counter indication is that you might get test mocks, test resource definitions, 
etc creeping in.

Explicitely making it an own mojo is a safe bet.
For my project both ways would work, but for others it might not be sufficient 
to always get all the test classpath.

LieGrue,
strub




----- Original Message -----
> From: Mauro Talevi <mauro.tal...@aquilonia.org>
> To: dev@jbehave.codehaus.org
> Cc: 
> Sent: Friday, March 29, 2013 9:25 AM
> Subject: Re: [jbehave-dev] JBEHAVE-454 ClassPath question
> 
> Actually, I would argue that having behaviour verification in a separate 
> module in compile scope should be considered a better practice.
> 
> Integration tests usually have cross-cutting concerns that span multiple 
> Maven modules, whereas the tests in a module's src/test directory are 
> meant to test the functionality of that module.
> 
> But, regardless, having the resolution scope set to test will cater for 
> both use cases.
> 
> If the user decides to use the plugin in scope compile, this is included 
> in the test one, and the only minor drawback possibly is that you'll get 
> a few extra test-scope dependencies pulled in the resolution, but given 
> that JBehave is a testing tool and the Maven modules are usually the 
> leaves of a build reactor I don't see counter-indications.
> 
> Cheers
> 
> On 28/03/2013 19:39, Mark Struberg wrote:
>> 
>>  Not sure if this is a good idea. There might be pure behave projects which 
> have their behave tests defined as compile scope in src/main/java. I guess 
> it's not the standard case but might theoretically have happened. But you 
> know this part much better than I do. All I could do is to point you to the 
> implication of @requiresDependencyResolution, to define the expected usage is 
> your task ;)
>> 
>>  txs and LieGrue,
>>  strub
>> 
>> 
>>>  ________________________________
>>>  From: Mauro Talevi <mauro.tal...@aquilonia.org>
>>>  To: dev@jbehave.codehaus.org
>>>  Sent: Thursday, March 28, 2013 12:46 AM
>>>  Subject: Re: [jbehave-dev] JBEHAVE-454 ClassPath question
>>> 
>>> 
>>>  We could make all mojos use  @requiresDependencyResolution test
>>> 
>>>  I don't see any counterindication atm.  Let me think about it ...
>>> 
>>>  Cheers
>>> 
>>>  On 26/03/2013 10:21, Mark Struberg wrote:
>>> 
>>>  hmm, maybe it got filtered out by the spam filter? I attached the 
> git-format-patch output. LieGrue,
>>  strub ----- Original Message -----
>>>>  From: Mauro Talevi <mauro.tal...@aquilonia.org> To: 
> dev@jbehave.codehaus.org Cc:
>>  Sent: Monday, March 25, 2013 11:07 PM
>>  Subject: Re: [jbehave-dev] JBEHAVE-454 ClassPath question Where have you 
> sent the patch? On 25/03/2013 21:15, Mark Struberg wrote:
>>>>>     Hi Mauro!   Just sent a patch for git-am. The main issue is 
> that you need a
>>     @requiresDependencyResolution test   in the Mojo if if should also 
> pickup test classpath entries. Otherwise you
>>>>>  will only get the runtime classpath.
>>>>>     txs and LieGrue,
>>     strub   ----- Original Message -----
>>>>>>     From: Mauro Talevi <mauro.tal...@aquilonia.org>  
> To: dev@jbehave.codehaus.org   Cc:
>>     Sent: Monday, March 25, 2013 12:36 AM
>>     Subject: Re: [jbehave-dev] JBEHAVE-454 ClassPath question   HI,   can 
> you send us a simple mvn project reproducing the issue?   The issue
>>     JBEHAVE-454 should be closed AFAICS.    You're issue could be
>>>>>  related to
>>>>>     something else.   Cheers   On 24/03/2013 20:01, Mark 
> Struberg wrote:
>>>>>>>      Hi folks!    I have a question regarding
>>>>>  http://jira.codehaus.org/browse/JBEHAVE-454
>>>>>      This issue is still marked as 'open' and I 
> currently
>>>>>  facing the
>>>>>     same problem.
>>>>>>>      I _do_ have <scope> set to test, but still 
> the jbehave
>>>>>  story gets
>>>>>     executed with the EmbedderClassLoader which does _not_ 
> contain any test
>>     classpath.
>>>>>>>      My problem starts when I start a dependency 
> injection container
>>>>>>>     (OpenWebBeans, Spring, Weld or embedded TomEE) 
> because they try to scan
>>>>>  the
>>>>>     ClassPath and subsequently fail to @Inject my beans.
>>>>>>>      There is also a 2nd thingy, which is more a 
> question kind of: for
>>>>>  CDI I
>>>>>     need to assign/open some Contexts to a new Thread. Is there 
> some hook
>>>>>  where I
>>>>>     can perform these tasks? I've seen the stories get 
> executed
>>>>>  asynchronously
>>>>>     via Futures.
>>>>>>>      Is there interest in cleaning this up, or is even 
> someone working
>>>>>  on it
>>>>>     atm?
>>>>>>>      I could help a bit on the maven part if this helps 
> - though I
>>>>>  might need
>>>>>     some help on the jbehave side...
>>>>>>>      txs and LieGrue,
>>      strub
>>>>> 
> ---------------------------------------------------------------------
>>>>>      To unsubscribe from this list, please visit:        
> http://xircles.codehaus.org/manage_email
>>>>>>>     
> ---------------------------------------------------------------------
>>     To unsubscribe from this list, please visit:      
> http://xircles.codehaus.org/manage_email
>>>>>>     
> ---------------------------------------------------------------------
>>     To unsubscribe from this list, please visit:      
> http://xircles.codehaus.org/manage_email
>>>>> 
> ---------------------------------------------------------------------
>>  To unsubscribe from this list, please visit:    
> http://xircles.codehaus.org/manage_email
>>>>> 
>>>>> 
> ---------------------------------------------------------------------
>>  To unsubscribe from this list, please visit: 
> http://xircles.codehaus.org/manage_email
>>> 
>>> 
>>  ---------------------------------------------------------------------
>>  To unsubscribe from this list, please visit:
>> 
>>       http://xircles.codehaus.org/manage_email
>> 
>> 
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe from this list, please visit:
> 
>     http://xircles.codehaus.org/manage_email
>

---------------------------------------------------------------------
To unsubscribe from this list, please visit:

    http://xircles.codehaus.org/manage_email


Reply via email to