Jono Bacon Stepping Down From Canonical

[TheBlackCat]

I just wish people understood why Ubuntu made these decisions

Nobody understands why Ubuntu made the decision. Their stated technical reasons were all false, and they haven't provided any serious alternative explanations since then other than something vague about management. Lots of people have come up with their own guesses, but the only explanations that could remotely be considered official are "they didn't do their homework" and "management said so".

[chrisb]

I just wish people understood why Ubuntu made these decisions

Nobody understands why Ubuntu made the decision.

Jono Bacon did give some reasons in his Q&A sessions. Someone made a decent summary of his points from those sessions:

• Wayland does too much. Having perpetually unused features in your software stack is poor software design.
  
• Wayland's team would not be flexible enough to offer a gutted version of Wayland to accommodate adequately, respectfully.
  
• Mir is to Wayland, what LightDM is to GDM/KDM.
  
• Ubuntu has very dire deadlines that they need to meet with phone manufacturers and the like. Having control over a project makes it easier to infuse extra resources to guarantee that these deadlines are met.

Whether his points are correct or not is a matter of opinion. He has the right to have his own opinion. And Mark Shuttleworth has the right to have his own opinion, and as the owner of Canonical he has the right to decide what software Canonical will use. He thought that Ubuntu would be more successful if they developed their own solution. It's that simple.

On his April Fools Q&A Jono Bacon said

Mir is great. You know what.. you know why Mir is great? It's because it means that Canonical can actually do something, and get stuff done, without having to care about people like all those whiny Wayland developers, by the way. So, yeah, we're sticking with Mir.

Perhaps there is some truth to that.

An interesting historical point that people forget is that initally Canonical explicitly rejected the option of creating their own display server.

we evaluated the cost of building a new display manager, informed by the lessons learned in Wayland. We came to the conclusion that any such effort would only create a hard split in the world which wasn?t worth the cost of having done it. There are issues with Wayland, but they seem to be solvable, we?d rather be part of solving them than chasing a better alternative. So Wayland it is.

Canonical rejected the option of creating a new display server and threw their weight behind Wayland. But it did not work out, so they changed strategy. And now lots of people get very emotional, and feel the need to pressure Canonical into going back to Wayland.

The continuing argument is like a celebrity breakup... "why did Brad dump Jennifer? They were the perfect couple! Why would he do it? What reasons?" The simple answer is "sometimes it just doesn't work out". Brad didn't want to be with Jennifer. Canonical didn't want to be with Wayland. It's sad, but it happens. As with all breakups, people get emotional, start picking sides, and start accusing the other side of being unreasonable, instead of just accepting that other people have different opinions, based on different experiences and perceptions of the world, and that they will do what they think is right, not what you think is right.

[Vim_User]

On his April Fools Q&A Jono Bacon said



Perhaps there is some truth to that.

How can they even know? There was never any attempt of Canonical/Ubuntu to work with the Wayland developers.

[NothingMuchHereToSay]

How can they even know? There was never any attempt of Canonical/Ubuntu to work with the Wayland developers.

I've heard that there was always a lot of hostility between Canonical and other developers, I guess they didn't wanna take that chance. Hell, it goes all the way back to Upstart in 2006, which probably hasn't been blown as out of proportion as most everything else that's unjustified.

[Vim_User]

I've heard that there was always a lot of hostility between Canonical and other developers, I guess they didn't wanna take that chance.

That comes down to: Hey, we assumed that the Wayland developers may be a bunch of whiners, but instead of looking if that may be true we just reassured them of our support and secretly developed an in-house solution.
If that would be true than any critical views against Canonical/Mir would be earned to the fullest.

[anda_skoa]

Canonical rejected the option of creating a new display server and threw their weight behind Wayland. But it did not work out, so they changed strategy.

Please exuse my rather limited understanding of English idioms (not my mother tongue), but doesn "throwing ones weight behind A" mean something like "working towards
A's goals"?
As in actively contributing in some form?

Also, doesn't "did not work out" not imply that something was tried?

There seem to have been no contributions of significant weight by Canonical to Wayland other than the quoted initial announcement, so either was they didn't try (which would make the second phrase weird) or their weight was so insignificant that it did not even show up as a contribution.

Hmm, or are you being sarcastic?

Cheers,
_

[TheBlackCat]

Jono Bacon did give some reasons in his Q&A sessions. Someone made a decent summary of his points from those sessions:

• Wayland does too much. Having perpetually unused features in your software stack is poor software design.
  
• Wayland's team would not be flexible enough to offer a gutted version of Wayland to accommodate adequately, respectfully.
  
• Mir is to Wayland, what LightDM is to GDM/KDM.
  
• Ubuntu has very dire deadlines that they need to meet with phone manufacturers and the like. Having control over a project makes it easier to infuse extra resources to guarantee that these deadlines are met.

Whether his points are correct or not is a matter of opinion.

Wow, I hadn't seen that, thankfully. At least their previous mistakes were technical ones. These are in "not even wrong" territory.

1+2. How, exactly, do they expect Wayland devs to somehow prevent Canonical from releasing a partial implementation of an MIT-licensed protocol? It is just silly. That is ignoring the fact that the security issues Mir has had indicates the Mir devs don't really know which features are important and which aren't.

3. So Mir is a flexible, general-purpose, extensible system designed to work across a large number of desktop environments while Wayland is a system designed and optimized solely for use by a single desktop environment? Sorry, could you please explain any rationale by which that comparison could possibly work?

4. Ignoring the fact that Wayland has done a better job meeting Mir's deadlines than Mir has (there are already Wayland phones out, for example), Wayland supports extensions for exactly these sorts of use-cases. Canonical wouldn't have to wait for Wayland devs to catch up, they could just release a version using their own extensions and (optionally) merge them upstream at their convenience. Heck, even Weston has extensions that aren't part of the official Wayland protocol, and if I recall correctly KDE plasma workspaces devs were or still are planning on using extensions for their own internal stuff. So they wouldn't have to wait for Wayland devs at all.

Canonical rejected the option of creating a new display server and threw their weight behind Wayland.

No, they threw their voice behind Wayland, but they did nothing to actually help, or even attempt to make sure their needs were met.

But it did not work out, so they changed strategy.

No, Canonical made no attempt to worked anything out, and they still cannot provide any remotely plausible reason why Mir helps them achieve their goals in any way. Judging by how their competitors in the mobile environment are faring, if anything it is holding them back.

instead of just accepting that other people have different opinions, based on different experiences and perceptions of the world, and that they will do what they think is right, not what you think is right.

"Everyone is entitled to his own opinion, but not his own facts."

Whether Wayland has technical issues that make it unsuitable for Mir isn't a matter of opinion, it is a matter of fact. Whether Canonical would be allowed to ship a partial implementation of Wayland isn't a matter of opinion, it is a matter of fact. Whether Wayland devs could prevent Canonical from responding quickly to their own internal needs isn't a matter of opinion, it is a matter of fact. Whether Canonical made an attempt to work with upstream Wayland to make sure their needs are satisfied isn't a matter of opinion, it is a matter of fact. All the reasons that Canonical has given for using Mir are fact-based issues, and they are factually incorrect.