Some strange behavior when the mouse leaves the Emperor side during sample selection

[fedarko]

It's a bit hard to describe, but when the cursor crosses over the dividing line between Empress and Emperor while selecting samples in Emperor, the selection gets kind of messed up. Samples remain highlighted, but nothing gets colored in the Empress plot. The Empress plot's existing colorings (e.g. feature metadata) get cleared sometimes but not always when this happens -- it seems to be influenced by how the selection is "ended" (e.g. while holding shift?, while still hovering over the selected samples?, etc.). Samples seem to remain highlighted indefinitely, also?

In any case, clearing the sample selection by making a selection containing no samples fixes things, as far as I can tell.

It would probably good to have the Emperor callbacks? / Emperor selection code detect if the cursor leaves the plot and then just automatically cancel the selection in that case.

Screen capture of some of these problems:

(I don't think this is very important, but I'm sure someone will run into it eventually...)