Page MenuHomePhabricator

LevelWindow component should use the datamanager selection events
Closed, WontfixPublic

Description

The LevelWindow handling is quite complex (and messed up) and old. A lot of effort is taken to search for the image that should be displayed in the level window widget. I think this could be strongly simplified by just using the selection mechanism of the data manager view and beeing able to manually select any image with the context menu.

This could lead to a substantial speedup, as one user on the mailing list is reporting that the LevelWindow code is time consuming: http://thread.gmane.org/gmane.comp.lib.mitk.user/813/focus=822

(I don't know if this is 3M relevant)

Event Timeline

Marco, could you reassign this? Jochen, do you have enough ressources to look into this issue? Usually I would ask Daniel S, but he's not there for the rest of the wee.

Undoable for 3M3. The selection mechanism can`t be used as long as no one want MITK-core to have openCherry dependencies. Think this is RFD for next MITK meeting

there should be a solution for 1.0

Please discuss in MITK meeting

i think daniel stein worked on this. @daniel: please comment, can we close this bug?

Just thought about a way to let the DataManger module access an appropriate method in mitkLevelWindowManager directly to tell about the selected Node.

If this is also not possible the actual behaviour has to remain. In my opinion it is already the best solution because for me it would be strange to change the Level / Window of an image which is not the topmost shown in the render windows but hidden by another one with a higher layer...

I disagree. The layer property is invisible to the user and unintuitive to change. All other views process the data object that is selected in the data manager. Only the LevelWindow slider chooses its data object by other criteria. This is unintuitive for the user.

By the way, multiple images do not have to overlay/overlap each other. They can be placed next to each other. Or different render windows could display different images...

And then there is still the reported performance problem of the selection mechanism.

I suggest to leave the current behavior for the 1.0 release, but to start a clean reimplementation in the next months (maybe as a party bug).

(In reply to comment #8)

I suggest to leave the current behavior for the 1.0 release, but to start a
clean reimplementation in the next months (maybe as a party bug).

Actually Markus and Markus have already started planning for a nicer render window with a clean interface for handling data, coupling views for scrolling/zooming, simple interactions (panning, zooming, level window).

kislinsk claimed this task.
kislinsk added a subscriber: kislinsk.
This task was automatically closed because it wasn't updated at least since July 2016 (over 2 years). Please re-open this task if you think that it is still relevant. This most probably means that you will resolve it.