ParaView issueshttps://gitlab.kitware.com/paraview/paraview/-/issues2020-05-21T20:20:15-04:00https://gitlab.kitware.com/paraview/paraview/-/issues/19850File output from Catalyst kills ParaView2020-05-21T20:20:15-04:00W. Alan ScottFile output from Catalyst kills ParaViewJeff M. wrote an exodus file from Catalyst. When we try to open this file with ParaView, it dies a horrible death. File attached.
Here's what Jeff wrote to me:
Hey Alan. To create this exodus file I ran SPARC and did a catalyst scri...Jeff M. wrote an exodus file from Catalyst. When we try to open this file with ParaView, it dies a horrible death. File attached.
Here's what Jeff wrote to me:
Hey Alan. To create this exodus file I ran SPARC and did a catalyst script which does an xyslice and then writes out the xy slice to a .exo file using the paraview exodus writer. I ran on 1 process to do this, so this is not a multiprocess issue.
When I try to load the file in paraview 5.8 (gui) there are a bunch of warnings about bad point ids followed by a core dump. That was on snip - a linux cluster. On my windows machine, it just hangs.
When I do something else in SPARC (extract subset with seed) and write out an exodus file, paraview 5.8 gui loads it no problem.
When I load a .cgns file from some of Watney's data into the gui, make a slice, write it out (from the gui) as an exodus file, I can load that exodus file no problem.
I have my suspicions as to why this is happening but the issue I am writing you about here is: ParaView shouldn't core dump when it tries to read this file, especially after it has already detected problems with the file.
File is UUR, under the HiFire SAND number.
[exodus_output_from_xyslice1_0008.exo](/uploads/f5bdac662901afdd04860e3e8f6e6bd3/exodus_output_from_xyslice1_0008.exo)5.8.1Utkarsh AyachitUtkarsh Ayachithttps://gitlab.kitware.com/paraview/paraview/-/issues/16960Change labels on GUI widgets for stream tracer seed points2020-05-21T20:10:13-04:00Kenneth MorelandChange labels on GUI widgets for stream tracer seed pointsThe stream tracer filter has some strange labels regarding items that define the seed points. I would like to see the following naming changes:
* Change the seed type named `High Resolution Line Source` to be simply `Line Source`. Ratio...The stream tracer filter has some strange labels regarding items that define the seed points. I would like to see the following naming changes:
* Change the seed type named `High Resolution Line Source` to be simply `Line Source`. Rationale: The line is usually used for no more than around 100 samples. That is by no means "high resolution." Besides, there is no reason the line source could not be dropped to 10 samples or less. Thus, the line source can be either high or low resolution. Declaring it as high resolution is just confusing. Note that the name `High Resolution Line Source` is also used in other filters (like plot over line). For the same reasons I would like to see them all changed.
* Change the seed type named `Point Source` to `Point Cloud`. Rationale: point source suggests a single point as the source. A point cloud more accurately depicts some volume in space with scattered samples of points. It will come as no surprise to users that the cloud is defined in a sphere.
* Change the group name `Sphere Parameters` to `Point Cloud Parameters`. Rationale: Match the new name for the seed type.5.9 (Fall 2020)https://gitlab.kitware.com/paraview/paraview/-/issues/17539Add New Macro should be Import New Macro2020-05-21T20:03:08-04:00W. Alan ScottAdd New Macro should be Import New MacroPlease change Macros/ Add New Macro to be Macros/ Import New Macro.Please change Macros/ Add New Macro to be Macros/ Import New Macro.https://gitlab.kitware.com/paraview/paraview/-/issues/19378[Plugin] UI_FILES must be in the main directory so autouic can find them2020-05-21T20:01:34-04:00Mathieu Westphal (Kitware)[Plugin] UI_FILES must be in the main directory so autouic can find themIn the current implementation, UI_FILES must be in the directory where the paraview_add_plugin takes place.
Hence the following **does not work** :
```
set(ui_files
Ressource/myFile.ui)
paraview_add_plugin(Plugin
...
UI_FILES ...In the current implementation, UI_FILES must be in the directory where the paraview_add_plugin takes place.
Hence the following **does not work** :
```
set(ui_files
Ressource/myFile.ui)
paraview_add_plugin(Plugin
...
UI_FILES ${ui_files})
```
while this works perfectly
```
set(ui_files
myFile.ui)
paraview_add_plugin(Plugin
...
UI_FILES ${ui_files})
```
Here is a slightly modified version of the SierraPlotToolsPlugin that shows this behavior.
[SierraPlotToolsStandAlone.tgz](/uploads/f8023e2248b599c438ae93af26e15b3b/SierraPlotToolsStandAlone.tgz)
```
-- Build files have been written to: /home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/build
[ 5%] Automatic MOC and UIC for target SierraPlotTools
AutoUic error
-------------
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqPlotVariablesDialog.cxx"
The file includes the uic file "ui_pqVariablePlot.h",
but the user interface file "pqVariablePlot.ui"
could not be found in the following locations
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqVariablePlot.ui"
AutoUic error
-------------
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqPlotter.cxx"
The file includes the uic file "ui_pqSierraToolsRichTextDocs.h",
but the user interface file "pqSierraToolsRichTextDocs.ui"
could not be found in the following locations
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqSierraToolsRichTextDocs.ui"
AutoUic error
-------------
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqSierraPlotToolsDataLoadManager.cxx"
The file includes the uic file "ui_pqSierraPlotToolsDataLoadManager.h",
but the user interface file "pqSierraPlotToolsDataLoadManager.ui"
could not be found in the following locations
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqSierraPlotToolsDataLoadManager.ui"
AutoUic error
-------------
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqSierraPlotToolsManager.cxx"
The file includes the uic file "ui_pqSierraPlotToolsActionHolder.h",
but the user interface file "pqSierraPlotToolsActionHolder.ui"
could not be found in the following locations
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqSierraPlotToolsActionHolder.ui"
AutoUic error
-------------
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqSierraPlotToolsManager.cxx"
The file includes the uic file "ui_pqVariablePlot.h",
but the user interface file "pqVariablePlot.ui"
could not be found in the following locations
"/home/glow/work/paraview/paraviewFirst/paraview/Plugins/SierraPlotToolsStandAlone/Plugin/pqVariablePlot.ui"
make[2]: *** [Plugin/CMakeFiles/SierraPlotTools_autogen.dir/build.make:58: Plugin/CMakeFiles/SierraPlotTools_autogen] Error 1
make[1]: *** [CMakeFiles/Makefile2:122: Plugin/CMakeFiles/SierraPlotTools_autogen.dir/all] Error 2
make: *** [Makefile:130: all] Error 2
```5.9 (Fall 2020)Ben BoeckelBen Boeckelhttps://gitlab.kitware.com/paraview/paraview/-/issues/18906ParaViewTestingData-v?.?.?.tar.gz could use better name2020-05-21T20:00:13-04:00Cory Quammencory.quammen@kitware.comParaViewTestingData-v?.?.?.tar.gz could use better nameWe have two data packages produced during releases:
`ParaViewTestingData-v5.6.0.tar.gz`: contains the content store of ParaView's testing data
`ParaViewData-v5.6.0.tar.gz`: contains the actual test data files named with their original n...We have two data packages produced during releases:
`ParaViewTestingData-v5.6.0.tar.gz`: contains the content store of ParaView's testing data
`ParaViewData-v5.6.0.tar.gz`: contains the actual test data files named with their original names.
There are probably better names for these two items. I propose
`ParaViewTestingContentStore-v5.6.0.tar.gz`: for the test content store
`ParaViewTestingDataFiles-v5.6.0.tar.gz`: for the data files that can be loaded directly.Ben BoeckelBen Boeckelhttps://gitlab.kitware.com/paraview/paraview/-/issues/17121OSPRay Fails to Render DataSet in VTI file in ParaView 5.2.0 with VolumeMappe...2020-05-21T11:47:38-04:00John PatchettOSPRay Fails to Render DataSet in VTI file in ParaView 5.2.0 with VolumeMapper's Input has no scalar arrayload vti file like that attached 30cubedSample.vti, click apply in in properties box.
Select Pressure (or anything or don't) in scalar filed drop down box.
Select Volume Representation.
Then click "Enable OSPRay" in the properties ta...load vti file like that attached 30cubedSample.vti, click apply in in properties box.
Select Pressure (or anything or don't) in scalar filed drop down box.
Select Volume Representation.
Then click "Enable OSPRay" in the properties tab.
Error Message: ERROR: In /Users/kitware/buildbot-slave/b99e1795/source-paraview/VTK/Rendering/OSPRay/vtkOSPRayVolumeMapperNode.cxx, line 107
vtkOSPRayVolumeMapperNode (0x7fe4934a3c80): VolumeMapper's Input has no scalar array!
You can't seem to select a scalar array directly out of the reader. You can use the calculator to pass the arrary, and that will work and get volume rendered...
[30cubedSample.vti](/uploads/15aeb27e02982d7b0c372b8b3bdb9c15/30cubedSample.vti)5.3David E. DeMarleDavid E. DeMarlehttps://gitlab.kitware.com/paraview/paraview/-/issues/12949Segfault setting options in 2D view2020-05-20T21:15:20-04:00Kitware RobotSegfault setting options in 2D view**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=12949). Further discussion may take place here.**
---
To reproduce this (no data needs to be loaded in):
Start ParaView
Chan...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=12949). Further discussion may take place here.**
---
To reproduce this (no data needs to be loaded in):
Start ParaView
Change to 2D view
Change an option and apply (e.g show axis)
Change to 3D view
Change back to 2D view
Change an option and apply (e.g show axis)
The segmentation fault happens when apply is clicked.https://gitlab.kitware.com/paraview/paraview/-/issues/17002Regression: Error while generating diagnostics (sys import missing)2020-05-20T21:12:54-04:00Elias PippingRegression: Error while generating diagnostics (sys import missing)Steps to reproduce:
- Create a `PythonAnnotation` and enter `x` into the expression field
Instead of the very helpful list of variables that are currently in scope (I'm assuming here you don't have a variable called `x`), which I get w...Steps to reproduce:
- Create a `PythonAnnotation` and enter `x` into the expression field
Instead of the very helpful list of variables that are currently in scope (I'm assuming here you don't have a variable called `x`), which I get with 5.1.2, with 5.2.0-RC3 instead I get
```
Traceback (most recent call last):
File "<string>", line 7, in <module>
File "<string>", line 5, in vtkPythonAnnotationFilter_EvaluateExpression
File "/Applications/ParaView-5.2.0-RC3.app/Contents/Python/paraview/annotation.py", line 93, in execute
"Variables in current scope are %s \n" % (expression, ns.keys()), file=sys.stderr)
NameError: global name 'sys' is not defined
```https://gitlab.kitware.com/paraview/paraview/-/issues/18210ParaView crash on Clip2020-05-20T20:48:51-04:00Mathieu Westphal (Kitware)ParaView crash on ClipParaView crash when clipping a specific dataset.
How to reproduce :
* Open ParaView
* Open [quadbog.vtu](/uploads/f372e06f9ec1f2f405ee5588d7123d8c/quadbog.vtu)
* Add a clip, Apply
* Crash
```
Thread 1 "paraview" received signal S...ParaView crash when clipping a specific dataset.
How to reproduce :
* Open ParaView
* Open [quadbog.vtu](/uploads/f372e06f9ec1f2f405ee5588d7123d8c/quadbog.vtu)
* Add a clip, Apply
* Crash
```
Thread 1 "paraview" received signal SIGSEGV, Segmentation fault.
0x00007fffe64408e1 in vtkDataSetSurfaceFilter::InsertQuadInHash (this=0x55555bc29a70, a=1842, b=1844, c=1845, d=1843, sourceId=528) at /home/glow/work/paraview/paraviewThird/paraview/VTK/Filters/Geometry/vtkDataSetSurfaceFilter.cxx:2311
2311 if (quad->numPts == 4 && c == quad->ptArray[2])
(gdb) bt
#0 0x00007fffe64408e1 in vtkDataSetSurfaceFilter::InsertQuadInHash (this=0x55555bc29a70, a=1842, b=1844, c=1845, d=1843, sourceId=528) at /home/glow/work/paraview/paraviewThird/paraview/VTK/Filters/Geometry/vtkDataSetSurfaceFilter.cxx:2311
#1 0x00007fffe643e355 in vtkDataSetSurfaceFilter::UnstructuredGridExecute (this=0x55555bc29a70, dataSetInput=0x55555c62eca0, output=0x55555c5a2620) at /home/glow/work/paraview/paraviewThird/paraview/VTK/Filters/Geometry/vtkDataSetSurfaceFilter.cxx:1795
#2 0x00007ffff108ef6d in vtkPVGeometryFilter::UnstructuredGridExecute (this=0x55555bc2ea70, input=0x55555c62eca0, output=0x55555c5a2620, doCommunicate=1) at /home/glow/work/paraview/paraviewThird/paraview/ParaViewCore/VTKExtensions/Rendering/vtkPVGeometryFilter.cxx:1448
#3 0x00007ffff108a6b0 in vtkPVGeometryFilter::ExecuteBlock (this=0x55555bc2ea70, input=0x55555c62ea20, output=0x55555c5a2620, doCommunicate=1, updatePiece=0, updateNumPieces=1, updateGhosts=0, wholeExtent=0x55555c5fe340)
at /home/glow/work/paraview/paraviewThird/paraview/ParaViewCore/VTKExtensions/Rendering/vtkPVGeometryFilter.cxx:490
#4 0x00007ffff108a98a in vtkPVGeometryFilter::RequestData (this=0x55555bc2ea70, request=0x55555be9b780, inputVector=0x55555b283ba0, outputVector=0x55555bc30fb0) at /home/glow/work/paraview/paraviewThird/paraview/ParaViewCore/VTKExtensions/Rendering/vtkPVGeometryFilter.cxx:553
```https://gitlab.kitware.com/paraview/paraview/-/issues/18626macOS: UI values can get out of sync with property values2020-05-20T16:54:41-04:00Cory Quammencory.quammen@kitware.commacOS: UI values can get out of sync with property values* Launch ParaView, built-in server.
* Add Sphere source. Click Apply.
* Click on Radius property text field. Type 1.0, but do not press enter or change the focus to a different UI widget.
* Click Apply.
The sphere size will remain the s...* Launch ParaView, built-in server.
* Add Sphere source. Click Apply.
* Click on Radius property text field. Type 1.0, but do not press enter or change the focus to a different UI widget.
* Click Apply.
The sphere size will remain the same. However, the radius property will remain at 1.0.
When clicking Apply, the current property values in the UI should be set as the property values before the proxy is updated.
Original issue here: #18575.https://gitlab.kitware.com/paraview/paraview/-/issues/19916Export Inspector: allow selecting of root directory for pvserver configs2020-05-20T16:04:18-04:00Utkarsh AyachitExport Inspector: allow selecting of root directory for pvserver configsThis is extracted from [#19073](https://gitlab.kitware.com/paraview/paraview/-/issues/19073#note_720701)
step 1: let's get the "export now" working, however clunky. For remote connections, we accept two separate Root Directories, one fo...This is extracted from [#19073](https://gitlab.kitware.com/paraview/paraview/-/issues/19073#note_720701)
step 1: let's get the "export now" working, however clunky. For remote connections, we accept two separate Root Directories, one for client side and one for server side. Data Extracts will export using the server-side root directory while Image Extracts including cinema/png/etc. will use the client side root directory.5.8.1Utkarsh AyachitUtkarsh Ayachithttps://gitlab.kitware.com/paraview/paraview/-/issues/15291hstack() not working2020-05-20T12:49:34-04:00Kitware Robothstack() not working**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=15291). Further discussion may take place here.**
---
(Tested with official 4.3.1 Linux binaries)
As described in the User ...**This issue was created automatically from an original [Mantis Issue](http://paraview.org/Bug/view.php?id=15291). Further discussion may take place here.**
---
(Tested with official 4.3.1 Linux binaries)
As described in the User Guide on page 102 the following should work in the Python Calculator:
hstack ([ velocity_x , velocity_y , velocity_z ])
Firstly, hstack() isn't available as a function, although it can be accessed as numpy.hstack(). Secondly, I don't know what it wrong with it but its behaviour is very weird.
Setup:
1. Create a Wavelet source
2. Create a Python Calculator
Testing some Python expressions:
EXPRESSION:
hstack([1,2,3])
RESULT:
Traceback (most recent call last):
File "<string>", line 4, in <module>
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 102, in execute
retVal = compute(inputs, expression, ns=variables)
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 74, in compute
retVal = eval(expression, globals(), mylocals)
File "<string>", line 1, in <module>
NameError: name 'hstack' is not defined
Traceback (most recent call last):
File "<string>", line 4, in <module>
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 102, in execute
retVal = compute(inputs, expression, ns=variables)
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 74, in compute
retVal = eval(expression, globals(), mylocals)
File "<string>", line 1, in <module>
NameError: name 'hstack' is not defined
EXPRESSION:
numpy.hstack([1,2,3])
RESULT:
works
EXPRESSION:
numpy.hstack([RTData, RTData, RTData])
RESULT:
Traceback (most recent call last):
File "<string>", line 4, in <module>
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 102, in execute
retVal = compute(inputs, expression, ns=variables)
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 74, in compute
retVal = eval(expression, globals(), mylocals)
File "<string>", line 1, in <module>
NameError: name 'hstack' is not defined
Traceback (most recent call last):
File "<string>", line 4, in <module>
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 105, in execute
retVal, self.GetArrayName())
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/vtk/numpy_interface/dataset_adapter.py", line 618, in append
narray = narray.flatten() * numpy.ones((arrLength, components))
MemoryError
EXPRESSION:
numpy.hstack([1, 2, RTData])
RESULT:
works
EXPRESSION:
numpy.hstack([1, RTData, RTData])
RESULT:
Traceback (most recent call last):
File "<string>", line 4, in <module>
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/calculator.py", line 105, in execute
retVal, self.GetArrayName())
File "/home/paulm/software/ParaView-4.3.1-Linux-64bit/lib/paraview-4.3/site-packages/paraview/vtk/numpy_interface/dataset_adapter.py", line 618, in append
narray = narray.flatten() * numpy.ones((arrLength, components))
MemoryError
https://gitlab.kitware.com/paraview/paraview/-/issues/19663typographic error in colormap name2020-05-20T09:50:23-04:00Jean M. Favretypographic error in colormap namein Remoting/Views/ColorMaps.json
"Asymmtrical Earth Tones (6_21b)" should be "Asymmetrical Earth Tones (6_21b)"in Remoting/Views/ColorMaps.json
"Asymmtrical Earth Tones (6_21b)" should be "Asymmetrical Earth Tones (6_21b)"Ben BoeckelBen Boeckelhttps://gitlab.kitware.com/paraview/paraview/-/issues/19893Please delete the Sierra Plot Tools plugin2020-05-19T21:59:21-04:00W. Alan ScottPlease delete the Sierra Plot Tools pluginPlease delete the Sierra Plot Tools plugin. It was created about 10 years ago, isn't used, and is mostly redundant with other controls within ParaView.Please delete the Sierra Plot Tools plugin. It was created about 10 years ago, isn't used, and is mostly redundant with other controls within ParaView.5.10 (Fall 2021)Vicente Boleavicente.bolea@kitware.comVicente Boleavicente.bolea@kitware.comhttps://gitlab.kitware.com/paraview/paraview/-/issues/19003Default shortcut clash2020-05-19T21:57:32-04:00NeysSchreinerDefault shortcut clashThe default shortcut for Camera Redo clashes with global Find (ctrl+F). This produces an error: Ambiguous shortcut overload. Changing the shortcut fixes the error.The default shortcut for Camera Redo clashes with global Find (ctrl+F). This produces an error: Ambiguous shortcut overload. Changing the shortcut fixes the error.Nicolas VuailleNicolas Vuaillehttps://gitlab.kitware.com/paraview/paraview/-/issues/19097Export Inspector: bad Qt layout2020-05-19T21:54:31-04:00Utkarsh AyachitExport Inspector: bad Qt layoutThe items under **Global Settings** don't stretch horizontally like widgets on **Properties** panel do. This indicates that there's some incorrect layouting at play.
![image](/uploads/27ee7d5fb5f9c342713df640c08aae0a/image.png)The items under **Global Settings** don't stretch horizontally like widgets on **Properties** panel do. This indicates that there's some incorrect layouting at play.
![image](/uploads/27ee7d5fb5f9c342713df640c08aae0a/image.png)5.8.1Utkarsh AyachitUtkarsh Ayachithttps://gitlab.kitware.com/paraview/paraview/-/issues/19880Glyph filter sometimes produces misplaced arrows2020-05-19T21:27:14-04:00DavidGlyph filter sometimes produces misplaced arrowsThis issue is from the discussion in [a thread on the ParaView forums.](https://discourse.paraview.org/t/misplaced-arrows-from-glyph-filter-in-5-8-0-under-linux/4160)
It's arising under Linux, 64-bit, 5.8.0 build downloaded from paravie...This issue is from the discussion in [a thread on the ParaView forums.](https://discourse.paraview.org/t/misplaced-arrows-from-glyph-filter-in-5-8-0-under-linux/4160)
It's arising under Linux, 64-bit, 5.8.0 build downloaded from paraview.org, and also occurs on a nightly build from 28th April.
Summary: After reading a VTU file with Int32 as the types for the connectivity and offset arrays, the Glyph filter when run on cell data sometimes produces arrows in incorrect places where there are no cells. It is not consistent; changing, e.g., the glyph scaling and re-applying will change the location of the arrows that are produced. If the VTU file uses Int64 for the types of connectivity and offset but is otherwise identical, then the issue does not arise.
Steps to reproduce:
1. Open the attached cleaned32.vtu and apply
2. Select current and surface with edges, reduce opacity to 0.2 so the glyphs can be seen
3. Filters->Glyph, select orientation array current_dir, no scale array, scale factor 5, all points, apply
4. If need be, switch back and forth between scale factor 4 + apply and scale factor 5 + apply a few times. At some point you’ll see the misplaced arrows.
Example picture of the problem (not always reproduced exactly):
![misplaced](/uploads/0290b0da9995140be66da28ca10cedc8/misplaced.png)
The issue does not occur if you use cleaned.vtu (which has Int64 as the offset and connectivity type).
[cleaned32.vtu](/uploads/3d56ba8a7a05414a0518d4d08b8e81d9/cleaned32.vtu)
[cleaned.vtu](/uploads/e1fbb08e9787764983b1e44322c9bdab/cleaned.vtu)5.8.1https://gitlab.kitware.com/paraview/paraview/-/issues/19921Intermittent errors in Cell Centers2020-05-19T21:26:49-04:00Kenneth MorelandIntermittent errors in Cell Centers[A thread on discourse](https://discourse.paraview.org/t/paraview-cell-centers-possible-unstable-program-behavior/4310) has brought up a problem with the `Cell Centers` filter. The filter seems to be unstable in that it intermittently cr...[A thread on discourse](https://discourse.paraview.org/t/paraview-cell-centers-possible-unstable-program-behavior/4310) has brought up a problem with the `Cell Centers` filter. The filter seems to be unstable in that it intermittently creates bad centers. Here is a video from some data provided by the user. The geometry of the source does not change, but you can see the centers "sparkling" as in each frame you see an error in the computation.
![cell-centers-errors](/uploads/e671ea41df315c77b82106a5326128a6/cell-centers-errors.mp4)
I believe the issue is that this data stores its connectivity array as int32 values. In this condition, it causes the `vtkCellCenters` filter to get cell connections in a thread-unsafe way. That in turn causes threads to interfere with each other.
Diving in more deeply, I point to this code in `vtkUnstructuredGrid.cxx`, which is supposed to be thread safe.
```cpp
void vtkUnstructuredGrid::GetCell(vtkIdType cellId, vtkGenericCell* cell)
{
int cellType = static_cast<int>(this->Types->GetValue(cellId));
cell->SetCellType(cellType);
vtkIdType numPts;
const vtkIdType* pts;
this->Connectivity->GetCellAtId(cellId, numPts, pts);
```
The problem is with this last line. `GetCellAtId` is sometimes thread safe, but not always thread safe. It is thread safe if the connectivity array is a simple array of type `vtkIdType`. However, if the array is of a different type, the values get copied to an object-local space (which is shared among threads), and this becomes very thread unsafe.5.8.1Kenneth MorelandKenneth Morelandhttps://gitlab.kitware.com/paraview/paraview/-/issues/19908Cell size filter dies with CTH AMR2020-05-19T21:22:57-04:00W. Alan ScottCell size filter dies with CTH AMRThe cell size filter is dying when run on a CTH AMR dataset. Here is how to replicate:
* master 5/1/2020 (and 5.8.0), Linux, local server.
* Load spcta_a.4.[0-3]. All vars on. Apply. (I believe this is called something like dave's sm...The cell size filter is dying when run on a CTH AMR dataset. Here is how to replicate:
* master 5/1/2020 (and 5.8.0), Linux, local server.
* Load spcta_a.4.[0-3]. All vars on. Apply. (I believe this is called something like dave's small cth at Kitware).
* cell size filter. Unselect everything other than Compute Volume. Apply.
Crash.
This is an important filter for CTH, since it could be used to calculate density.
@cory.quammen Could we possibly move this one to 5.8.1?5.8.1Cory Quammencory.quammen@kitware.comCory Quammencory.quammen@kitware.comhttps://gitlab.kitware.com/paraview/paraview/-/issues/19729Find Data: poor widget alignment2020-05-19T15:33:46-04:00Utkarsh AyachitFind Data: poor widget alignmentexpanding the Find Data dialog vertically results in the following:
![image](/uploads/d0b045dfb3480a95eeead60d1339e846/image.png)
Also note the bad horizontal aligment. The **Block ID** and **DISPL...** combo-boxes should be aligned on...expanding the Find Data dialog vertically results in the following:
![image](/uploads/d0b045dfb3480a95eeead60d1339e846/image.png)
Also note the bad horizontal aligment. The **Block ID** and **DISPL...** combo-boxes should be aligned on the left with the widgets above/below it.
![image](/uploads/bdead53e6daf8394dd0776b91b66dd3a/image.png)Vicente Boleavicente.bolea@kitware.comVicente Boleavicente.bolea@kitware.com