RFC: Should KDE fork the wlr-layer-shell protocol?

Related to: #229

Plasma uses the upstream version of the layer-shell protocol as of now. But there are a few missing things that we would like to have in the protocol. Some had been successfully upstreamed, some got stuck, and some got rejected (but we still would like to have those features in the protocol). We would also like to move things faster to meet the Plasma schedule.

The main advantage of forking the layer shell protocol under the KDE namespace is that it's going to let us have more freedom over the protocol design to meet Plasma needs and iterate over things quicker.

The main disadvantage is that it is going to contribute to some protocol fragmentation.