And another thing: update requires clean builds of DCMQI and DCMTK, so continuous clients will probably fail

Ok, I updated my fork to use the latest DCMQI with our modifications as well as a new jsoncpp version which solves the compilation issues. As a side effect, we have to update DCMTK, but I think it should be worth it before the release.

I have seen the fixes deactivating QNetwork in 1881c497af3e4b810d131a2c05bb819e485bf4e7 . However this also removed the
-n option command line option to disable networking. This option is used by the plugin generator tests. I could remove it there, but @kislinsk I think we should discuss the general strategy.

Pushed new branch T23293-modernization-fixup.

Pushed new branch T23293-more-modernizations-and-warning-fixes.

Same for threshold in FiberExtractionRoi

The patch:

Pushed new branch T23775-warnings-no-errors-for-apps.

I will add a little feature: a WARNINGS_NO_ERRORS option for the mitkFunctionCreateCommandLineApp CMake function

if the ID is unique, this should maybe named getUID . And I wonder if we should make the IDType a proper class, did you discuss this?

I included this in my "clang" branch since I thought it was a local problem. You can cherry pick the fix, don't know when I can merge

Pushed new branch T23268-more-clang-fixes.

Does a DCMTK.patch file in MITKs CMakeExternal seems appropriate in this case?

Pushed new branch T23327-update-to-itk-412.

Pushed new branch T22775-dcmtk-update.