ebrahim
(Ebrahim Ebrahim)
October 21, 2026, 5:09pm
1
Title: 2026.10.27 Weekly Meeting
Next Tuesday, we will be having our next weekly hangout at 10:00 AM ET until 10:30 AM ET.
(Note the different end time!)
Anyone is welcome to join at this link: Google Meet meeting
Agenda:
Please post to this thread to put a topic on the agenda! We will try to prioritize agenda items during the meeting.
Thanks
Sam and Ebrahim
ebrahim
(Ebrahim Ebrahim)
October 21, 2026, 5:10pm
2
This one will end earlier at 10:30 AM ET .
ebrahim
(Ebrahim Ebrahim)
October 26, 2026, 4:57pm
3
Bump: this meeting will be tomorrow
Note the earlier end time
mau_igna_06
(Mauro I. Dominguez)
October 26, 2026, 9:40pm
4
Iβd like to request reviewers to check this PRs so, if needed, questions can be answered on the meeting as Iβd like to get this work merged:
master β mauigna06:add_mesh_checker_feature
opened 05:48PM - 23 Oct 26 UTC
Now it's possible to check meshes are valid before doing the boolean operation:
β¦
<img width="2556" height="1437" alt="Screenshot from 2026-10-23 12-30-14" src="https://github.com/user-attachments/assets/49e05a49-1288-499b-bb2c-c71d87047706" />
Boolean operations require closed, manifold, consistently oriented input meshes without degenerate or self-intersecting faces, and give invalid results otherwise. The new Input check section checks the input models before the result is computed. "Check inputs" shows a report per model, "Check inputs before applying" (on by default) asks whether to compute the result if any input has problems, and "Show defects in the views" (on by default) shows the defects of each input in the model "<name> defects".
kigumi finds the defects exactly with vtkKigumiMeshDefectsFilter, including self-intersections, which VTK and vtkbool cannot detect in a single mesh. VTK filters (vtkFeatureEdges, vtkPolyDataConnectivityFilter) and a numpy check of the edge directions and the signed volume, as in the vtkbool tests, check the mesh as the selected backend sees it, after merging points with the vtkbool tolerance if vtkbool is selected. Inside-out meshes, and sliver faces that vtkbool cannot handle, are also reported. Models are checked in the coordinate system of the output, as the Boolean operation uses them. CombineModelsLogic.checkInputs() provides the same from Python.
vtkKigumiMeshDefectsFilter and kigumi/vtk_polydata.h are copied from kigumi fork commit 3d7048e.
master β mauigna06:fix-plane-cut
opened 08:31PM - 18 Sep 26 UTC
- Fixes missing end cap on first run of Dynamic Modeler plane cut tool
- Rewrit⦠es the way caps are created to pass a challenging test that fails with current implementation
And if possible also this one:
main β mauigna06:patch-6
> Extension has a reasonable name (not too general, not too narrow, suggests wha⦠t the extension is for). The extension name should not start with Slicer (unless it explicitly provides a bridge between Slicer and a tool or library), because it would make it more difficult to find extensions if the name of many started with the same word.
Where do you write the extension name? Do you mean module name?
> Repository is associated with 3d-slicer-extension GitHub topic
The repository of the extension is hosted on Gitlab
Thanks a lot,
Mauro
ebrahim
(Ebrahim Ebrahim)
October 27, 2026, 2:41pm
6
Notes from the meeting:
Mauroβs pull requests
Extension review
Steve working on extension review dashboard so that reviewers can more easily manage AI review of ExtensionsIndex.
ARM mac
Sam: ARM mac builds are working well now
Also build time on computron has been taken from 10 hours to 2 hours. There were a lot of issues.
The next major task is Qt6 migration, setting up Qt6 across platforms on the build machines.
Qt 6 static build
Andras: the Qt static build launcher has been updated to run from GitHub for all six platforms (arm64 and x64 across Windows, Linux, and Mac)
opened 12:34PM - 21 Apr 26 UTC
## Summary
While using the intel app launcher on arm64 machines mostly works, s⦠ome low-level python packages can misbehave.
## Steps to reproduce
Launch like this:
```
/opt/s2/Slicer-build/Slicer
```
Type this in the python console:
```
>>> import os
>>> os.system("arch")
0
```
and in the terminal you see this:
```
Python console user input: import os
Python console user input: os.system("arch")
i386
0
```
If instead you create a launcher script:
```
./Slicer-build/Slicer --launcher-generate-exec-wrapper-script
```
And launch slicer with the generated script:
```
bash /private/var/folders/yk/qm79wxyd1x38xpn3n046gykh0000gn/T/Slicer-JlUjaY.sh /opt/s2/Slicer-build/bin/Slicer.app/Contents/MacOS/./Slicer
```
then arch is printed as `arm64` as expected.
This became an issue when trying to use `ctypes` on a mac build, which failed like this because it have the wrong architecture.
```
[Python] File "/opt/s2/python-install/lib/python3.12/ctypes/__init__.py", line 397, in __getitem__
[Python] func = self._FuncPtr((name_or_ordinal, self))
[Python] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[Python] AttributeError: dlsym(0x9d937ca0, objc_msgSendSuper_stret): symbol not found
```
## Environment
- Slicer version: Slicer-5.11
- Operating system: Only mac arm systems
AI Policy discussion
main β Thibault-Pelletier:ai-policy
opened 08:23AM - 27 Oct 26 UTC
Adapted from VTK's AI policy:
https://gitlab.kitware.com/vtk/vtk/-/blob/master/A⦠I_POLICY.md
## AI disclosure
- [ ] I did not use AI to generate any content in this pull request
- [x] I used AI to generate content in this pull request, so I have read the [AI policy](https://github.com/Slicer/Slicer/blob/main/AI_POLICY.md) carefully and disclose usage in detail below:
Used to adapt VTK's policy to format it to Github / Slicer usage (and read entirely)
Will have to cover this next time β most of the PR is agreed upon and the main stickining point is whether to require AI co-authorship or to require the opposite.