The core now has a method requestCapture, which receives a
CaptureRequest object with all the needed information.
This reduces code duplication in future features.
More flexible tool API
Minor code format fixes
Clipboard freeze fixed(?)
Arrow correcly growing close to the start point
Improve maintainability
Add undo/redo stack
Updated dbus API. Now it sends a signal with every capture, it may
be captureFailed or a captureTaken which contains the raw image
bytes in png format. You have to add an id to the screenshot calls
so it will be returned as a way to know the origin of the signal.
