Ok, we can keep them in this release.

Discussed with @wiesenfa that this issue is not critical for release v1.0.

Recap: The initially chosen license (GPLv3) is not compatible with GPLv2 only licenses of dependencies (e.g. ggplot2, relations)

as name property, the name of the RT node is provided correctly.
I created a working example with Pic3D for comparison.
Looking at both caaaaa.xml files, the only difference I notice is that the corrupted one is encoded with UTF-8 but otherwise it looks similar (as far as I see).

Which node is it? you can look up the name if you look into caaaaa.xml

<?xml version="1.0" encoding="UTF-8"?>
<Version Writer="C:\Development\MITK-dev\src\Modules\SceneSerialization\src\mitkSceneIO.cpp" Revision="$Revision: 17055 $" FileVersion="1"/>

sorry just saw the message. I put it in the shared folder.

could you provied (e.g. via nextcloud) such an invalid scene file?

As Ralf pointed out, it is actually not related to the GenericProperty as such but it seems that the scene does not even contain any nrrd files after saving and that is why it breaks

seems to not fail on windows machines. but does on my Mac (and thus might also on linux). minor issue, defer to post 1.0.0

Do you have still problems with crashes? If yes, this is almost certainly not due to the GenericProperty thing, as they have no relevance for dose visualization and scenes without dose do not crash but also cannot load the GenericProperty.

GenericProperty associated to key 'MITK.IO.reader.DICOM.PixelSpacingInterpretation'
GenericProperty associated to key 'MITK.IO.reader.DICOM.ReaderImplementationLevel'
GenericProperty associated to key 'dicomseriesreader.PixelSpacingInterpretation'
GenericProperty associated to key 'dicomseriesreader.ReaderImplementationLevel' still need to be changed

The idea of the web app is to lower the entry level for users that are not familiar with and/or don't want to install R (as this is already a hurdle for some of them).

Closed. @neher or who ever might stumple upon the prolem again, is welcome to open it again and provide data. :)

Since solution 1 worked well for me and seems to be more generic, I would vote for 1.

In T22412#93055, @kislinsk wrote:

Also crashes with normal 4D images. In Multilabel segmentation 2D region crowing, too.

@neher or @jaegerp can someone provide test data? Otherwise I would close this task for now, because I cannot check or reproduce the problem without data.

Can D447 landed and this issue closed? If nothing speaks against it, it would be good, before I turning the rest of the seg tools upside down. Thanks.

@thomass This is the issue we talked about last week when you hat problem to use predicates to find some nodes by DICOM tags values.

1. Could you provide also a SS example. As it should also work with SS and I would like to know the reason.
2. A MAC version will be available with the next public MITK release (planned for next january).
3. Thanks for the example data. I have some questions regarding it:
   • a) acquesition 2592 containes 2 slices (with 21 time points/frequencies each). Is it meant to be one volume with 2 slices or are the 2 series but due to a multple slice sequence packed into one folder?
   • b) what is the list.txt (see https://phabricator.mitk.org/w/mitk/cest/cest_user_manual/#list-txt-frequency-offsets-%CE%B4%CF%89) you used so far?

Deleted branch bugfix/T28101-CMake3.19.

Pushed new branch bugfix/T28101-CMake3.19.

Will be fixed very soon in CMake 3.19.2: https://gitlab.kitware.com/cmake/cmake/-/issues/21529

It works for me on the latest develop branch (#921705).

I tried it with my current develop (not the latest due to ongoing QT issues - wip) and it resolves the last warning (predescibed dose) but not the others. According to the D449 this is the one you changed, so as expected I suppose :) As soon as I get the latest develop up and running I will check the diff further.

@thomass Please check the patch provided by https://phabricator.mitk.org/D449. It should solve you problem I hope. Would be cool if you could give feedback.