A BlueBerry.cleaned MitkWorkbench does not start since the VTK 6.2 upgrade. This error is reproducable and seems to occur somewhere in the VTK factory mechanism.
Description
Description
Event Timeline
Comment Actions
The factory which provokes the crash is created by the RenderWindow of ACVD.
When blueberry loads the remeshing plugin for the first time, a vtkRenderingContextOpenGLObjectFactory will be registered by the renderwindow.
Once it is registered, it will crash as soon as some vtkObject tries to use it to instantiate itself.
The factory is never actually needed, so preventing ACVD from creating it would solve the problem.
Comment Actions
User kolbch has pushed new remote branch:
bug-18878-UpdateRenderwindowAfterDepartmentLogoChange
Comment Actions
(In reply to Git Admin from comment #2)
User kolbch has pushed new remote branch:
bug-18878-UpdateRenderwindowAfterDepartmentLogoChange
ignore this. Wrong bug.
Comment Actions
As we planned to implement our own version of the ACVD algorithm to make it more robust to errors and since we are in the feature freeze phase, I'll fix this bug with a workaround: The renderwindow of the ACVD library is replaced by a stub.
Comment Actions
[e3df00]: Merge branch 'bug-18878-PatchACVDWorkaround'
Merged commits:
2015-04-08 14:14:25 Stefan Kislinskiy [8314c0]
Replaced ACVD RenderWindow class with stub.
Patch was applied before the tarball was generated.