[Paraview] Extension from Plugin-Reader does not apper - QT 4.4 issue?

[Fabian Wein]

Hi,

I want to write a Reader for Paraview (I guess a plugin is better than
directly
hacking into the Paraview source).

As I have to base on an vtkUnstructuredGridAlgorithm I cloned the
vtkGAMBITReader
as it was the shortest (Only change of name, no functionality yet).
I follwed the Plugin-HowTO and with the help of the XML files
in the examples section and was able to compile a plugin. The
extension I use is .h5
(but we do not want to use the HDF5 reader but our own stuff).
SYSTEM_HDF5 is disabled.

When I load the Plugin no .h5 extension appears in the file open dialog.

I have several (starters) questions:

* Shall the extension appear there?
* I use OpenSuse 11.0 64bit with QT 4.4 - Paraview compiles nice after
commenting
  the WARNING() in the cmake file.
  - Can this be the reason for the problem?
  - What actually is the QT 4.4 issue? It compiles ...
* When I load a plugin again, is it reloaded?
* Is vtkGAMBITReader a good start? I found no appropriate extension in
file open.

I just started with the mailing list and did not find something
appropriate yet.

Thanks a lot,

Fabian

P.S.: I have no experience with ParaView yet (we have first to read
our data :))
___
ParaView mailing list
ParaView@paraview.org
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Ghost cells for standalone Paraview

[Fabian Wein]

Hi Ken,

thanks a lot - you solved the issue! :)

It works fine with the clip filter - memory consumption seems to be in
the green range!

Best regards from ol' Europe,

Fabian

 OK, I understand now.  What you are asking for is not exactly what
 VTK/ParaView considers ghost cells.  You want add a layer of points
 around the boundary of an image data (i.e. 3D uniform regular grid) with
 minimum values to force the contour filter to create closed surfaces.  I
 don’t know of an easy way to do this in ParaView (someone please correct
 me if I am wrong).
 
 I think you can get the effect you want by using the clip by scalars
 feature.  Instead of using the contour filter, use the clip filter.
  Change the “Clip Type” to “Scalar” and set the value to the isosurface
 you want to see.  The advantage of this approach is that the data you
 see will show the surface right at the edge of the volume instead of the
 arbitrary rounding at the edges you see in the Amira screenshots.  The
 disadvantage is that the underlying data is a solid unstructured
 geometry which could take up much more memory than the original image
 data.  If memory is an issue, let me know and I can describe
 memory-light version that uses a combination of the contour, extract
 surface, and clip filters.
___
ParaView mailing list
ParaView@paraview.org
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Questions about Multiblock

[Fabian Wein]

Hi,

I export VTK multiblock files. I have the following questions:
* How can I assign names to the datasets? I googled and found a hint
to use name in
  the pvd file but this is ignored.
* Is it possible to write the blocks to a single file?

* Paraview crashes when I load the pvd file:
VTKFile type=Collection version=0.1
  Collection
   DataSet timestep=1 name=this is a name part=0
file=piezo_0.vtm/
  /Collection
/VTKFile

But loading the vtm file works. Is there some logging about the crash?

* Is it possible to writte several timesteps within one XML file?

Thanks a lot,

Fabian
___
ParaView mailing list
ParaView@paraview.org
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Questions about Multiblock

[Fabian Wein]

Hi Berk,

  * How can I assign names to the datasets? I googled and found a hint
  to use name in
   the pvd file but this is ignored.
  
  This is currently not supported. It shouldn't be too hard to
  implement. You can file a feature request at http://paraview.org/Bug.

I just assigned - thanks for your feedback.

Fabian

___
ParaView mailing list
ParaView@paraview.org
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Current state of LiveData

[Fabian Wein]

Hi,

we have own reader implemented which I want to extend for live data.

I found the comment from 2008 that support for live data ist was planned.

What is the current state, what do I have to do to implement live data
by myself?

I plan to read my data by myself via a thread (a good idea) and then
signal
the new data to paraview.

The key question is, how to signal this to paraview?

Thanks a lot,

Fabian

P.S.: It's kind of urgent :(
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] In which version are the Python GUI tools?

[Fabian Wein]

Hi,

I found
http://www.itk.org/Wiki/Python_GUI_Tools

and also the In-Situ talk.

I just don't have the trace/... buttons in my pathon console
for my Paraview 3.6.1 (Linux, 64-bit)

???

Thanks, Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Current state of LiveData

[Fabian Wein]

Hi David,

 1) You have to recognize when data is available and complete, and you
 need some sort of atomicity guarantee. Ie, you don't want a reader
 trying to read a half written file, and you don't want the some
 process modifying the file while the reader is halfway through reading
 it.
 
 2) The reader has to be marked Modified() when there is something new
 to read. Otherwise the downstream pipeline won't know that what it
 last produced isn't up-to-date, and won't call RequestData() on it.
 
 3) The downstream end (mapper/renderwindow/view/writer or whatever)
 has to be told to Update()/Render() so that the reader's new results
 show up.

Thanks for the information!
I'm working along this guideline, there are several problems I'm fighting
with but this seems exactly to be the information I needed.

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Live Data does update but doesn't refresh

[Fabian Wein]

Thanks to David's help I come to a certain extend to an extension of my
reader to do live updates.

Background:

I want to to in-situ optimization, on request the optimization shall be
restarted with new parameters and I want to see the most current result.

The optimization results are interpreted as a FEM time series where an
interation is expressed by a time step.

* I start reading a time series (0...100)
  and give this back by RequestInformation.

* Assume the users sees time step 2

* No I start new optimization and my reader thread.
 - The thread within my reader reads the current iteration data.
 - I want to force Paraview to display the most current data.
 - The actual time step number/ value displayed within paraview is
   irrelevant for me as long as the image is ok.

* My thread calls Modified() and after that Update()
  - This leads to a call of RequestInformation() and RequestData().
  - Here I'm lost: I tried several things but nothing is repainted.
  - When I increase the timestep in the visualization with the mouse,
the data returned by the last RequestData() is displayed.

But how can i force Paraview to display the new data after RequestData
without my mouse interaction?

Shall I return a different time range within RequestInformation() ?

I'm working on this for two days now - having no progress but I'm sure
it is easy ... * please help ! * :))

Fabian




___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] new thread: Live Data does update but doesn't refresh

[Fabian Wein]

Thanks to David's help I come to a certain extend to an extension of my
reader to do live updates.

Background:

I want to to in-situ optimization, on request the optimization shall be
restarted with new parameters and I want to see the most current result.

The optimization results are interpreted as a FEM time series where an
interation is expressed by a time step.

* I start reading a time series (0...100)
  and give this back by RequestInformation.

* Assume the users sees time step 2

* No I start new optimization and my reader thread.
 - The thread within my reader reads the current iteration data.
 - I want to force Paraview to display the most current data.
 - The actual time step number/ value displayed within paraview is
   irrelevant for me as long as the image is ok.

* My thread calls Modified() and after that Update()
  - This leads to a call of RequestInformation() and RequestData().
  - Here I'm lost: I tried several things but nothing is repainted.
  - When I increase the timestep in the visualization with the mouse,
the data returned by the last RequestData() is displayed.

But how can i force Paraview to display the new data after RequestData
without my mouse interaction?

Shall I return a different time range within RequestInformation() ?

I'm working on this for two days now - having no progress but I'm sure
it is easy ... * please help ! * :))

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Please help!! How to force Render() from a reader

[Fabian Wein]

Dear Kitware experts and users, :)

I'm still fighting with my live data reader and my deadline is
approaching :(

In my vtkMultiBlockDataSetAlgorithm derived reader
I implement

RequestInformation()
RequestData()
and
RequestUpdateExtent()

and call Modfied() and Update() from a thread.
Sofar all works, I trigger RequestUpdateExtent() for a valid
time-value not
read up to now. This time value is request by RequestData() and I return
the new data - there is just no rendering.

I have no idea about how to force it.

Thank you very much!

Fabian Wein

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Python Render() has no effect

[Fabian Wein]

I'm still struggling with my live data reader.

A thread calls Update() on my algorithm. This leada to a RequestData()
and I give back the current data - independend of the requested time step.

It's just not displayed. I tried with python

Render()
GetRenderView().StillRender()
Show()

all has no effect. How can I do this?

Please help, I invested a week in extending my reader but this single
piece
is missing.

Thanks,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] How to invalidate data

[Fabian Wein]

Dear last resource :)

My last idea to make my live data reader work is to use animation.

Independed on what time value is requested I give back the current data.

I'll need a repeated loop for that - but once the time values are read,
they are cached and not requested again from my reader.

How can this be solved? Do I something wrong?

Thanks a lot,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Please help!! How to force Render() from a reader

[Fabian Wein]

Hi Takuya,

 I am not sure what you mean by a thread (a client side code?),
 but perhaps you can simply write
 
   pqApplicationCore::instance()-render();

This has no effect. Meanwhile I think this is because of the client/server
structure.

Update within the reader forces RequestUpdateExtent and RequestData
for the new time value - but it is not dispayed in the time GUI next
to the
VCR area.

A reader is a server part, isn't it? I guess I need also a gui part.

Is there a complete code example anywhere?

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Python Render() has no effect

[Fabian Wein]

Hi,

 You should be able to do it from C++. If you have a
 pqProxy/pqPipelineSource, you can call getProxy() to get the proxy
 object.

Sorry, I don't understand.

I actually have a reader derived from vtkMultiBlockDataSetAlgorithm

 What happens if you call MarkModified(None) on your source proxy?
 This made me research what a source proxy is (I tried to do it without
 python, just with C++) but now it seems that python is the choice.

With python I can do the following within the python shell:

# this is atually my reader
reader = servermanager.sources.CFSreader
view = GetActiveView()
view.ViewTime = 2
view.StillRender()

voila! it works.

But even when it runs in a thread, paraview is blocked.

I finally found a solution. I use the Animation tool. Paraview
is also blocked but so far it is ok.

BTW,
http://www.vtk.org/Wiki/ParaView:AnimationTutorial
seems to be heavily outdated - It was luck that we found the Duration
property.

Anyway - I hope my workaround is sufficient.

Thanks!

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Please help!! How to force Render() from a reader

[Fabian Wein]

Hi Takuya,

 Perhaps you are right in that you need a gui (client) part.
 (PV experts: correct me if I am wrong)
 
 Is there a complete code example anywhere?
 
 Here is what I have in the gui (client) part for the pesudo live
 updating of my time-aware reader (I am using a QTimer for periodically
 polling a running case in a simple way). It's sort of brutal
 implementation though.
 
 If you are intersted in the full version, have a look at
 http://openfoamwiki.net/index.php/Contrib_Parallelized_Native_OpenFOAM_Reader_for_ParaView


Thanks for the support. It helped! But due to time constraints I'll go
with the animator.

Cheers,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] How to invalidate data

[Fabian Wein]

Hi,

The animation idea for in-situ does not work as smooth as required.

How can I mark data modified from my read in C++ such that it is requested
by the animation and time stepping again?

The animation cache cannot be set below 10K.

It would be a great help to know that this is not possible from my
data reader.
Do I need python?

Thanks, Fabian

 My last idea to make my live data reader work is to use animation.
 
 Independed on what time value is requested I give back the current data.
 
 I'll need a repeated loop for that - but once the time values are read,
 they are cached and not requested again from my reader.
 
 How can this be solved? Do I something wrong?
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Communication from Python to own C++ reader

[Fabian Wein]

Hi,

it's me again - I have the impression that nothing that I try really
works :(
But I'm getting closer :)

I get access to my reader from python via
r = GetSources().items()[0][1]

I want to get an integer value from my reader.

The only way I know to do this, is:

*.h
  vtkSetMacro(CurrentPolledStep, int);
  vtkGetMacro(CurrentPolledStep, int);
  int CurrentPolledStep;

and *.xml
 !-- Allow AnimationReader.py to check if there is a new in-situ
result --
 IntVectorProperty  name=CurrentPolledStep
   command=SetCurrentPolledStep
   number_of_elements=1
   default_values=1
   BooleanDomain name=int/
   Documentation
 Current step determined by the reader step.
   /Documentation
 /IntVectorProperty

I actually don't want to have it in the GUI but that's the only way I
found.

It is also displayed as a checkbox in the GUI but according to
classvtkSMBooleanDomain.html
it shall take any int value.

I also have a TimeStep value from a range.

In python I always get the value set in the GUI but not the current
value in my
C++ object.

I call Modified() within C++ after setting and
r.UpdatePipelineInformation()
in python but
r.TimeStep
and
r.CurrentPolledStep
don't change in python.

Any hint?

Thanks,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] How to invalidate data

[Fabian Wein]

Hi David,

 Sorry Fabian, I've been quite busy at Vis and haven't been able to help.
I can understand! For my tight deadline we skipped paraview and
implemented a
simple standalone live-data-visualization tody in Qt for a
presentation next week
- hence I'm not under pressure any more.

But I will need to solve the issue in Paraview later, more relaxed :)

I finally came to such a scenario as you described - I just don't know
(yet)
how to implement it.

In python it seems to be a bad idea as paraview is not responding
while the python
shell is running (even when it waits most of the time on sleep()).

At least from python I also did not find, how to call
RequestInformation() and
RequestUpdateExtent() from my vtkAlgorithm based reader. Do you know?

You ment this methods by standard information property, right?

Writing a reader is not that complicated and most probably all a user
needs
to implement in general.
But a live-data aware animation needs some kind of GUI (start,
pause, stop),
propably it might even allow concurrent user interactions like zooming.
I guess this is rather difficult to implement as plugin?!

Indeed I think it would be best to have an enhanced Animation feature
that does
not only request as fast as possible but polls RequestInformation(),
RequestUpdateExtent().

I believe such an extension would make performing live-data/
realtime-vis/ in-situ
as requested by some users much easier to implement (by just making
the readers sensitive)
but without handling the full VTK pipeline.

I just submitted a feature request: 9724

Thanks for any help/ comments,

Fabian

 Warning, I did this back in the 2.6/3.0 timeframe.
 The animation has changed since then, as has data information
 gathering, so the same path may not still work.
 
 It looks like in the simplest path for live data in paraview it would:
  have the client periodically poll the reader via a standard
 information property
  that property calls a function on the server's root node that:
makes the vtk reader safely check for new data (simplest way is to
 look for new file N, then update to N-1)
update the temporal range/timesteps information
call modified on itself to make the lazy VTK pipeline flow far enough
  an answer of there is new data, max time is now t would mark the
 live reader's source proxy as modified and...
  when in an auto accept mode automatically set the application's time
 to t and call StillRender on the active view


___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] Cannot find SaveData for python scripting

[Fabian Wein]

Hi,

I searched and googled but could not find the python call for SaveData().

I import my data by my own reader and want to export it as VTK file by
scripting.

Any hints? Is there a doxygen documentation or something like that for
the
complete python interface? I just have the examples from the web and
the online
python help.

Thanks,

Fabian

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] Cannot find SaveData for python scripting

[Fabian Wein]

Hi Adriano,

 To save a state: servermanager.SaveState() 
 
 To load a state: servermanager.LoadState()
 
 If there is a command you are looking for that doesn't appear using dir(),
 then try dir(servermanager) as it may be hiding in there somewhere.

Thanks a lot.

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] TecPlot problem or question for alternative format

[Fabian Wein]

Hi,

I got a code which exports data (LBM) in a very simple format:

---
TITLE= LBM data
VARIABLES = X, Y, RHO, U, V, VOR
 ZONE T = Z
 I=  351, J=  122, K=1,  ZONETYPE=ORDERED
 DATAPACKING=POINT
 DT=(SINGLE SINGLE SINGLE SINGLE SINGLE SINGLE)
11  0.000D+00  0.000D+00
0.000D+00  0.000D+00
21  0.000D+00  0.000D+00
0.000D+00  0.000D+00
...
---
The format of point data is: x, y, several scalar data values

I was told this is TecPlot and the guy uses TecPlot for visualization.

I exchanged the format D+ by e+ and opened the file with the CVS
version from 03/25/2010.

The file is recognized by the TecPlot reader but the geometry of the
visualization is not correct.

I understand that TecPlot is quite complicated (I saw just the number
of pages in the data format docu).

I assume that it is more easy for me to convert the simple data to a
structure that PV 3.6.2 can read.
What is the most simple PV format for this?

Thanks,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] TecPlot problem or question for alternative format

[Fabian Wein]

 It seems the Tecplot file does not provide the section of x and y
 corrdinates (before the solution) at all, or you just skipped it at your
 e-mail for simplification purposes?

No, I did not skip anything, this is the data I have. The X and Y
coordinates
are just data values as the following RHO, U, V, ...

So the TecPlot file is not valid?

Which simple file format for 3.6.2 do you suggest to convert the data to?

Thanks,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] TecPlot problem or question for alternative format

[Fabian Wein]

Zhanping,

 Try removing 'K = 1' from the zone header to see it the Tecplot reader
 works. As Berk suggested, you may converted the data to legacy VTK.

Thanks for your reply, it did not work. I'll create a VTK file by
myself, it's
not that much.

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] programmable access to vector components (via Python)

[Fabian Wein]

I have vector data with 6 elements.

The vector is called 'mechTensor' and I can visualize all components 
and the
magnitude where the components are called XX, YY, ZZ, XY, YZ, XZ. Btw, 
this

is the Voigt notation for stresses and strains.

I do not know, where the component names are assigned, they are not 
defined in
the hdf5 file. Maybe this is done in our proprietary reader but I 
could not

find it there up to now.

I cannot access the components in the calculator. Having the equation
mechTensor_XX where the string is selected from the Scalars menue, I 
get the error:


vtkFunctionParser (0x17f8b50): Syntax error: operator expected;  see 
position 12


I guess that the parser has a problem with the double character after 
the underline ?!


Then I tried the Python Calculator. Here I do not know how to access 
the components.


inputs[0].CellData['mechTensor'][0]

Results in

Cell array result with 6 components, has only 1 tuples but there are 
2 cells


I have indeed only two elements, i.e. two times the 6 components vector.

I have a bunch of questions: :)

* am I right with the assumption, that the calculator does not like 
double characters

  after the underline?
* Is this XX, YY, ... coming from Paraview?
* How do I access the vector components via python?
* Where can I find more information about the python interface?

Thanks a lot for your help!

Fabian

P.S.: Same behavior with ParaView 3.10, 3.12 and 3.14
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] programmable access to vector components (via Python)

[Fabian Wein]

Hi Sam,

thanks for your reply.

The ensight file canot be read:

vtkPGenericEnSightReader (0x1e74fc0): Assuming binary file

But as far as I understand, the XX, ... comes from Paraview ?!

Can you apply Stress_XX in the calculator?

I converted my file to VTK and it appears that Paraview adds the XX, ...
BTW, this Voigt interpretation is in my case the wrong interpretation.
My 6-element vector is neither stress nor strain but the upper part of a
3x3 Voigt elasticity tensor.

Also with the VTK file I cannot use the components in the calculator.
This seems to be a Bug ?!

Fabian


On 05/23/2012 04:11 PM, Samuel Key wrote:

Greetings Fabian,



ParaView using its EnSight binary-formatted results reader accepts
symmetric second-order tensor node/cell variables with a user-assigned
name. I have attached an EnSight case file so you can see how it is
done. The EnSight case file is an ASCII-text file that contains a
file table of contents, user-assigned variable names, time-step
values, and directions on the relationship between the files in
order to know what to read when.

I checked one of my EnSight data sets and Stress appears in the
calculator with the scaler components Stress_XX, Stress_XY,
Stress_XZ, Stress_YY, ...

I hope this helps; I know you are not looking to change your results
file format, but I can supply you with Fortran-90/95 routines that
output simulation results in an EnSight Gold binary format.

By the way, the EnSight reader in ParaView supports polyhedral finite
elements -- and it works; I also use the EnSight part-construct to
output the results by-material.

Sam Key

On 5/23/2012 5:28 AM, Fabian Wein wrote:

I have vector data with 6 elements.

The vector is called 'mechTensor' and I can visualize all components
and the
magnitude where the components are called XX, YY, ZZ, XY, YZ, XZ.
Btw, this
is the Voigt notation for stresses and strains.

I do not know, where the component names are assigned, they are not
defined in
the hdf5 file. Maybe this is done in our proprietary reader but I
could not
find it there up to now.

I cannot access the components in the calculator. Having the equation
mechTensor_XX where the string is selected from the Scalars menue,
I get the error:

vtkFunctionParser (0x17f8b50): Syntax error: operator expected; see
position 12

I guess that the parser has a problem with the double character
after the underline ?!

Then I tried the Python Calculator. Here I do not know how to access
the components.

inputs[0].CellData['mechTensor'][0]

Results in

Cell array result with 6 components, has only 1 tuples but there
are 2 cells

I have indeed only two elements, i.e. two times the 6 components
vector.

I have a bunch of questions:

* am I right with the assumption, that the calculator does not like
double characters
after the underline?
* Is this XX, YY, ... coming from Paraview?
* How do I access the vector components via python?
* Where can I find more information about the python interface?

Thanks a lot for your help!

Fabian

P.S.: Same behavior with ParaView 3.10, 3.12 and 3.14


VTKFile type=UnstructuredGrid version=0.1 byte_order=LittleEndian
  UnstructuredGrid
Piece NumberOfPoints=6 NumberOfCells=2
  PointData
DataArray type=UInt32 Name=origNodeNums format=ascii 
RangeMin=1 RangeMax=6
  1 2 3 4 5 6
/DataArray
DataArray type=Float64 Name=mechDisplacement 
NumberOfComponents=3 format=ascii RangeMin=0 RangeMax=6.5143401362
  0 0 0 1.0621182716 -1.540229249 0
  0 0 0 -1.0621182716 -1.540229249 0
  2.6584411185 -5.9472109623 0 -2.6584411185 -5.9472109623 0
/DataArray
  /PointData
  CellData
DataArray type=UInt32 Name=origElemNums format=ascii 
RangeMin=1 RangeMax=2
  1 2
/DataArray
DataArray type=UInt32 Name=elemTypes format=ascii RangeMin=6 
RangeMax=6
  6 6
/DataArray
DataArray type=Float64 Name=mechTensor NumberOfComponents=6 
format=ascii RangeMin=0.12158454416 RangeMax=0.43860303129
  0.4199107952 0.076653611027 0.10085517859 1.9035494669e-17 
1.7654630288e-15 1.8058899887e-18
  0.079614651163 0.079321092678 0.046394751379 4.0167328192e-17 
-1.6973418431e-18 9.7245440946e-21
/DataArray
DataArray type=Float64 Name=mechTensorTrace format=ascii 
RangeMin=0.20533049522 RangeMax=0.59741958481
  0.59741958481 0.20533049522
/DataArray
  /CellData
  Points
DataArray type=Float32 Name=Points NumberOfComponents=3 
format=ascii RangeMin=0 RangeMax=3.6055512755
  0 2 0 1.5 2 0
  0 0 0 1.5 0 0
  3 2 0 3 0 0
/DataArray
  /Points
  Cells
DataArray type=Int64 Name=connectivity format=ascii RangeMin=0 
RangeMax=5
  3 1 0 2 5 4
  1 3
/DataArray
DataArray type=Int64 Name=offsets format=ascii RangeMin=4 
RangeMax=8
  4 8
/DataArray
DataArray type=UInt8 Name=types format=ascii

Re: [Paraview] programmable access to vector components (via Python)

[Fabian Wein]

Hi Sam,

thanks for your reply.

The ensight file canot be read:

vtkPGenericEnSightReader (0x1e74fc0): Assuming binary file

But as far as I understand, the XX, ... comes from Paraview ?!

Can you apply Stress_XX in the calculator?

I converted my file to VTK and it appears that Paraview adds the XX, ...
BTW, this Voigt interpretation is in my case the wrong interpretation.
My 6-element vector is neither stress nor strain but the upper part of a
3x3 Voigt elasticity tensor.

Also with the VTK file I cannot use the components in the calculator.
This seems to be a Bug ?!

Fabian


On 05/23/2012 04:11 PM, Samuel Key wrote:

Greetings Fabian,



ParaView using its EnSight binary-formatted results reader accepts
symmetric second-order tensor node/cell variables with a user-assigned
name. I have attached an EnSight case file so you can see how it is
done. The EnSight case file is an ASCII-text file that contains a
file table of contents, user-assigned variable names, time-step
values, and directions on the relationship between the files in
order to know what to read when.

I checked one of my EnSight data sets and Stress appears in the
calculator with the scaler components Stress_XX, Stress_XY,
Stress_XZ, Stress_YY, ...

I hope this helps; I know you are not looking to change your results
file format, but I can supply you with Fortran-90/95 routines that
output simulation results in an EnSight Gold binary format.

By the way, the EnSight reader in ParaView supports polyhedral finite
elements -- and it works; I also use the EnSight part-construct to
output the results by-material.

Sam Key

On 5/23/2012 5:28 AM, Fabian Wein wrote:

I have vector data with 6 elements.

The vector is called 'mechTensor' and I can visualize all components
and the
magnitude where the components are called XX, YY, ZZ, XY, YZ, XZ.
Btw, this
is the Voigt notation for stresses and strains.

I do not know, where the component names are assigned, they are not
defined in
the hdf5 file. Maybe this is done in our proprietary reader but I
could not
find it there up to now.

I cannot access the components in the calculator. Having the equation
mechTensor_XX where the string is selected from the Scalars menue,
I get the error:

vtkFunctionParser (0x17f8b50): Syntax error: operator expected; see
position 12

I guess that the parser has a problem with the double character
after the underline ?!

Then I tried the Python Calculator. Here I do not know how to access
the components.

inputs[0].CellData['mechTensor'][0]

Results in

Cell array result with 6 components, has only 1 tuples but there
are 2 cells

I have indeed only two elements, i.e. two times the 6 components
vector.

I have a bunch of questions:

* am I right with the assumption, that the calculator does not like
double characters
after the underline?
* Is this XX, YY, ... coming from Paraview?
* How do I access the vector components via python?
* Where can I find more information about the python interface?

Thanks a lot for your help!

Fabian

P.S.: Same behavior with ParaView 3.10, 3.12 and 3.14


VTKFile type=vtkMultiBlockDataSet version=1.0 byte_order=LittleEndian
  vtkMultiBlockDataSet
DataSet index=0 file=vtk_1/vtk_1_0_0.vtu
  
/DataSet

  /vtkMultiBlockDataSet
/VTKFile
VTKFile type=UnstructuredGrid version=0.1 byte_order=LittleEndian
  UnstructuredGrid
Piece NumberOfPoints=6 NumberOfCells=2
  PointData
DataArray type=UInt32 Name=origNodeNums format=ascii 
RangeMin=1 RangeMax=6
  1 2 3 4 5 6
/DataArray
DataArray type=Float64 Name=mechDisplacement 
NumberOfComponents=3 format=ascii RangeMin=0 RangeMax=6.5143401362
  0 0 0 1.0621182716 -1.540229249 0
  0 0 0 -1.0621182716 -1.540229249 0
  2.6584411185 -5.9472109623 0 -2.6584411185 -5.9472109623 0
/DataArray
  /PointData
  CellData
DataArray type=UInt32 Name=origElemNums format=ascii 
RangeMin=1 RangeMax=2
  1 2
/DataArray
DataArray type=UInt32 Name=elemTypes format=ascii RangeMin=6 
RangeMax=6
  6 6
/DataArray
DataArray type=Float64 Name=mechTensor NumberOfComponents=6 
format=ascii RangeMin=0.12158454416 RangeMax=0.43860303129
  0.4199107952 0.076653611027 0.10085517859 1.9035494669e-17 
1.7654630288e-15 1.8058899887e-18
  0.079614651163 0.079321092678 0.046394751379 4.0167328192e-17 
-1.6973418431e-18 9.7245440946e-21
/DataArray
DataArray type=Float64 Name=mechTensorTrace format=ascii 
RangeMin=0.20533049522 RangeMax=0.59741958481
  0.59741958481 0.20533049522
/DataArray
  /CellData
  Points
DataArray type=Float32 Name=Points NumberOfComponents=3 
format=ascii RangeMin=0 RangeMax=3.6055512755
  0 2 0 1.5 2 0
  0 0 0 1.5 0 0
  3 2 0 3 0 0
/DataArray
  /Points
  Cells
DataArray type=Int64 Name=connectivity format=ascii RangeMin=0 
RangeMax=5
  3 1 0 2 5 4

Re: [Paraview] programmable access to vector components (via Python)

[Fabian Wein]

It appears to me you have found a limitation (or bug?) in the
Calculator filter. My comments are inserted below.


I just submitted a bug report: http://paraview.org/Bug/view.php?id=13204


files, however, I have never used it. I graduated from VTK
XML-structured ASCII-text results files to the EnSight
binary-formatted specification in order to get compact, manageable
file-sets for my modestly large explicit transient dynamic simulations.


We are happy with HDF5 but this requires a own reader plugin for
our format which we wrote some years ago. With HDF5 we have also access
via Matlab, with is a plus for us.


If the error message reflects a limitation in the Calculator, then
the Calculator only processes scalars and vectors (1st-order tensors),
and that in itself is no small accomplishment.


My mechTensor is just a vector with 6 components, so the limitation is
with the size of the vector.

This answers two of my questions:


* am I right with the assumption, that the calculator does not like double 
characters after the underline?
* Is this XX, YY, ... coming from Paraview?


Anyone who can help me with the following two?


* How do I access the vector components via python?
* Where can I find more information about the python interface?


Thanks a lot,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Re: [Paraview] programmable access to vector components (via Python)

[Fabian Wein]

@Utkarsh: I agree, it seems to be the same bug

@Berk: Thank you very much, that pushed me to solve the problem.

Currently it does not work work with our own filter (CellData is not 
known) but

when I export to vtk and read it works.

Currently I still work on the details of the python algorithm but in 
principal it works.


Anyway, any hint on what is missing for our own plugin reader?

Fabian


I think this might be same as the bug: http://paraview.org/Bug/view.php?id=12951

Utkarsh

On Fri, May 25, 2012 at 4:13 PM, Berk Geveciberk.gev...@kitware.com  wrote:

Which version of ParaView is this? This bug sounds very familiar to me and I
vaguely remember sitting down with Utkarsh to fix it. If it is an older
version, can you try 3.14?

As for the Python notation, take a look at:

http://paraview.org/Wiki/ParaView/Users_Guide/Python_Programmable_Filter
http://paraview.org/Wiki/ParaView/Users_Guide/Python_Calculator

You need to use something like: inputs[0].CellData['mechTensor'][:, n]
where n is whichever component that you are accessing. This notation is
based on Numpy. You can learn more about it here:

Guide to NumPy - Trelgol

Note that NumPy is much more powerful and efficient that the Array
Calculator so I'd recommend using the Python Calculator or the Python
Programmable Filter whenever possible.


-berk


On Fri, May 25, 2012 at 10:03 AM, Fabian Wein
fabian.w...@am.uni-erlangen.de  wrote:


It appears to me you have found a limitation (or bug?) in the
Calculator filter. My comments are inserted below.



I just submitted a bug report: http://paraview.org/Bug/view.php?id=13204



files, however, I have never used it. I graduated from VTK
XML-structured ASCII-text results files to the EnSight
binary-formatted specification in order to get compact, manageable
file-sets for my modestly large explicit transient dynamic simulations.



We are happy with HDF5 but this requires a own reader plugin for
our format which we wrote some years ago. With HDF5 we have also access
via Matlab, with is a plus for us.



If the error message reflects a limitation in the Calculator, then
the Calculator only processes scalars and vectors (1st-order tensors),
and that in itself is no small accomplishment.



My mechTensor is just a vector with 6 components, so the limitation is
with the size of the vector.

This answers two of my questions:



* am I right with the assumption, that the calculator does not like
double characters after the underline?
* Is this XX, YY, ... coming from Paraview?



Anyone who can help me with the following two?



* How do I access the vector components via python?
* Where can I find more information about the python interface?



Thanks a lot,

Fabian

___
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at:
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview




___
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at:
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview


___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

[Paraview] superbuild fails on OS X

[Fabian Wein]

I have the latest OS X and qt4 and qt5 installed by brew.

I use the current superbuild, which covers PV 4.4 (why not 5.0?) but build 
fails.

I use the default setting in my build and just set
CMAKE_INSTALL_PREFIX and ENABLE_paraview

make 
…
-- Installing: 
/Users/fwein/code/ParaViewSuperbuild/build/paraview/src/paraview-build/CMakeFiles/__macos_install/plugins/libPacMan.dylib
-- Installing: 
/Users/fwein/code/ParaViewSuperbuild/build/paraview/src/paraview-build/CMakeFiles/__macos_install/bin/paraview-config
-- Installing: 
/Users/fwein/code/ParaViewSuperbuild/build/paraview/src/paraview-build/CMakeFiles/__macos_install/bin/pvserver
-- Installing: 
/Users/fwein/code/ParaViewSuperbuild/build/paraview/src/paraview-build/CMakeFiles/__macos_install/bin/pvdataserver
-- Installing: 
/Users/fwein/code/ParaViewSuperbuild/build/paraview/src/paraview-build/CMakeFiles/__macos_install/bin/pvrenderserver
[ 62%] Completed 'paraview'
[100%] Built target paraview

Note there are only four files in bin

make install
….
[100%] Built target paraview
Install the project...
-- Install configuration: ""
CMake Error at cmake_install.cmake:39 (file):
  file INSTALL cannot find
  
"/Users/fwein/code/ParaViewSuperbuild/build/install/Applications/paraview.app".
make: *** [install] Error 1

Any suggestion or hint where to look in the log files is appreciated.

Thanks,

Fabian

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] superbuild fails on OS X [solved]

[Fabian Wein]

I found it: Qt needs to be enabled!

ENABLE_qt4 results in the error below, with USE_SYSTEM_qt4 (installed with 
brew) it works!

While I understand that there is no gui without qt and that it makes sense to 
build paraview without gui, I suggest to make it 
a little clearer:

The help for ENABLE_qt* is just „Request to build project qt*“. Maybe an „Qt is 
necessary for the gui“ could be added. Also the error message from make install 
could be enriched with a check for Qt and a message.


/Users/fwein/code/ParaViewSuperbuild/build/qt4/src/qt4/src/gui/painting/qpaintengine_mac.cpp:345:19:
 error: use of undeclared identifier 'CMGetProfileByAVID'
CMError err = CMGetProfileByAVID((CMDisplayIDType)displayID, 
);
  ^
/Users/fwein/code/ParaViewSuperbuild/build/qt4/src/qt4/src/gui/painting/qpaintengine_mac.cpp:348:9:
 error: use of undeclared identifier 'CMCloseProfile'
CMCloseProfile(displayProfile);
^
In file included from 
/Users/fwein/code/ParaViewSuperbuild/build/qt4/src/qt4/src/gui/painting/qpaintengine_mac.cpp:44:
In file included from ../../include/QtGui/private/qpaintengine_mac_p.h:1:
../../include/QtGui/private/../../../../qt4/src/gui/painting/qpaintengine_mac_p.h:252:9:
 warning: private field 'mBytesPerLine' is not used [-Wunused-private-field]
int mBytesPerLine;
^
3 warnings and 2 errors generated.
make[5]: *** [.obj/release-shared/qpaintengine_mac.o] Error 1
make[4]: *** [release-install] Error 2
make[3]: *** [sub-gui-install_subtargets-ordered] Error 2
CMake Error at /Users/fwein/code/ParaViewSuperbuild/build/pv-qt4-build.cmake:33 
(message):
  Failed with exit code 2



> Am 18.01.2016 um 10:25 schrieb Fabian Wein <fabian.w...@fau.de>:
> 
> I have the latest OS X and qt4 and qt5 installed by brew.
> 
> I use the current superbuild, which covers PV 4.4 (why not 5.0?) but build 
> fails.
> 
> I use the default setting in my build and just set
> CMAKE_INSTALL_PREFIX and ENABLE_paraview
> 
> make 
> …
> [100%] Built target paraview
> 
> make install
> ….
> [100%] Built target paraview
> Install the project...
> -- Install configuration: ""
> CMake Error at cmake_install.cmake:39 (file):
>  file INSTALL cannot find
>  
> "/Users/fwein/code/ParaViewSuperbuild/build/install/Applications/paraview.app".
> make: *** [install] Error 1
> 
> Any suggestion or hint where to look in the log files is appreciated.

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] Superbild for 5.0.0

[Fabian Wein]

When will the superbuild for 5.0.0 be available?

Thanks and regards,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] qt5 issues on OSX

[Fabian Wein]

Hello,

I have an up-to-date OS X system (10.11) and try to compile PVSB by myself 
using the system clang compiler. 
I use psvb v5.1.2 and only enable paraview and qt. I compile with export 
VERBOSE=1

It is known, that qt4 does not compile any more with clang on OSX.

The standard qt5 version in versions.cmake is 5.4.1and results in a compile 
error (see below)

Changing qt5 in versions.cmake to the current 5.7.0 requires to remove 
-skip qtquick1 
-skip qtwebkit
from qt5.cmake.

It results in another compile error (see below) which is strange as qt5.7.0 
compiles if I just compile the sources (with default configure and the 
configure from pvsb, see also below). 

According to 
https://forum.qt.io/topic/64255/compile-errors-updating-my-osx-and-qt/6 
one should configure -std=c++11 instead of -std=c++1z which I didn’t manage yet.

With the latest qt 5.6 I made the same observations as with qt 5.7.0.

With qt 5.5.1 ist works.

However I was not able to compile qt 5.5.1 on Linux with gcc 6.1 (but had no 
issues with 5.7.0 beside the two skips to be removed). So I’ll use 5.5.1 for 
OSX with clang and 5.7.0 with gcc 6.1 on Linux.

Anyway, psvb doesn’t compile on OS X 10.11 with standard clang but with the 
proper change in versions.cmake for qt5 it works.



Error of standard qt 5.4.1 within pvsb:

Users/fwein/tmp/pvsb_qt5_org/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/src/tools/qlalr/lalr.cpp:241:1:
 error: declaration of anonymous struct must be a definition
struct _Nullable: public std::unary_function
^
/Users/fwein/tmp/pvsb_qt5_org/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/src/tools/qlalr/lalr.cpp:241:56:
 error: expected unqualified-id
struct _Nullable: public std::unary_function
   ^
/Users/fwein/tmp/pvsb_qt5_org/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/src/tools/qlalr/lalr.cpp:303:98:
 error: expected expression
  NameList::iterator nn = std::find_if (rule->rhs.begin (), 
rule->rhs.end (), std::not1 (_Nullable (this)));

 ^
/Users/fwein/tmp/pvsb_qt5_org/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/src/tools/qlalr/lalr.cpp:638:107:
 error: expected expression
  NameList::iterator first_not_nullable = std::find_if (dot, 
rule->rhs.end (), std::not1 (_Nullable (this)));

  ^
4 errors generated.



Error of qt 5.7.0 within pvsb:

/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/clang++
 -c -fPIC -arch x86_64 -mmacosx-version-min=10.11 
--sysroot=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.11.sdk
 -ffunction-sections -O2 -fPIC -std=c++1z -isysroot 
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.11.sdk
 -mmacosx-version-min=10.8 -fno-exceptions -Wall -W -DQT_NO_MTDEV 
-DQT_NO_LIBUDEV -DQT_NO_EVDEV -DQT_NO_TSLIB -DQT_NO_LIBINPUT -DQT_BOOTSTRAPPED 
-DQT_LITE_UNICODE -DQT_NO_CAST_TO_ASCII -DQT_NO_CODECS -DQT_NO_DATASTREAM 
-DQT_NO_LIBRARY -DQT_NO_QOBJECT -DQT_NO_SYSTEMLOCALE -DQT_NO_THREAD 
-DQT_NO_UNICODETABLES -DQT_NO_USING_NAMESPACE -DQT_NO_DEPRECATED 
-DQT_NO_TRANSLATION -DQT_CRYPTOGRAPHICHASH_ONLY_SHA1 -DQT_NO_FOREACH 
-DQT_NO_CAST_FROM_ASCII -DQT_BUILD_BOOTSTRAP_LIB -DQT_BUILDING_QT 
-DQT_ASCII_CAST_WARNINGS -DQT_MOC_COMPAT -DQT_USE_QSTRINGBUILDER 
-DQT_DEPRECATED_WARNINGS -DQT_DISABLE_DEPRECATED_BEFORE=0x05 
-DQT_NO_EXCEPTIONS -D_LARGEFILE64_SOURCE -D_LARGEFILE_SOURCE -DQT_NO_DEBUG 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/src/tools/bootstrap
 -I. 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include/QtCore
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include/QtCore/5.7.0
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include/QtCore/5.7.0/QtCore
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include/QtXml
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include/QtXml/5.7.0
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/include/QtXml/5.7.0/QtXml
 -I../../../include -I../../../include/QtCore 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5-build/qtbase/include/QtXml
 
-I/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/mkspecs/macx-clang
 -o .obj/qlogging.o 
/Users/fwein/tmp/pvsb_qt5_7_0/ParaViewSuperbuild/build/qt5/src/qt5/qtbase/src/corelib/global/qlogging.cpp
In file included from 

Re: [Paraview] Superbuild repo change?

[Fabian Wein]

> I did a fresh pull of the ParaView superbuild repository, and now I can’t do 
> a clean build of PV 5.2.0 on my Mac (OSX 10.9.5).   
> 
> First, I had to remove the entries for “qt” and “scipy“ in the CMakeLists.txt:
..

> After removing those entries, the cmake completes and the build starts and 
> passes out in qt4.   
> I’m not sure why it’s complaining about OS X 10.7 or later?)

For recent OS X qt4 is known to not compile. Use qt5 and it works (macOS 10.12)

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] Strange error building paraview-superbuild als nightly build

[Fabian Wein]

Hello,

I have a strange error. The issue is not very important, just a nice 
to have.


I have a simple project building paraview-superbuild as external 
project which includes applying two simple patches
(add building of the libs boost_filesystem and hdf5_cpp) and our 
external plugin.


When run the project manually (cmake ..) it works reproducible 
perfectly smooth on several Linux machines and Mac.


When I run the project via ctest for a nightly build I have errors 
building qt5. When I do 'make' in my project or
the paraview-superbuild project all runs without any notice of a 
error, simply in install/bin no paraview stuff built.


I diffed Testing/Temporary/LastBuild_20170118-1244.log with the output 
from my manual build.
All is more or less the same but sometimes differently ordered due to 
parallal building (which I don't trigger actively)


The real difference start in the ctest log in line ~36500
with
-
[ 88%] Performing install step for 'qt5'
In file included from 
/home/fwein/code/cfs_paraview_build/build/superbuild/qt5/src/qtbase/src/corelib/tools/qlocale_tools.cpp:42:0:
/home/fwein/code/cfs_paraview_build/build/superbuild/qt5/src/qtbase/src/corelib/tools/qdoublescanprint_p.h:151:53: 
fatal error: double-conversion/double-conversion.h: No such file or 
directory

 #include 
 ^
compilation terminated.
g++: error: .obj/qlocale_tools.o: No such file or directory
mv: cannot stat 'libQt5Core.so.5.7.1': No such file or directory
install: cannot stat '../../lib/libQt5Core.so.5.7.1': No such file or 
directory
strip: 
'/home/fwein/code/cfs_paraview_build/build/install/lib/libQt5Core.so.5.7.1': 
No such file
/home/fwein/code/cfs_paraview_build/build/superbuild/qt5/src/qtbase/src/network/kernel/qnetworkproxy_libproxy.cpp:47:19: 
fatal error: proxy.h: No such file or directory

 #include 
   ^
compilation terminated.
g++: error: .obj/qnetworkproxy_libproxy.o: No such file or directory
mv: cannot stat 'libQt5Network.so.5.7.1': No such file or directory
...


In the manual built it is

[ 88%] Performing install step for 'qt5'
strip:/home/fwein/code/cfs_paraview_build/build/install/bin/fixqt4headers.pl: 
File format not recognized
strip:/home/fwein/code/cfs_paraview_build/build/install/bin/syncqt.pl: 
File format not recognized
Some of the required modules 
(android|ios|winrt|osx_webview_experimental|qtHaveModule(webengine)) 
are not available.

Skipped.
[ 90%] Completed 'qt5'
[ 90%] Built target qt5
Scanning dependencies of target paraview
[ 90%] Creating directories for 'paraview'
[ 91%] Performing download step (download, verify and extract) for 
'paraview'

-- verifying file...
...
--

I cannot see how this is caused by my ctest script
-
SET(CTEST_SOURCE_DIRECTORY "$ENV{HOME}/code/cfs_paraview")
SET(CTEST_BINARY_DIRECTORY "$ENV{HOME}/code/cfs_paraview_build")
SET(CTEST_CMAKE_GENERATOR "Unix Makefiles")
SET(CTEST_PROJECT_NAME "Paraview metabuild")

set(CTEST_SITE "eamc061")
set(CTEST_BUILD_NAME "Paraview 5.2 metabuild")
set(BUILDNAME "Paraview 5.2 metabuild Buildname what for?")

# somehow CTEST_START_WITH_EMPTY_BINARY_DIRECTORY does not work?! So 
do it also manually

file(REMOVE_RECURSE "${CTEST_BINARY_DIRECTORY}")
file(MAKE_DIRECTORY "${CTEST_BINARY_DIRECTORY}")
SET(CTEST_START_WITH_EMPTY_BINARY_DIRECTORY TRUE)

SET(BUILDTYPE "RELEASE")

message("Start dashboard...")
ctest_start(Experimental)

message("  Update")
find_program(CTEST_SVN_COMMAND NAMES svn)
set(CTEST_UPDATE_COMMAND "${CTEST_SVN_COMMAND}")
ctest_update(SOURCE "${CTEST_SOURCE_DIRECTORY}" RETURN_VALUE res)

message("  Configure")
ctest_configure(BUILD "${CTEST_BINARY_DIRECTORY}" RETURN_VALUE res)

message("  Build")
ctest_build(BUILD "${CTEST_BINARY_DIRECTORY}" RETURN_VALUE res)

message("  Test")
ctest_test(BUILD "${CTEST_BINARY_DIRECTORY}/build" RETURN_VALUE res)

message("  Submit")
-

The CMakeCache.txt in my build directory and the paraview-superbuild 
build directory are identical


Anyone has any idea? I can live without nightly builds but it would be 
nice for me to understand.


Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] packing superbuild 5.2 fails on openSUSE due to ldconfig

[Fabian Wein]

ctest -R cpack fails on openSUSE but works on Ubuntu

CMake Error at 
/home/fwein/code/cfs_paraview/metabuild/build/cpack/paraview/TGZ/build/cmake_install.cmake:46 
(message):

  Failed to install pvdataserver:

  Traceback (most recent call last):

File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 507, in 

  main(sys.argv[1:])
File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 499, in main
  _install_binary(main_exe, is_excluded, bundle_dest, 
opts.libdir, installed, manifest, dry_run=opts.dry_run)
File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 403, in _install_binary

  deps = binary.dependencies.values()
File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 198, in dependencies

  deplib = Library.create_from_reference(dep, self)
File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 259, in create_from_reference

  paths.extend(cls.default_search_paths())
File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 231, in default_search_paths

  cls.__search_cache = pipe().split('\n')
File 
"/home/fwein/code/cfs_paraview/metabuild/src/paraview-superbuild/superbuild/cmake/scripts/fixup_bundle.unix.py", 
line 23, in __call__
  command = subprocess.Popen(command_args, stdin=last_input, 
stdout=subprocess.PIPE)

File "/usr/lib64/python2.7/subprocess.py", line 390, in __init__
  errread, errwrite)
File "/usr/lib64/python2.7/subprocess.py", line 1024, in 
_execute_child

  raise child_exception

  OSError: [Errno 2] No such file or directory

The reason is that fixup_bundle.unix.py:23 is called with

command_args=['ldconfig', '-v', '-N', '-X']

ldconfig is on openSUSE and Ubuntu /sbin/ldconfig but sbin in on 
openSUSE not in the user path. For me a link in $HOME/bin helped but 
maybe using the full path

would work on all unix systems?!

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] Strange error building paraview-superbuild als nightly build

[Fabian Wein]

thank you very much for the hint. I solved it via
set(CTEST_BUILD_COMMAND „/usr/bin/make")
in the .ctest file.

Fabian

> Am 19.01.2017 um 17:26 schrieb Ben Boeckel <ben.boec...@kitware.com>:
> 
> On Thu, Jan 19, 2017 at 11:35:30 +0100, Fabian Wein wrote:
>> I have a strange error. The issue is not very important, just a nice 
>> to have.
> 
> I suspect this is CTest's fault. Could you see if there is a `-i` flag
> in your DartConfiguration.tcl (or anywhere else in the cache) associated
> with the `make` command? The fix is to pass
> `-DMAKE_COMMAND:STRING=/path/to/make` to CMake during the configure,
> otherwise it "helpfully" passes `-i` (act like errors don't happen)
> which causes these kinds of problems. I suspect that Qt's configure is
> passing things it should not be passing and then setting preprocessor
> definitions it should not be.
> 
> Note that it may be helpful to use an external Qt using the
> `standalone-qt` project in `superbuild/standalone-qt` so that you do not
> need to build it every night (it changes infrequently enough that this
> isn't too much of a hassle).
> 
> --Ben



Dr. Fabian Wein
ZISC - Zentralinstitut für wissenschaftliches Rechnen
Universität Erlangen-Nürnberg 
D-91052 Erlangen, Germany
fabian.w...@fau.de

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] PV 5.3-RC.2 build error on macOS 10.12

[Fabian Wein]

I raised the following issue: 
https://gitlab.kitware.com/paraview/paraview-superbuild/issues/47

First I build the current superbuild master on macOS 10.12 which built PV 5.2 
with qt 5.8.0 without problems. Then I switched the source to PV 5.3-RC.2 and 
got the following error

 89%] Building CXX object 
Qt/Core/CMakeFiles/pqCore.dir/moc_pqRepresentation.cpp.o
/Users/fwein/code/cfs_paraview/manual_53/build/superbuild/paraview/build/Qt/Core/moc_pqQVTKWidget.cpp:85:8:
 error: use of undeclared identifier 'QVTKWidget'; did you mean 'pqQVTKWidget'?
{ ::staticMetaObject, qt_meta_stringdata_pqQVTKWidget.data,
   ^~
   pqQVTKWidget

Regards,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] PV 5.3-RC.2 build error on macOS 10.12

[Fabian Wein]

Works with a clean build, should have checked by myself :(

Regards,

Fabian

> Am 26.02.2017 um 15:51 schrieb Utkarsh Ayachit <utkarsh.ayac...@kitware.com>:
> 
> Is this a clean build? I'd suggest doing  a clean build. I suspect the
> moc file didn't get regenerated or something like that.
> 
> On Sun, Feb 26, 2017 at 4:25 AM, Fabian Wein <fabian.w...@fau.de> wrote:
>> I raised the following issue: 
>> https://gitlab.kitware.com/paraview/paraview-superbuild/issues/47
>> 
>> First I build the current superbuild master on macOS 10.12 which built PV 
>> 5.2 with qt 5.8.0 without problems. Then I switched the source to PV 
>> 5.3-RC.2 and got the following error
>> 
>> 89%] Building CXX object 
>> Qt/Core/CMakeFiles/pqCore.dir/moc_pqRepresentation.cpp.o
>> /Users/fwein/code/cfs_paraview/manual_53/build/superbuild/paraview/build/Qt/Core/moc_pqQVTKWidget.cpp:85:8:
>>  error: use of undeclared identifier 'QVTKWidget'; did you mean 
>> 'pqQVTKWidget'?
>>{ ::staticMetaObject, qt_meta_stringdata_pqQVTKWidget.data,
>>   ^~
>>   pqQVTKWidget
>> 
>> Regards,
>> 
>> Fabian
>> ___
>> Powered by www.kitware.com
>> 
>> Visit other Kitware open-source projects at 
>> http://www.kitware.com/opensource/opensource.html
>> 
>> Please keep messages on-topic and check the ParaView Wiki at: 
>> http://paraview.org/Wiki/ParaView
>> 
>> Search the list archives at: http://markmail.org/search/?q=ParaView
>> 
>> Follow this link to subscribe/unsubscribe:
>> http://public.kitware.com/mailman/listinfo/paraview

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] ParaView 5.2 Release Candidate 2 available for download

[Fabian Wein]

Are you interested in superbuild tests, too? Or do you want to skip that issue 
for the time 5.2 is released?

On macOS 10.12 Sierra I have problems building qt5:

superbuild fails with the output below but building qt5 manually seems to work 
(at least it builds for a longer time up to now)

On Linux (openSUSE tumbleweed with gcc 6.2.1) hdf5 fails the rest is still 
building …


/Users/fwein/tmp/pvsb/superbuild/qt5/src/qtbase/configure -top-level 
-opensource -confirm-license -release -prefix /Users/fwein/tmp/pvsb/install -I 
/Users/fwein/tmp/pvsb/install/include -L /Users/fwein/tmp/pvsb/install/lib 
-skip qtconnectivity -skip qtlocation -skip qtmultimedia -skip qtsensors -skip 
qtserialport -skip qtsvg -skip qtwayland -skip qtwebchannel -skip qtwebengine 
-skip qtwebsockets -nomake examples -nomake tests -no-dbus -qt-libjpeg -qt-pcre 
-no-alsa -no-pulseaudio -system-zlib -no-openssl -no-alsa -no-pulseaudio 
-system-zlib -no-openssl -sdk 
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.12.sdk
 -qt-libpng -sdk 
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.12.sdk
 -qt-libpng

This is the Qt Open Source Edition.

You are licensed to use this software under the terms of
the GNU Lesser General Public License (LGPL) versions 3.
You are also licensed to use this software under the terms of
the GNU General Public License (GPL) versions 2.

You have already accepted the terms of the Open Source license.

Preparing build tree...
Creating qmake...
Done.
Running configuration tests...
Failed to process makespec for platform 'macx-clang'
Turn on verbose messaging (-v) to see the final report.
CMake Error at /Users/fwein/tmp/pvsb/superbuild/sb-qt5-configure.cmake:45 
(message):
  Failed with exit code 101
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] ParaView 5.2 Release Candidate 2 available for download

[Fabian Wein]

Did not work:

-DCMAKE_OSX_SDK=macosx10.12

in cmake ../paraview-superbuild -DENABLE_cxx11=ON -DENABLE_hdf5=ON 
-DENABLE_qt5=ON -Dsuperbuild_download_location=~/tmp/pvsb_downloads 
-DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_OSX_DEPLOYMENT_TARGET=10.12 
-DCMAKE_OSX_SYSROOT=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.12.sdk
 -DCMAKE_OSX_SDK=macosx10.12

results in 

erforming build step for 'qt5'
Info: creating cache file 
/Users/fwein/tmp/pvsb/superbuild/qt5/build/qtbase/.qmake.cache
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib:
 file: ../../../lib/libQt5Bootstrap.a(qvector.o) has no symbols
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib:
 file: ../../../lib/libQt5Bootstrap.a(qvector.o) has no symbols
clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
deployment target of OS X 10.9
clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
deployment target of OS X 10.9
Makefile:13164: warning: overriding commands for target 
`.moc/qeventdispatcher_cf.moc'
Makefile:12994: warning: ignoring old commands for target 
`.moc/qeventdispatcher_cf.moc'
clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
deployment target of OS X 10.9
Undefined symbols for architecture x86_64:
  "std::__1::__vector_base_common::__throw_length_error() const", 
referenced from:
  void std::__1::vector<QMetaMethodBuilderPrivate, 
std::__1::allocator 
>::__push_back_slow_path(QMetaMethodBuilderPrivate&&)
 in qmetaobjectbuilder.o
  void std::__1::vector<QMetaPropertyBuilderPrivate, 
std::__1::allocator 
>::__push_back_slow_path(QMetaPropertyBuilderPrivate&&)
 in qmetaobjectbuilder.o
  void std::__1::vector<QMetaEnumBuilderPrivate, 
std::__1::allocator 
>::__push_back_slow_path(QMetaEnumBuilderPrivate&&) in 
qmetaobjectbuilder.o
  "void std::__1::__sort<std::__1::__less<int, int>&, int*>(int*, int*, 
std::__1::__less<int, int>&)", referenced from:
  QString::multiArg(int, QString const**) const in qstring.o
  QSortFilterProxyModelPrivate::_q_sourceHeaderDataChanged(Qt::Orientation, 
int, int) in qsortfilterproxymodel.o
  QSortFilterProxyModel::removeRows(int, int, QModelIndex const&) in 
qsortfilterproxymodel.o
  "std::bad_alloc::bad_alloc()", referenced from:
  qBadAlloc() in qglobal.o
  QVector<QList >::reallocData(int, int, 
QFlags) in qglobal.o
ld: symbol(s) not found for architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[6]: *** [../../lib/QtCore.framework/QtCore] Error 1
make[5]: *** [sub-corelib-make_first] Error 2
make[4]: *** [sub-src-make_first] Error 2
make[3]: *** [module-qtbase-make_first] Error 2
CMake Error at /Users/fwein/tmp/pvsb/superbuild/sb-qt5-build.cmake:43 (message):
  Failed with exit code 2

I have the impressions that my other issues and comments on the gitlab issue 
tracker are removed, so I stop posting reports of attempts there.



> CMAKE_OSX_SDK needs to be something like `macosx10.9`.  I forget the command 
> to make XCode list the valid sdk versions available.  For your case I would 
> guess `macosx10.12` is the correct value.  Qt looks up the path to the SDK 
> from that string.
> 
> HTH,
> Shawn
> 
> On Fri, Oct 21, 2016 at 7:07 AM, Fabian Wein <fabian.w...@fau.de> wrote:
> I added a lot of the stuff to the issue tracker.
> 
> I’m currently lost with building qt5 on macOS 10.12. It’s a qmake issue, I 
> tracked it down to
> 
> Creating qmake... Running configuration tests... Failed to process makespec 
> for platform ‚macx-clang' Info: creating stash file 
> /Users/fwein/tmp/pvsb/superbuild/qt5/build/qtbase/.qmake.stash Project ERROR: 
> QMAKE_MAC_SDK can only contain short-form SDK names (eg. macosx, iphoneos) 
> CMake Error at /Users/fwein/tmp/pvsb/superbuild/sb-qt5-configure.cmake:43 
> (message):
> 
> And it has a connection to the CMAKE_OSX_SDK I do not know what to set. 
> CMAKE_OSX_SDK=macosx does not help.
> 
> Any qt5 configuration expert reading?
> 
> > Am 20.10.2016 um 17:18 schrieb Ben Boeckel <ben.boec...@kitware.com>:
> >
> > On Thu, Oct 20, 2016 at 15:13:51 +0300, Gena Bug via ParaView wrote:
> >> Now, with RC2, fontconfig warns me about blank section:
> >>
> >> Fontconfig warning: line 160: blank doesn't take any effect anymore.
> >> please remove it from your fonts.conf
> >
> > We're now using a newer fontconfig, so the fontconfig may be newer than
> > your system, so this is just fontconfig being more pedantic than your
> > distro. Looking at fontconfig, there's no environment variable to
> > suppress these warnings.
> >
>

Re: [Paraview] ParaView 5.2 Release Candidate 2 available for download

[Fabian Wein]

But your guess was right:

xcodebuild -showsdks
iOS SDKs:
iOS 10.0-sdk iphoneos10.0

iOS Simulator SDKs:
Simulator - iOS 10.0-sdk iphonesimulator10.0

macOS SDKs:
macOS 10.12 -sdk macosx10.12

tvOS SDKs:
tvOS 10.0   -sdk appletvos10.0

tvOS Simulator SDKs:
Simulator - tvOS 10.0   -sdk appletvsimulator10.0

watchOS SDKs:
watchOS 3.0 -sdk watchos3.0

watchOS Simulator SDKs:
Simulator - watchOS 3.0 -sdk watchsimulator3.0



> Am 21.10.2016 um 16:11 schrieb Fabian Wein <fabian.w...@fau.de>:
> 
> Did not work:
> 
> -DCMAKE_OSX_SDK=macosx10.12
> 
> in cmake ../paraview-superbuild -DENABLE_cxx11=ON -DENABLE_hdf5=ON 
> -DENABLE_qt5=ON -Dsuperbuild_download_location=~/tmp/pvsb_downloads 
> -DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_OSX_DEPLOYMENT_TARGET=10.12 
> -DCMAKE_OSX_SYSROOT=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.12.sdk
>  -DCMAKE_OSX_SDK=macosx10.12
> 
> results in 
> 
> erforming build step for 'qt5'
> Info: creating cache file 
> /Users/fwein/tmp/pvsb/superbuild/qt5/build/qtbase/.qmake.cache
> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib:
>  file: ../../../lib/libQt5Bootstrap.a(qvector.o) has no symbols
> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib:
>  file: ../../../lib/libQt5Bootstrap.a(qvector.o) has no symbols
> clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
> deployment target of OS X 10.9
> clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
> deployment target of OS X 10.9
> Makefile:13164: warning: overriding commands for target 
> `.moc/qeventdispatcher_cf.moc'
> Makefile:12994: warning: ignoring old commands for target 
> `.moc/qeventdispatcher_cf.moc'
> clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
> deployment target of OS X 10.9
> Undefined symbols for architecture x86_64:
>  "std::__1::__vector_base_common::__throw_length_error() const", 
> referenced from:
>  void std::__1::vector<QMetaMethodBuilderPrivate, 
> std::__1::allocator 
> >::__push_back_slow_path(QMetaMethodBuilderPrivate&&)
>  in qmetaobjectbuilder.o
>  void std::__1::vector<QMetaPropertyBuilderPrivate, 
> std::__1::allocator 
> >::__push_back_slow_path(QMetaPropertyBuilderPrivate&&)
>  in qmetaobjectbuilder.o
>  void std::__1::vector<QMetaEnumBuilderPrivate, 
> std::__1::allocator 
> >::__push_back_slow_path(QMetaEnumBuilderPrivate&&) 
> in qmetaobjectbuilder.o
>  "void std::__1::__sort<std::__1::__less<int, int>&, int*>(int*, int*, 
> std::__1::__less<int, int>&)", referenced from:
>  QString::multiArg(int, QString const**) const in qstring.o
>  
> QSortFilterProxyModelPrivate::_q_sourceHeaderDataChanged(Qt::Orientation, 
> int, int) in qsortfilterproxymodel.o
>  QSortFilterProxyModel::removeRows(int, int, QModelIndex const&) in 
> qsortfilterproxymodel.o
>  "std::bad_alloc::bad_alloc()", referenced from:
>  qBadAlloc() in qglobal.o
>  QVector<QList >::reallocData(int, int, 
> QFlags) in qglobal.o
> ld: symbol(s) not found for architecture x86_64
> clang: error: linker command failed with exit code 1 (use -v to see 
> invocation)
> make[6]: *** [../../lib/QtCore.framework/QtCore] Error 1
> make[5]: *** [sub-corelib-make_first] Error 2
> make[4]: *** [sub-src-make_first] Error 2
> make[3]: *** [module-qtbase-make_first] Error 2
> CMake Error at /Users/fwein/tmp/pvsb/superbuild/sb-qt5-build.cmake:43 
> (message):
>  Failed with exit code 2
> 
> I have the impressions that my other issues and comments on the gitlab issue 
> tracker are removed, so I stop posting reports of attempts there.
> 
> 
> 
>> CMAKE_OSX_SDK needs to be something like `macosx10.9`.  I forget the command 
>> to make XCode list the valid sdk versions available.  For your case I would 
>> guess `macosx10.12` is the correct value.  Qt looks up the path to the SDK 
>> from that string.
>> 
>> HTH,
>> Shawn
>> 
>> On Fri, Oct 21, 2016 at 7:07 AM, Fabian Wein <fabian.w...@fau.de> wrote:
>> I added a lot of the stuff to the issue tracker.
>> 
>> I’m currently lost with building qt5 on macOS 10.12. It’s a qmake issue, I 
>> tracked it down to
>> 
>> Creating qmake... Running configuration tests... Failed to process makespec 
>> for platform ‚macx-clang' Info: creating stash file 
>> /Users/fwein/tmp/pvsb/superbui

Re: [Paraview] ParaView 5.2 Release Candidate 2 available for download

[Fabian Wein]

Hi David,

you are right, I apply the cmake option for superbuild. This was my first 
attempt. With an additional -verbose on configure I get

Running configuration tests… 
Failed to process makespec for platform ‚macx-clang‘ 
Info: creating stash file 
/Users/fwein/tmp/pvsb/superbuild/qt5/build/qtbase/.qmake.stash 
Project ERROR: QMAKE_MAC_SDK can only contain short-form SDK names (eg. macosx, 
iphoneos) CMake Error at 
/Users/fwein/tmp/pvsb/superbuild/sb-qt5-configure.cmake:43 (message):

But I do not know how to find and change the QMAKE_MAC_SDK value as it seems to 
come from qmake.


Maybe we have to wait for qt 5.7.1? But when I build qt manually with the 
borrowed configure from the superbuild but without -sdk it works. The only 
difference I see is that superbuild uses another configure?! 

Using the path for CMAKE_OSX_SDK results with an immediate error in configure. 
using macosx or macosx10.12 results within make in the error reported before, 
using macos10.12 (from macOS 10.12) results in make in

Project ERROR: Could not resolve SDK Path for 'macos10.12'
Project ERROR: Could not resolve SDK PlatformPath for 'macos10.12'
Project ERROR: Could not resolve SDK SDKVersion for ‚macos10.12'



> Hi Fabian,
> 
> Your attempt to set CMAKE_OSX_SDK=10.12 was probably on the superbuild 
> project, and it is not passed to the Qt subproject as QT_MAC_SDK. Try passing 
> the proper value to CMAKE_OSX_SDK (i.e., the full path which is usually 
> /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.12.sdk)
>  and see if the superbuild properly determines the value of QMAKE_MAC_SDK 
> from that.
> 
>   David
> 
>> But your guess was right:
>> 
>> xcodebuild -showsdks
>> iOS SDKs:
>>  iOS 10.0-sdk iphoneos10.0
>> 
>> iOS Simulator SDKs:
>>  Simulator - iOS 10.0-sdk iphonesimulator10.0
>> 
>> macOS SDKs:
>>  macOS 10.12 -sdk macosx10.12
>> 
>> tvOS SDKs:
>>  tvOS 10.0   -sdk appletvos10.0
>> 
>> tvOS Simulator SDKs:
>>  Simulator - tvOS 10.0   -sdk appletvsimulator10.0
>> 
>> watchOS SDKs:
>>  watchOS 3.0 -sdk watchos3.0
>> 
>> watchOS Simulator SDKs:
>>  Simulator - watchOS 3.0 -sdk watchsimulator3.0
>> 
>> 
>> 
>>> Am 21.10.2016 um 16:11 schrieb Fabian Wein <fabian.w...@fau.de>:
>>> 
>>> Did not work:
>>> 
>>> -DCMAKE_OSX_SDK=macosx10.12
>>> 
>>> in cmake ../paraview-superbuild -DENABLE_cxx11=ON -DENABLE_hdf5=ON 
>>> -DENABLE_qt5=ON -Dsuperbuild_download_location=~/tmp/pvsb_downloads 
>>> -DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_OSX_DEPLOYMENT_TARGET=10.12 
>>> -DCMAKE_OSX_SYSROOT=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.12.sdk
>>>  -DCMAKE_OSX_SDK=macosx10.12
>>> 
>>> results in 
>>> 
>>> erforming build step for 'qt5'
>>> Info: creating cache file 
>>> /Users/fwein/tmp/pvsb/superbuild/qt5/build/qtbase/.qmake.cache
>>> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib:
>>>  file: ../../../lib/libQt5Bootstrap.a(qvector.o) has no symbols
>>> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib:
>>>  file: ../../../lib/libQt5Bootstrap.a(qvector.o) has no symbols
>>> clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
>>> deployment target of OS X 10.9
>>> clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
>>> deployment target of OS X 10.9
>>> Makefile:13164: warning: overriding commands for target 
>>> `.moc/qeventdispatcher_cf.moc'
>>> Makefile:12994: warning: ignoring old commands for target 
>>> `.moc/qeventdispatcher_cf.moc'
>>> clang: warning: libstdc++ is deprecated; move to libc++ with a minimum 
>>> deployment target of OS X 10.9
>>> Undefined symbols for architecture x86_64:
>>> "std::__1::__vector_base_common::__throw_length_error() const", 
>>> referenced from:
>>>void std::__1::vector<QMetaMethodBuilderPrivate, 
>>> std::__1::allocator 
>>> >::__push_back_slow_path(QMetaMethodBuilderPrivate&&)
>>>  in qmetaobjectbuilder.o
>>>void std::__1::vector<QMetaPropertyBuilderPrivate, 
>>> std::__1::allocator 
>>> >::__push_back_slow_path(QMetaPropertyBuilderPrivate&&)
>>>  in qmetaobjectbuilder.o
>>>void std::__1::vector<QMeta

[Paraview] building qt via paraview-superbuild on macOS 10.12 still fails

[Fabian Wein]

Yesterday qt 5.7.1 was released but I still have issues building qt via pvsb 
while building qt everywhere manually with the very same configure options just 
works. 

configuring qt manually results in linker calls like
clang++ -headerpad_max_install_names -stdlib=libc++ -Wl,-syslibroot …

When I verbose the build out of pvsb the options „-headerpad_max_install_names 
-stdlib=libc++“ are missing. Adding them manually for the single link case, 
linking works.

I have no idea why the identical configuring from pvsb/cmake leads to different 
results on macOS 10.12.

I added the details to 
https://gitlab.kitware.com/paraview/paraview-superbuild/issues/29

Any help is very much appreciated,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] Question on compiling external plugins with 5.2

[Fabian Wein]

Hello,

we have a legacy implementation of a PV reader plugin which is patched into the 
PV code.

I want to refactor this approach to an external plugin.

In https://gitlab.kitware.com/paraview/paraview-superbuild it says:
-
External plugins
The superbuild supports building more plugins into ParaView using the 
paraviewexternalplugins project. As an example, to build two external plugins a 
and b, the following settings should be used:
• enable_paraviewexternalplugins:BOOL=ON: Enables building using external 
plugins.
• paraview_PLUGINS_EXTERNAL:STRING=a;b: The list of plugins to build.
• paraview_PLUGIN_a_PATH:PATH=/path/to/plugin/a: The path to plugin a's source 
directory. It must contain a plugins.cmake to be picked up by ParaView.
——
Sounds good, I assumed I have to add this via -D to the superbuild cmake, 
however it tells me
CMake Warning:
  Manually-specified variables were not used by the project:
enable_paraviewexternalplugins

And indeed I could not find anything about this, not in the code and not with 
google.

what do I miss? I there somewhere an example or documentation? Is this 
something new with 5.2?

Thanks for any help,

Fabian

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] ParaViewConfig.cmake created too late with ENABLE_paraviewpluginsexternal

[Fabian Wein]

I think there is still an issue building plugins via 
-DENABLE_paraviewpluginsexternal:BOOL=ON when building 
paraview-superbuild.


cmake ../paraview-superbuild -DCMAKE_BUILD_TYPE:STRING=Release 
-DBUILD_TESTING=OFF -DENABLE_cxx11=ON -DENABLE_qt5=ON 
-DENABLE_boost:BOOL=ON -DENABLE_hdf5:BOOL=ON -DENABLE_zlib:BOOL=ON 
-DENABLE_png:BOOL=ON 
-Dsuperbuild_download_location=$HOME/code/cfsdepscache/paraview 
-DENABLE_paraviewpluginsexternal:BOOL=ON 
-Dparaview_PLUGINS_EXTERNAL:STRING=CFSReader 
-Dparaview_PLUGIN_CFSReader_PATH:PATH=$METABUILD/../plugins/CFSReader


results in the error

CMake Error at ~/code/cfs_paraview/plugins/CFSReader/CMakeLists.txt:43 
(FIND_PACKAGE):
  By not providing "FindParaView.cmake" in CMAKE_MODULE_PATH this 
project has
  asked CMake to find a package configuration file provided by 
"ParaView",

  but CMake did not find one.

  Could not find a package configuration file provided by "ParaView" 
with any

  of the following names:

ParaViewConfig.cmake
paraview-config.cmake

  Add the installation prefix of "ParaView" to CMAKE_PREFIX_PATH or set
  "ParaView_DIR" to a directory containing one of the above files.  If
  "ParaView" provides a separate development package or SDK, be sure 
it has

  been installed.

And indeed ParaViewConfig.cmake does not exist yet in 
ParaView_DIR=~/code/cfs_paraview/metabuild/build/superbuild/paraview/build


It works when I do the cmake ../paraview-superbuild first without 
-DENABLE_paraviewpluginsexternal:BOOL=ON ...

and then a second time with (as shown above).

However, we want to build the paraview-superbuild with our plugin 
automatically (e.g. in our nightly builds) with an own cmake project 
where paraview-superbuild is an external project.


Do I something wrong? Is this a bug? Any hint on how to do a workaround?

Thanks!

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] Question on compiling external plugins with 5.2

[Fabian Wein]

Then I do something wrong:

cmake ../paraview-superbuild -DCMAKE_BUILD_TYPE:STRING=Release 
-DBUILD_TESTING=OFF -DENABLE_cxx11=ON -DENABLE_qt5=ON -DENABLE_boost:BOOL=ON 
-DENABLE_hdf5:BOOL=ON -DENABLE_zlib:BOOL=ON -DENABLE_png:BOOL=ON 
-Dsuperbuild_download_location=$HOME/code/cfsdepscache/paraview 
-DCMAKE_OSX_SDK=macosx10.12 -DENABLE_paraviewexternalplugins:BOOL=ON 
-Dparaview_PLUGINS_EXTERNAL:STRING=CFSReader 
-Dparaview_PLUGIN_CFSReader_PATH:PATH=$METABUILD/../plugins/CFSReader

…
CMake Warning:
  Manually-specified variables were not used by the project:

ENABLE_paraviewexternalplugins
…

I should have written, that I tried some variants before. 

Fabian



> Am 02.01.2017 um 20:01 schrieb Ben Boeckel <ben.boec...@kitware.com>:
> 
> On Fri, Dec 30, 2016 at 23:48:52 +0100, Fabian Wein wrote:
>> • enable_paraviewexternalplugins:BOOL=ON: Enables building using external 
>> plugins.
> 
> Sorry, that should be ENABLE_paraviewexternalplugins. PR pushed:
> 
>https://gitlab.kitware.com/paraview/paraview-superbuild/merge_requests/253
> 
> --Ben

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] Reconfiguring external project

[Fabian Wein]

Hello,

I successfully build my external project via 
cmake ../paraview-superbuild … —DENABLE_paraviewpluginsexternal:BOOL=ON  …
my CMakeLists.txt of my external project ist still work in progress.

How can I make the cmake ../paraview-superbuild to execute by plugin’s 
CMakeLists.txt again?

Removing ./superbuild/paraviewpluginsexternal-prefix

does not help and I found no other stamp file source.

Thanks,

Fabian

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] git question on paraview-superbuild master

[Fabian Wein]

I have a question, which is probably related to my lack of experience 
with git:


https://gitlab.kitware.com/paraview/common-superbuild/blob/master/versions.cmake

contains the line
set(qt5_ver "${qt5_ver_series}.1")

However, when I follow 
https://gitlab.kitware.com/paraview/paraview-superbuild/ and do


git clone --recursive 
https://gitlab.kitware.com/paraview/paraview-superbuild.git

cd paraview-superbuild
git fetch origin
git checkout master
git submodule update

I get with
grep qt5_ver_series superbuild/versions.cmake
the old content
set(qt5_ver "${qt5_ver_series}.0")

However, when I do
cd superbuild
git checkout master

it works:
set(qt5_ver "${qt5_ver_series}.1")

Do I miss something?
Does the checkout always be repeated in superbuild? Then I request to 
change

https://gitlab.kitware.com/paraview/paraview-superbuild/

What is the relationship of paraview-superbuild and common-superbuild?

Thanks,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] changing color of coordinate system arrows

[Fabian Wein]

Is it possible to change the colors of the coordinate system arrows? 
The yellow one is not that clear visible on a white background.


Thanks,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] changing color of coordinate system arrows

[Fabian Wein]

I opened https://gitlab.kitware.com/paraview/paraview/issues/17340
I did not see the option for „minor enhancement“.

Thanks :)

Fabian

> Not currently, I am afraid. They are not exposed to be user settable. Feel 
> free to report a issue on the issue tracker, however. It should be fairly 
> straight forward to support.
> 
> Utkarsh
> 
> On Wed, Mar 29, 2017 at 11:00 AM, Fabian Wein <fabian.w...@fau.de> wrote:
> Is it possible to change the colors of the coordinate system arrows? The 
> yellow one is not that clear visible on a white background.
> 
> Thanks,
> 
> Fabian
> ___
> Powered by www.kitware.com
> 
> Visit other Kitware open-source projects at 
> http://www.kitware.com/opensource/opensource.html
> 
> Please keep messages on-topic and check the ParaView Wiki at: 
> http://paraview.org/Wiki/ParaView
> 
> Search the list archives at: http://markmail.org/search/?q=ParaView
> 
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview
> 

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] ParaView 5.3.0 released

[Fabian Wein]

> On behalf of the ParaView team, I am happy to announce that ParaView
> 5.3.0 has been released and is ready to download from
> 
> http://www.paraview.org/download
> 
> Release notes for ParaView 5.3.0 are available at
> 
> https://blog.kitware.com/paraview-5-3-0-release-notes/

Please add the tag to the superbuild.

BTW: My whole office is grateful for the reload (F5) and autoscaled warping of 
PV 5.2 (at least compared to PV 5.0)!! :)

Thanks to the folks at kitware for the really cool software!

Fabian




Dr. Fabian Wein
ZISC - Zentralinstitut für wissenschaftliches Rechnen
Universität Erlangen-Nürnberg 
D-91052 Erlangen, Germany
fabian.w...@fau.de

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] ParaView 5.3.0 released

[Fabian Wein]

Sorry,

this doesn't work for me:

I start with a clean directory
mkdir pv53
cd pv53
git clone --recursive 
https://gitlab.kitware.com/paraview/paraview-superbuild.git

cd paraview-superbuild
git fetch --tags origin
git checkout v5.3.0

-> error: pathspec 'v5.3.0' did not match any file(s) known to git.

Fabian

On 03/15/17 02:21, Cory Quammen wrote:

git fetch --tags origin
git checkout v5.3.0

On Tue, Mar 14, 2017 at 6:06 PM, Fabian Wein <fabian.w...@fau.de> wrote:

The superbuild was tagged with v5.3.0 this morning.



I just did

git clone —recursive https://gitlab.kitware.com/paraview/paraview-superbuild.git
cd paraview-superbuild
git fetch origin
git checkout v5.3.0

and got
error: pathspec 'v5.3.0' did not match any file(s) known to git.

Do I miss something?

Thanks,

Fabian


Best regards,
Cory

On Tue, Mar 14, 2017 at 5:52 PM, Fabian Wein <fabian.w...@fau.de> wrote:

On behalf of the ParaView team, I am happy to announce that ParaView
5.3.0 has been released and is ready to download from

http://www.paraview.org/download

Release notes for ParaView 5.3.0 are available at

https://blog.kitware.com/paraview-5-3-0-release-notes/


Please add the tag to the superbuild.

BTW: My whole office is grateful for the reload (F5) and autoscaled warping of 
PV 5.2 (at least compared to PV 5.0)!! :)

Thanks to the folks at kitware for the really cool software!

Fabian


___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] error when building ffmpeg

[Fabian Wein]

Hi,

for me it appears you are not using superbuild. I suggest you try to build PV 
with superbuild:
https://gitlab.kitware.com/paraview/paraview-superbuild/

Fabian

> Am 12.04.2017 um 19:13 schrieb Hedieh Ebrahimi :
> 
> Hi all, 
> 
> when buiding paraview I get the following error:
> 
>  12%] Building CXX object 
> VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o
> /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx: In member 
> function ‘int vtkFFMPEGWriterInternal::Start()’:
> /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> error: expected type-specifier before ‘CodecID’
>c->codec_id = static_cast(this->avOutputFormat->video_codec);
>  ^
> /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> error: expected ‘>’ before ‘CodecID’
> /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> error: expected ‘(’ before ‘CodecID’
> /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> error: ‘CodecID’ was not declared in this scope
> /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:72: 
> error: expected ‘)’ before ‘;’ token
>c->codec_id = static_cast(this->avOutputFormat->video_codec);
> ^
> VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/build.make:74: recipe for target 
> 'VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o' failed
> make[2]: *** [VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o] 
> Error 1
> CMakeFiles/Makefile2:7935: recipe for target 
> 'VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/all' failed
> make[1]: *** [VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/all] Error 2
> Makefile:149: recipe for target 'all' failed
> make: *** [all] Error 2
> 
> I have trying to build on opensuse leap 42.2. and my ffmpeg version is 3.2.4
> 
> Could anybody give me a hint on how to fix this?
> 
> Thank you very much in Advance, 
> Hedieh Ebrahimi
> Consultant 
> 
> 
> 
> 
> 
> 
> ___
> Powered by www.kitware.com
> 
> Visit other Kitware open-source projects at 
> http://www.kitware.com/opensource/opensource.html
> 
> Please keep messages on-topic and check the ParaView Wiki at: 
> http://paraview.org/Wiki/ParaView
> 
> Search the list archives at: http://markmail.org/search/?q=ParaView
> 
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] error when building ffmpeg

[Fabian Wein]

Superbuild allows to build distributable packages via ctest -R cpack and we 
share these. So I never had to investigate about shared/ static libs. I would 
suggest that only static and only one executable doesn’t work, alone because of 
the plugins. But I really don’t know.

> Hi Fabian, 
> 
> Thanks for your swift reply.
> I have many opensuse machines all using the same opensuse version. I would 
> like to build the executable so that I can just copy it to the other machines 
> and run it. ( statistically linked). 
> 
> Is BUILD_SHARED_LIBS=0 the only settings that will make this work or there 
> are more steps I would need to take?
> 
> I´d appreciate your help, 
> 
>  
> 
> Hedieh Ebrahimi
> Consultant 
> 
> 
> 
> 
> 
> 
> 
> On 13 April 2017 at 11:48, Fabian Wein <fabian.w...@fau.de> wrote:
> You can still use superbuild.
> * Manually you build PV then modify the code and make again.
> * Programmatically you can patch superbuild such that patches with your 
> changes are applied to PV. Before PV 5.2 I did this myself. It is easy as 
> superbuild applies some patches by itself you can use as template.
> 
> Fabian
> 
> > The reason I am not using superbuild is because I would need to modify the 
> > paraview source code. Is there any way to be able to modify the paraview 
> > source code using superbuild?
> >
> > Thanks,
> >
> > Hedieh Ebrahimi
> > Consultant
> >
> >
> >
> >
> >
> >
> >
> > On 13 April 2017 at 09:11, Fabian Wein <fabian.w...@fau.de> wrote:
> > Hi,
> >
> > for me it appears you are not using superbuild. I suggest you try to build 
> > PV with superbuild:
> > https://gitlab.kitware.com/paraview/paraview-superbuild/
> >
> > Fabian
> >
> > > Am 12.04.2017 um 19:13 schrieb Hedieh Ebrahimi 
> > > <hedieh.ebrah...@amphos21.com>:
> > >
> > > Hi all,
> > >
> > > when buiding paraview I get the following error:
> > >
> > >  12%] Building CXX object 
> > > VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o
> > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx: In 
> > > member function ‘int vtkFFMPEGWriterInternal::Start()’:
> > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > error: expected type-specifier before ‘CodecID’
> > >c->codec_id = static_cast(this->avOutputFormat->video_codec);
> > >  ^
> > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > error: expected ‘>’ before ‘CodecID’
> > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > error: expected ‘(’ before ‘CodecID’
> > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > error: ‘CodecID’ was not declared in this scope
> > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:72: 
> > > error: expected ‘)’ before ‘;’ token
> > >c->codec_id = static_cast(this->avOutputFormat->video_codec);
> > > ^
> > > VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/build.make:74: recipe for target 
> > > 'VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o' failed
> > > make[2]: *** 
> > > [VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o] Error 1
> > > CMakeFiles/Makefile2:7935: recipe for target 
> > > 'VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/all' failed
> > > make[1]: *** [VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/all] Error 2
> > > Makefile:149: recipe for target 'all' failed
> > > make: *** [all] Error 2
> > >
> > > I have trying to build on opensuse leap 42.2. and my ffmpeg version is 
> > > 3.2.4
> > >
> > > Could anybody give me a hint on how to fix this?
> > >
> > > Thank you very much in Advance,
> > > Hedieh Ebrahimi
> > > Consultant
> > >
> > >
> > >
> > >
> > >
> > >
> > > ___
> > > Powered by www.kitware.com
> > >
> > > Visit other Kitware open-source projects at 
> > > http://www.kitware.com/opensource/opensource.html
> > >
> > > Please keep messages on-topic and check the ParaView Wiki at: 
> > > http://paraview.org/Wiki/ParaView
> > >
> > > Search the list archives at: http://markmail.org/search/?q=ParaView
> > >
> > > Follow this link to subscribe/unsubscribe:
> > > http://public.kitware.com/mailman/listinfo/paraview
> >
> >
> 
> 

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] error when building ffmpeg

[Fabian Wein]

Please keep the thread on the list.

I guess there is some kind of build directory in your build directory where it 
should work.

Fabian

> Hi Fabien, 
> 
> When I run ctest  -R cpack in the build directory I get "not tests found".
> How can I fix this?
> 
> Thanks in Advance, 
> 
> Hedieh Ebrahimi
> Consultant 
> 
> 
> 
> 
> 
> 
> 
> On 13 April 2017 at 12:03, Fabian Wein <fabian.w...@fau.de> wrote:
> Superbuild allows to build distributable packages via ctest -R cpack and we 
> share these. So I never had to investigate about shared/ static libs. I would 
> suggest that only static and only one executable doesn’t work, alone because 
> of the plugins. But I really don’t know.
> 
> > Hi Fabian,
> >
> > Thanks for your swift reply.
> > I have many opensuse machines all using the same opensuse version. I would 
> > like to build the executable so that I can just copy it to the other 
> > machines and run it. ( statistically linked).
> >
> > Is BUILD_SHARED_LIBS=0 the only settings that will make this work or there 
> > are more steps I would need to take?
> >
> > I´d appreciate your help,
> >
> >
> >
> > Hedieh Ebrahimi
> > Consultant
> >
> >
> >
> >
> >
> >
> >
> > On 13 April 2017 at 11:48, Fabian Wein <fabian.w...@fau.de> wrote:
> > You can still use superbuild.
> > * Manually you build PV then modify the code and make again.
> > * Programmatically you can patch superbuild such that patches with your 
> > changes are applied to PV. Before PV 5.2 I did this myself. It is easy as 
> > superbuild applies some patches by itself you can use as template.
> >
> > Fabian
> >
> > > The reason I am not using superbuild is because I would need to modify 
> > > the paraview source code. Is there any way to be able to modify the 
> > > paraview source code using superbuild?
> > >
> > > Thanks,
> > >
> > > Hedieh Ebrahimi
> > > Consultant
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > On 13 April 2017 at 09:11, Fabian Wein <fabian.w...@fau.de> wrote:
> > > Hi,
> > >
> > > for me it appears you are not using superbuild. I suggest you try to 
> > > build PV with superbuild:
> > > https://gitlab.kitware.com/paraview/paraview-superbuild/
> > >
> > > Fabian
> > >
> > > > Am 12.04.2017 um 19:13 schrieb Hedieh Ebrahimi 
> > > > <hedieh.ebrah...@amphos21.com>:
> > > >
> > > > Hi all,
> > > >
> > > > when buiding paraview I get the following error:
> > > >
> > > >  12%] Building CXX object 
> > > > VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o
> > > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx: In 
> > > > member function ‘int vtkFFMPEGWriterInternal::Start()’:
> > > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > > error: expected type-specifier before ‘CodecID’
> > > >c->codec_id = 
> > > > static_cast(this->avOutputFormat->video_codec);
> > > >  ^
> > > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > > error: expected ‘>’ before ‘CodecID’
> > > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > > error: expected ‘(’ before ‘CodecID’
> > > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:29: 
> > > > error: ‘CodecID’ was not declared in this scope
> > > > /home/hedieh/sofware/paraview/VTK/IO/FFMPEG/vtkFFMPEGWriter.cxx:192:72: 
> > > > error: expected ‘)’ before ‘;’ token
> > > >c->codec_id = 
> > > > static_cast(this->avOutputFormat->video_codec);
> > > > 
> > > > ^
> > > > VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/build.make:74: recipe for 
> > > > target 'VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o' 
> > > > failed
> > > > make[2]: *** 
> > > > [VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/vtkFFMPEGWriter.cxx.o] Error 1
> > > > CMakeFiles/Makefile2:7935: recipe for target 
> > > > 'VTK/IO/FFMPEG/CMakeFiles/vtkIOFFMPEG.dir/all' failed
> > > > make[1]: *** [VTK/IO/FFMPEG/CMa

Re: [Paraview] ParaView 5.3.0 and Qt5 linking error

[Fabian Wein]

Hi,

in case you want to use superbuild, gcc 6.3.1 compiles qt 5.8.0 from PV 5.3 
for me without any problems. Why do you want to use precompiled qt? 
On a fast Linux I build PV 5.3 superbuild with qt in ~30 min.

If you want to compile qt-everywhere manually be warned, a plain configure
took hours and many GBs for me - apparently all combinations of debug/relase/…
were built. 

> I am here working on an Arch Linux system. What kind of standard library do 
> you mean? I compiled GCC (6.2.0) and MPI (1.10.4) by myself.
> 
> Btw you are right, the precompiled Qt libraries probably need older standard 
> libraries. Maybe the GCC-6.2.0 is the problem. I also tried the precompiled 
> Qt5 library in version 5.8.0. It does not work. The system Qt5 libraries (fom 
> Arch repo) seem to be functional though.
> 
> Thx for your response!
> 
> Tobi
> 
> 
> On Friday, April 21, 2017 2:56:27 PM CEST Ben Boeckel wrote:
>> On Fri, Apr 21, 2017 at 14:38:30 +0200, Tobias D wrote:
>>> Following error is thrown:
>>> [100%] Linking CXX executable
>>> ../bin/pvserver../lib/libvtkPVServerManagerApplication- pv5.3.so.1:
>>> undefined reference to `__cxa_throw_bad_array_new_length@Qt_5'collect2:
>>> error: ld returned 1 exit status
>> 
>> 
>> 
>>> - Qt 5.6.2 (precompiled from qt-project.org)
>> 
>> It looks like Qt5 is linking a different (or expecting a newer) standard
>> library than you have available on your machine. What distro are you
>> using?
>> 
>> --Ben
> 
> 
> ___
> Powered by www.kitware.com
> 
> Visit other Kitware open-source projects at 
> http://www.kitware.com/opensource/opensource.html
> 
> Please keep messages on-topic and check the ParaView Wiki at: 
> http://paraview.org/Wiki/ParaView
> 
> Search the list archives at: http://markmail.org/search/?q=ParaView
> 
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] superbuild: fontconfig too old to compile (current does)

[Fabian Wein]

I've pushed an MR for the common-superbuild here:


Ben,

thank you! So the current master will have the upgrade within a few days?

Cheers,

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

[Paraview] superbuild: fontconfig too old to compile (current does)

[Fabian Wein]

Hello to the list,

I have openSUSE tumbleweed with gcc 7.2.1

fontconfig  2.12.1 does not compile with the error at the end of this 
mail.


Building the current fontconfig 2.12.6 works for me, maybe the version
could be upgraded?

Regards,

Fabian

Here is the error:

Performing build step for 'fontconfig'
Making all in fontconfig
Making all in fc-blanks
Making all in fc-case
Making all in fc-lang
Making all in fc-glyphname
Making all in src
  CC   fcatomic.lo
  CC   fcblanks.lo
  CC   fccache.lo
  CC   fccfg.lo
  CC   fccharset.lo
  CC   fccompat.lo
  CC   fcdbg.lo
  CC   fcdefault.lo
  CC   fcdir.lo
  CC   fcformat.lo
  CC   fcfreetype.lo
  CC   fcfs.lo
  CC   fcinit.lo
  CC   fclang.lo
  CC   fclist.lo
  CC   fcmatch.lo
  CC   fcmatrix.lo
  CC   fcname.lo
  CC   fcobjs.lo
  CC   fcpat.lo
fcmatch.c:324:63: error: 'PRI_CHAR_WIDTH_STRONG' undeclared here (not 
in a function); did you mean 'PRI_WIDTH_STRONG'?
 #define FC_OBJECT(NAME, Type, Cmp) { FC_##NAME##_OBJECT, Cmp, 
PRI_##NAME##_STRONG, PRI_##NAME##_WEAK },

   ^
fcobjs.h:54:1: note: in expansion of macro 'FC_OBJECT'
 FC_OBJECT (CHAR_WIDTH,  FcTypeInteger, NULL)
 ^
fcmatch.c:324:84: error: 'PRI_CHAR_WIDTH_WEAK' undeclared here (not in 
a function); did you mean 'PRI_WIDTH_WEAK'?
 #define FC_OBJECT(NAME, Type, Cmp) { FC_##NAME##_OBJECT, Cmp, 
PRI_##NAME##_STRONG, PRI_##NAME##_WEAK },


  ^
fcobjs.h:54:1: note: in expansion of macro 'FC_OBJECT'
 FC_OBJECT (CHAR_WIDTH,  FcTypeInteger, NULL)
 ^
gmake[6]: *** [Makefile:633: fcmatch.lo] Error 1
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] superbuild: fontconfig too old to compile (current does)

[Fabian Wein]

I could not find the build error in the link.

The earliest version that builds for my gcc 7.2.1 is fontconfig-2.12.4, maybe 
this is a „compromise“?

Thanks, Fabian

> Am 25.10.2017 um 15:13 schrieb Ben Boeckel <ben.boec...@kitware.com>:
> 
> On Wed, Oct 25, 2017 at 13:35:51 +0200, Fabian Wein wrote:
>> thank you! So the current master will have the upgrade within a few days?
> 
> It's not so simple :( . There's a build failure with older compilers:
> 
>https://open.cdash.org/viewNotes.php?buildid=5114526
> 
> --Ben

___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] Parview Build Faulure on Mac OS 10.12.6

[Fabian Wein]

First question: do you use the superbuild?

On 12/22/17 00:37, Beach, Timothy A. (GRC-LTE0)[Vantage Partners, LLC] wrote:

Helo,  I’ve been working through some issues compiling on a new MAC but this 
one has me stuck.   I used CMAKE 3.10 to generate xcodebuild files for paraview 
5.4. Any ideas?

Thanks

Tim

/make: *** No rule to make target 
`/Users/…/lib/cmake/paraview-5.4/Modules/vtkCommonCoreHierarchy.txt', needed by 
`/Users/…/VTK/Wrapping/Python/vtkAbstractArrayPython.cxx'.  Stop./

/Command /bin/sh failed with exit code 2/

//

*/** BUILD FAILED **/*//


___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
https://paraview.org/mailman/listinfo/paraview

Re: [Paraview] undefined symbol: PyUnicodeUCS2_AsEncodedString

[Fabian Wein]

Hi Shawn,

It looks to me like you built against a Python that was configured with UCS2 
unicode objects and are linking to a Python that was configured with UCS4 
unicode objects.  Python is not ABI-compatible with other Pythons that were 
built with a different kind of unicode support.


I aggree with your observation. The point is, that not me is linking but it is 
the paraview-superbuild which failes. With USE_SYSTEM_python=OFF I would expect 
that my system python is not touched?!

BTW, I can confirm that it works with USE_SYSTEM_python=ON, so a trivial 
workaround exists.

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
https://paraview.org/mailman/listinfo/paraview

[Paraview] undefined symbol: PyUnicodeUCS2_AsEncodedString

[Fabian Wein]

I build latest head of paraview-superbuild with Python enabled and 
USE_SYSTEM_PYTHON=OFF on a openSUSE Tumbleweed system

I get the following error
Scanning dependencies of target CinemaPython
[  2%] Copying files for Python package 'cinema_python'
[  2%] Compiling Python package 'cinema_python'
Traceback (most recent call last):
  File "/usr/lib64/python2.7/runpy.py", line 174, in _run_module_as_main
[  2%] Building CXX object 
VTK/ThirdParty/xdmf2/vtkxdmf2/libsrc/CMakeFiles/vtkxdmf2.dir/XdmfHeavyData.cxx.o
[  2%] Building CXX object 
ThirdParty/protobuf/vtkprotobuf/src/CMakeFiles/protoc_compiler.dir/google/protobuf/stubs/strutil.cc.o
"__main__", fname, loader, pkg_name)
  File "/usr/lib64/python2.7/runpy.py", line 72, in _run_code
exec code in run_globals
  File "/usr/lib64/python2.7/compileall.py", line 16, in 
import struct
  File "/usr/lib64/python2.7/struct.py", line 1, in 
from _struct import *
ImportError: 
/home/fwein/code/cfs_paraview/build_python/build/install/lib64/python2.7/lib-dynload/_struct.so:
 undefined symbol: PyUnicodeUCS2_AsEncodedString
gmake[8]: *** [ThirdParty/cinema/CMakeFiles/CinemaPython.dir/build.make:61: 
ThirdParty/cinema/cinema_python.build-complete] Error 1
gmake[7]: *** [CMakeFiles/Makefile2:1192: 
ThirdParty/cinema/CMakeFiles/CinemaPython.dir/all] Error 2
gmake[7]: *** Waiting for unfinished jobs

To my knowledge one need to decide if Python is compiled with 16 or 32 bit 
encoding,

on my system I have 32 bit:

/usr/bin/python2
Python 2.7.14 (default, Oct 12 2017, 15:50:02) [GCC] on linux2
Type "help", "copyright", "credits" or "license" for more information.

import sys
print(sys.maxunicode)

1114111

When I try the Python built from PV, it has 16 bits:

/home/fwein/code/cfs_paraview/build_python/build/superbuild/python/build/python
Python 2.7.14 (default, Feb  2 2018, 15:33:16)
[GCC 7.3.0] on linux2
Type "help", "copyright", "credits" or "license" for more information.

import sys
print(sys.maxunicode)

65535

So the mixing from own built python and system python makes problems.

Is this mixing a bug?

Currently, I don't need python in PV, and for my colleague SYSTEM_PYTHON 
worked. So just to let you know.

BTW, what is the status with python3?

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
https://paraview.org/mailman/listinfo/paraview

Re: [Paraview] FEM solver interfaced with Paraview

[Fabian Wein]

 of the following questions have probably ever been asked, but:


 1. What is the best format to create a result file readable by Paraview, 
especially for huge file? after some readings, VTK XML format seems to be the 
most interesting (parallelization capabilities) compared to the legacy VTK one, 
am I right or do ou have any other suggestion?
 2. I’m still trying to figure out if xdmf and/or hdf5 format are relevant, and 
if they are relevant for my application: any feedback on it?

The workflow I’m imaginating sounds like:

    
  Intermediate calculations

Raw data (ascii file) from my solver --> intermediate file in xdmf or hdf5 format 
--> Paraview input file (depending on the reader)


To my understanding there is no xdmf OR hdf5: 
http://xdmf.org/index.php/Main_Page

We also have an own FEM Code. We export directly hdf5 files (w/o xdmf) which we 
read with an own paraview reader plugin.

Now we would probably try xdmf and see if we can circumvent having an own 
reader plugin.

hdf5 has libs for all languages (C/C++, Python, Matlab) and e.g. the handling 
in Python is quite convenient. Almost a one-liner.

The hdf5 C/C++ lib for writing the data is not the best software I know but it 
is ok.

I would not export to ascii first but directly use the hdf5 libs.

Fabian
___
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
https://paraview.org/mailman/listinfo/paraview