On 7/15/05, Vincent Massol <[EMAIL PROTECTED]> wrote:
> Hi Kenney,
> 
> > -----Original Message-----
> > From: Kenney Westerhof (JIRA) [mailto:[EMAIL PROTECTED]
> > Sent: jeudi 14 juillet 2005 22:33
> > To: [email protected]
> > Subject: [jira] Created: (MNG-591) Integration tests lifecycle
> 
> [snip]
> 
> > [Btw, I'm doing this while creating a cactus plugin, for the moment using
> > cargo in the TestSuite itself to deploy.]
> 
> Cool :-)
> 
> > The idea is that the integration test sources go in src/itest/*;
> 
> Don't you mean src/test/itest/*?

Is it always considered a build failure if integration tests do not pass 100%?

If so, then src/test/* is the right place for this stuff, otherwise
integration tests should be separated out into something like
src/itest/*.

M2 has done a great job of reducing the need for pattern-based
customization of directory contents processing.  Take java vs
resources source directories as an example.

The same rule of thumb would seem to apply here, assuming that
integration tests are not required to pass at 100% on every build.
 
> I think I'd prefer src/test/cactus actually because there can possibly be
> different types of integration tests for the same project and each would
> need to be run with its own testing framework (which might not be junit
> BTW).

That's a good point about having multiple possible execution engines
for integration testing, including junit.

According to the above feedback then, the cactus integration tests
would reside in either src/test/cactus or src/itest/cactus.

[snip]

Kind Regards,
John Fallows.

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to