I was not yet able to merge the changes into master as t he dashboard is currently red

The bugfix is not really relevant as the MITK Workbench does not use this feature, but as the bug is already fixed and tested and ready for merge...

[5f67db]: COMP: Merge branch 'bug-17365-no-planarfigure-flicker-with-mesa'

Merged commits:

2014-06-18 13:31:00 Markus Engel [c2dc94]
dummy commit

[c307a7]: Merge branch 'bug-17365-no-planarfigure-flicker-with-mesa'

Merged commits:

2014-02-27 14:36:28 Markus Engel [3b1a75]
always using 'GL_LINE_STRIP' for rendering PlanarFigures

I have re-fixed the problem by always using 'GL_LINE_STRIP' as this GL method does not cause the flickering effect when using MESA.

I have integrated the fix into my snapshot.

[ff5720]: Merge branch 'bug-17276-crash-in-PlanarFigureSegmentationController-ma

Merged commits:

2014-02-20 09:26:11 Markus Engel [2e25e2]
initializing filters if they are NULL

I have fixed this potential Crash by making sure that the filters are correctly initialized before trying to call methods on them...

Yes, it's safe. There really should be no issues connected to these changes..

[1ea5b2]: Merge branch 'bug-17196-more-configuration-for-displayinteractor-maste

Merged commits:

2014-02-13 16:20:26 Markus Engel [e28b4c]
added more configurability to DisplayInteractor

[1c0765]: COMP: Merge branch 'bug-17190-planarfigureinteraction-with-old-figures

Merged commits:

2014-02-24 09:15:13 Markus Engel [12aac5]
added missing property to fix PlanarFigureIOTest

OK, turns out we really need the check for this property here. Without it, the PlanarFigure is not correctly created when moving the mouse even slightly during a double-click.

I have removed the check for the property as described in the description. Interaction with PlanarFgigures that have been serialized without the property "initallyplaced" works fien again.

Works like a charm. The fix has not been closed because I was not able to integrate it into master yet (Dashboard locked).

[510c65]: Merge branch 'bug-17190-planarfigureinteraction-with-old-figures'

Merged commits:

2014-02-24 08:28:23 Markus Engel [3e07d1]
fixed additional interaction bug in PlanarFigureInteractor

I have added the missing properties with commit 6f00a85f1bcfb13aa03a5293204a6263caec6cd1 and integrated it into my snapshot.

Well, as far as I know, the refactoring is finished. Daniel has invested a massive amount of time in testing and bugfixing and everything seems to work real fine right now.

So far, I have checked the new classes for general code-related issues:

The measures taken to improve the performance of the interaction framework did not have the impact I had hoped they would.

New remote branch pushed: bug-16651-tools-have-been-adapted-to-new-interaction-framework

New remote branch pushed: bug-16625-improved-circle-interaction

[d46d53]: Merge branch 'bug-16625-improved-circle-interaction-master'

Merged commits:

2013-12-05 16:13:38 Markus Engel [2bf22a]
made PlanarCircle editable with any point on the circle

New remote branch pushed: bug-16505-mitkFunctionInstallAutoLoadModules_dependencies-rebased

[04b173]: Merge branch 'bug-16449-correct-planarfigureinteractor-tolerance-maste

Merged commits:

2013-11-19 13:30:04 Markus Engel [165073]
using the same tolerance as the PF-Mapper in the PF-Interactor

Exception has been fixed

New remote branch pushed: bug-16456-crash-in-planarfigure-interpolation

[54c053]: Merge branch 'bug-16456-crash-in-planarfigure-interpolation-master'

Merged commits:

2013-11-19 15:00:24 Markus Engel [30c4f8]
fixed logic to remove planarFigures from interpolation

This issue is present in the current version.

I adapted the distance-tolerance of the interactor to match that of the mapper. Thus a PlanarFigure that is rendered is also editable by the interactor.

New remote branch pushed: bug-16449-planarFigureInteractor-correct-tollerance

New remote branch pushed: bug-16449-correct-planarfigureinteractor-tolerance-master

[508875]: Merge branch 'bug-16236-final-version-for-master'

Merged commits:

2014-09-30 15:17:00 Markus Engel [2f8ccf]
fixed compile errors

Ticket will be closed due to inactivity. Nothing is being done here anymore

[47395c]: COMP: Merge branch 'bug-16236-final-version-for-master'

Merged commits:

2014-10-07 17:37:09 Markus Engel [60eef7]
using int instead of unsigned int in HotSpotTest

[71be84]: COMP: Merge branch 'bug-16236-final-version-for-master'

Merged commits:

2014-10-07 17:59:35 Markus Engel [ff4f47]
fixed another singed/unsigned comparison warning/error

[33cd2c]: COMP: Merge branch 'bug-16236-final-version-for-master'

Merged commits:

2014-10-07 17:54:24 Markus Engel [115647]
trying to get rid of warning with signed/unsigned comparison

OK, I will integrate this now..

The description of this bug really sounds like the issue that should have been fixed with T15856.

You are right, I had forgotten to add a missing action. I pushed it, now it should work.

OK, I think now I understand what you are trying to do here.

I just had a look at the new branch.

Sorry for the late reply, I checked everything again. Seems to work fine for me now.

First of all, thank you for working on this bug. I admit that I have not tried to move a whole PlanarCircle by moving the central point.

New remote branch pushed: bug-15817-listeners-specific-to-renderer

Teh feature has been implemented as proposed in the wiki specification

OK, now that everything works fine here this bug can be closed..

[09b4ab]: Merge branch 'bug-15817-listeners-specific-to-renderer'

Merged commits:

2013-08-08 15:45:37 Markus Engel [71597e]
creating correct filtering for renderer specific listeners

I have added a new method ImageWriter::SetUseCompression(bool) that can be used to dis-/enable the compression for the writing of images.

New remote branch pushed: bug-15749-make-compression-configuratble

New remote branch pushed: bug-15749-make-compression-configurable-master

[b1bc1b]: Merge branch 'bug-15749-configurable-compression'

Merged commits:

2014-06-18 13:01:46 Markus Engel [f37186]
use of compression is configurable

Simple fix to speedup sorting of DICOM files. By declaring Tag objects as static, they do not need to be constructed in each run.

The implementation of the conditions is finished, Dashboard is green.

Reopen to push some changes that were made after review with Daniel.

I have further improved the code and extended the documentation.

[cde4c2]: COMP: Merge branch 'bug-15697-stateMachineConditions'

Merged commits:

2013-08-07 16:36:36 Markus Engel [e7b41b]
removed const in std::vector to satisfy gcc

(In reply to Sascha Zelzer from comment #5)

Here it comes ;-) Thanks for working on it.

(In reply to Markus Engel from comment #4)

1. Is the new StateMachineCondition class supposed to be used

programmatically by external users (e.g. adding new conditions at runtime

...

private. The two classes StateMachineFactory and StateMachineContainer are
friend classes now.

In principal, sounds good to me. However, removing the MITK_CORE_EXPORT
macro from classes already contained in a previous MITK release will also
break third-party code. If the following classes should be removed from the
public MITK API, please mark them as deprecated but keep the
MITK_CORE_EXPORT macro. We can remove the macro after the next release:

• StateMachineState
• StateMachineAction
• StateMachineTransition

Additionally, if the StateMachineTransition class is not going to be
exported, using private methods and friend classes doesn't make so much
sense any more (except for protecting ourselves against internal API misuse).

OK, I have changed the code according to your review.

[8f7e44]: Merge branch 'bug-15697-stateMachineConditions'

Merged commits:

2013-08-07 14:17:57 Markus Engel [b77739]
moved checking of conditions to new method, results are cached for potential following checks

Thank you for the great review! I have taken each point into consideration and tried to adapt the changes accordingly:

New remote branch pushed: bug-15697-stateMachineConditions

Branch has been integrated, this bug is closed...

I have fixed the two issues described above.
The interaction direction has been corrected, I have also adapted the default interaction config in the XML files as well. That is actually why nobody noticed the bug, the default XML config says 'leftright' in order to have 'updown'.

New remote branch pushed: bug-15652-statemachine-fixes

[61e494]: Merge branch 'bug-15652-statemachine-fixes'

Merged commits:

2013-08-09 11:38:31 Markus Engel [11a9fe]
corrected interaction direction in XML config as well

I have fixed the two issues described above.
The interaction direction has been corrected, I have also adapted the default interaction config in the XML files as well. That is actually why nobody noticed the bug, the default XML config says 'leftright' in order to have 'updown'.

New remote branch pushed: bug-15647-planarfigureinteraction-dependency-cleanup

I have removed the unnecessary depenency to the segmentation module from the CmakeLists.txt.

[3fc51d]: Merge branch 'bug-15647-planarfigureinteraction-dependency-cleanup'

Merged commits:

2013-07-18 08:18:07 Markus Engel [9a44bd]
removed unnecessary dependency to segmentation module

That's true.

binary segmentation with error

Unittest that fails due to the error

Screenshots of Pic3D pre and post cast

I have compared the ::Print() output of both input and output images and found out that they differ in pixeltype.

I have opened Pic3D both pre and post Cast-Test and created screenshots. The changes are obvious!

I attach the unittest that fails due to the error. It has a Qt dependency!

extended unittest has been integrated into master..