directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Tencé, Vincent" <vte...@optimuminformatique.com>
Subject RE: Philosophy of failing tests
Date Fri, 25 Feb 2005 17:08:33 GMT
> It's not a question of test size.  It's a problem of of the 
> code that is 
> being tested.  Fixing it could be a big change.

That's what I meant by too big a test: one that requires a big change in the
code base to make it pass.
 
> So it's your opinion that unit tests are more like regression tests?

I unit test to dirve development of code, so to me it's more like a
development technique. Nevertheless, you get extra benefits from this. One
of those being building a test suite that prevents regression.

- Vincent


> Tencé, Vincent wrote:
> 
> >Then maybe the test is too big and should be broken down in 
> smaller tests.
> >Anyhow, until the test passes, the code shouldn't go into svn IMHO.
> >
> >-- Vincent
> >
> >  
> >
> >>-----Original Message-----
> >>From: Alan D. Cabrera [mailto:adc@toolazydogs.com]
> >>Sent: February 25, 2005 11:34 AM
> >>To: Apache Directory Developers List
> >>Subject: Re: Philosophy of failing tests
> >>
> >>
> >>It's the *making it pass* phase that may take some time.
> >>
> >>
> >>Regards,
> >>Alan
> >>
> >>Tencé, Vincent wrote:
> >>
> >>    
> >>
> >>>Sounds good ;-) 
> >>>
> >>>With a cycle of:
> >>>
> >>>writing the test, making it pass, refactoring
> >>>
> >>>we should not have any failing test in the repo.
> >>>
> >>>-- Vincent
> >>>
> >>> 
> >>>
> >>>      
> >>>
> >>>>-----Original Message-----
> >>>>From: Alan D. Cabrera [mailto:adc@toolazydogs.com]
> >>>>Sent: February 25, 2005 11:15 AM
> >>>>To: Apache Directory Developers List
> >>>>Subject: Re: Philosophy of failing tests
> >>>>
> >>>>
> >>>>We have some whipper snapper from France who is putting in more 
> >>>>extensive tests, discovering problems w/ our code as he goes 
> >>>>along.  Not 
> >>>>to worry though, we have given him committer karma and now he 
> >>>>has to fix 
> >>>>the problems he finds!  ;)
> >>>>
> >>>>
> >>>>Regards,
> >>>>Alan
> >>>>
> >>>>
> >>>>Tencé, Vincent wrote:
> >>>>
> >>>>   
> >>>>
> >>>>        
> >>>>
> >>>>>No such thing can exist! That's mine. How do we get failing 
> >>>>>     
> >>>>>
> >>>>>          
> >>>>>
> >>>>unit tests in
> >>>>   
> >>>>
> >>>>        
> >>>>
> >>>>>the first place in SVN? If we do however, I agree they 
> >>>>>          
> >>>>>
> >>should be kept
> >>    
> >>
> >>>>>failing. It'll make visible that we have a problem and put 
> >>>>>     
> >>>>>
> >>>>>          
> >>>>>
> >>>>pressure on the
> >>>>   
> >>>>
> >>>>        
> >>>>
> >>>>>team to deal with it.
> >>>>>
> >>>>>- Vincent
> >>>>>
> >>>>>
> >>>>>
> >>>>>     
> >>>>>
> >>>>>          
> >>>>>
> >>>>>>-----Original Message-----
> >>>>>>From: Alan D. Cabrera [mailto:adc@toolazydogs.com]
> >>>>>>Sent: February 25, 2005 11:03 AM
> >>>>>>To: Apache Directory Developers List
> >>>>>>Subject: Philosophy of failing tests
> >>>>>>
> >>>>>>
> >>>>>>What is everyone's philosophy on failing unit tests?  
> I, for one, 
> >>>>>>believe that they should be left failing until they are fixed.
> >>>>>>
> >>>>>>
> >>>>>>Regards,
> >>>>>>Alan
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>>  
> >>>>>>
> >>>>>>       
> >>>>>>
> >>>>>>            
> >>>>>>
> 

Mime
View raw message