[Paraview] [Paraview-developers] Paraview with python problem

Thu Dec 19 09:10:14 EST 2013

Cool. Glad it's working. I am guessing the Windows build of Python
expects things to be under certain locations for it to work without
setting the variables explicitly.

Utkarsh

On Thu, Dec 19, 2013 at 1:40 AM, Lodron, Gerald
<Gerald.Lodron at joanneum.at> wrote:
> I solved the issue:
>
> I only copied the python libs, but python needs two environment variables. When I start paraview with following environment variables it works:
> set PYTHONPATH=D:\Develop\Libraries90\DigPython\2.7.3.3\python-2.7.3\Lib
> set PYTHONHOME=D:\Develop\Libraries90\DigPython\2.7.3.3\python-2.7.3
>
> uff, thanks for help
> best regards, happiest Gerald of the world....
>
> -----Ursprüngliche Nachricht-----
> Von: Lodron, Gerald
> Gesendet: Donnerstag, 19. Dezember 2013 06:49
> An: 'Utkarsh Ayachit'
> Cc: Paraview Developer (Paraview-developers at paraview.org); Paraview User (paraview at paraview.org)
> Betreff: WG: [Paraview-developers] Paraview with python problem
>
> Uh, now i am close in debugging:
>
> It closes in vtkPythonInterpreter::Initialize(); at vtkMatplotlibMathTextUtilities.cxx at 115
>
> I FOUND SOMETHING WITH PYTHON; YEAH
>
> Now I debugged further deeper:
>
> Inside bool vtkPythonInterpreter::Initialize(int initsigs /*=0*/) of vtkPythonInterpreter.cxx the Py_InitializeEx(initsigs); call closes my app, so there is something wrong with my python... I do not have python in debug mode compiled so I am stopping debugging here, but now I know where to look....
>
>
>
> -----Ursprüngliche Nachricht-----
> Von: Lodron, Gerald
> Gesendet: Mittwoch, 18. Dezember 2013 23:03
> An: 'Utkarsh Ayachit'
> Cc: Paraview Developer (Paraview-developers at paraview.org); Paraview User (paraview at paraview.org)
> Betreff: AW: [Paraview-developers] Paraview with python problem
>
> Hi again,
>
> so now i made a completely new empty build directory and i compiled it in 32 and 64 bit in release and debug mode (whereby I only had release libs of python) and all 4 variants close like before after applying any source......
> So I can really say that there is no dll/lib mismatch....
>
> Which python lib are you exactly using, I use one of my colleges which works for them. May there is something wrong inside them....
>
> Best regards,
> Yours tired Gerald.....
>
>
>
>
> -----Ursprüngliche Nachricht-----
> Von: Utkarsh Ayachit [mailto:utkarsh.ayachit at kitware.com]
> Gesendet: Mittwoch, 18. Dezember 2013 17:08
> An: Lodron, Gerald
> Cc: Paraview Developer (Paraview-developers at paraview.org); Paraview User (paraview at paraview.org)
> Betreff: Re: [Paraview-developers] Paraview with python problem
>
> Try a clean build :). I am sure that will resolve this issue (well, 90% sure). Nothing should close the application without exceptions or errors, unless the libraries are mismatched -- which could happen in case of a broken build. I'd strongly suggest doing a clean build.
>
> On Wed, Dec 18, 2013 at 10:55 AM, Lodron, Gerald <Gerald.Lodron at joanneum.at> wrote:
>> So, things seem to come to an end:
>>
>> (I selected the Sphere source and pressd Apply)
>>
>> vtkObjectBase *vtkCubeAxesRepresentationClientServerNewCommand()
>> {
>>   return vtkCubeAxesRepresentation::New(); }
>>
>> Of vtkCubesAxesRepresentationClientServer @ 13
>>
>> Where the New() seems to close my app. I then debugged to the
>> constructor of vtkCubeAxesRepresentation  at
>>
>> this->CubeAxesActor = vtkCubeAxesActor::New(); on line 50
>>
>> and then I came to the constructor of
>> vtkCubeAxesActor.cxx on line 131:
>>
>>     this->XAxes[i] = vtkAxisActor::New();
>>
>>
>> I must now go home and will restart the debugging tomorrow but any idea why the standard New function of vtk could fail?
>>
>> Best regards,
>> Yours
>> Gerald
>>
>>
>> -----Ursprüngliche Nachricht-----
>> Von: Lodron, Gerald
>> Gesendet: Mittwoch, 18. Dezember 2013 16:21
>> An: 'Utkarsh Ayachit'
>> Cc: 'Paraview Developer'; 'Paraview User'
>> Betreff: WG: [Paraview-developers] Paraview with python problem
>>
>> Ok forgot my last mail, it is not the vtkSMProxyManager, it is the
>> vtkSMSessionProxyManager where it jumps in and the name string of
>> name.toAscii().data(); is correct
>>
>> It closes in vtkSMSessionProxyManager:: RegisterProxy at proxy->CreateVTKObjects();, I will debug forward....
>>
>> -----Ursprüngliche Nachricht-----
>> Von: Lodron, Gerald
>> Gesendet: Mittwoch, 18. Dezember 2013 16:14
>> An: 'Utkarsh Ayachit'
>> Cc: Paraview Developer (Paraview-developers at paraview.org); Paraview
>> User (paraview at paraview.org)
>> Betreff: AW: [Paraview-developers] Paraview with python problem
>>
>>
>> It closes in
>>
>> pxm->RegisterProxy("representations", pcTest, reprProxy);
>>
>> at pqObjectBuilder.cxx at line 527, and when I set a breakpoint into
>> registerProxy of vtkSMProxyManager it gets never hit. It seems that
>> the name.toAscii().data() is corrupted, but name is a vaild string
>> "DataRepresentation1". When I write
>>
>> const char* pcTest = name.toAscii().data();
>>   //pxm->RegisterProxy("representations", name.toAscii().data(), reprProxy);
>>   pxm->RegisterProxy("representations", pcTest, reprProxy);
>>
>> and watch at pcTest it is "
>> îþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþ
>> îþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþîþ
>> îþîþîþîþîþîþîþîþîþîþÓ6
>>
>> ¡~ hêè Xlø " so it seems that:
>>
>> Name.toAscii().data() is not allowed in that combination OR The memory is completely corrupted from somewhere else......
>>
>>
>>
>>
>>
>>
>> -----Ursprüngliche Nachricht-----
>> Von: Utkarsh Ayachit [mailto:utkarsh.ayachit at kitware.com]
>> Gesendet: Mittwoch, 18. Dezember 2013 15:52
>> An: Lodron, Gerald
>> Cc: Paraview Developer (Paraview-developers at paraview.org); Paraview
>> User (paraview at paraview.org)
>> Betreff: Re: [Paraview-developers] Paraview with python problem
>>
>> Weird. Let's try this, but a break point in that method. It'll only be called once after you hit "apply". Then step into the code, including createDataRepresentation(..). If you cannot step, I think it's a problem with mismatched/improperly built libraries or something. If the call stack starts making no sense, it's definitely that, and I'd recommend a clean build -- start with an empty binary directory.
>>
>> Utkarsh
>>
>> On Wed, Dec 18, 2013 at 9:45 AM, Lodron, Gerald <Gerald.Lodron at joanneum.at> wrote:
>>> Hi,
>>>
>>> I am currently debugging, it closes (not crashing!) in
>>>
>>> pqDataRepresentation* pqDisplayPolicy::newRepresentation(pqOutputPort* port,
>>>   pqView* view) const
>>> {
>>>   return pqApplicationCore::instance()->getObjectBuilder()->
>>>     createDataRepresentation(port, view); } Of pqDisplayPolicy.cxx
>>>
>>> Here i cannot debug in for any reason (i come somewhere into assembler of qt and when I step out it closes). I will try debugging more ...
>>> PS: I also checked the successful RequestData return of the sources so the source is correctly applied, it closes afterwards...
>>>
>>> Gerald
>>>
>>>
>>> -----Ursprüngliche Nachricht-----
>>> Von: Utkarsh Ayachit [mailto:utkarsh.ayachit at kitware.com]
>>> Gesendet: Mittwoch, 18. Dezember 2013 15:40
>>> An: Lodron, Gerald
>>> Cc: Paraview User (paraview at paraview.org); Paraview Developer
>>> (Paraview-developers at paraview.org)
>>> Betreff: Re: [Paraview-developers] Paraview with python problem
>>>
>>> Gerald,
>>>
>>> Since you compiled from source, can you do debug build and then run through a debugger to see why it's aborting? I don't think I've seen this behaviour before. I wonder if its related to the dependency issue. Was this a clean build or were you doing an incremental build?
>>>
>>> Utkarsh
>>>
>>> On Wed, Dec 18, 2013 at 9:19 AM, Lodron, Gerald <Gerald.Lodron at joanneum.at> wrote:
>>>> Hi
>>>>
>>>>
>>>>
>>>> I successfully compiled paraview 4.1.0 with python 2.7.3 enabled in 32 bit.
>>>> Paraview also starts but when I generate ANY data (over open file or
>>>> over
>>>> sources->Sphere or box or anything else etc.) it closes after
>>>> sources->pressing
>>>> Apply. Does anyone has a clue what could cause that strange behaviour?
>>>>
>>>>
>>>>
>>>> Thanks for advice...
>>>>
>>>>
>>>>
>>>> Best regards
>>>>
>>>> Gerald Lodron
>>>>
>>>>
>>>>
>>>> _________________________________________________________________
>>>>
>>>>
>>>>
>>>> JOANNEUM RESEARCH Forschungsgesellschaft mbH
>>>>
>>>>
>>>>
>>>> DIGITAL - Institute for Information and Communication Technologies
>>>>
>>>> Steyrergasse 17, 8010 Graz, Austria
>>>>
>>>>
>>>>
>>>> phone: +43 316 876-1751
>>>>
>>>> fax: +43 316 8769-1751
>>>>
>>>> e-mail: gerald.lodron at joanneum.at
>>>>
>>>> web: www.joanneum.at/digital
>>>>
>>>> _________________________________________________________________
>>>>
>>>>
>>>>
>>>> This message and any attached files are confidential and intended
>>>> solely for the addressee(s). Any publication, transmission or other
>>>> use of the information by a person or entity other than the intended
>>>> addressee(s) is prohibited. If you receive this in error please
>>>> contact the sender and delete the material. The sender does not
>>>> accept liability for any errors or omissions as a result of the transmission.
>>>>
>>>>
>>>>
>>>> Please consider the environment before printing this page.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Paraview-developers mailing list
>>>> Paraview-developers at paraview.org
>>>> http://public.kitware.com/mailman/listinfo/paraview-developers
>>>>