There are 2 main types of blocks, Full and Partial. The former hides all visual aspects of the object, whereas the latter reveals at least half of the object, while still providing sufficient blocking. The shape type of the object affects the contours of a full block, and the breadth of a partial block.
As a convenience, whenever an object is moved to the Object layer, it automatically becomes "opaque", and thus, blocks vision (partial type).
You can set how an object blocks either through the token editor dialog, or through script, using the new API functions provided for this system. Please review our wiki for more information about these functions.
Dynamic Vision Blocking is a work in progress, and there are already some identified issues we are working on as of this writing, not to mention, we also have a lot more ideas to implement. Highest on the list is to finish implementation of custom vision blocking. We want to be careful about doing it right, and, unfortunately, we have little time to work on it due to our other project goals, so we ask everyone to be patient till we get around to finishing this.
Given that everything is quite fluid about it now, where things can change a lot between versions, caution is advised. Do keep a backup of your files, so you have something to fall back on, should issues arise.
We hope you like this new addition, and thank you for supporting our project.