Canonical reveals plans to launch Mir display server (The H)

I think Canonical's previous experience starting their own in-house projects is now long enough to give everyone enough information to make their own judgments with regard to whether Canonical can pull this off or not in the time frame critical for Canonical, which is to say the 14.04 LTS release. There is a game clock ticking away.

I look at the history of Unity and I have some pretty grave concerns about Canonical being able to execute well enough to make this happen in the time that the have. Mir and toolkit support for it has to be ready for OEMs to trust by 14.04 or Canonical is absolutely screwed.

As an idea, Unity, has seen 4 _different_ public codebase implementations in the span of 2 or 3 years. Mutter, qt4, compiz plugin and now qt5. Do you remember the big reveal that Canonical did with Dell for the very first Unity/Windows dual boot demo based on the mutter-derived Unity? Do you remember how quickly everyone backed away from even walking to talk about that implementation and Dell's plans for it? I do. Oh I do.

Now it looks like compiz is dead as a project as Canonical was carrying the burden of maintaining it just to keep the Unity plugin alive. From accounts I've read, compiz as maintained for Ubuntu is no longer taking contributed enhancements because Canonical's engineering team is full-in on the qt5 Unity codebase.

None of the Unity development history says to me that Canonical is going to get Mir right on the first implementation. I sure as hell hope that what we are seeing made public is not the first implementation with 2 years to go on the clock to execute a workable solution. Canonical is definitely does better with iterative failure as a roadmap to improvement.

To the great horror of Monty Python fans, I must admit that when I think about Shuttleworth explaining the history of Unity, I think of him saying the following:

"When I first came here, this was all swamp. Everyone said I was daft to build a user interface on a swamp, but I built in all the same, just to show them. It sank into the swamp. So I built a second one. That sank into the swamp. So I built a third. That burned down, fell over, then sank into the swamp. But the fourth one stayed up. And that's what you're going to get, Lad, the strongest user interface in all of Linuxland."

The only question I have is how many manhours can Canonical put into Mir in order to speed up the process of build/burn/fall/sink/rebuild of the first couple of implementation attempts. 8 months of private development, are we seeing the 4th implementation of Mir or the 1st?

Personally I'm going to be watching linaro's response to Mir very closely. How soon will linaro pick up Mir and try to test it against their existing ARM hardware validation matrix? And how will the public linaro discussion over mir development through all of this year go?

-jef