ParaView issueshttps://gitlab.kitware.com/paraview/paraview/-/issues2019-03-05T10:38:17-05:00https://gitlab.kitware.com/paraview/paraview/-/issues/11778Mac install symlinks to server tools2019-03-05T10:38:17-05:00Kitware RobotMac install symlinks to server tools**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11778). Further discussion may take place here.**
---
In 3.10 we now ship the server tools (pvserver, pvpython, pvbatch, ...) in...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11778). Further discussion may take place here.**
---
In 3.10 we now ship the server tools (pvserver, pvpython, pvbatch, ...) in the bin directory of the app bundle instead of as a separate download. To make these more accessible we should add a menu option (like CMake) to install symlinks to the server tools in /usr/bin.https://gitlab.kitware.com/paraview/paraview/-/issues/11777DISPLAY environment variable (from *.pvx files for CAVE) being set after disp...2018-11-28T09:40:46-05:00Kitware RobotDISPLAY environment variable (from *.pvx files for CAVE) being set after display is opened**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11777). Further discussion may take place here.**
---
The error (unable to open display) is occurring before the environment var...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11777). Further discussion may take place here.**
---
The error (unable to open display) is occurring before the environment variables (from the pvx file) for the pvservers is being set. The vtkPVDisplayInformation::CopyFromObject is being called prior to the environment variables being set resulting in vtkSMMultiProcessRenderView::EndCreateVTKObjects setting RemoteRenderAvailable=false. I have not yet figured out how to do it, but the best solution would involve the PVX info being passed to all of the pvservers when they are started (assuming the pvx file is passed to the pvserver command) and if so, then the environment variable(s) should be set immediately - maybe the vtkSMClientServerRenderSyncManagerHelper::InitializeRenderSyncManager should run before testing the displays assuming it is actually telling the pvservers how (and when) to set their environments. I've tested paraview-3.8.1 on our render cluster resulting in same errors. If we had only 1 X screen or X display and only 1 GPU per node, I agree we could set the DISPLAY environment variable. But, being we have 2 GPUs per render nodes in the CAVE and 2 GPUs per compute node in our render cluster, paraview is not working the way it should and it is a bug that should be fixed. Below is some basic output from printf statements I set in various functions to show what is and is not happening. It appears the HOSTNAME variable is not set when the pvservers are started so simply accept that "host: (null)" is a pvserver on one of the render nodes whereas "host:console" is the head node where the client is running.https://gitlab.kitware.com/paraview/paraview/-/issues/117762d axis needs enhancing2018-08-28T09:39:51-04:00Kitware Robot2d axis needs enhancing**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11776). Further discussion may take place here.**
---
We need to enhance the 2d axis. I am using sources/ mandlebrot as an exam...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11776). Further discussion may take place here.**
---
We need to enhance the 2d axis. I am using sources/ mandlebrot as an example.
I can think of a few options, such as:
* Lock the axis to reasonable numbers. With mandelbrot, use a Y range from -1.5 to 1.5.
* Let the user to select the range, and thus the size of the axis.
* Allow user to add a grid pattern.
* Allow user to turn on/off different axes (for instance, upper and right.)
https://gitlab.kitware.com/paraview/paraview/-/issues/117752d axis is confusing2018-08-28T09:39:51-04:00Kitware Robot2d axis is confusing**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11775). Further discussion may take place here.**
---
It is confusing how to turn on the show axis for a 2d view. Lets add a no...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11775). Further discussion may take place here.**
---
It is confusing how to turn on the show axis for a 2d view. Lets add a note in the display tab, for 2d views, that tells a user to go to the Edit View Options.
An example dataset is sources/ mandlebrot.https://gitlab.kitware.com/paraview/paraview/-/issues/11772vtp and vtk cannot handle more than order 120 million triangles2020-04-10T16:59:24-04:00Kitware Robotvtp and vtk cannot handle more than order 120 million triangles**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11772). Further discussion may take place here.**
---
We think 32 bit integers are used in vtp and vtk files to represent the nu...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11772). Further discussion may take place here.**
---
We think 32 bit integers are used in vtp and vtk files to represent the number of coordinates and polygons. In particular, the offset is suspect. In our case, when the number of points and triangles approaches respectively 60 million and 120 million, paraview cannot render the object. We get the following error:
ERROR: In /Users/partyd/Kitware/ParaView-3.8/src/VTK/IO/vtkXMLUnstructuredDataReader.cxx, line 589
vtkXMLPolyDataReader (0x10a504c00): Cannot read cell offsets from Polys in piece 0 because the "offsets" array is not long enough.
Some versions of paraview are able to read the vtp files on unix boxes. It depends on the build. Here is the header of the vtp file:
[dgd]% head test.vtp
<?xml version="1.0"?>
<VTKFile type="PolyData" version="0.1" byte_order="LittleEndian">
<PolyData>
<Piece NumberOfPoints=" 61342156" NumberOfVerts="0" NumberOfLines="0" NumberOfStrips="0" NumberOfPolys=" 118579532">
<Points>
<DataArray type="Float32" NumberOfComponents="3" format="appended" offset="0"/>
</Points>
<Polys>
<DataArray type="Int32" Name="connectivity" format="appended" offset=" 736105876"/>
<DataArray type="Int32" Name="offsets" format="appended" offset=" 2159060264"/>
https://gitlab.kitware.com/paraview/paraview/-/issues/11765Python trace bug for heterogeneous multiblock datasets2020-05-26T21:29:26-04:00Kitware RobotPython trace bug for heterogeneous multiblock datasets**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11765). Further discussion may take place here.**
---
Load a multiblock dataset with only unstructured grid leafs. Paraview see...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11765). Further discussion may take place here.**
---
Load a multiblock dataset with only unstructured grid leafs. Paraview sees a multiblock dataset containing only unstructured grid data, so it creates an unstructured grid display. Next you enable more blocks, so that multiblock dataset now contains unstructured grid and polydata leafs. Save the python trace now. The trace script records volume rendering properties of the unstructured grid display. When the trace script is replayed, paraview sees the multiblock dataset has unstructured grid and polydata leafs, so it creates a geometry display instead of an unstructured grid display. The geometry display lacks the volume rendering properties.
You could reproduce this bug using an extract block filter with an input multiblock dataset that contains unstructured grid and polydata.Mathieu Westphal (Kitware)Mathieu Westphal (Kitware)https://gitlab.kitware.com/paraview/paraview/-/issues/11764Deprecation of VTK XML Inline Binary Option2019-03-05T10:38:17-05:00Kitware RobotDeprecation of VTK XML Inline Binary Option**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11764). Further discussion may take place here.**
---
As we move to Large file support the inline binary option becomes less and...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11764). Further discussion may take place here.**
---
As we move to Large file support the inline binary option becomes less and less attractive. We should deprecate the ability to write and read this type of file, and instead push people to use appended compressed instead.https://gitlab.kitware.com/paraview/paraview/-/issues/11762Xdmf file with hyperslab itemtype crashes Paraview2023-06-15T08:38:35-04:00Kitware RobotXdmf file with hyperslab itemtype crashes Paraview**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11762). Further discussion may take place here.**
---
The attached file, which is supposed to cut a 2x2x2 cube out of a larger 3...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11762). Further discussion may take place here.**
---
The attached file, which is supposed to cut a 2x2x2 cube out of a larger 3x3x3 cube, causes Paraview to crash. ("*** glibc detected *** ... realloc(): invalid next size: 0x0000000008e1a800 ***")
To extract the subgrid [0,1]^3 from the larger [0,2]^3 grid, the start-stride-count in the first DataItem is set to
0 0 0
1 1 1
2 2 2
The same behavior occurs when trying to access the opposite corner [1,2]^3 (the one with values in the [8,15] range) using
1 1 1
1 1 1
2 2 2
instead. The crash can apparently be avoided using ranges whose count values are at least 3 (which should obviously be too large for the 2x2x2 target grid), and even grossly out-of-bound values such as
12 11 321
1 1 1
6 13 13
are accepted without complaints, and result in
1 2 0 3 4 0 0 0
i.e. a copy of the beginning of the source array.https://gitlab.kitware.com/paraview/paraview/-/issues/11759Prism - crash when disconnecting server.2018-08-28T09:39:51-04:00Kitware RobotPrism - crash when disconnecting server.**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11759). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11759). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView trunk, local server, Windows.
Disconnect. It will crash.https://gitlab.kitware.com/paraview/paraview/-/issues/11758Prism - Cube axis labels are too small2018-08-28T09:39:51-04:00Kitware RobotPrism - Cube axis labels are too small**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11758). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11758). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView trunk, local server, Windows.
Turn on cube axis on the SESAME surface. Note that the labels are tiny. We need to scale it properly, and further, add the ability to modify the font size and color.https://gitlab.kitware.com/paraview/paraview/-/issues/11757Prism - log scale cube axis remains linear2018-08-28T09:39:51-04:00Kitware RobotPrism - log scale cube axis remains linear**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11757). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11757). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView trunk, local server, Windows.
Turn on cube axis.
Notice that the cube axis shows linear change, but the data on the SESAME surface is log. The cube axis should also log scaled.
https://gitlab.kitware.com/paraview/paraview/-/issues/11756Prism - Link selection bug2018-08-28T09:39:51-04:00Kitware RobotPrism - Link selection bug**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11756). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11756). Further discussion may take place here.**
---
Ask me for my tutorial on Prism, as well as data to run Prism.
ParaView trunk, local server, Windows.
Turn on contours, as well as showing the tables curves.
Go to timestep 2.
Now, with the SESAME surface as well as the 3D view showing, try to select cells through on the SESAME surface. You should be able to select the data - not the contours or curves.
https://gitlab.kitware.com/paraview/paraview/-/issues/11754Save screenshot goes crazy with Lock Viewsize2019-03-05T10:38:17-05:00Kitware RobotSave screenshot goes crazy with Lock Viewsize**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11754). Further discussion may take place here.**
---
Lets try to get this one done for 3.10.1, since it makes lock screensize u...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11754). Further discussion may take place here.**
---
Lets try to get this one done for 3.10.1, since it makes lock screensize useless.
ParaView trunk, local server, Windows client.
Sources/ Sphere.
Split screen horizontal. Make the sphere visible in the second viewport.
Tools/ Lock Viewsize.
Save Screenshot. Uncheck "Save only selected view". Continue to saving the screenshot.
The screenshot will look like junk.https://gitlab.kitware.com/paraview/paraview/-/issues/11747Windows crashes deleting macros2018-08-28T09:39:51-04:00Kitware RobotWindows crashes deleting macros**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11747). Further discussion may take place here.**
---
ParaView trunk, local server, windows.
Load a macro. (I am using a trivia...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11747). Further discussion may take place here.**
---
ParaView trunk, local server, windows.
Load a macro. (I am using a trivial one created with the trace recorder.)
Macros/ Delete this macro. ParaView will crash.
(I am doing this remote desktop, but I don't think it would matter.)https://gitlab.kitware.com/paraview/paraview/-/issues/11745Trace recorder does not turn off clip plane tool.2018-08-28T09:39:51-04:00Kitware RobotTrace recorder does not turn off clip plane tool.**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11745). Further discussion may take place here.**
---
ParaView trunk, Windows, local server.
* Open disk_out_ref.exo, apply.
*...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11745). Further discussion may take place here.**
---
ParaView trunk, Windows, local server.
* Open disk_out_ref.exo, apply.
* Python shell. Trace. Start Trace.
* Clip, turn Show Plane off, Accept.
* Stop Trace. Save Trace.
* Delete the clip filter.
Now, run the saved trace. It will give the following bugs:
<<The clip widget is still visible, and the Show Plane button is still checked.>>
I talked to Utkarsh, and he feels that it should be possible to just turn the plane tool off through the python (with some hack), assuming the user has done that during the trace. Lets try this, and only try for a full solution (being able to turn on and off) if we have users complain. https://gitlab.kitware.com/paraview/paraview/-/issues/11744Python trace does not create save animations2019-03-05T10:38:23-05:00Kitware RobotPython trace does not create save animations**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11744). Further discussion may take place here.**
---
This is very important functionality, and needs to be implemented as soon ...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11744). Further discussion may take place here.**
---
This is very important functionality, and needs to be implemented as soon as reasonably possible.
To test, do the following:
Start Trace.
Open Can.exo.
Save Animation.
Stop Trace.
(You can read the python that files aren't being saved...)
Save trace as macro.
Run macro.
No new animation will be saved.https://gitlab.kitware.com/paraview/paraview/-/issues/11738Viewport size is wrong2019-03-05T10:38:23-05:00Kitware RobotViewport size is wrong**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11738). Further discussion may take place here.**
---
I am seeing a really strange viewport clear bug. I will attach a screensh...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11738). Further discussion may take place here.**
---
I am seeing a really strange viewport clear bug. I will attach a screenshot. This occurs more often with the normal glyph filter, but seems to happen all of the time, in a harder to replicate pattern.
I have two monitors on my computer, the main one is the bigger one.
* ParaView trunk, remote server (doesn't have problems local server), Windows client (probably needs to be Windows 7).
* On Windows 7, start ParaView in the larger monitor. Click the <windows key> + <left arrow>. This has now placed ParaView in the left half of the window.
* Open disk_out_ref.exo. Apply.
* Normals Glpyhs filter. Apply.
* Move disk_out_ref.exo towards the bottom of the 3d window (<SHFT>, <right arrow>).
* Rotate. You should see that the bottom of the window has bad frame buffer junk.
Thinking about it, this could very likely be a driver issue.
https://gitlab.kitware.com/paraview/paraview/-/issues/11733Error reading EnSight sos file2023-06-15T08:35:15-04:00Kitware RobotError reading EnSight sos file**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11733). Further discussion may take place here.**
---
I installed Paraview 3.8.1 (ParaView-3.8.1-Win64-x86_MPI.exe) and I
tried ...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11733). Further discussion may take place here.**
---
I installed Paraview 3.8.1 (ParaView-3.8.1-Win64-x86_MPI.exe) and I
tried to load an EnSight Server of Server file. This results in the
error messages:
==========================================================
ERROR: In ..\..\..\..\src\Servers\Filters\vtkGenericEnSightReader2.cxx, line 356
vtkGenericEnSightReader2 (0000000007B96730): Unable to open file:
D:\2_proc\NO_Measured_data\/Case02_v10023_DOM_0.case
ERROR: In ..\..\..\..\src\Servers\Filters\vtkGenericEnSightReader2.cxx, line 874
vtkGenericEnSightReader2 (0000000007B96730): Error determining EnSightVersion
ERROR: In ..\..\..\..\src\VTK\Filtering\vtkExecutive.cxx, line 756
vtkCompositeDataPipeline (00000000076F3BC0): Algorithm
vtkGenericEnSightReader2(0000000007B96730) returned failure for
request: vtkInformation (0000000007C30E80)
Debug: Off
Modified Time: 76728
Reference Count: 1
Registered Events: (none)
Request: REQUEST_INFORMATION
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
ERROR: In ..\..\..\..\src\Servers\Filters\vtkGenericEnSightReader2.cxx, line 356
vtkGenericEnSightReader2 (0000000007B96490): Unable to open file:
D:\2_proc\NO_Measured_data\/Case02_v10023_DOM_1.case
ERROR: In ..\..\..\..\src\Servers\Filters\vtkGenericEnSightReader2.cxx, line 874
vtkGenericEnSightReader2 (0000000007B96490): Error determining EnSightVersion
ERROR: In ..\..\..\..\src\VTK\Filtering\vtkExecutive.cxx, line 756
vtkCompositeDataPipeline (00000000076F3AE0): Algorithm
vtkGenericEnSightReader2(0000000007B96490) returned failure for
request: vtkInformation (0000000007C32640)
Debug: Off
Modified Time: 76776
Reference Count: 1
Registered Events: (none)
Request: REQUEST_INFORMATION
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
ERROR: In ..\..\..\..\src\Servers\Filters\vtkPVEnSightMasterServerReader2.cxx,
line 255
vtkPVEnSightMasterServerReader2 (0000000006451F00): Error reading case
file on at least one node.
ERROR: In ..\..\..\..\src\VTK\Filtering\vtkExecutive.cxx, line 756
vtkCompositeDataPipeline (0000000007CA00E0): Algorithm
vtkPVEnSightMasterServerReader2(0000000006451F00) returned failure for
request: vtkInformation (0000000007C30D80)
Debug: Off
Modified Time: 76681
Reference Count: 1
Registered Events: (none)
Request: REQUEST_INFORMATION
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
==========================================================
My sos file looks like:
#
# Example Server-of-server case file
# parts in <> have to be filled in by the user
#
FORMAT
type: master_server gold
SERVERS
number of servers: 2
#Server 1
machine id: PCAME2
executable: "D:\Program Files (x86)\CEI\bin\ensight92_server.bat"
casefile: D:\2_proc\NO_Measured_data\DOM_0\Case02_v10023_DOM_0.case
#Server 2
machine id: PCAME2
executable: "D:\Program Files (x86)\CEI\bin\ensight92_server.bat"
casefile: D:\2_proc\NO_Measured_data\DOM_1\Case02_v10023_DOM_1.case
==========================================================
The both case files are located according to the plce as defined in
the sos file.
Loading the data in EnSight is no problem.https://gitlab.kitware.com/paraview/paraview/-/issues/11718Slice Offsets don't work as expected.2019-03-05T10:38:23-05:00Kitware RobotSlice Offsets don't work as expected.**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11718). Further discussion may take place here.**
---
* Wavelet
* Slice
* Delete All Offsets
* Add 10 offsets using "New Ran...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11718). Further discussion may take place here.**
---
* Wavelet
* Slice
* Delete All Offsets
* Add 10 offsets using "New Range" (default parameters).
* Apply
* delete first 2 and last 2 values. HIt apply. You'll see 6 planes. So far so good.
* now delete middle 4. Hit apply. You should see only 2 planes. You see far more.
I've verified that ParaView is calling the API correctly on vtkCutter. Looks like vtkCutter is somehow not doing the right thing.https://gitlab.kitware.com/paraview/paraview/-/issues/11716Reflect filter does not flip vectors unless there are returned by inCD->GetVe...2019-03-05T10:38:23-05:00Kitware RobotReflect filter does not flip vectors unless there are returned by inCD->GetVectors()**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11716). Further discussion may take place here.**
---
I thought the usage of SetActiveVectors or SetActiveScalars() was consider...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=11716). Further discussion may take place here.**
---
I thought the usage of SetActiveVectors or SetActiveScalars() was considered obsolete in ParaView, since there cannot be more than one such DataArray at a time. Filters such as vtkReflectionFilter which do a GetVectors() to find the vector to flip ignore all other vectors which are simply copied via outPD->CopyData(inPD, i, ptId);
this is a bug, and results in wrong streamlines and all other vector displays in reflected surfaces in ParaView.Mathieu Westphal (Kitware)Mathieu Westphal (Kitware)