ParaView issueshttps://gitlab.kitware.com/paraview/paraview/-/issues2024-03-07T20:23:59-05:00https://gitlab.kitware.com/paraview/paraview/-/issues/22421Molecule representation: crash when creating new view with ParaView 5.12-RC12024-03-07T20:23:59-05:00Francois MazenMolecule representation: crash when creating new view with ParaView 5.12-RC1The second output of the PDB Reader has to be a Molecule representation. When I create a new 3D render view and try to show the second output (forcing molecule representation), then ParaView crashes.
To reproduce with ParaView 5.12-RC1:...The second output of the PDB Reader has to be a Molecule representation. When I create a new 3D render view and try to show the second output (forcing molecule representation), then ParaView crashes.
To reproduce with ParaView 5.12-RC1:
- Open the file 3GQP.pdb available here: https://data.kitware.com/#item/5afd92b68d777f15ebe1aad4
- Click apply, the "Output-1" is displayed with Molecule representation in the 3D render view
- Create a new view (new layout or split current view)
- Click "Render View" for the view type
- Click on the eye of the "Output-1" item in the pipeline browser
- Crash in `pqColorMapEditor::renderViews` because the `representation` variable is null
ParaView 5.11.0 does not crash and the new render view switch to Molecule representation as expected.5.13 (Summer 2024)Cory Quammencory.quammen@kitware.comCory Quammencory.quammen@kitware.comhttps://gitlab.kitware.com/paraview/paraview/-/issues/22408pvserver: make it so render windows in Windows do not need to always be on top2023-12-04T15:56:43-05:00Cory Quammencory.quammen@kitware.compvserver: make it so render windows in Windows do not need to always be on top5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22406CAVEInteraction: let roles be listed in order declared in Python interactor s...2023-12-04T15:43:02-05:00Cory Quammencory.quammen@kitware.comCAVEInteraction: let roles be listed in order declared in Python interactor script and not alphabetized5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22405CAVEInteraction: in Python interactor, object/attribute pulldowns should be b...2023-12-04T14:56:19-05:00Cory Quammencory.quammen@kitware.comCAVEInteraction: in Python interactor, object/attribute pulldowns should be blankAccess to all proxies and their properties is already available within the script.Access to all proxies and their properties is already available within the script.5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22389Paraview 5.12.0-RC1: Loading State File Error for vtkAppendSelection in Paral...2024-03-07T20:23:59-05:00Zongze YangParaview 5.12.0-RC1: Loading State File Error for vtkAppendSelection in Parallel Server when Using State File Created with `Find Data` and `Plot Over Time`The encountered error is as follows:
```
ERROR: In vtkDemandDrivenPipeline.cxx, line 677
vtkPVCompositeDataPipeline (0x600005419f00): Input port 0 of algorithm vtkAppendSelection (0x600005627c00) has 0 connections but is not optional.
`...The encountered error is as follows:
```
ERROR: In vtkDemandDrivenPipeline.cxx, line 677
vtkPVCompositeDataPipeline (0x600005419f00): Input port 0 of algorithm vtkAppendSelection (0x600005627c00) has 0 connections but is not optional.
```
However, it appears that this error does not impact the rendering results.
## Steps to Reproduce the Issue:
1. Initiate the server (optional):
```bash
mpiexec -n 2 pvserver
```
2. Load the file `can.ex2` (Examples -> can.ex2)
3. Navigate to View -> Find Data:
Locate the point (4, 4, 1) with a tolerance of 1, and then proceed to `Plot Over Time`.
4. Save the state. (The state files are attached.)
The error consistently manifests when using the state file in conjunction with the mpi-server, irrespective of whether the state file was created with or without the mpi-server.
[test-mpi2.pvsm](/uploads/90562b4c7d2a365bb3efa7d09be7a71b/test-mpi2.pvsm)
[test-nompi.pvsm](/uploads/3662d6a453176671d9a3b584ea3036f1/test-nompi.pvsm)5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22385ParaView 5.12.0-RC1: non-responsive floating windows : ‘Output Messages’ and ...2024-03-07T20:23:59-05:00Lourenço Sassetti MendesParaView 5.12.0-RC1: non-responsive floating windows : ‘Output Messages’ and ‘Choose Preset’This occurs using WSL2, under Windows 11, using both Ubuntu 22.04 and OpenSuse 15.5. It’s very easy to replicate.
It has nothing to do with specific states or files. Also occurs with “paraview --dr”.
1) The ‘Output Messages’ window when...This occurs using WSL2, under Windows 11, using both Ubuntu 22.04 and OpenSuse 15.5. It’s very easy to replicate.
It has nothing to do with specific states or files. Also occurs with “paraview --dr”.
1) The ‘Output Messages’ window when undocked, can’t be docked any more or resized, but it's still possible to close or select the checkboxes. If it pops up automatically, while using ParaView, it becomes completely non-responsive. The only option is to close it in the ‘View’ menu. It’s annoying because you’re not able to choose where it will pop up. The only way I found to dock it again is resetting ParaView: ‘Edit’ > ‘Reset To Default Settings’ and restart ParaView. This issue occurs both in versions 5.11.2 and 5.12.0.
2) In v5.12.0, the ‘Color Map Editor’ > ‘Choose Preset’ window is always non-responsive to mouse events, and even with TAB you can’t move to all checkboxes. It gets stuck in the default color presets table displayed, and you can’t select the other groups of presets (linear, diverging, etc). This issue occurs only in version 5.12.0.5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22384ParaView 5.12-RC1 uses `~/.local` folder for python library instead of bundle...2024-03-07T20:23:59-05:00Francois MazenParaView 5.12-RC1 uses `~/.local` folder for python library instead of bundled python libsUsing the binary release of 5.12-RC1 of ParaView, it looks like the internal python runtime uses libraries from `~/.local` instead of the libraries available in the ParaView installation folder.
It's visible for the matplotlib path in t...Using the binary release of 5.12-RC1 of ParaView, it looks like the internal python runtime uses libraries from `~/.local` instead of the libraries available in the ParaView installation folder.
It's visible for the matplotlib path in the about dialog:
![image](/uploads/397b510ffe2eb436cacfad8d78fe9335/image.png)
When I remove my `~/.local` folder the expected lib is loaded:
![image](/uploads/478369cc715d4154b05895929bdb31ad/image.png)
When using the local libraries, it creates weird errors when using the FindData panel:
```
Error: Failed to evaluate Expression '(mag(B_A) == 0)'. The following exception stack should provide additional developer specific information. This typically implies a malformed expression. Verify that the expression is valid.
Warning: In vtkPythonSelector.cxx, line 104
vtkPythonSelector (0xa2d9380): Could not invoke 'python_selector.execute()'
Traceback (most recent call last):
File "/home/francois/Documents/Paraview/ParaView-5.12.0-RC1-MPI-Linux-Python3.10-x86_64/lib/python3.10/site-packages/paraview/detail/python_selector.py", line 80, in execute
maskArray = calculator.compute(inputs, query, ns=elocals)
File "/home/francois/Documents/Paraview/ParaView-5.12.0-RC1-MPI-Linux-Python3.10-x86_64/lib/python3.10/site-packages/paraview/detail/calculator.py", line 167, in compute
retVal = eval(subEx, globals(), mylocals)
File "<string>", line 1, in <module>
NameError: name 'B_A' is not defined
Generic Warning: In vtkSelection.cxx, line 752
No values to evaluate because there was an error in the selection evaluator.
ERROR: In vtkExecutive.cxx, line 729
vtkPVCompositeDataPipeline (0x8dd5320): Algorithm vtkPVExtractSelection (0x94b0d50) returned failure for request: vtkInformation (0xa2b9b70)
Debug: Off
Modified Time: 1243144
Reference Count: 1
Registered Events: (none)
Request: REQUEST_DATA
FROM_OUTPUT_PORT: 0
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
```
I can't reproduce this issue with official binary of 5.11.0.5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22372OptiX/VisRTX Raytracing Crashes in v5.12.0-RC12024-03-07T20:23:57-05:00Ryan Krattigerryan.krattiger@kitware.comOptiX/VisRTX Raytracing Crashes in v5.12.0-RC1Related issue: https://gitlab.kitware.com/paraview/paraview/-/issues/20991
Using the VisRTX ray tracer leads to crashes in ParaView. This behavior was a previously a known issues when using OSMesa off-screen rendering. This behavior now...Related issue: https://gitlab.kitware.com/paraview/paraview/-/issues/20991
Using the VisRTX ray tracer leads to crashes in ParaView. This behavior was a previously a known issues when using OSMesa off-screen rendering. This behavior now (as of 5.12.0-RC1) also leads to crashes when using the NVidia EGL server which was previously able to run the OptiX ray tracer in 5.11.
Steps to reproduce:
1. Download/extract/start ParaView with EGL server (or osmesa or client builtin server)
2. Create a sphere (defaults)
3. Enable raytracing
4. Select OptiX Raytracer (defaults)
This error prints until ParaView crashes
```console
( 366.059s) [paraview ] FrameBuffer.h:79 ERR| VISRTX Error: Could not get color texture.
( 366.059s) [paraview ] FrameBuffer.h:92 ERR| VISRTX Error: Could not get depth texture.
```
---
System and Software
* GPU: Quadro T2000 Mobile / Max-Q
* NVidia Drivers (from nvidia-smi)
```console
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 545.23.06 Driver Version: 545.23.06 CUDA Version: 12.3 |
|-----------------------------------------+----------------------+----------------------+
```
* OS: Ubunutu 20.04 LTS
CC: @cory.quammen @vbolea5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22369ParaView 5.12.0-RC1: Incorrect time information and a 'question mark' next to...2024-03-07T20:23:57-05:00Jaswant Panchumarti (Kitware)ParaView 5.12.0-RC1: Incorrect time information and a 'question mark' next to time step value1. Download any Windows ParaView 5.12 RC1 from https://www.paraview.org/files/v5.12/
2. Go to Help -> Example Visualizations
3. Click on the first example named "Exodus II file with timesteps, Clip filter"
4. After it loads, see the time...1. Download any Windows ParaView 5.12 RC1 from https://www.paraview.org/files/v5.12/
2. Go to Help -> Example Visualizations
3. Click on the first example named "Exodus II file with timesteps, Clip filter"
4. After it loads, see the time toolbar, it shows incorrect time step and a question mark.
Surprisingly, I see the issue even in 5.11.2 on Linux.
![question-mark-in-time-combo-box](/uploads/91f32e6f66bed64863936ea25a9548d8/question-mark-in-time-combo-box.png)5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22367ParaView 5.12.0-RC1: Warning about missing property OSPRayMaterial2024-03-07T20:23:57-05:00Cory Quammencory.quammen@kitware.comParaView 5.12.0-RC1: Warning about missing property OSPRayMaterial* Start ParaView, built-in server
* Add Sphere source
* Color by Normals - X
* Check **Enable Ray Tracing** view property
* Switch **Back End** to "OSPRay pathtracer"
* You will get the warning:
```
Generic Warning: In vtkSMPropertyHelp...* Start ParaView, built-in server
* Add Sphere source
* Color by Normals - X
* Check **Enable Ray Tracing** view property
* Switch **Back End** to "OSPRay pathtracer"
* You will get the warning:
```
Generic Warning: In vtkSMPropertyHelper.cxx, line 570
Failed to locate property: OSPRayMaterial
```5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/22128Colormap opacity enabled after switch to categorical and back2024-03-07T20:23:57-05:00Aron HelserColormap opacity enabled after switch to categorical and backOn current PV master, 5/3/2023, with the new colormapping UI reorganization, there is an unexpected behavior enabling opacity mapping.
To reproduce:
* load Source .. Data Objects .. HyperTreeGrid (or any dataset with something that can b...On current PV master, 5/3/2023, with the new colormapping UI reorganization, there is an unexpected behavior enabling opacity mapping.
To reproduce:
* load Source .. Data Objects .. HyperTreeGrid (or any dataset with something that can be a categorical color map)
* color by `vtkEdgeFlags`
* show the Color Map Editor. Notice "Enable Opacity Mapping for Surfaces" is disabled
* enable "Interpret Values as Categories". Immediately disable.
* "Enable Opacity Mapping for Surfaces" is now enabled, and my surface has disappeared.
Note that if I now explicitly disable "Enable Opacity Mapping for Surfaces" at this point, it sticks - I can switch to categorical and back without a problem. Maybe this points to default value handling somehow?
I also see that when "Interpret Values as Categories" is on, an "Enable Opacity Mapping for Surfaces" option is present and enabled by default.
CC @jaswant.panchumarti @cory.quammen5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21843client-server connection arguments should be quoted (or something changed)2024-03-07T20:23:56-05:00Andrew Bauerclient-server connection arguments should be quoted (or something changed)When setting the ssh command in the client-server window we should add in quotes or something else to avoid issues with spaces in paths. The example below is for Putty's `plink` executable which was installed to `C:\Program Files\HPCMP K...When setting the ssh command in the client-server window we should add in quotes or something else to avoid issues with spaces in paths. The example below is for Putty's `plink` executable which was installed to `C:\Program Files\HPCMP Kerberos\plink.exe`. The spaces in this path are an example of the problem with creating a client-server connection. A user can currently get around this by adding quotes at the beginning and end of the full path but users can easily forget to do this. Diagnosing and correcting the problem is beyond many user's knowledge of ParaView and should be fixed automatically internally to ParaView.
![B188B04A-9086-46CC-B25F-E6D160C6DA37](/uploads/60c357fa25c94c45a41b67f94792374f/B188B04A-9086-46CC-B25F-E6D160C6DA37.jpeg)
Initially assigning this to @cory.quammen, really just to notify him of this issue. This seems like a simple fix so I'm hoping that it can go into the v5.12 release. It's not critical to be in v5.12 though.5.13 (Summer 2024)Jaswant Panchumarti (Kitware)Jaswant Panchumarti (Kitware)https://gitlab.kitware.com/paraview/paraview/-/issues/21841pvserver interactive points -> tag missmatch -> exit2023-03-28T21:48:46-04:00Dani Schmidpvserver interactive points -> tag missmatch -> exitI use a ParaView 5.11 client (windows/linux give same result) and pvserver on a linux machine. Generally works very well, but with certain types of interactions such as ‘Hover Points On’ I get tag missmatch and pvserver exits. This issue...I use a ParaView 5.11 client (windows/linux give same result) and pvserver on a linux machine. Generally works very well, but with certain types of interactions such as ‘Hover Points On’ I get tag missmatch and pvserver exits. This issues is confirmed by mwestphal (see https://discourse.paraview.org/t/pvserver-interactive-points-tag-missmatch-exit). I have not found a way to reproduce this with a synthetic dataset, so I provide an example dataset. Here are the steps to reproduce the issue:
1. Get the example dataset from https://www.dropbox.com/s/o8pomdziflyfypu/test.tar.gz?dl=1
2. Start pvserver on linux and connect from windows 5.11 ParaView
3. Load pvd data file. 3D model, quite a few variables, several timesteps
4. ‘Hover Points On’ works at this stage
5. I have two materials in my model, the respective domains are marked by phase=0 and phase=1. I only want to see the parts of the model with phase=1 and use a threshold filter set to 0.1 and 1.
6. If I now try to use ‘Hover Points On’ on the shown phase=1 part of the model then I get the tag missmatch. pvserver then states ‘Exiting…’ but gets stuck. Using Ctrl-C at this stage spits out a stack trace. You find the whole command line output below.
Thanks
Dani
```
Client connected.
( 572.906s) [pvserver ]vtkSocketCommunicator.c:849 ERR| vtkSocketCommunicator (0x1268d70): Tag mismatch: got 1, expecting 10.
( 572.942s) [pvserver ]vtkSocketCommunicator.c:849 ERR| vtkSocketCommunicator (0x1268d70): Tag mismatch: got 1, expecting 41235.
( 572.942s) [pvserver ]vtkSocketCommunicator.c:849 ERR| vtkSocketCommunicator (0x1268d70): Tag mismatch: got 1, expecting 41235.
( 585.484s) [pvserver ]vtkSocketCommunicator.c:783 ERR| vtkSocketCommunicator (0x1268d70): Could not receive tag. 1
( 585.484s) [pvserver ]vtkTCPNetworkAccessMana:295 ERR| vtkTCPNetworkAccessManager (0x1178c60): Some error in socket processing.
Exiting…
^C
Loguru caught a signal: SIGINT
Stack trace:
21 0x40196d /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/pvserver-real() [0x40196d]
20 0x7f5af397ad0a __libc_start_main + 234
19 0x40188e /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/pvserver-real() [0x40188e]
18 0x7f5af24e5654 vtkInitializationHelper::Finalize() + 100
17 0x7f5af13a8ae2 vtkProcessModule::Finalize() + 98
16 0x7f5aedd199fa /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/libvtkCommonCore-pv5.11.so.1(+0x6ae9fa) [0x7f5aedd199fa]
15 0x7f5aedbc0b29 vtkCallbackCommand::Execute(vtkObject*, unsigned long, void*) + 25
14 0x7f5aeb3131b6 /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libvtkRemotingViews-pv5.11.so.1(+0x1e81b6) [0x7f5aeb3131b6]
13 0x7f5ad7b44a89 vtkOSOpenGLRenderWindow::~vtkOSOpenGLRenderWindow() + 9
12 0x7f5ad7b44a1c vtkOSOpenGLRenderWindow::~vtkOSOpenGLRenderWindow() + 28
11 0x7f5ad7b449e3 vtkOSOpenGLRenderWindow::Finalize() + 83
10 0x7f5ad7b44914 vtkOSOpenGLRenderWindow::DestroyOffScreenWindow() + 36
9 0x7f5ad610866b OSMesaDestroyContext + 27
8 0x7f5ad61373e8 /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0x1a93e8) [0x7f5ad61373e8]
7 0x7f5ad6135e38 /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0x1a7e38) [0x7f5ad6135e38]
6 0x7f5ad6c33acb /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0xca5acb) [0x7f5ad6c33acb]
5 0x7f5ad6ade9c5 /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0xb509c5) [0x7f5ad6ade9c5]
4 0x7f5ad6d9b544 /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0xe0d544) [0x7f5ad6d9b544]
3 0x7f5ad6c611a5 /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0xcd31a5) [0x7f5ad6c611a5]
2 0x7f5ad6c34a6b /home/schmid/ParaView-5.11.0-RC2-osmesa-MPI-Linux-Python3.9-x86_64/bin/…/lib/…/lib/libOSMesa.so.8(+0xca6a6b) [0x7f5ad6c34a6b]
1 0x7f5aed6577b2 pthread_cond_wait + 482
0 0x7f5af398fd60 /lib/x86_64-linux-gnu/libc.so.6(+0x38d60) [0x7f5af398fd60]
( 613.056s) [pvserver ] :0 FATL| Signal: SIGINT
```5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21675Python. Function CreationRepresentation -> Registering Error ?2024-03-07T20:23:55-05:00Camille YvinPython. Function CreationRepresentation -> Registering Error ?I have a library that use the function CreateRepresentation from servermanager
In Paraview 5.6.0, the following minimal example works :
```python
from paraview.simple import *
from paraview.servermanager import *
rv = GetRenderView()
...I have a library that use the function CreateRepresentation from servermanager
In Paraview 5.6.0, the following minimal example works :
```python
from paraview.simple import *
from paraview.servermanager import *
rv = GetRenderView()
box = sources.Box(registrationGroup="sources", Center=[2., 2., 2.])
display = CreateRepresentation(box, rv, registrationGroup='representations')
display.SetRepresentationType('Surface')
```
In Paraview 5.10.0, the previous example does not work. I have this message :
```python
File "<string>", line 7, in <module>
File "/home/applis/paraview/5.10.0/lib/python3.9/site-packages/paraview/servermanager.py", line 2332, in CreateRepresentation
setattr(proxy, items[0], items[1])
NameError: name 'items' is not defined
```
So I corrected the lines 2331 and 2332 of servermanager.py from
```
for param in extraArgs.items():
setattr(proxy, items[0], items[1])
```
to
```
for param in extraArgs.items():
setattr(proxy, param[0], param[1])
```
Then when I run the minimal example, I have the following message:
```
Traceback (most recent call last):
File "<string>", line 7, in <module>
File "/home/applis/paraview/5.10.0/lib/python3.9/site-packages/paraview/servermanager.py", line 2332, in CreateRepresentation
setattr(proxy, param[0], param[1])
File "/home/applis/paraview/5.10.0/lib/python3.9/site-packages/paraview/servermanager.py", line 489, in __setattr__
raise AttributeError("Attribute %s does not exist. " % name +
AttributeError: Attribute registrationGroup does not exist. This class does not allow addition of new attributes to avoid mistakes due to typos. Use add_attribute() if you really want to add this attribute.
````
The proxy does not seem to have the attribute `registrationGroup`. In the documentation of the function, it seems possible to use this attribute :
```python
def CreateRepresentation(aProxy, view, **extraArgs):
"""Creates a representation for the proxy and adds it to the render
module.
This method can also be used to initialize properties by passing
keyword arguments where the key is the name of the property.In addition
registrationGroup and registrationName (optional) can be specified (as
keyword arguments) to automatically register the proxy with the proxy
manager.
This method tries to create the best possible representation for the given
proxy in the given view. Additionally, the user can specify proxyName
(optional) to create a representation of a particular type."""
global rendering
if not aProxy:
raise RuntimeError ("proxy argument cannot be None.")
if not view:
raise RuntimeError ("view argument cannot be None.")
if "proxyName" in extraArgs:
display = CreateProxy("representations", extraArgs['proxyName'], None)
del extraArgs['proxyName']
else:
display = view.SMProxy.CreateDefaultRepresentation(aProxy.SMProxy, 0)
if display:
display.UnRegister(None)
if not display:
return None
proxy = _getPyProxy(display)
proxy.Input = aProxy
for param in extraArgs.items():
setattr(proxy, param[0], param[1])
proxy.UpdateVTKObjects()
view.Representations.append(proxy)
return proxy
```
When I delete the 'registrationGroup' keyword when I called the function CreateRepresentation in the minimal example, it is okay but I have the following message when I exit Paraview (this message appears only when I called the function CreateRepresentation) :
```
# Loguru caught a signal: SIGABRT
# Stack trace:
# 8 0x152dc69dc073 /home/applis/paraview/5.10.0/bin/../lib/../lib/libvtkIOGeometry-pv5.10.so.1(+0x38073) [0x152dc69dc073]
# 7 0x152deabb6037 __cxa_finalize + 247
# 6 0x152dea774136 std::vector<std::string>::~vector() + 198
# 5 0x152deabfeeec /lib64/libc.so.6(+0x82eec) [0x152deabfeeec]
# 4 0x152deabfd5ec /lib64/libc.so.6(+0x815ec) [0x152deabfd5ec]
# 3 0x152deabf64e7 /lib64/libc.so.6(+0x7a4e7) [0x152deabf64e7]
# 2 0x152deab9ddb5 abort + 295
# 1 0x152deabb337f gsignal + 271
# 0 0x152deabb3400 /lib64/libc.so.6(+0x37400) [0x152deabb3400]
# ( 317.581s) [paraview ] :0 FATL| Signal: SIGABRT
```
I think that when the Representation is created, it is not corectly registered which explains this SIGABRT signal. Is someone have an idea of the way that I can correct it ?
Best regards
Camille5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21633[5.11.0-RC2] CGNS reload + different base selection => extract block incorrect2024-03-07T20:23:54-05:00Menno Deij - van Rijswijk[5.11.0-RC2] CGNS reload + different base selection => extract block incorrectSteps to reproduce:
1. load `Example_NFace_n.cgns` from ExternalData/Testing/Data
2. Select `Base_Surface_Elements` and then Apply
3. Unselect `Base_Surface_Elements` and select `Base_Volume_Elements`, then right-click the source and ch...Steps to reproduce:
1. load `Example_NFace_n.cgns` from ExternalData/Testing/Data
2. Select `Base_Surface_Elements` and then Apply
3. Unselect `Base_Surface_Elements` and select `Base_Volume_Elements`, then right-click the source and choose `"Reload files > Existing files"`
3. Verify that selected base is `Base_Volume_Elements`
4. Add `Extract Block` filter
5. `Extract Block` filter is populated with blocks from the `Base_Surface_Elements`, not the currently selected `Base_Volume_Elements`
I expected the Extract Block filter to have the block from Base_Volume_Elements to choose.
@MicK7 - Any idea why this happens?5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21617pasting clipboard filename [paraview 5.10.0 / windows]2024-03-07T20:23:54-05:00Theo Igapasting clipboard filename [paraview 5.10.0 / windows]If I want to open a data file by pasting the clipboard to the "Open file" window, I get always a `file:///` prefix in the window, that has to be manually deleted.
Other programs do not show this issue.
![sequence](/uploads/2654756a716d...If I want to open a data file by pasting the clipboard to the "Open file" window, I get always a `file:///` prefix in the window, that has to be manually deleted.
Other programs do not show this issue.
![sequence](/uploads/2654756a716d4cdc18bb8869a844988c/sequence.mp4)5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21592Bump up dataset size threshold for when lower LOD is used2024-03-07T20:23:53-05:00Cory Quammencory.quammen@kitware.comBump up dataset size threshold for when lower LOD is usedIt's set to 20MB, which is very low and likely to make ParaView's interactive rendering capabilities look poor to new users. Recommend bumping to 256MB at least, maybe larger.It's set to 20MB, which is very low and likely to make ParaView's interactive rendering capabilities look poor to new users. Recommend bumping to 256MB at least, maybe larger.5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21570Add trame to the superbuild2024-03-07T20:23:52-05:00Ben BoeckelAdd trame to the superbuildFiling here instead of the superbuild for easier project management purposes.
The current plan:
- use a venv (or equivalent) to manage trame with ParaView 5.11
Future ideas (tentative 5.12):
- have an option for "latest" trame for de...Filing here instead of the superbuild for easier project management purposes.
The current plan:
- use a venv (or equivalent) to manage trame with ParaView 5.11
Future ideas (tentative 5.12):
- have an option for "latest" trame for development/deployment test purposes
- add trame to the superbuild with an option for "some specific version" to ship with ParaView itself (probably cannot update external however); this would pin "latest" as of tagging and/or testing checkpoints
/cc @wascott @sebastien.jourdain @cory.quammen5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/215245.11-RC1: Setting default reader does not work as expected2024-03-07T20:23:51-05:00Menno Deij - van Rijswijk5.11-RC1: Setting default reader does not work as expectedWhen I open a CGNS file there are two readers: CGNS Series reader and IOSS reader. PV 5.11 now has a "Set reader as default" button.
When I click this, the file opens as expected.
But the next time, I still get the dialog to choose a ...When I open a CGNS file there are two readers: CGNS Series reader and IOSS reader. PV 5.11 now has a "Set reader as default" button.
When I click this, the file opens as expected.
But the next time, I still get the dialog to choose a reader.
I expected not to see the dialog again.5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/21519Crash after Reset Session2024-03-07T20:23:51-05:00Artem BodrinCrash after Reset SessionParaView crashes after reset session if last selected item in Pipeline browser was output port.
Affected versions: 5.9.1, 5.10.1
The reason of crash is that `Current` class member of `vtkSMProxySelectionModel` is not reset to null on ...ParaView crashes after reset session if last selected item in Pipeline browser was output port.
Affected versions: 5.9.1, 5.10.1
The reason of crash is that `Current` class member of `vtkSMProxySelectionModel` is not reset to null on reset session (as a result of proxy unregistration). `vtkSMProxySelectionModel` is the one that registered as "ActiveSources".
Line of crash is the one highlighted with breakpoint on video.
For 5.9.1 sources it is vtkSMProxySelectionModel:414.
`Current` class member declared as `vtkSmartPointer<vtkSMProxy> Current;`, but it seems like there should be a weak pointer.
![reset_session_crash](/uploads/ef21d9e93d01eda29a57381f39d85941/reset_session_crash.mp4)
UPD. Making `Current` class member as `vtkWeakPointer` in my local sources did the job.
UPD2. Another weak spot is here:
``` c++
bool vtkSMParaViewPipelineController::UnRegisterPipelineProxy(vtkSMProxy* proxy)
{
PREPARE_FOR_UNREGISTERING(proxy);
vtkSMSessionProxyManager* pxm = proxy->GetSessionProxyManager();
const char* _proxyname = pxm->GetProxyName("sources", proxy);
if (_proxyname == NULL)
{
return false;
}
SM_SCOPED_TRACE(Delete).arg("proxy", proxy);
const std::string proxyname(_proxyname);
// ensure proxy is no longer active.
vtkSMProxySelectionModel* selmodel = pxm->GetSelectionModel("ActiveSources");
assert(selmodel != NULL);
if (selmodel->GetCurrentProxy() == proxy)
{
selmodel->SetCurrentProxy(NULL, vtkSMProxySelectionModel::CLEAR_AND_SELECT);
}
// remove proxy from TimeKeeper.
vtkSMProxy* timeKeeper = this->FindTimeKeeper(proxy->GetSession());
vtkSMTimeKeeperProxy::RemoveTimeSource(timeKeeper, proxy,
/*unsuppress_input*/ (proxy->GetProperty("TimestepValues") != NULL ||
proxy->GetProperty("TimeRange") != NULL));
// unregister dependencies.
this->UnRegisterDependencies(proxy);
// this will remove both proxy-list-domain helpers and animation helpers.
this->FinalizeProxy(proxy);
// unregister the proxy.
pxm->UnRegisterProxy("sources", proxyname.c_str(), proxy);
return true;
}
```
You will never get to `selmodel->SetCurrentProxy(NULL, vtkSMProxySelectionModel::CLEAR_AND_SELECT);` because `vtkSMOutputPort` proxies are not registered under "sources" AFAIR (or does not have a name).5.13 (Summer 2024)Cory Quammencory.quammen@kitware.comCory Quammencory.quammen@kitware.com