Hello all,
I’m close to finishing up preview snapping and would like some feedback on what direction to take it in.
Questions:
Should snapping be more like the 3D viewport / node editors, where the location of your mouse cursor when you press G dictates the snap target? Or should it be more like the timeline, where the initial mouse position does not affect the calculated snap target?
(I’m currently going with the latter option since it’s similar to other NLEs, but would be interested to know if people would benefit strongly by being able to dictate snap targets more “immediately” like this. If so, timeline snapping could be refactored too, to incorporate this change).
Would people want a flag that you can select to decide the source snap point based on the initial mouse position? I’m planning on making all corners and centers of selected strips to be snap sources, but with this change you would be able to select, say, only the bottom left corner as a snap source w/ the flag set and your mouse cursor closest to that corner when you start the transformation.
Would it be helpful to color separate snap types differently (using theme prefs)? Would also like some input from UI people on this. E.g., if snapping to borders, then use one color, if snapping to centers, use another.
Definately latter. That behavior in 3D viewport exists because silhouette is view-dependant and points which can be used as snap base are scattered in 3D view. For 2D images it doesnt make sense to have that complexity. It will be bad UX
Personally, I would do it in same way as it is done in VSE timeline.
This would be useful when using gizmos. For example if you snap one side of image to something and then want to snap it’s other side to something else.
Not sure about this. In 3D viewport I think you can snap to imaginary lines based on simple geometry, these items are color coded. Not sure if they are themable. For simple snap points I wouldn’t use too much colors as long as it is evident what is snapping to what.