- User Since
- Aug 1 2016, 12:10 PM (356 w, 1 d)
Dec 19 2016
Nov 18 2016
Aug 2 2016
Pull request: https://github.com/MITK/MITK/pull/132
Pull request that implements the first half of this task (integrating Coverity into the superbuild):
Work-in-progress changes that are dependent on 18393:
I'm not working on it, but I know there are several stakeholders who are very much still interested in it.
I have a work-in-progress solution here that addresses this issue by rendering VTK to an OpenGL Framebuffer Object (FBO) and sharing that FBO with Qt Quick using QQuickFramebufferObject:
One additional note: Run applications with environment variable QML_BAD_GUI_RENDER_LOOP=1 in order to use single-threaded Qt Quick rendering so that you don't need to worry about the thread safety problems, which is a separate issue (T18394)
Pull request: https://github.com/MITK/MITK/pull/88
Can I add a few small patches into this bug?
Sounds good! The CMake generator expressions are a good idea if they'll work here.
Pull request: https://github.com/MITK/MITK/pull/87
Pull request: https://github.com/MITK/MITK/pull/85
Pull request: https://github.com/MITK/MITK/pull/79
Hi Matt, You shouldn't have to do another pull request. In fact, you can (and even should) interactively rebase your topic branch, squashing micro-commits as you see necessary to get a nice set of logical commits ready to be merged into master. Looking forward to getting this enhancement in MITK master...
Thanks! With that explanation it all makes sense. I was coming from the perspective making the Qt build configuration work the same regardless of Qt4 or Qt5 (using QT_QMAKE_EXECUTABLE), but you're right that it doesn't make sense. And anyway, at some point (probably in the distant future?) MITK will drop Qt4 support and only the (better) Qt5 configuration will be needed.
No problem. Thanks for merging and finishing the work! I have a couple questions:
Never mind about the cmake 2.8.12 dependency - I understand now.
GitHub pull request: https://github.com/MITK/MITK/pull/62
GitHub pull request: https://github.com/MITK/MITK/pull/63
GitHub pull request:
Github pull request:
Github pull request:
Github pull request:
Great, thanks for merging! I'll make sure to re-enable the MITK pre-commit hook that I had meant to just temporarily disable.
Github pull request: https://github.com/MITK/MITK/pull/49
Just checking on status of this bug. Have there been any issues getting the pull request merged? I'm hoping the upgrade is transparent for all Qt4-based code.
I've tested with Qt5, but not Qt4. I can test with Qt4, but I figured it would be easier for you since that's your standard development environment (I think?).
Is that patch something general enough to be contributed to qtsolutions?
Uhoh... sounds like a problem since this code is used by nearly every MITK-based application that exists today (But also all the more reason to make sure it gets a thorough review before merging into master)
Thanks for reviewing, Michael. Is there someone else who needs to review and approve the changes to get this merged into master?
I looked into the why question. It seems the reason for the parent node is because the mapper that renders the PlaneGeometryData nodes as crosshairs needs a fast way to know about the other PlaneGeometryData nodes in the scene in order to paint the crosshair lines with a gap at the point where the planes intersect with each other.
Well, some mechanism for knowing about other visible PlaneGeometryData nodes is needed. Perhaps Daniel Maleike wasn't aware of this when he made a TODO note in the QmlMitkFourRenderWindowWidget.cpp about this issue, but I still think there is a more flexible solution that doesn't sacrifice performance.
Here's a pull request that implements a very simple mechanism for knowing about other visible PlaneGeometryData nodes.
FYI - Just updated pull request 83
Pull request for leak fixes in ImageReadAccessor:
Fix memory leak in mitkImageVtkMapper2DLevelWindow unit test:
Another leak fix:
Fix for memory leak in VtkPropRenderer::Initialize2DvtkCamera:
Pull request for several leaks found while running unit tests against a build of VTK with VTK_DEBUG_LEAKS enabled:
Pull request: https://github.com/MITK/MITK/pull/32
Sounds good. I'll watch for T15963 to be closed/merged.
Pull request: https://github.com/MITK/MITK/pull/31
New remote branch pushed: bug-16060-InfiniteLoopInNavigationDataEvaluationFilter
New remote branch pushed: bug-16027-NavigationData-PrintSelf-missing-Orientation
Pull request: https://github.com/MITK/MITK/pull/30
Pull request: https://github.com/MITK/MITK/pull/29
The previous pull request had not been rebased against master for a
while so it was missing a fix for a new DataNode method that has the
same const-correctness problem on the renderer parameter.
Pull request: https://github.com/MITK/MITK/pull/27
Pull request: https://github.com/MITK/MITK/pull/28
New remote branch pushed: bug-15911-AddMissingBaudRates
Pull request: https://github.com/MITK/MITK/pull/26
Actually, I'm just going to split things up properly - there's not that much.
New remote branch pushed: bug-15906-NavigationDataTransformFilterAddPrecomose
Pull request: https://github.com/MITK/MITK/pull/25
Github pull request: https://github.com/MITK/MITK/pull/21
Pull request: https://github.com/MITK/MITK/pull/22
Pull request: https://github.com/MITK/MITK/pull/23
New remote branch pushed: bug-15895-NavigationDataObjectVisualizationFilter-fixes