RE: [Webtest] haltonfailure bug?done (build 237) -----Original Message----- From: webt...@lists.canoo.com [mailto:webt...@lists.canoo.com]On Behalf Of Dierk Koenig Sent: Mittwoch, 16. Oktober 2002 15:10 To: webt...@lists.canoo.com Subject: RE: [Webtest] haltonfailure bug?

I agree, the doc needs to be changed. Thanx. -----Original Message----- From: webt...@lists.canoo.com [mailto:webt...@lists.canoo.com]On Behalf Of dani...@pronto.com.au Sent: Mittwoch, 16. Oktober 2002 4:36 To: webt...@gate.canoo.com Subject: RE: [Webtest] haltonfailure bug?

Well, this looks like an error in the documentation.

haltonfailure: Defines that the excution of the current <testSpec> stops if a failure of one of the test steps is detected, e.g. a <verify...> test did not pass.

This implies that when it is set to "false" execution of the current <testSpec> DOES NOT STOP if a failure of one of the test steps is detected, ie, all subsequent tests are executed.

From the documentation, it would appear that haltonerror allows ANT to continue, and haltonfailure allows the testspec to continue.

> -----Original Message----- > From: Dierk Koenig [mailto:dier...@canoo.com] > Sent: Wednesday, 16 October 2002 12:41 AM > To: webt...@gate.canoo.com > Subject: RE: [Webtest] haltonfailure bug?

Hi Dan,

haltonfailure=false means to proceed with your ANT script, although your current test spec fails. The processing will proceed with your next test spec, not your next test step in the failing test spec.

I guess you have a rather large test spec with lots of steps that get executed in sequence. The solution would be to break this into smaller test specs.

We found that it is useful to have one test spec per "scenario", i.e. one path through your "use case", talking in UML terms.

cheers Mittie