The Apps are pushed into:
personal/kilgus/StreamingTestApps

Closed. I explained all my findings to the MITK group on the last retreat.

To measure the FPS, I always used this code snipped:

[909511]: Merge branch 'bug-19085-FixOpenCL'

Merged commits:

2015-05-20 13:28:59 Thomas Kilgus [245a8f]
Use correct macro for OclBinaryThresholdImageFilter.

If I start a second App via command line, the second one crashes instead of giving the focus to the first App.

Although I did not look into the code, this should be pretty easy to implement. Did you do any work already or looked into it? Would you like to contribute it yourself? I would gladly do a code review and the merge.

Backface rendering of last release and current master

Backface rendering of last release and current master

This seems to be a VTK issue. I added a bug report in their Mantis bug tracker:
http://www.vtk.org/Bug/view.php?id=15433

For the upcoming release (15-03) this will be declared as a known issue. Each user can temporarily fix the anti aliasing with his graphic driver settings.

Bisect result:

Stack trace of the crash

I reproduced this bug with Peter by removing the line:

node->AddProperty("Image.Displayed Component", mitk::IntProperty::New(0), renderer, overwrite);

in mitkImageVtkMapper2D.cpp

Stack trace of the crash

On which platforms does this occur? Only Arch Linux? I have no issues with Ubuntu 14.04.

[96ce99]: Merge branch 'bug-18762-DocuGeometryTools'

Merged commits:

2015-02-24 13:02:29 Thomas Kilgus [0d7c4b]
Added docu for geometry tools plugin.

GeometryTools Plugin docu should be ok now.

[179afe]: Merge branch 'bug-18728-FixExportMacro'

Merged commits:

2015-02-10 14:59:33 Thomas Kilgus [b47b41]
COMP: Fixed export macro.

[a2558b]: Merge branch 'bug-18728-StructureWidget'

Merged commits:

2015-02-10 10:02:41 Thomas Kilgus [65694d]
Added StructureParameterWidget.

I think we should add a NULL check in the Update() (e.g. if input->GetVtkPolyData() == NULL) return;), similar like the ImageVtkMapper2D does it. (I assumed there was also one in the surface mapper, but apparently there is not.)

I would like to merge this bug asap due to lots of potential for conflicts. Unfortunately, we modified the core (only mitkIOUtil).

[b6d54e]: Merge branch 'bug-18651-RemoveLegacyIO'

Merged commits:

2015-01-29 13:12:45 Thomas Kilgus [4c2eb7]
Converted ManualSegmentationToSurfaceFilterTest to CPPUNIT and added reference data for the filter.

On my branch the following tests still fail:

Todo: Write a test for this feature. I remember it was broken once already and we still do not have any test.

I also tried to save Pic3D as .png. This should save everything as a set of 2D slices, however, the following occurs:

The old values are still (wrongly) set in:
void QmitkStdMultiWidgetEditorPreferencePage::Update()

Very easy bug. Passing this to Andi in order to distribute it during our bugsquashing. This is a good starting bug for new people.

I will take care of this.

Branch is ready to merge once the flags are set.

I changed the interface and added documentation.

[7029bd]: Merge branch 'bug-18622-FixGetValue'

Merged commits:

2015-04-08 15:47:19 Thomas Kilgus [1f2c8c]
Fixed the interface error and added documentation for the wrapper methods.

I made a mistake in my previous comment the interface is changed from:
void mitk::LookupTable::GetColor(int x, double rgba[3])
to
void mitk::LookupTable::GetColor(double x, double rgb[3])

This is some serious issue in MITK-ToF, which never used to work. People should know that this bug exists. Passing to Eric.

[e72c59]: Merge branch 'bug-18578-FixWarning'

Merged commits:

2015-01-16 10:48:41 Thomas Kilgus [3da38f]
COMP: removed unused member.

[90b332]: Merge branch 'bug-18578-FixWarning'

Merged commits:

2015-01-16 11:08:22 Thomas Kilgus [3fa666]
COMP: mitkSurfaceVtkMapper2DTest is only build if the cmake flag is set.

[68193a]: Merge branch 'bug-18578-AngleFix'

Merged commits:

2015-01-16 10:25:27 Thomas Kilgus [25425b]
Changing Ascii code to keyboard degree sign.

[84dd55]: Merge branch 'bug-18578-FixWarning'

Merged commits:

2015-01-16 12:11:49 Thomas Kilgus [d536ad]
COMP: Fixed UpdateVtkTransform.

I discovered that the PointSetVtkMapper2D has the same issues. There it is solved with a manual transformation of each point. This can be done via the UpdateVtkTransform method of VtkMapper if a vtkAssembly would be used. However, the vtkTextActors have to be replaced by some 3D actor in this case. Christoph and I will have a look at this next Wednesday.

The solution is explained in the core change request and in the documentation of the SurfaceVtkMapper2D.

I wrote a core change request and will push the solution now.

we recently switched from QVTKWidget to QVTKWidget2 and now by default anti-aliasing is disabled.

git bisect identified the following commit:

Closed. Just a nice to have.

Closed. Just a nice to have.

This sounds like a nice add-on to me, however, that never worked, did it?

Do you have any contribution regarding this?

[77f574]: Merge branch 'bug-18528-UnifyStdMultiProperties'

Merged commits:

2014-12-22 13:28:29 Thomas Kilgus [7ba26c]
Even more multi widget refactoring.

1. is done.

[77f327]: Merge branch 'bug-18528-GLMappersModule'

Merged commits:

2014-12-19 15:21:02 Thomas Kilgus [46fc06]
MultiWidget and gradientbackground refactoring.

I wrote a core change request.

mitkCoreGLfree

mitkCoreGLfree

Todo:
-RenderWindowFrame uses a vtkMitkRectangleProp to render rectangles with OpenGL. Replace with VTK!

[9f3384]: Merge branch 'bug-18508-FixTest'

Merged commits:

2015-08-19 16:45:57 Thomas Kilgus [6c4b4e]
COMP: Fixed base geometrytest.

[1fccae]: Merge branch 'bug-18508-CombineDifferentAffineInteractorIntoOne'

Merged commits:

2015-08-19 16:18:48 Thomas Kilgus [3650db]
Fixed rotation and translation in mouse mode. Rotation mode crashes for images.

Last todo's are:
-Scaling in mouse mode
-Fix crash of the rotation with images in mouse mode

Passing to Jasmin. We will try to finish this today :)!

[def5d4]: Merge branch 'bug-18508-GeoToolsIntegration'

Merged commits:

2014-12-22 14:43:44 Thomas Kilgus [9ac42e]
Cleaning up for first master merge.

Well its a feature. Check it out yourself in the plugin org.mitk.gui.qt.geometrytools. I wanted to add a little bit more stuff, but I am still not sure how to do it.

I added the following line to Core\Code\CMakeLists.txt and therefore need a core change request:

This bug was intended to (re)-write an interactor to modify geometry and make it accessible via a plugin for the user.

[f52dba]: Merge branch 'bug-18423-ToFScreenShotMaker'

Merged commits:

2015-01-05 15:04:10 Thomas Kilgus [7257f2]
Cleaning up.

[ba1a25]: Merge branch 'bug-18405-TinyXMLPatch'

Merged commits:

2014-11-12 15:58:38 Thomas Kilgus [dd9d7a]
Added old tiny xml patch to the new patch.

I'm pretty sure that this related to changes in SceneIO (either reader/writer changes or compatibility changes I made a few months ago). So it's not Rendering ;).

I will have a look at this.

[28ca1d]: Merge branch 'bug-18384-AdaptingErrorMsgForRotatedPlanarFigures'

Merged commits:

2014-11-05 16:02:05 Thomas Kilgus [9c6d27]
Adapted error message for rotated planar figure statistics.

Works for me in Ubuntu 14.04 current master with two different data sets. Please re-open with detailed a description how to reproduce the bug.

In the QmitkImageStatisticsCalculationThread there is already a concept for passing error messages from the calculating thread to the main GUI, however, it was removed in T15359 for some reason. I'm not sure if this concept can be used, so I adapt the default message to be a little more generic as quick fix from
"Planar figure is outside the images bounds."
to
"Planar figure is on a rotated image plane or outside the image bounds."

As the feature never worked, I created a new feature request in T18400 for Andy's proposal.

Here is my stack trace:

I found more meshes on my machine where this option crashes. It works, however, with "ball.stl" from MITK-Data.

To be continued in T18350.

Resolved -> Thirdparty

Once this is merged, the bug here can be closed.

Unfortunately, my QtCreator won't produce a stack trace and I don't know why.

I pushed a simple fix for to avoid the crash. The live wire tool does still not work well on this data set.