# Missing prebuilds for IGT extensions for Preview Release 5.9

**URL:** <https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143>\
**Category:** Development\
**Tags:** slicerigt, openigtlink, extensions, preview\
**Created:** [August 20, 2025, 7:08am UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143 "2025-08-20T07:08:36Z")\
**Posts on this page:** 14\
**Page:** 1

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 20, 2025, 7:08am UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/1 "2025-08-20T07:08:36Z")

</div>

Hello, I’m developing slicelet for image guided therapy, particularly a scripted slicelet for simpler R&D. Ideally we’d like to use the Preview Release of Slicer 5.9 that ships with a more recent version of Python 3.12, however, it seems like a number of core IGT extensions are missing as of revision 33856 built 2025-08-19:

- [https://extensions.slicer.org/catalog/IGT/33856/linux](https://extensions.slicer.org/catalog/IGT/33856/linux)

Inspecting the Extensions-Nightly build status for IGT extensions of interest:

- SlicerIGSIO
  - [SlicerPreview - Build Summary](https://slicer.cdash.org/builds/3899313)
  - [GitHub - IGSIO/SlicerIGSIO at 1a89776c9f1c8bbbad62000561aa892afe1e7077](https://github.com/IGSIO/SlicerIGSIO/tree/1a89776c9f1c8bbbad62000561aa892afe1e7077)

- SlicerIGT
  - [SlicerPreview - Build Summary](https://slicer.cdash.org/builds/3899419)
  - [GitHub - SlicerIGT/SlicerIGT at cc9678a9b6f91aa9ffde3ff0176cad357335616f](https://github.com/SlicerIGT/SlicerIGT/tree/cc9678a9b6f91aa9ffde3ff0176cad357335616f)

- SlicerOpenIGTLink
  - [SlicerPreview - Build Summary](https://slicer.cdash.org/builds/3899228)
  - [GitHub - openigtlink/SlicerOpenIGTLink at f806007ccc4b5c7cb6ea14979670a468b39c5a45](https://github.com/openigtlink/SlicerOpenIGTLink/tree/f806007ccc4b5c7cb6ea14979670a468b39c5a45)

The build errors for both SlicerIGSIO and SlicerOpenIGTLink are quite nondescript, while cmake error for SlicerIGT from the missing SlicerIGSIO simple enough.

```auto
[2%] Creating directories for 'YASM'
[5%] Performing download step (git clone) for 'YASM'
Cloning into 'YASM'...
Already on 'master'
Your branch is up to date with 'origin/master'.
gmake[5]: *** read jobs pipe: Bad file descriptor. Stop.
gmake[5]: *** Waiting for unfinished jobs....
[8%] Generate version-YASM.txt and license-YASM.txt

```

Any idea what may be blocking the build server here? Only one test is failing atm:

> **IGT Extension Test Results:**
>
> ```shell
> ruffsl@box ~/w/s/S/inner-build> make test
> Running tests...
> Test project /home/ruffsl/ws/slicer/SlicerIGSIO-Release/inner-build
> Start 1: vtkEncodeUncompressedSequenceTest
> 1/3 Test #1: vtkEncodeUncompressedSequenceTest ................ Passed 0.25 sec
> Start 2: qSlicerMetafileImporterModuleGenericTest
> 2/3 Test #2: qSlicerMetafileImporterModuleGenericTest ......... Passed 2.08 sec
> Start 3: qSlicerMetafileImporterModuleWidgetGenericTest
> 3/3 Test #3: qSlicerMetafileImporterModuleWidgetGenericTest ... Passed 2.33 sec
> 
> 100% tests passed, 0 tests failed out of 3
> 
> Label Time Summary:
> qSlicerMetafileImporterModule = 4.42 sec*proc (2 tests)
> qSlicerVideoUtilModule = 0.25 sec*proc (1 test)
> 
> Total Test time (real) = 4.67 sec
> 
> ```
> 
> ```shell
> ruffsl@box ~/w/s/SlicerIGT-Release> make test
> Running tests...
> Test project /home/ruffsl/ws/slicer/SlicerIGT-Release
> Start 1: qSlicerBreachWarningModuleGenericTest
> 1/36 Test #1: qSlicerBreachWarningModuleGenericTest .................................. Passed 1.67 sec
> Start 2: qSlicerBreachWarningModuleWidgetGenericTest
> 2/36 Test #2: qSlicerBreachWarningModuleWidgetGenericTest ............................ Passed 1.85 sec
> Start 3: py_BreachWarningSelfTest
> 3/36 Test #3: py_BreachWarningSelfTest ............................................... Passed 9.96 sec
> Start 4: qSlicerCollectPointsModuleGenericTest
> 4/36 Test #4: qSlicerCollectPointsModuleGenericTest .................................. Passed 1.56 sec
> Start 5: qSlicerCollectPointsModuleWidgetGenericTest
> 5/36 Test #5: qSlicerCollectPointsModuleWidgetGenericTest ............................ Passed 1.74 sec
> Start 6: qSlicerCreateModelsModuleGenericTest
> 6/36 Test #6: qSlicerCreateModelsModuleGenericTest ................................... Passed 1.56 sec
> Start 7: qSlicerCreateModelsModuleWidgetGenericTest
> 7/36 Test #7: qSlicerCreateModelsModuleWidgetGenericTest ............................. Passed 1.82 sec
> Start 8: qSlicerFiducialRegistrationWizardModuleGenericTest
> 8/36 Test #8: qSlicerFiducialRegistrationWizardModuleGenericTest ..................... Passed 2.35 sec
> Start 9: qSlicerFiducialRegistrationWizardModuleWidgetGenericTest
> 9/36 Test #9: qSlicerFiducialRegistrationWizardModuleWidgetGenericTest ............... Passed 2.55 sec
> Start 10: py_nomainwindow_qSlicerFiducialsToModelRegistrationModuleGenericTest
> 10/36 Test #10: py_nomainwindow_qSlicerFiducialsToModelRegistrationModuleGenericTest ... Passed 2.43 sec
> Start 11: py_FiducialsToModelRegistration
> 11/36 Test #11: py_FiducialsToModelRegistration ........................................ Passed 4.24 sec
> Start 12: py_Guidelet
> 12/36 Test #12: py_Guidelet ............................................................ Passed 3.26 sec
> Start 13: qSlicerLandmarkDetectionModuleGenericTest
> 13/36 Test #13: qSlicerLandmarkDetectionModuleGenericTest .............................. Passed 1.66 sec
> Start 14: qSlicerLandmarkDetectionModuleWidgetGenericTest
> 14/36 Test #14: qSlicerLandmarkDetectionModuleWidgetGenericTest ........................ Passed 1.77 sec
> Start 15: vtkLandmarkDetectionTest
> 15/36 Test #15: vtkLandmarkDetectionTest ............................................... Passed 0.28 sec
> Start 16: qSlicerLandmarkDetectionModuleGenericTest
> 16/36 Test #16: qSlicerLandmarkDetectionModuleGenericTest .............................. Passed 1.56 sec
> Start 17: qSlicerLandmarkDetectionModuleWidgetGenericTest
> 17/36 Test #17: qSlicerLandmarkDetectionModuleWidgetGenericTest ........................ Passed 1.79 sec
> Start 18: py_nomainwindow_qSlicerModelRegistrationModuleGenericTest
> 18/36 Test #18: py_nomainwindow_qSlicerModelRegistrationModuleGenericTest .............. Passed 2.43 sec
> Start 19: py_ModelRegistration
> 19/36 Test #19: py_ModelRegistration ................................................... Passed 4.13 sec
> Start 20: qSlicerPathExplorerModuleGenericTest
> 20/36 Test #20: qSlicerPathExplorerModuleGenericTest ................................... Passed 2.49 sec
> Start 21: qSlicerPathExplorerModuleWidgetGenericTest
> 21/36 Test #21: qSlicerPathExplorerModuleWidgetGenericTest ............................. Passed 2.65 sec
> Start 22: vtkPivotCalibrationTest
> 22/36 Test #22: vtkPivotCalibrationTest ................................................ Passed 0.29 sec
> Start 23: qSlicerPivotCalibrationModuleGenericTest
> 23/36 Test #23: qSlicerPivotCalibrationModuleGenericTest ............................... Passed 1.68 sec
> Start 24: qSlicerPivotCalibrationModuleWidgetGenericTest
> 24/36 Test #24: qSlicerPivotCalibrationModuleWidgetGenericTest ......................... Passed 1.87 sec
> Start 25: py_nomainwindow_qSlicerTextureModelModuleGenericTest
> 25/36 Test #25: py_nomainwindow_qSlicerTextureModelModuleGenericTest ................... Passed 2.41 sec
> Start 26: py_TextureModel
> 26/36 Test #26: py_TextureModel ........................................................ Passed 9.46 sec
> Start 27: qSlicerTransformProcessorModuleGenericTest
> 27/36 Test #27: qSlicerTransformProcessorModuleGenericTest ............................. Passed 1.50 sec
> Start 28: qSlicerTransformProcessorModuleWidgetGenericTest
> 28/36 Test #28: qSlicerTransformProcessorModuleWidgetGenericTest ....................... Passed 1.69 sec
> Start 29: qSlicerUltrasoundSnapshotsModuleGenericTest
> 29/36 Test #29: qSlicerUltrasoundSnapshotsModuleGenericTest ............................ Passed 1.65 sec
> Start 30: qSlicerUltrasoundSnapshotsModuleWidgetGenericTest
> 30/36 Test #30: qSlicerUltrasoundSnapshotsModuleWidgetGenericTest ...................... Passed 1.84 sec
> Start 31: py_nomainwindow_qSlicerViewpointModuleGenericTest
> 31/36 Test #31: py_nomainwindow_qSlicerViewpointModuleGenericTest ...................... Passed 2.51 sec
> Start 32: py_Viewpoint
> 32/36 Test #32: py_Viewpoint ........................................................... Passed 3.03 sec
> Start 33: qSlicerWatchdogModuleGenericTest
> 33/36 Test #33: qSlicerWatchdogModuleGenericTest ....................................... Passed 1.70 sec
> Start 34: qSlicerWatchdogModuleWidgetGenericTest
> 34/36 Test #34: qSlicerWatchdogModuleWidgetGenericTest ................................. Passed 1.78 sec
> Start 35: py_nomainwindow_qSlicerSequenceReplayModuleGenericTest
> 35/36 Test #35: py_nomainwindow_qSlicerSequenceReplayModuleGenericTest ................. Passed 2.38 sec
> Start 36: py_SequenceReplay
> 36/36 Test #36: py_SequenceReplay ...................................................... Passed 3.30 sec
> 
> 100% tests passed, 0 tests failed out of 36
> 
> Label Time Summary:
> qSlicerBreachWarningModule = 3.53 sec*proc (2 tests)
> qSlicerCollectPointsModule = 3.31 sec*proc (2 tests)
> qSlicerCreateModelsModule = 3.38 sec*proc (2 tests)
> qSlicerFiducialRegistrationWizardModule = 4.90 sec*proc (2 tests)
> qSlicerLandmarkDetectionModule = 7.07 sec*proc (5 tests)
> qSlicerPathExplorerModule = 5.14 sec*proc (2 tests)
> qSlicerPivotCalibrationModule = 3.84 sec*proc (3 tests)
> qSlicerTransformProcessorModule = 3.18 sec*proc (2 tests)
> qSlicerUltrasoundSnapshotsModule = 3.49 sec*proc (2 tests)
> qSlicerWatchdogModule = 3.48 sec*proc (2 tests)
> 
> Total Test time (real) = 90.85 sec
> 
> ```
> 
> ```shell
> ruffsl@box ~/w/s/S/inner-build> make test
> Running tests...
> Test project /home/ruffsl/ws/slicer/SlicerOpenIGTLink-Release/inner-build
> Start 1: vtkMRMLConnectorCommandSendAndReceiveTest
> 1/8 Test #1: vtkMRMLConnectorCommandSendAndReceiveTest .........***Failed 15.28 sec
> Start 2: vtkMRMLConnectorImageSendAndReceiveTest
> 2/8 Test #2: vtkMRMLConnectorImageSendAndReceiveTest ........... Passed 10.36 sec
> Start 3: qSlicerOpenIGTLinkRemoteModuleGenericTest
> 3/8 Test #3: qSlicerOpenIGTLinkRemoteModuleGenericTest ......... Passed 2.09 sec
> Start 4: qSlicerOpenIGTLinkRemoteModuleWidgetGenericTest
> 4/8 Test #4: qSlicerOpenIGTLinkRemoteModuleWidgetGenericTest ... Passed 2.32 sec
> Start 5: qSlicerOpenIGTLinkRemoteModuleGenericTest
> 5/8 Test #5: qSlicerOpenIGTLinkRemoteModuleGenericTest ......... Passed 2.02 sec
> Start 6: qSlicerOpenIGTLinkRemoteModuleWidgetGenericTest
> 6/8 Test #6: qSlicerOpenIGTLinkRemoteModuleWidgetGenericTest ... Passed 2.35 sec
> Start 7: qSlicerPlusRemoteModuleGenericTest
> 7/8 Test #7: qSlicerPlusRemoteModuleGenericTest ................ Passed 2.12 sec
> Start 8: qSlicerPlusRemoteModuleWidgetGenericTest
> 8/8 Test #8: qSlicerPlusRemoteModuleWidgetGenericTest .......... Passed 2.32 sec
> 
> 88% tests passed, 1 tests failed out of 8
> 
> Label Time Summary:
> SlicerOpenIGTLink = 25.64 sec*proc (2 tests)
> qSlicerOpenIGTLinkRemoteModule = 8.77 sec*proc (4 tests)
> qSlicerPlusRemoteModule = 4.44 sec*proc (2 tests)
> 
> Total Test time (real) = 38.86 sec
> 
> The following tests FAILED:
> 1 - vtkMRMLConnectorCommandSendAndReceiveTest (Failed)
> Errors while running CTest
> Output from these tests are in: /home/ruffsl/ws/slicer/SlicerOpenIGTLink-Release/inner-build/Testing/Temporary/LastTest.log
> Use "--rerun-failed --output-on-failure" to re-run the failed cases verbosely.
> make: *** [Makefile:91: test] Error 8
> ruffsl@box ~/w/s/S/inner-build [2]> make test ARGS="--rerun-failed --output-on-failure"
> Running tests...
> Test project /home/ruffsl/ws/slicer/SlicerOpenIGTLink-Release/inner-build
> Start 1: vtkMRMLConnectorCommandSendAndReceiveTest
> 1/1 Test #1: vtkMRMLConnectorCommandSendAndReceiveTest ...***Failed 0.42 sec
> SUCCESS: connected to server
> error: [/home/ruffsl/ws/slicer/SlicerOpenIGTLink-Release/inner-build/bin/SlicerOpenIGTLinkCxxTests] exit abnormally - Report the problem.
> 
> 0% tests passed, 1 tests failed out of 1
> 
> Label Time Summary:
> SlicerOpenIGTLink = 0.42 sec*proc (1 test)
> 
> Total Test time (real) = 0.42 sec
> 
> The following tests FAILED:
> 1 - vtkMRMLConnectorCommandSendAndReceiveTest (Failed)
> Errors while running CTest
> make: *** [Makefile:91: test] Error 8
> 
> ```

* * *

Following the docs, I’ve gone ahead and compiled 3D Slicer from source using Ubuntu 24.04, including the IGT extensions from the HEADs of their respective default branches, as linked above. All seems to build and run fine as show (unsure if SlicerIGSIO is tested here):

> **[simplescreenrecorder-2025-08-19\_19.39.42.webm](https://drive.google.com/file/d/1PGeVYoqFwHGzAL4wC2pgmfMllQj_oYk9/view?usp=sharing)**
>
> Google Drive file.

Example scene attached, sans the `CTChest.nrrd` from downloaded sample data:

> **[4up\_Tracking\_RIO.zip](https://drive.google.com/file/d/18sx1h9j2NFOkWeo37aKfhZh5Q1Wj3Bhd/view?usp=sharing)**
>
> Google Drive file.

And synthetic transform publisher for animating the example demo:

```python
"""
============================
Position sending server
============================

Simple application that starts a server that provides a stream of random positions.

from slicer.util import pip_install
pip_install("pyigtl")

Adapted from:
https://github.com/lassoan/pyigtl/blob/d28132d8e9e1cef4ac531d795f5284c8aa739408/examples/example_position_server.py
https://github.com/lassoan/pyigtl/blob/d28132d8e9e1cef4ac531d795f5284c8aa739408/pyigtl/tests/test_basic_comm.py#L87-L93
"""

import pyigtl # pylint: disable=import-error

from time import sleep
import numpy as np
from scipy.spatial.transform import Rotation as R

server = pyigtl.OpenIGTLinkServer(port=18944)

transverse180 = np.array([
    [-1.0, 0.0, 0.0, 0.0],
    [0.0, 1.0, 0.0, 0.0],
    [0.0, 0.0, -1.0, 0.0],
    [0.0, 0.0, 0.0, 1.0]
])

timestep = 0
while True:
    if not server.is_connected():
        # Wait for client to connect
        sleep(0.1)
        continue

    # Generate periodic trajectory: Lissajous curve on sphere
    timestep += 1
    radius = 50.0
    center = np.array([0.0, 0.0, -100.0])
    # Use two periodic angles for smooth looping
    phi = (timestep * 0.001) % (2 * np.pi) # azimuthal
    theta = (timestep * 0.002) % (2 * np.pi) # polar
    x = center[0] + radius * np.sin(theta) * np.cos(phi)
    y = center[1] + radius * np.sin(theta) * np.sin(phi)
    z = center[2] + radius * np.cos(theta)
    position = np.array([x, y, z])

    # Orientation: always point toward center
    forward = center - position
    forward = forward / np.linalg.norm(forward)
    up = np.array([0.0, 1.0, 0.0])
    if np.abs(np.dot(forward, up)) > 0.99:
        up = np.array([1.0, 0.0, 0.0])
    rot_obj, _ = R.align_vectors([forward, up], [[0, 0, 1], [0, 1, 0]])
    rot_matrix = rot_obj.as_matrix()
    # Build 4x4 transform
    transform = np.eye(4)
    transform[:3, :3] = rot_matrix
    transform[:3, 3] = position

    transform_message = pyigtl.TransformMessage(transform, device_name='ToolToRAS')
    server.send_message(transform_message, wait=True)
    transform_message = pyigtl.TransformMessage(transverse180, device_name='TipToTool')
    server.send_message(transform_message, wait=True)
    # Since we wait until the message is actually sent, the message queue will not be flooded

```

---

<div class="post-metadata">

**Author:** ![RafaelPalomar](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/rafaelpalomar/32/1436_2.png) [@RafaelPalomar](https://discourse.slicer.org/u/RafaelPalomar)\
**Post date:** [August 20, 2025, 3:48pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/2 "2025-08-20T15:48:53Z")

</div>

@ruffsl the infrastructure for building extensions on GNU/Linux has been updated recently. Some extensions seem to have been affected by this update ( [Slicer Build Environment Upgraded to `qt5-almalinux8-gcc14` - #6 by fedorov](https://discourse.slicer.org/t/slicer-build-environment-upgraded-to-qt5-almalinux8-gcc14/43802/6) ). Our colleagues from Kitware are aware of this issue and will check on it soon.

While extensions such as SlicerSOFA and DCMQI seem to fail only in the GNU/Linux build system, some other extension may fail legitimately because they need an update to adapt to the new build system. To test in which case the IGT extensions fall into, it is not enough to build them in your own system. A good test to discriminate whether the extension needs an update or the build system needs and adjustment, you would need to build Slicer and the extensions using Slicer building environment. @ruffsl, would you have the availability to give the extensions a try with [GitHub - Slicer/SlicerBuildEnvironment: A repository of scripts to set up a Slicer build environment.](https://github.com/Slicer/SlicerBuildEnvironment) using the `qt5-almalinux8-gcc14` environment? This would help speeding up the diagnostic and eventual resolution of the problem.

@jcfr, @Sam_Horvath, @pieper, @lassoan

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 21, 2025, 2:01am UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/3 "2025-08-21T02:01:25Z")

</div>

> [@RafaelPalomar](#):
>
> would you have the availability to give the extensions a try with [GitHub - Slicer/SlicerBuildEnvironment: A repository of scripts to set up a Slicer build environment.](https://github.com/Slicer/SlicerBuildEnvironment) using the `qt5-almalinux8-gcc14` environment?

I gave it a go using the reference in the readme, see my exact script included below:

> **Build Script**
>
> ```bash
> #!/usr/bin/env bash
> 
> ################################################################################
> # MARK: Build Slicer
> ################################################################################
> 
> SLICER_BUILD_ENV_NAME=qt5-almalinux8-gcc14 
> SLICER_BUILD_ENV_VERSION=5.9
> SLICER_BRANCH=main
> 
> ROOT_DIR=/tmp/Slicer-$SLICER_BUILD_ENV_VERSION
> mkdir -p $ROOT_DIR
> 
> cd ${ROOT_DIR}
> 
> slicer_build_env_script=/tmp/bin/slicer-buildenv-$SLICER_BUILD_ENV_NAME-$SLICER_BUILD_ENV_VERSION
> mkdir -p $(dirname $slicer_build_env_script)
> 
> # Download sources
> git clone https://github.com/Slicer/Slicer -b $SLICER_BRANCH
> 
> # Download corresponding build environment and generate convenience script
> docker run --rm slicer/buildenv-qt5-almalinux8-gcc14 > $slicer_build_env_script
> chmod u+x $slicer_build_env_script
> 
> # Configure Slicer
> $slicer_build_env_script cmake \
> -BSlicer-build \
> -SSlicer \
> -GNinja \
> -DCMAKE_BUILD_TYPE:STRING=Release
> 
> # Build Slicer
> $slicer_build_env_script cmake --build Slicer-build
> 
> # Package Slicer
> $slicer_build_env_script cmake --build Slicer-build/Slicer-build --target package
> 
> ################################################################################
> # MARK: Build Extension
> ################################################################################
> 
> EXTENSION_NAME=SlicerOpenIGTLink
> 
> # Download extension source
> git clone https://github.com/openigtlink/SlicerOpenIGTLink.git ${EXTENSION_NAME}
> 
> # Configure the extension
> $slicer_build_env_script cmake \
> -B${EXTENSION_NAME}-build \
> -S${EXTENSION_NAME} \
> -GNinja \
> -DCMAKE_BUILD_TYPE:STRING=Release \
> -DSlicer_DIR:PATH=/work/Slicer-build/Slicer-build
> 
> # Build the extension
> $slicer_build_env_script cmake --build ${EXTENSION_NAME}-build
> 
> # Package the extension
> $slicer_build_env_script cmake --build ${EXTENSION_NAME}-build --target package
> 
> ```

However, the build for the extension failed:

```auto
[24/26] No install step for 'inner'
[26/26] Completed 'inner'
ninja: error: unknown target 'package'

```

> <https://pastebin.com/MYY5NxNs>

* * *

I will note that one thing that tripped my up from the docs while packaging the exetentions from my local source builds on my Ubuntu host was that `make package` would not work as expected from the toplevel build folder for for either `SlicerIGSIO` and `SlicerOpenIGTLink`,

- [https://slicer.readthedocs.io/en/latest/developer\_guide/extensions.html#create-an-extension-package](https://slicer.readthedocs.io/en/latest/developer_guide/extensions.html#create-an-extension-package)

Instead I had to `cd` into `SlicerOpenIGTLink-Release/inner-build` and `SlicerIGSIO-Release/inner-build` to find the Makefile with the expected `package` function call.

> <https://github.com/IGSIO/SlicerIGSIO/blob/1a89776c9f1c8bbbad62000561aa892afe1e7077/CMakeLists.txt#L14-L15>

> <https://github.com/openigtlink/SlicerOpenIGTLink/blob/f806007ccc4b5c7cb6ea14979670a468b39c5a45/CMakeLists.txt#L14-L15>

This isn’t something I encountered when building `SlicerIGT` from source.

> <https://github.com/SlicerIGT/SlicerIGT/blob/cc9678a9b6f91aa9ffde3ff0176cad357335616f/SuperBuild.cmake#L58-L59>

* * *

Also, it looks like I can’t run the finished Slicer build from outside the docker image on my host:

```shell
ruffsl@box /t/S/S/Slicer-build> ./Slicer --version
/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./SlicerApp-real: error while loading shared libraries: libCTKQtTesting.so.0.1: cannot open shared object file: No such file or directory

```

Although I suspect that’s as expected without these new different QT 5 libs installed locally?

---

<div class="post-metadata">

**Author:** ![RafaelPalomar](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/rafaelpalomar/32/1436_2.png) [@RafaelPalomar](https://discourse.slicer.org/u/RafaelPalomar)\
**Post date:** [August 23, 2025, 5:33pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/4 "2025-08-23T17:33:21Z")

</div>

@ruffsl, thanks for having a look at it so quickly. And very good investigation job!

> [@ruffsl](#):
>
> `ninja: error: unknown target 'package'`

This is expected since the SlicerOpenIGTLink is a superbuild extension. Superbuild extensions build first an ecosystem of libraries to support the build of the extension, which as you found out happens in `inner-build`. The script should probably account for that and issue:

```auto
$slicer_build_env_script cmake --build ${EXTENSION_NAME}-build/inner-build --target package

```

–

> [@ruffsl](#):
>
> Also, it looks like I can’t run the finished Slicer build from outside the docker image on my host:
> 
> ```auto
> ruffsl@box /t/S/S/Slicer-build> ./Slicer --version
> /tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./SlicerApp-real: error while loading shared libraries: libCTKQtTesting.so.0.1: cannot open shared object file: No such file or directory
> 
> ```
> 
> Although I suspect that’s as expected without these new different QT 5 libs installed locally?

Yes, that’s what I think. To fully verify you could try executing 3D Slicer inside the container, but I don’t think it is needed. You have successfully build the SlicerOpenIGTLink, with the Slicer environment, which does not happen in the Slicer infrastructure.

Let’s report all of this as an issue in 3D Slicer (if there’s not one yet) and continue the discussion from there. Send me a private with your github handle, if you want to be notified and participate there. I would be useful if you could give it a try to SlicerIGSIO and SlicerIGT too, as they seem to fail in different ways (in reality, SlicerIGT seems to fail due to SlicerIGSIO not being built).

Thank you for the great insight you have provided on the problem.

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 25, 2025, 7:02pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/5 "2025-08-25T19:02:49Z")

</div>

> [@RafaelPalomar](#):
>
> The script should probably account for that and issue:

```diff
-$slicer_build_env_script cmake --build ${EXTENSION_NAME}-build --target package
+$slicer_build_env_script cmake --build ${EXTENSION_NAME}-build/inner-build --target package

```

A case by case change, or something should `slicer_build_env_script` auto detect?

That worked for packaging `SlicerOpenIGTLink` via `qt5-almalinux8-gcc14`:

```auto
[24/26] No install step for 'inner'
[26/26] Completed 'inner'

real 0m36.421s
user 0m0.046s
sys 0m0.053s
[0/1] Run CPack packaging tool...
CPack: Create package using TGZ
CPack: Install projects
CPack: - Install project: SlicerOpenIGTLink []
CPack: - Install project: OpenIGTLink []
CPack: - Install project: OpenIGTLinkIO []
CPack: Create package
CPack: - package: /work/SlicerOpenIGTLink-build/inner-build/33872-linux-amd64-SlicerOpenIGTLink-gitf806007-2025-07-26.tar.gz generated.

```

> [@RafaelPalomar](#):
>
> would be useful if you could give it a try to SlicerIGSIO and SlicerIGT too

Looks like they both build packages, but as before, can’t run this Slicer on host to test:

```auto
[31/34] No install step for 'inner'
[34/34] Completed 'inner'

real 1m11.141s
user 0m0.031s
sys 0m0.056s
[0/1] Run CPack packaging tool...
CPack: Create package using TGZ
CPack: Install projects
CPack: - Install project: SlicerIGSIO []
CPack: - Install project: IGSIO []
CPack: Create package
CPack: - package: /work/SlicerIGSIO-build/inner-build/33872-linux-amd64-SlicerIGSIO-git1a89776-2023-10-03.tar.gz generated.

```

```diff
EXTENSION_NAME=SlicerIGT

# Download extension source
git clone https://github.com/SlicerIGT/SlicerIGT.git ${EXTENSION_NAME}

# Configure the extension
$slicer_build_env_script cmake \
  -B${EXTENSION_NAME}-build \
  -S${EXTENSION_NAME} \
  -GNinja \
  -DCMAKE_BUILD_TYPE:STRING=Release \
  -DSlicer_DIR:PATH=/work/Slicer-build/Slicer-build \
+ -DSlicerIGSIO_DIR:PATH=/work/SlicerIGSIO-build/inner-build

# Build the extension
time $slicer_build_env_script cmake --build ${EXTENSION_NAME}-build | tee extension_build.log

# Package the extension
$slicer_build_env_script cmake --build ${EXTENSION_NAME}-build --target package

```

```auto
real 0m30.716s
user 0m0.045s
sys 0m0.062s
[0/1] Run CPack packaging tool...
CPack: Create package using TGZ
CPack: Install projects
CPack: - Install project: SlicerIGT []
CPack: Create package
CPack: - package: /work/SlicerIGT-build/33872-linux-amd64-SlicerIGT-gitcc9678a-2025-08-19.tar.gz generated.

```

> [@RafaelPalomar](#):
>
> Let’s report all of this as an issue in 3D Slicer (if there’s not one yet) and continue the discussion from there.

On which repo issue tracker in particular? Of these?

- [GitHub - Slicer/Slicer: Multi-platform, free open source software for visualization and image computing.](https://github.com/Slicer/Slicer)
- [GitHub - Slicer/DashboardScripts: Collection of scripts used to configure/build/test/package 3D Slicer and associated extensions](https://github.com/Slicer/DashboardScripts)
- [GitHub - Slicer/SlicerBuildEnvironment: A repository of scripts to set up a Slicer build environment.](https://github.com/Slicer/SlicerBuildEnvironment)

> [@RafaelPalomar](#):
>
> your github handle, if you want to be notified and participate there.

My user handle on GitHub is the same as here:

- [ruffsl (Ruffin) · GitHub](https://github.com/ruffsl)

---

<div class="post-metadata">

**Author:** ![lassoan](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/lassoan/32/13_2.png) [@lassoan](https://discourse.slicer.org/u/lassoan)\
**Post date:** [August 25, 2025, 7:40pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/6 "2025-08-25T19:40:18Z")

</div>

If you used the official Slicer build environment to build the package then you can install that package in the matching official Slicer Preview Release.

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 25, 2025, 10:50pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/7 "2025-08-25T22:50:27Z")

</div>

> [@lassoan](#):
>
> you can install that package in the matching official Slicer Preview Release.

The source checkout of Slicer I used with the build environment was the current `HEAD` of the `main` branch (`b5a65a49a08171276f6a9624999450c0e8ec0116`). How can discover what commit was used in the compilation for a given build jobs. E.g:

- [SlicerPreview - Build Summary](https://slicer.cdash.org/builds/3905813)
- [SlicerPreview - Uploads](https://slicer.cdash.org/build/3905813/files)

It looks like the `Help -> About 3D Slicer` menu does display a valid commit sha in the version string (but not in `Slicer --launcher-version` oddly enough):

```auto
5.9.0-2025-08-22 r33871 / c7e7ed1

```

- [ENH: Match color label title font size to legend font size · Slicer/Slicer@c7e7ed1 · GitHub](https://github.com/Slicer/Slicer/commit/c7e7ed13f809b6af859af7aae3e7736d72057a5f)

But I can’t find the corresponding job for commit `c7e7ed1`, nor what commit a given job corresponds without downloading and inspecting each separate binary archive.

I’d just like to be sure I’m using the exact matching source trees for apples to apples.

That said, it looks to be installing & working, so I suppose the extensions built on top of `b5a65a4` are close enough to `c7e7ed1` in terms of API/ABI at the moment.

* * *

This all aside, it is concerning that the official Slicer build environment when used locally is not outputting an equivalent running binary as the remote buildfarm does.

---

<div class="post-metadata">

**Author:** ![jamesobutler](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/jamesobutler/32/7511_2.png) [@jamesobutler](https://discourse.slicer.org/u/jamesobutler)\
**Post date:** [August 26, 2025, 12:29am UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/8 "2025-08-26T00:29:41Z")

</div>

The Slicer Preview across all platforms is built daily starting at 11pm ET, so the latest commit on `main` is what is used for the build. You can see at [SlicerPreview](https://slicer.cdash.org/index.php?project=SlicerPreview) the commit hash that was checked out for each platform. As of writing this `c7e7ed1`was the latest commit when it was triggered last night. The `b5a65a4` was the next commit which came several hours later in the morning ET. Slicer Previews are built once daily by the factory machines and are not built and packaged for each commit to `main`.

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 26, 2025, 3:26pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/9 "2025-08-26T15:26:53Z")

</div>

> [@jamesobutler](#):
>
> Slicer Previews are built once daily by the factory machines and are not built and packaged for each commit to `main`.

Now understood; not every commit may be scheduled for daily build.

Thanks for pointing out the `Revision` column in the `Nightly-Packages` table includes the Git commit sha. I was fruitlessly looking for the sha under the job’s own page. From the dashboard URLs, I see the build stores this revision info under the slug `/update`.

```diff
-https://slicer.cdash.org/builds/3906947
+https://slicer.cdash.org/builds/3906947/update

```

However, where is this revision info for extensions jobs under `Extensions-Nightly`?

- [https://slicer.cdash.org/builds/3907251/update](https://slicer.cdash.org/builds/3907251/update)

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 26, 2025, 4:48pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/10 "2025-08-26T16:48:00Z")

</div>

> [@ruffsl](#):
>
> This all aside, it is concerning that the official Slicer build environment when used locally is not outputting an equivalent running binary as the remote buildfarm does.

In searching for resources on the missing `libCTKQtTesting.so` library, in found:

> [@Slicer-5.2.2-linux-amd64/Slicer does not launch](https://discourse.slicer.org/t/slicer-5-2-2-linux-amd64-slicer-does-not-launch/31318/4):
>
> Hi @tbillah - it looks like you didn’t run the ldd inside the shell launched by the Slicer launcher. Try this: ubuntu@sdp-console:/home/pieper$ cd ~/Downloads/Slicer-5.4.0-linux-amd64 ubuntu@sdp-console:~/Downloads/Slicer-5.4.0-linux-amd64$ ./Slicer --launch ldd bin/SlicerApp-real linux-vdso.so.1 (0x00007ffc31a93000) libqSlicerApp.so =\> /home/ubuntu/Downloads/Slicer-5.4.0-linux-amd64/bin/../lib/Slicer-5.4/libqSlicerApp.so (0x00007fc7d156e000) libqSlicerBaseQTApp.so =\> …

An sure enough, `libCTKQtTesting.so` was merely just the first shared lib not found:

```shell
~/t/S/S/Slicer-build> ./Slicer --launch ldd ./bin/SlicerApp-real | head
        linux-vdso.so.1 (0x0000773d3e11d000)
        libqSlicerApp.so => /home/ruffsl/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./libqSlicerApp.so (0x0000773d3e073000)
        libqSlicerBaseQTApp.so => /home/ruffsl/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./libqSlicerBaseQTApp.so (0x0000773d3e01f000)
        libqSlicerModulesCore.so => /home/ruffsl/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./libqSlicerModulesCore.so (0x0000773d3dfff000)
        libqSlicerBaseQTCLI.so => /home/ruffsl/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./libqSlicerBaseQTCLI.so (0x0000773d3df9d000)
        libqSlicerBaseQTGUI.so => /home/ruffsl/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./libqSlicerBaseQTGUI.so (0x0000773d3dc00000)
        libqMRMLWidgets.so => /home/ruffsl/tmp/Slicer-5.9/Slicer-build/Slicer-build/bin/./libqMRMLWidgets.so (0x0000773d3d800000)
        libCTKQtTesting.so.0.1 => not found
        libCTKVisualizationVTKWidgets.so.0.1 => not found
        libCTKScriptingPythonWidgets.so.0.1 => not found

```

After downloading and inspecting the buildfarm artifacts for the same commit, noticed the file tree was different compared to my local cmake build folder, and that I was attempting to run the `Slicer` binary directly, rather than from the extracted launcher binary from the generated `.tar.gz` archive nested inside the very same local build.

Sure enough, the packaged version is working as expected without any missing libs:

```auto
~/t/S/S/Slicer-build/Slicer-5.9.0-2025-08-25-linux-amd64$ ./Slicer \
    --launch ldd ./bin/SlicerApp-real | grep "not found"
~/t/S/S/Slicer-build/Slicer-5.9.0-2025-08-25-linux-amd64$

```

Is this the Cmake macro triggered by this corresponding command? Cpack is new to me.

```auto
$slicer_build_env_script cmake --build Slicer-build/Slicer-build --target package

```

> <https://github.com/Slicer/Slicer/blob/284df9431ec0dd02db225494de9067898a6a43a4/CMake/SlicerPackageAndUploadTarget.cmake#L142-L146>

---

<div class="post-metadata">

**Author:** ![jamesobutler](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/jamesobutler/32/7511_2.png) [@jamesobutler](https://discourse.slicer.org/u/jamesobutler)\
**Post date:** [August 26, 2025, 4:49pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/11 "2025-08-26T16:49:44Z")

</div>

#### **33881-SlicerIGT-gitcc9678a-g+±64bits-Qt5.15-Release**

^This was building commit [cc9678a](https://github.com/SlicerIGT/SlicerIGT/commit/cc9678a9b6f91aa9ffde3ff0176cad357335616f) of SlicerIGT repo using the SlicerPreview build of Slicer core that was built.

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 26, 2025, 4:55pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/12 "2025-08-26T16:55:59Z")

</div>

How is that association traced? Or just fuzzy search via timestamps? Git only timestamps commits creation, not when the commit was push to a remote branch.

---

<div class="post-metadata">

**Author:** ![jamesobutler](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/jamesobutler/32/7511_2.png) [@jamesobutler](https://discourse.slicer.org/u/jamesobutler)\
**Post date:** [August 26, 2025, 5:01pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/13 "2025-08-26T17:01:21Z")

</div>

SlicerIGT’s extension index entry specifies to use the latest of the `master` branch as stated [here](https://github.com/Slicer/ExtensionsIndex/blob/3d062ee7cd56253dc70305d20b8fee1031ba865f/SlicerIGT.json#L8). So at build time it will pull the latest which in that specific case was [cc9678a](https://github.com/SlicerIGT/SlicerIGT/commit/cc9678a9b6f91aa9ffde3ff0176cad357335616f). Git has both authored date and commit date, and in this situation was pulling the latest of the `master` branch at build time.

---

<div class="post-metadata">

**Author:** ![ruffsl](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/ruffsl/32/80858_2.png) [@ruffsl](https://discourse.slicer.org/u/ruffsl)\
**Post date:** [August 28, 2025, 2:42pm UTC](https://discourse.slicer.org/t/missing-prebuilds-for-igt-extensions-for-preview-release-5-9/44143/14 "2025-08-28T14:42:20Z")

</div>

> [@ruffsl](#):
>
> > [@RafaelPalomar](#):
> >
> > Let’s report all of this as an issue in 3D Slicer (if there’s not one yet) and continue the discussion from there.
> 
> On which repo issue tracker in particular? Of these?
> 
> - [GitHub - Slicer/Slicer: Multi-platform, free open source software for visualization and image computing.](https://github.com/Slicer/Slicer)
> - [GitHub - Slicer/DashboardScripts: Collection of scripts used to configure/build/test/package 3D Slicer and associated extensions](https://github.com/Slicer/DashboardScripts)
> - [GitHub - Slicer/SlicerBuildEnvironment: A repository of scripts to set up a Slicer build environment.](https://github.com/Slicer/SlicerBuildEnvironment)

@RafaelPalomar , was a ticket created, our should I add it?