Real Life Examples

Jan 14, 2009

Sun Microsystems' Java Pet Store J2EE BluePrint Application

This application was built in order to serve as a comparison model between the .NET and Java programming technologies. Financed by the Middleware company, the private competition was carried out by both Microsoft and Sun individually. Four weeks of programming the application required only 25% of the code used initially in Java.  When the application was tested in supervised conditions, the .Net app was 1000% faster than the upgraded version of the Java app. The version of the programming languages used were .NET 1.1 for a Windows 2003 OS, and J2EE for a Windows 2000. The results speak for themselves:

  • Preparing time prior to the actual test (for optimizing and upgrading the code):

        - .Net 1.1 – 2 weeks
        - J2EE– 10 weeks

  • Number of lines of code

        - .Net 1.1– 2.096
        - J2EE – 14.004

  • The cost/execution rate considered for each development language, calculated through the division of the maximum actions per second that the server could process.

        - .Net 1.1 – $316
        - J2EE - $1.305

  • Pages rendered per second

        - .Net 1.1 – 1.400
        - J2EE– 600

  • Number of simultaneous users supported by servers:

        - .Net 1.1 – 6.000
        - J2EE/Windows 2000 – 4.000

  • Number of simultaneous actions executed per second:

        - .Net 1.1– 117
        - J2EE– 59

The Nile Criterion

The Nile criterion is a server application targeted at complete e-commerce solutions, which was widely employed in private testing conditions in order to test the standards of different application server products. The Nile criterion provides a general framework that includes standard components found in the majority of web apps.

The criterion demonstrated that the same application developed using ASP.NET executed better than the same app programmed on a J2EE application server when output caching was enabled, and that the Microsoft .NET version of the Nile criterion outrun the J2EE version by more than 400% when the same feature was not enabled.


Next articles in this series:

1. Features and Performance
2. Marketing and International Standards
3. Case Analysis on .NET
4. Real Life Examples
5. .NET Future