Leider kann Ihr Browser nicht alle Designelemente dieser Seite korrekt darstellen. Damit Sie trotzdem die kompletten Inhalte sehen können, bieten wir Ihnen diese abgerüstete Version an. Um das korrekte Design der Seite zu sehen, probieren Sie bitte einen anderen Browser.



Why wasn't this tested before?
This question often arises shortly before a deadline.
Our answer: Because testing was not easy enough.

Especially creating and maintaining a test structure is often used as an excuse.
JUnitDoclet lowers the step toward JUnit.
It generates skeletons of TestCases based on your application source code.
And it supports you to reorganize tests during refactoring.
Suddenly all the excuses don't count any longer.

This tool is provided by ObjectFab GmbH under GNU LGPL. It is just one result of many coaching and development projects, the ObjectFab team did. Please read the whole story and who else was/is involved. We'd like to thank Robert Watkins in particular, please read about his role.

drucken top


JUnitDoclet - The Story

The idea of JUnitDoclet was raised in 2001 by Steffen Gemkow, who was consulting as a Java- and OO-Coach at that time. But Steffen never really had the time to implement his idea. Steffen talked with friends at his company about JUnitDoclet. Among them: Andreas Braig. He liked the idea and promised to write an article about it, once JUnitDoclet would be there. In March 2002 Steffen went on vacation and when he returned, he had created version 0.1 of JUnitDoclet. Andreas and Steffen did some refinements and asked Uwe Petschke, Frank Finger, Falk Lehmann and Martin Uhlig to review version 0.3.

Steffen was developing a framework for the programming contest "The Way Out" in April 2002. Of course he used JUnitDoclet, now in version 0.4. It was quite stable, but very limited in terms of customization. JUnitDoclet 0.4 shipped as a stealth release with the framework of "The Way Out" and got some attention. In July, when Steffen was coaching a team of C++-Developers on their way to serverside Java, he introduced not only ANT and JUnit to them, but JUnitDoclet as well. The acceptance of JUnit paired with JUnitDoclet was at a level, Steffen hasn't seen before. Andreas wrote the article as promised and pushed Steffen to make JUnitDoclet more adaptable. He even wrote a major part of the TestCases covering all the new features (compared to 0.4).

We discovered, that there was a project called JUnitDoclet at SourceForge already. Robert Watkins from Australia had created some tool with the same name. When he learned about the other JUnitDoclet, he was surprised by the wealth of features implemented and offered to change the name of the tool he created, so that this JUnitDoclet can be hosted on Sourceforge. Thank you Robert. This is much more than we expected.

In summer of 2005 Fabrice Bellingard suggested a plugin for Eclipse (called: e-junitdoclet) which he offered to write. It is available as Beta 1, so please go ahead and check it out.

JUnitDoclet is the result of a team effort. Credits go to Steffen Gemkow, Andreas Braig, Uwe Petschke, Falk Lehmann, Frank Finger, Martin Uhlig, Robert Watkins and Fabrice Bellingard.

drucken top