| <!doctype html PUBLIC "-//W3C//DTD HTML 3.2 Final//EN"> |
| <html> |
| <head> |
| <meta http-equiv="content-type" content="text/html; charset=iso-8859-1"> |
| <meta http-equiv="content-style-type" content="text/css"> |
| <link rel="stylesheet" type="text/css" href="style.css"> |
| <title>ProGuard Results</title> |
| </head> |
| <body> |
| |
| <h2>Results</h2> |
| |
| <b>ProGuard</b> successfully processes any Java bytecode, ranging from small |
| midlets to entire run-time libraries. It primarily reduces the size of the |
| processed code, with some potential increase in efficiency as an added bonus. |
| The improvements obviously depend on the original code. The table below |
| presents some typical results: |
| <p> |
| |
| <table> |
| |
| <tr> |
| <th width="28%">Input Program</th> |
| <th width="12%">Original size</th> |
| <th width="12%">After shrinking</th> |
| <th width="12%">After optim.</th> |
| <th width="12%">After obfusc.</th> |
| <th width="12%">Total reduction</th> |
| <th width="12%">Time</th> |
| <th width="12%">Memory usage</th> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://java.sun.com/j2me/">Worm</a>, a sample midlet from Sun's JME</td> |
| <td align="center">10.3 K</td> |
| <td align="center">9.8 K</td> |
| <td align="center">9.6 K</td> |
| <td align="center">8.5 K</td> |
| <td align="center">18 %</td> |
| <td align="center">2 s</td> |
| <td align="center">19 M</td> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://www.javadocking.com/">Javadocking</a>, a docking library</td> |
| <td align="center">290 K</td> |
| <td align="center">281 K</td> |
| <td align="center">270 K</td> |
| <td align="center">201 K</td> |
| <td align="center">30 %</td> |
| <td align="center">12 s</td> |
| <td align="center">32 M</td> |
| </tr> |
| |
| <tr> |
| <td><b>ProGuard</b> itself</td> |
| <td align="center">648 K</td> |
| <td align="center">579 K</td> |
| <td align="center">557 K</td> |
| <td align="center">348 K</td> |
| <td align="center">46 %</td> |
| <td align="center">28 s</td> |
| <td align="center">66 M</td> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://www.clarkware.com/software/JDepend.html">JDepend</a>, a Java quality metrics tool</td> |
| <td align="center">57 K</td> |
| <td align="center">36 K</td> |
| <td align="center">33 K</td> |
| <td align="center">28 K</td> |
| <td align="center">51 %</td> |
| <td align="center">6 s</td> |
| <td align="center">24 M</td> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://java.sun.com/javase/6/">the run-time classes</a> from Sun's Java 6</td> |
| <td align="center">53 M</td> |
| <td align="center">23 M</td> |
| <td align="center">22 M</td> |
| <td align="center">18 M</td> |
| <td align="center">66 %</td> |
| <td align="center">16 min</td> |
| <td align="center">270 M</td> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://jakarta.apache.org/tomcat/index.html">Tomcat</a>, the Apache servlet container</td> |
| <td align="center">1.1 M</td> |
| <td align="center">466 K</td> |
| <td align="center">426 K</td> |
| <td align="center">295 K</td> |
| <td align="center">74 %</td> |
| <td align="center">17 s</td> |
| <td align="center">44 M</td> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://www.kclee.com/clemens/java/javancss/">JavaNCSS</a>, a Java source metrics tool</td> |
| <td align="center">632 K</td> |
| <td align="center">242 K</td> |
| <td align="center">212 K</td> |
| <td align="center">152 K</td> |
| <td align="center">75 %</td> |
| <td align="center">20 s</td> |
| <td align="center">36 M</td> |
| </tr> |
| |
| <tr> |
| <td><a target="other" href="http://ant.apache.org/">Ant</a>, the Apache build tool</td> |
| <td align="center">2.4 M</td> |
| <td align="center">401 K</td> |
| <td align="center">325 K</td> |
| <td align="center">242 K</td> |
| <td align="center">90 %</td> |
| <td align="center">23 s</td> |
| <td align="center">61 M</td> |
| </tr> |
| |
| </table> |
| <p> |
| Results were measured with ProGuard 4.0 on a 2.6 GHz Pentium 4 with 512 MB |
| of memory, using Sun JDK 1.5.0 in Fedora Core 3 Linux. |
| <p> |
| The program sizes include companion libraries. The shrinking step produces the |
| best results for programs that use only small parts of their libraries. The |
| obfuscation step can significantly shrink large programs even further, since |
| the identifiers of their many internal references can be replaced by short |
| identifiers. |
| <p> |
| The Java 6 run-time classes are the most complex example. The classes perform |
| a lot of introspection, interacting with the native code of the virtual |
| machine. The 1500+ lines of configuration were largely composed by automated |
| analysis, complemented by a great deal of trial and error. The configuration |
| is probably not complete, but the resulting library successfully serves as a |
| run-time environment for running applications like ProGuard and the ProGuard |
| GUI. |
| <p> |
| For small inputs, timings are governed by the reading and parsing of the jars. |
| For large inputs, the optimization step becomes more important. For instance, |
| processing the Java 6 run-time classes without optimization only takes 2 |
| minutes. |
| <p> |
| Memory usage (the amount of physical memory used by ProGuard while processing) |
| is governed by the basic java virtual machine and by the total size of the |
| library jars and program jars. |
| <hr> |
| <address> |
| Copyright © 2002-2009 |
| <a href="http://www.graphics.cornell.edu/~eric/">Eric Lafortune</a>. |
| </address> |
| |
| </body> |
| </html> |
