What's your opinion on Wayland? Specifically, how likely is it that the reliability and consistency of the graphical parts of desktop Linux will improve if the applications I use the most are ported to Wayland?
It looks like it's shaping up very nicely to solve the problems it was designed to solve--namely, making every frame of image on the screen pixel-perfect. However, I don't know if it's going to solve other outstanding problems (or regress on some fronts).
Despite the pains, X allows window managers to arbitrate over window management policy relatively effectively. From what I see so far, it seems that this will be a less clear-cut scheme in Wayland, given its clientside window management and decoration setup. Individual applications will have more freedom to misbehave in terms of custom titlebars etc. and refusing to cooperate with management actions. In an environment where GTK, Qt, wrapped X programs, and other toolkits will be coexisting, there's a lot of room for implementations to behave differently in ways that will confuse and upset users. Whether this will actually be the case is yet to be seen, but I've never seen three different software projects implement a feature in exactly the same way without sharing code.
Wayland may improve the speed of compositing on the Linux desktop, but it's very unlikely to improve consistency or reliability. Wayland applications draw their own window decorations (rather than a window manager), which obviously doesn't encourage consistency. And last I heard, when a Wayland application hangs, its window will become completely immovable. No, really.
What's your opinion on Wayland? Specifically, how likely is it that the reliability and consistency of the graphical parts of desktop Linux will improve if the applications I use the most are ported to Wayland?