Paul Michael Reilly <[EMAIL PROTECTED]> writes: > > You qualify. :-) >
Wait lurking some times on this list and you will find real maven hackers :-) > > The first testing tier will be based on the classes and will be mostly > traditional unit testing, but I will observe that unit testing has > never given me much bang for the buck. Rather it is the integration Not sure this is the right place to start such a discussion but I actually found unit testing to be very rewarding if taken seriously and with the "right" approach. IMHO, unit testing is the cheapest path towards assessing conformance of a piece of software with its specs, whether written, implied, formal or otherwise. As a developer, I *always* found that writting my unit tests gave me insights on my code, design and implementations. And I have not written a test class without finding at least one error or mistake. > level testing, driven by JUnit, or a JUnit derived framework, like > Abbot, that I have found to pay big dividends. But this particular > App has highlighted opportunities where traditional unit testing may > offer bigger dividends. A second testing phase based on the main jar > artifact is a definite win. Sure. Good unit test should be coupled with integration and system test to achieve suitable coverage and confidence level. I try now to systematically create an it module for my projects that contains such high-level tests. I did not found the idea but borrowed it from maven codebase itself which is structured that way. > > So I like your suggestions and appreciate you taking the time to make > them. Always a pleasure and never a waste of time. regards, -- OQube < software engineering \ génie logiciel > Arnaud Bailly, Dr. \web> http://www.oqube.com --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
