BLOOM

BLOOM

BLOOM is a comprehensive system aimed at using visualization for software understanding. This system includes a variety of data analysis techniques, a visual query language for defining what should be visualized and how it should appear, and a powerful and flexible visualization framework containing a broad range of 2D and 3D visualizations. Our goal is to let the programmer define and then create a sophisticated software visualization in under five minutes.

PUBLICATIONS:

Bee/Hive: A Software Visualization Back End, IEEE Workshop on Software Visualization, May 2001.

An Overview of BLOOM, PASTE '01, June 2001.

Languages for Dynamic Instrumentation, Proc. WODA (ICSE Workshop on Dynamic Analysis) 2003, May 2003.

Event-Based Performance Analysis, Proc. IWPC 2003, May 2003.

The BLOOM Software Visualization System, in Software Visualization -- From Theory to Practice, MIT Press, 2003.

A Visual Query Language for Software Visualization, Proc. IEEE Symp. on Human Computing Languages and Environments, Sept. 2002.

Automated Fault Localization Using Potential Invariants, AADebug 2003.

Generating Java Trace Data, Proc. Java Grande, June 2000.

PICTURES OF VISUALIZATIONS:

Allocations by class

Resources used for different events

Dynamic call graph

Spiral views of the stack (sampled during execution)

Allocations by Object

Gprof-style profiling

File Change times

File-based performance display

Class usage automata

PICTURES OF THE VISUAL QUERY LANGUAGE:

PICTURES OF THE INTERACTIVE INTERFACE:

VIDEOS OF BLOOM

DOWNLOAD THE SOFTWARE