[Insight-developers] Tests failing in the Dashboard

Tue Oct 27 11:05:17 EDT 2009

Luis, Bill,

I have all of the changes ready to go to convert all of ITK to the  
templated rounding methods (pending build on victoria.nlm). I did want  
to run one more experiment before including itkMath into ITK, but that  
does not appear to be an option right now.

If there are no objection, and Luis is not currently working on  
cleaning up the last errors on the continuous builds (Luis?), I would  
like to commit this before lunch, to green up the dashboard.

I suspect that it should go relatively smoothly, except a few warning.  
And BCC failing on a a typename used as a template parameter. I do not  
know when BCC likes typename template arguments and when it doesn't,  
so I will just put out the fires when the show up on the continuous  
(unless I get a volunteer).

Sorry again for the problems I caused!

Brad

On Oct 27, 2009, at 6:45 AM, Luis Ibanez wrote:

> Bill,
>
> My commits were made after build with no compilation errors,
> (that's the Experimental build from macondo.kitware).
>
> but it was not from a clean build (not a build from scratch).
> I'm following up on the remaining issues.
>
>
>     Luis
>
>
> --------------------------------------------------------------------------------------
> On Mon, Oct 26, 2009 at 11:30 PM, Bill Lorensen <bill.lorensen at gmail.com 
> > wrote:
>> Luis,
>>
>> The dashboard wrath is getting worse after your changes. It will be a
>> while before it finishes, but the BillsLaptop borland continuous is
>> generating dozens of errors.
>>
>> I suggest we stop checking in code throughout the day until this  
>> issue
>> is resolved.
>>
>> Bill
>>
>> On Mon, Oct 26, 2009 at 10:44 PM, Luis Ibanez <luis.ibanez at kitware.com 
>> > wrote:
>>> Hi Brad,
>>>
>>> Thanks for the clarification....
>>>
>>> It looks that we will have to reconsider the call for the
>>> ImageBase::CopyInformation
>>> method...
>>>
>>>
>>> ---
>>>
>>> On a different note:
>>>
>>> It seems that some of your changes for itkMath were accidentally  
>>> committed
>>> to the repository. The Dashboard is looking pretty red at this  
>>> point.
>>>
>>>
>>> I'm committing some changes to calm the wrath of the Dashboard Gods,
>>>
>>> but
>>>
>>> the source code sacrifices will only take effect for the first  
>>> continuous builds
>>> of tomorrow morning....
>>>
>>> --
>>>
>>> These changes may not reflect what you intended to do. For example,
>>> I'm including itkMath.h in itkMacro.h, and in some other files, then
>>> removing from itkMacro.h the previous implementation of Math  
>>> functionalities.
>>>
>>> (This is just to get a Green Dashboard,
>>>  please feel free to reorganize according to your plans,
>>>  after things settle down.).
>>>
>>>
>>>
>>>      Luis
>>>
>>>
>>> ---------------------------------------------------------------------------------------------------------
>>> On Mon, Oct 26, 2009 at 2:54 PM, Bradley Lowekamp
>>> <blowekamp at mail.nih.gov> wrote:
>>>> Hello Luis,
>>>> The ImageBase::CopyInformation method does not support copying  
>>>> miss matching
>>>> dimensions of meta data.
>>>>
>>>> The new rounding work is not being used any where besides a  
>>>> couple of
>>>> tests.
>>>> To migrate from the old rounding methods to the new, I will need  
>>>> to modify
>>>> itkMacro.h and about 20 or so other Code files plus a few tests.  
>>>> When I am
>>>> ready to do this large commit I will inform this mailing list the  
>>>> day before
>>>> to help isolate potential issues.
>>>> Brad
>>>>
>>>>
>>>> On Oct 26, 2009, at 2:32 PM, Bill Lorensen wrote:
>>>>
>>>> This one is real. The Borland compiler is the only one that reports
>>>> divide by floating point 0.
>>>> http://www.cdash.org/CDash/testDetails.php?test=34297691&build=456734
>>>>
>>>> This one is directly caused by the CopyInformation add. Something  
>>>> to
>>>> do with the dynamic cast. Probably becuase input and output  
>>>> dimensions
>>>> in the test change.
>>>>
>>>> Bill
>>>>
>>>> On Mon, Oct 26, 2009 at 2:12 PM, Luis Ibanez <luis.ibanez at kitware.com 
>>>> >
>>>> wrote:
>>>>
>>>> In today's continuous builds we have several failing tests,
>>>>
>>>> but they are not consistent across builds.
>>>>
>>>> Suspect causes are:
>>>>
>>>> My commits of:
>>>>
>>>> http://www.cdash.org/CDash/viewUpdate.php?buildid=456734
>>>>
>>>> * itkUnaryFunctorImageFilter
>>>>
>>>> * itkOptMattesMutualInformationImageToImageMetric.txx
>>>>
>>>>
>>>> Please let me know if you see any other potential
>>>>
>>>> reasons for the failing tests.
>>>>
>>>> Otherwise, I will mostly suspect of a secondary effect
>>>>
>>>> of the change in the itkUnaryFunctorImageFilter...
>>>>
>>>>
>>>>    Thanks
>>>>
>>>>
>>>>          Luis
>>>>
>>>> _______________________________________________
>>>>
>>>> Powered by www.kitware.com
>>>>
>>>> Visit other Kitware open-source projects at
>>>>
>>>> http://www.kitware.com/opensource/opensource.html
>>>>
>>>> Kitware offers ITK Training Courses, for more information visit:
>>>>
>>>> http://kitware.com/products/protraining.html
>>>>
>>>> Please keep messages on-topic and check the ITK FAQ at:
>>>>
>>>> http://www.itk.org/Wiki/ITK_FAQ
>>>>
>>>> Follow this link to subscribe/unsubscribe:
>>>>
>>>> http://www.itk.org/mailman/listinfo/insight-developers
>>>>
>>>> _______________________________________________
>>>> Powered by www.kitware.com
>>>>
>>>> Visit other Kitware open-source projects at
>>>> http://www.kitware.com/opensource/opensource.html
>>>>
>>>> Kitware offers ITK Training Courses, for more information visit:
>>>> http://kitware.com/products/protraining.html
>>>>
>>>> Please keep messages on-topic and check the ITK FAQ at:
>>>> http://www.itk.org/Wiki/ITK_FAQ
>>>>
>>>> Follow this link to subscribe/unsubscribe:
>>>> http://www.itk.org/mailman/listinfo/insight-developers
>>>>
>>>> ========================================================
>>>>
>>>> Bradley Lowekamp
>>>>
>>>> Lockheed Martin Contractor for
>>>>
>>>> Office of High Performance Computing and Communications
>>>>
>>>> National Library of Medicine
>>>>
>>>> blowekamp at mail.nih.gov
>>>>
>>>>
>>>
>>

========================================================
Bradley Lowekamp
Lockheed Martin Contractor for
Office of High Performance Computing and Communications
National Library of Medicine
blowekamp at mail.nih.gov

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20091027/291e52d6/attachment.htm>