Given the behaviour of the MediaBrowser class at this point, what I have outlined is the most viable option.
If we want something else, the entire Media Browsing/Selection paradigm will need to be changed.
This is why I've asked people to make UI toys for an eventual replacement of Orbiter (the Clutter thread)
And yes, I do feel as if I have a right to have the final say on this issue, seeing as:
(1) I have put more hours into Orbiter, and the Media PlugIn, than all of you put together.
(2) I have almost two decades of UI experience.
(3) I have taken it upon myself, to ensure that the UI experience will eventually be completely consistent across EACH AND EVERY TARGET that Orbiter supports.
Have any of you stepped up to do this?
And no, don't whine and say, "But, you're bullying us!!" ...
Those of you who have actually put code into this system, and have worked with those of us who do the development work KNOW that we value actual work above any and all talk and discussion. _VERY RARELY_ have I explicitly put my own foot down in terms of how things should work. I can count them on my fingers,
and before you say, "But you should never do this!! It's against open source development!"
no, it's not.
Have any of you done Linux Kernel development?
Linus has final word on _EVERYTHING_ that goes into the main line. PERIOD. End of story. He earned this right with the sheer amount of work he has put into the kernel.
If you can thoroughly prove the point that having divergent queueing behaviour would be better than what I am saying should be done, then, I will go along with it.