c can constitute three kinds of product revenue of company X. A pie chart can be a suitable visualization     for the interpretation. In another scenario a,b,c may be the coefficients of a 2D line ax+by=c. A 2D line that satisfies the equation is the best visualization. In fact, such interpretations can go on and on     forever. It is unlikely that any fixed-mapping visualization will be available a priori to cover everything that might be of interest. The diversity of interpretations is what makes the problem difficult. In     general, the interpretations can be arbitrary. This problem, from the best of our knowledge, is rarely addressed and impossible to avoid in a debugging visualization scenario.