users@grizzly.java.net

Re: big questions

From: Igor Minar <iiminar_at_gmail.com>
Date: Wed, 21 Apr 2010 20:58:11 +0200

On Wed, Apr 21, 2010 at 6:14 PM, Oleksiy Stashok
<Oleksiy.Stashok_at_sun.com> wrote:
> Does anyone have any objections to me migrating our tests from junit to
> testng?  I much prefer testng and we get quite a bit more control and
> features for very little effort.  This is probably the least contentious of
> the two but I wanted to throw that out there.
>
> IMO it could be a good idea.
> Today I've reworked HTTP SuspendTest in order to test the HTTP response
> suspending in the HTTPS mode. And I like the way I was able to do it using
> testng @DataProvider, So instead of writing separate tests for HTTP and
> HTTPS, we have a single test set driven by "isSslEnabled" parameter.

I'm particularly in love with junit, but just FYI: you can write
parametrized tests with junit too

> Regarding Netbeans support - it could be better for sure. For example
> Netbeans test report doesn't distinguish parameterized tests, in my case in
> SuspendTest, when each test is being run twice in two different modes
> HTTP/HTTPS, Netbeans reports that the test was run just once, and if HTTP
> succeeded and HTTPS failed - Netbeans reports that the test was passed
> successfully.

the NB support for junits parametrized tests is wonky too. it shows
test execution as fooTest[0], fooTest[1] ... if you have fooTest that
is executed with two different parameter values.

/i