WebChar

WebChar is a system for automatically discovering performance and energy pitfalls in Web browsers. A poster and extended abstract on WebChar will appear in IISWC 2012.

Contact: Adrian Sampson

Publication

Automatic Discovery of Performance and Energy Pitfalls in HTML and CSS. Adrian Sampson, Calin Cascaval, Luis Ceze, Pablo Montesinos, and Dario Suarez Gracia. IISWC 2012.

The extended abstract that will appear at IISWC is only two pages. Full details are available in an eight-page technical report:

Automatic Discovery of Performance and Energy Pitfalls in HTML and CSS (Extended Technical Report). Adrian Sampson, Calin Cascaval, Luis Ceze, Pablo Montesinos, and Dario Suarez Gracia. TR UW-CSE-12-09-02.

Code and Data Release

The code for the WebChar analysis is available for download. The analysis is written in Python and requires SciPy and PyML. The library implements the feature calculation, correlation, and reporting for WebChar. The script wc.py is the main interface to the analysis.

The datasets used in the paper are also available, along with a set of Python scripts that are used to analyze them and perform statistical tests. The README file included in this archive explains how to run the analyses used in the paper.