Brent Barker | 11 Aug 18:10 2014

Report times are wrong when running multiple threads

When executing reports in multiple threads, the execution times are not correct. Instead of varying from 1-10 minutes, they say the execute in microseconds (which i know is not the case)

Are other seeing this?
Dan McLellan | 23 Jul 20:24 2014

Are there plans to provide a story-specific timeout feature in JBehave?

Hi Everyone,

It looks, from issue 1436 of this site, that JBehave's current position is 
that the StoryTimeouts feature is intended for the execution of all test 
cases in a job, rather than each individual story being run:

Per Mario Talevi:

As for the expected behaviour, do note that the story timeout is global, 
i.e. not per effective story execution time, in the sense that it's 
calculated (in "real" time) from the moment the execution of all stories 
starts.   After the elapsed time as reached the timeout, all story execution 
is cancelled.   We can try to improve on this for 4.0.


But the JBehave website itself seems to indicate that story-specific timeout 
control is already available in the current release, although the 
whereabouts of the class/methods to set those controls are not specified:

<excerpt from>
It is important to note that it's the entire story that is executed in one 
concurrent execution and that scenarios within the same story cannot be run 
concurrently. For each story execution a timeout in seconds can be set via 
the same configuration mechanism used to set the number of threads.
</excerpt from>


But I can't find any methods or variables in Embedder, EmbedderControls, 
StoryReporter to set timeouts for individual stories, or any "watchdog" 
functionality in StoryManager (or elsewhere) to enforce them.

It would be good to know:
1.  Is story-specific execution timeout currently available, and if so, how 
can it be set?

2.  If you can't currently set the timeout on individual stories, is any 
such feature currently being targeted for 4.0 or another future release?

Thank you.

- Dan McLellan

To unsubscribe from this list, please visit:

Brent Barker | 21 Jul 19:33 2014

Maping stories at the senario level

HI again,

I am using the story mapping feature with Maven. I am wondering if it is possible to map stories with meta tags at the scenario level, not only at the story level.

For example, I have a scenario inside of a story annotated with <at> bug. I'd like the map reports view to show that a scenario in the story has that <at> bug meta tag

Thanks in advance!
Brent Barker | 10 Jul 19:49 2014

Conditionally skip story using meta tag

Hi everyone!

Is there a easy way to conditionally skip a story using a meta tag? For example, I want to annotate my story with <at> bug BUG-3333 and then check if that bug is still open or not. If it is still open, skip the scenario/story.

Right now I am extending RegexStoryParser, calling super.ParseStory to get the story object, then if the bug meta property exists, I add the skip meta property to it. 

Is there a better way to do this? I am not sure how to feed a list of meta filters to the embedder, since I can not access it from the story configuration where I have access to the story object

Hans Schwäbli | 30 Jun 10:52 2014

Question regarding WebDriverScreenshotOnFailure

I have subclassed WebDriverScreenshotOnFailure and overridden afterScenarioFailure(UUIDExceptionWrapper).
I wondered why my class never was used. Then I discovered that I have to add the same annotation in the overriden method like it is in the superclass:
<at> AfterScenario(uponOutcome = Outcome.FAILURE)
Do you intend not to "inherit" this annotation to sub classes when afterScenarioFailure(UUIDExceptionWrapper) is overridden?
Brent Barker | 25 Jun 20:08 2014

Eclipse JBehave plugin - step method to story navigation

A feature that would be really nice and would give a huge advantage over other framework's plugins would be the ability to navigate from an annotated step method to the story/stories it is being used in. Currently I have only seen the ability to navigate from a story to a step.

Is this possible to do? 

Brent Barker | 24 Jun 01:24 2014

Configuring Threads in pom.xml

I'm trying to configure the number of threads to run via maven, however the number of threads does not seem to be picked up. If i directly edit the annotation in WeldAnnotatedStoryRunner, I am able to get the tests to run with multiple threads. 

Here is my configuration:

Is there something I am missing here?

Thanks in advance!
Jonathan Gray | 23 Jun 00:56 2014

Specifying report path in multi-module project


I have a multi-module project which looks something 
like the following (it happens to be Maven based).

- pom.xml
- app
+- pom.xml
- app-test
+- pom.xml

If I run "clean verify" against the project pom (it's not currently using
 inheritance from the parent) when it runs the maven plugin in the
 app-test project I can see the following

[INFO] Generating reports view to 'C:\project\target\jbehave' ...

I would expect it to be generating to C:\project\app-test\target\jbehave

Each of my Java tests are currently JUnitStory and running as embeddables.
The configuration is using LoadFromClasspath(...) which works fine.
The stories run correctly but if they've already failed then at times they 
won't succeed because the target (parent) hasn't been cleaned.

To unsubscribe from this list, please visit:

Brenner, Richard | 18 Jun 14:03 2014

jbehave-osgi p2 repository problem


I am trying to get jbehave osgi to run, ultimately to use swtbot tests within BDD tests.
I did the following (Windows 7, maven 3.2.1, Jdk 7):
git checkout jbehave-4.x
mvn clean install -s settings.xml
git clone
mvn clean install -P bundles -s settings.xml
mvn clean verify -P distribution-equinox -s settings.xml
The build says success but there is no p2 repository generated in the target as described here:
"...This will create the p2 repository using the Luna repository inside the the folder /org.jbehave.osgi.equinox.p2/target/repository."
The same happens when I do the same steps in Ubuntu Linux.
Any help would be appreciated.
Hans Schwäbli | 18 Jun 08:55 2014

Re: Filtering works not as expected with a scenario containing an examples table

When will there be a beta-8 version of the 4.x branch containing this fix?
I am asking because the examples for the JBehave article will need that, and the magazine is published on 2nd of July.

On Wed, May 21, 2014 at 8:50 AM, Mauro Talevi <mauro.talevi-hQ+s5KbX5Ykpi3HaZ6Sa4g@public.gmane.orgg> wrote:
Sold! To the German-speaking gentleman at the back of the room :-)

On 20/05/2014 21:00, Mirko Friedenhagen wrote:

I stand corrected, in this case JEDES is a better translation for ANY.
Regards Mirko
-- (

On Tue, May 20, 2014 at 10:14 AM, Hans Schwäbli
<> wrote:
Mirko, I suppose you are a native German speaker like me, right?

What is your message? That the current German translations for ANY and
FAILURE are the best?

Lets put it in the context:

* Ergebnis: BELIEBIGES
* Ergebnis: JEDES

Did you try that feature? You should really try and see how it behaves I

To me "Ergebnis: IRGENDWELCHE" it sounds unnatural. One can understand with
a bit thought what that might mean. But it seems not so intuitive like the
English word ANY.

"Ergebnis: JEDES" seems to express the correct meaning and is easy to
understand. Because whatever the result is, the steps in that block will be
added before or after the scenario. In German: In *jedem* Fall werden
Schritte vor oder nach dem Szenario hinzugefügt.

Or concerning BELIEBIGES: Bei einem *beliebigen* Testergebnis werden
Schritte vor oder nach dem Szenario hinzugefügt. Sound natural and easy to
understand to me.

But IRGENDWELCHE? Bei *irgendwelchen* Testergebnissen werden Schritte vor
oder nach dem Szenario hinzugefügt? This sounds very strange to me.

Concerning ISTQB, this is the definition of a failure:
Deviation of the component or system from its expected delivery, service or

Failures is translated by ISTQB as "Fehlerwirkungen".

So it is not "Ausfall". You may say it is also not "Fehler". But
"Fehlerwirkung is an artificial word originating from ISTQB. Noone I ever
met (except ISTQB teachers) ever uses this word but instead says "Fehler".

By the way, defect is translated as Fehlerzustand by ISTQB. This is also an
artificial word which noone uses except ISTQB teachers. In Germany we call
them: Bugs or simply Fehler, or much more academically and very seldom:

Besides that, ISTQB, although helpful to some degree in its basic teachings,
I consider it to be non-agile in its full extent. I take only the good from

On Tue, May 20, 2014 at 8:42 AM, Mauro Talevi <>
So, what's the consensus then with the keywords?

On 16/05/2014 18:42, Mauro Talevi wrote:

I'll defer to whatever you guys decide is best.  We can always change it

On 15/05/2014 18:27, Mirko Friedenhagen wrote:


I am not sure I agree :-). JEDES would be EVERY IMO.

According to ISTQB FEHLER would be the DEFECT which causes a FAILURE
(FEHLSCHLAG), which may lead to an AUSFALL (BREAKDOWN) of a server ;-)

Am 15.05.2014 12:34 schrieb "Hans Schwäbli"
I re-tested it and now it works. Thank you!

However I did not use that feature in-depth so there might be some other

I wondered a bit about outcome ANY. It seems to be like the finally-block
in Java. The German translation IRGENDWELCHE is maybe not the best for ANY.
Ergebnis: "BELIEBIGES" or "JEDES" seems to be better to me.

And Ergebnis: "AUSFALL" seems not to be the best translation too. I think
better would be Ergebnis "FEHLER".

Maybe some other German speaking guys can share their opinions about a
translation for ANY and FAILURE?

On Thu, May 15, 2014 at 12:08 AM, Mauro Talevi
<> wrote:
There was an issue with parsing with non-EN locales.   Now fixed, try
again with latest head.

On 14/05/2014 17:35, Hans Schwäbli wrote:

I quickly tested the lifecycle.


Gegeben im Lager sind 100 T-Shirts
Ergebnis: ERFOLG
Gegeben im Lager sind 200 T-Shirts
Gegeben im Lager sind 300 T-Shirts
Ergebnis: AUSFALL
Gegeben im Lager sind 400 T-Shirts
Szenario: Versandkosten fallen weg
Wenn ein Kunde 20 T-Shirts bestellt
Dann betragen die Versandkosten 7,5 Euro
Result is:

Gegeben im Lager sind 100 T-Shirts
Gegeben im Lager sind 200 T-Shirts
Gegeben im Lager sind 300 T-Shirts
Gegeben im Lager sind 400 T-Shirts

Szenario: Versandkosten fallen weg
Gegeben im Lager sind 100 T-Shirts
Wenn ein Kunde 20 T-Shirts bestellt
Dann betragen die Versandkosten 7,5 Euro
Gegeben im Lager sind 200 T-Shirts
Gegeben im Lager sind 300 T-Shirts
Gegeben im Lager sind 400 T-Shirts

It does not work as I expect it since it executes all three after steps
although it should only execute the one for "Ergebnis: ERFOLG" (Outcome:

On Friday or next week I can test that a bit more thoroughly.

On Tue, May 13, 2014 at 8:00 PM, Mauro Talevi
<> wrote:
Cool, we'll push out new beta soon.

Can you also take the Lifecycle After upon outcome functionality for a
spin while you're at it?

On 13/05/2014 13:42, Hans Schwäbli wrote:

I mixed up snapshot versions with beta-versions, sorry.

I tried now the snapshot version and it works now as expected
concerning the problem with the examples table.

Thank you!

But there is a problem with comments. I will write a posting just on

On Thu, May 8, 2014 at 10:51 AM, Mauro Talevi
<> wrote:
No, a new beta has not been deployed yet.   In the meantime, you can
use the latest 3.9.x or build the 4.0 snapshot from source.

On 8 May 2014, at 08:59, Hans Schwäbli <>

Thank you! Is it also deployed?
I did not find it here:
The last snapshot there is from 2nd of May.
The same snapshot date is on:

On Wed, May 7, 2014 at 11:25 PM, Mauro Talevi
<> wrote:
This issue is now fixed in head of 4.x branch.   It did not apply to

On 07/05/2014 10:55, Hans Schwäbli wrote:

I created such an example for jbehave-core now and attached it to
this posting. I still cannot work on a clone the Github project because of
company restrictions (I haven't yet received an answer why it is not working
inside the company proxy).

In case the mailing list does not support attachments I have also
sent them directly to Mauro.

To reproduce it you will need this in the Maven pom.xml:


<metaFilter>+component order -skip</metaFilter>


On Wed, May 7, 2014 at 11:03 AM, Hans Schwäbli
<> wrote:
I committed it here:

I will add such a scenario to the core examples. Until then you
could use that other example if you like. It is the example for the article
on JBehave by the way.

On Tue, May 6, 2014 at 3:04 PM, Mauro Talevi
<> wrote:
Yes, it looks likely to be unrelated to given stories and such.

Could you please add a scenario reproducing the behaviour to the
meta_filtering.story in the core examples (preferably in English)?

Does it work with 3.x?

On 06/05/2014 11:34, Hans Schwäbli wrote:

I already use StoryControls.doIgnoreMetaFiltersIfGivenStory(true).
And I removed the given story in the story.

But the result is the same.

Maybe tomorrow I can commit the whole project, so that you can
reproduce it.

On Tue, May 6, 2014 at 11:00 AM, Stephen de Vries
<stephendv <at>> wrote:

On 6 May 2014, at 10:51, Hans Schwäbli
<> wrote:

I have the example story, see below. It runs not as expected when
filtering by: +Komponente Bestellung -Skip


My guess is that the given story doesn’t have the same meta-tags.
Fix is to set: StoryControls.doIgnoreMetaFiltersIfGivenStory(true)


To unsubscribe from this list, please visit:

To unsubscribe from this list, please visit:

To unsubscribe from this list, please visit:

Brent Barker | 18 Jun 00:28 2014

Excluding "-" in meta filters

I have a meta filter on a story similar to this:

<at> bug JBEHAVE-281

The issue is when I try to filter on that, using 

"-bug JBEHAVE-281" 

it does not match, but only matches to similar items when I do 

"-bug JBEHAVE*"

Is there a way to escape the dash in the meta filter so I can filter for that specific bug?