Ubuntu Announces Mir, A X.Org/Wayland Replacement

[daniels]

Hairy from a technical POV or something else. And if its the technical stuff this means that Canonical will have to go through it to get it to work right?

Yep.

Also on the red hat/wayland comments. Are there any WL contributors from red hat?

No. They hang around and participate in discussion, but don't actively work on it.

[F i L]

I think people are seriously overestimating Canonical's amount of pull and influence here. After all, they are just a company with an admittedly sizable but hardly dominating chunk of Linux market-share, and pretensions of being the next Android. Compared to companies like Red Hat, Intel, AMD, etc they are pretty much insignificant, with almost no upstream contributions to their credit and limited influence on other distributions other than clones such as Linux Mint. I think they are going to have a hard time getting anyone to take this seriously.

It simply does not have that much of a leg to stand on. Whatever Red Hat backs is going to be the choice of the enterprise and embedded markets, as Red Hat has a strong profitable base and a whole community of reliable customers behind them, and you can bet that Red Hat is not going to want Canonical to dominate the graphics stack and will instead go with the upstream solution where they already have great influence. That means no matter what Wayland is far from dead, and if AMD and Nvidia want to be taken seriously they will have to support at least both Wayland and Mir (AMD especially since Catalyst is primarily for 3D workstations).

The main trouble spot is going to be the incompatibilities of duelling implementations. Valve's strong endorsement of Ubuntu is problematic here, although I have found Valve to be problematic from the very beginning.

Still, we did not need this mess. Whatever happens Canonical has thrown a sizable wrench into what was going to be a fairly smooth transition.

Well said. The only way i can see this making any sort of sense is if Canonical has already struck some sort of deal with Valve/NVidia/AMD or at least have faith they can sway core developers from those companies to lend developer support to the project. There's simply no way Canonical by themselves can match the existing efforts and experience of the Wayland developers.

The Display Server is the second most important thing to the Linux Kernel itself for making a consistent environment for developers. Moreover, one of Linux's biggest huddles has been proper driver support from the major card manufactures, especially in the case of AMD. If 'Mir' gains any support at all, at the very least it will only lead to those card manufactures holding off proper support (needed for any form of Linux-Gaming momentum) until a "victorious successor" to X11 has be realized by the community at large, and that could be many times longer than if Canonical had simply chosen to back Wayland and make their voice heard during the months that Mir was secretly under development (a fact that is concerning alone).

Canonical had better pull a magical fucking rabbit out of it's secret hat soon on this (like Nvidia/AMD devs contributing to the project), or admit they've made a mistake and accept Wayland. As a developer, this doesn't make me want to "support Ubuntu more" for so many reasons. Even this announcement is premature, and stinks of a simple PR stunt.

I mean really, did they honestly expect us to be excited about this news? I think that's what gets me the most. They've made this announcement with zero convincing arguments and evidence they haven't even attempted collaboration with Wayland's devs and they word it like they're doing us a favor. I wouldn't have minded so much if back in June 2012 (when the project started) they came out and said "hey, we have our own ideas on how to do a DS. Here's the code we're starting...". Then at least people would have had a chance to give input on their ideas early on and help them choose the best path... you know, kinda like an Open Source project.

[liam]

What's next, they're going to announce their own KERNEL?!

I guess you missed the first comment in the G+ thread
Canonical seems like it is really flailing. Trying everything possible to get traction. It's not surprising that they abandoned the Wayland "pledge". I think I even said this at the time, their "weight" behind projects amounts to very little.
Did anyone see the recent MWC? While people like CNET gave Ubuntu Touch their STAR OF THE SHOW, actual companies were much more interested in FFOS (for good reason but the point is that Ubuntu just doesn't carry much/any weight outside of a very small group).

[iniudan]

As for what some have been suggesting as to Canonical not releasing the source code to this project, it would be a bad choice for them to do something like that.
After all, they didn't close off Upstart.

Here the sources: http://bazaar.launchpad.net/~mir-team/mir/trunk/files

[Hamish Wilson]

Also on the red hat/wayland comments. Are there any WL contributors from red hat?

I never said it did, just that Wayland is the one that is likely to get their support. Thanks for the clarification though daniels.

Something else I realized is that this might actually cause developers to target X11 for a lot longer than if it were a simple transition to Wayland. We will now have three competing standards on Linux - Wayland, Mir, and X11. You could try to target all three, or one or two, but one of them has a huge advantage from a developers point of view over the others, and ironically that is X11. Why? Because both Wayland and Mir will have legacy support for X11 applications, meaning that if someone targets X11 it should mean it should work on all three implementations. It would be the easiest option to run on all three.

So anyone cheering this as a means to killing off X11 quicker might need to step back and consider this. Regardless, fun times are ahead. In a bad way.

[intellivision]

I'm not sure where you get the idea that people are making their software incompatible with Android - but i haven't gotten that impression at all, in any sense. - nor was i saying anything along those lines.

Android also is not built (the way Ubuntu is) around Gnu/Linux. ...and there is a big difference between Google and Canonical ~ aside from Google being highly profitable and *actually* making decent contributions to FOSS; google summer of code, linux kernel development, etc - while canonical is NOT a wildly profitable company and essentially just uses FOSS developers as free labor force.

Google also doesn't hype their FOSS community involvement and tout themselves as being the 'shepards/leaders of open-source' the way that Mark Shuttleworth and his band of morons tend to.

Anyway, i don't see ubuntu phone being successful (compared to google) and i do think upstream developers shouldn't support Canonical's efforts, nor support ubuntu-specific crap like Mir...

Actually you were saying something along those lines.
You stated that you hoped that "receive some real backlash for their exploitation of the FOSS community. Canonical is only involved with FOSS for free labor, nothing more..." when Google has been doing the exact same thing with Android and Chrome OS.
Should developers shun those platforms because they don't push their changes upstream?
In fact, remember how long it took to import the changes from the Android Linux kernel to the upstream one? Also remember how it fell back onto the community to push those changes, not Google?

If you do think that Ubuntu should be shunted for exploiting the open source community but not Android, I have something to teach you about double standards

[stiiixy]

Genuine, and I believe relevant question, that has yet to be addressed;

Can someone please explain to me how Ubuntu moving to Qt5/QML for the very front of the OS is going to affect the final outcome of a user's application experience? Could I still install my Rhythmbox and Amarok? Can I still connect with my digi-cam and phone to Digikam? Can I remove the Russian audio track from the only copy of The Magic Pudding I could find on the interwebs with Audacity?

When it comes to the end user, (I don't actually use Ubuntu, which not should matter, I'm just advocating 'ignorant' users whom could care less about the backend stuff), how will this ultimately affect me.

[daniels]

You stated that you hoped that "receive some real backlash for their exploitation of the FOSS community. Canonical is only involved with FOSS for free labor, nothing more..." when Google has been doing the exact same thing with Android and Chrome OS.
Should developers shun those platforms because they don't push their changes upstream?

Google are at least honest about their intentions and their approach towards open source; they don't invent technical justifications for decisions which were made long beforehand on totally different grounds.

[zanny]

Genuine, and I believe relevant question, that has yet to be addressed;

Can someone please explain to me how Ubuntu moving to Qt5/QML for the very front of the OS is going to affect the final outcome of a user's application experience? Could I still install my Rhythmbox and Amarok? Can I still connect with my digi-cam and phone to Digikam? Can I remove the Russian audio track from the only copy of The Magic Pudding I could find on the interwebs with Audacity?

When it comes to the end user, (I don't actually use Ubuntu, which not should matter, I'm just advocating 'ignorant' users whom could care less about the backend stuff), how will this ultimately affect me.

Amarok and Digikam are KDE, and thus qt, apps anyway. GTK apps run fine under KDE and you can set independent themes for them. Same policy with qt. You will probably have both toolkits installed, but apps from either still work, it just means Unity is written for qt with qml.

[ninez]

Actually you were saying something along those lines.
You stated that you hoped that "receive some real backlash for their exploitation of the FOSS community. Canonical is only involved with FOSS for free labor, nothing more..." when Google has been doing the exact same thing with Android and Chrome OS.

You may interpret what i said however you like, but that doesn't make your interpretation even close to correct.

1st. it is quite dishonest to be insinuating that Android/ChromeOS rely on the gnu/linux stack + applications @ anywhere near the level that Canonical does. that is simply false.
2nd. It is silly for you to also not see the difference of contributions to FOSS between Canonical and Google, be that via source code, GSOC (which go look at the various projects over the years students have been paid to work on - free labor, eh?) and/or $$$ through other funding ~ such as funding code-weavers to support Adobe Photoshop in Wine, Linux foundation gold member, etc, etc.
3rd. Google's use of the linux kernel for android (and everywhere else) has put linux into the mobile game big time. (sorry Canonical coming years later and 'piggy-backing' using android s/w components is not the same thing).
4th. there is also a difference in 'stated intentions', M.Shuttleworth has always presented Canonical as sheparding/leading the larger FOSS community, he litters his speech with a zillion happy-fuzzy buzzwords, which are essentially all BS / a scam. ~ unlike Mark/Canonical, Redhat was honest from the beginning (as a good example) - they have a very commercial enterprise using FOSS software, yet are heavy contributors to FOSS projects and have been for years and years. ~ Canonical is not and likely never will be either. ~ how much code does Canonical pump into the kernel? how about compilers like GCC? ...how about anything that isn't just 'self-serving'... Google does many things that benefit a wider audience, much like Redhat does...

Should developers shun those platforms because they don't push their changes upstream?

why are you asking a question which bears no relevance to anything? I never said someone should be shunned for not pushing changes upstream. There are sometimes valid reasons to not do so, for a variety of reasons.

In fact, remember how long it took to import the changes from the Android Linux kernel to the upstream one? Also remember how it fell back onto the community to push those changes, not Google?

the length of time really bears no relevance. Sure, Google did take a few years to adopt a more open approach to kernel development, but guess what?? ALL of the code that was commited, regardless of the time frame and who did it - was still code they wrote for the linux kernel and is in fact, their contribution (which is larger than anything Canonical has done or will do! and Ubuntu wouldn't even be running on a tablet, if google wasn't in the game)... Now sure, i can agree - it is problematic that it took such a long time and others did some of the commits - but it doesn't change the fact that google actually puts a lot of money, time, code, etc above and beyond anything Canonical does. (or will ever do). I think you also have to remember the context too - at that time, google has essentially been used to doing everything in-house, working with upstream linux kernel developers was something that was going to take time regardless.

If you do think that Ubuntu should be shunted for exploiting the open source community but not Android, I have something to teach you about double standards

You might want to work on that, you haven't said anything compelling. in fact, it's mostly the opposite. it seems you felt the need to drag in arguments that have nothing to do with the core issues and they were pretty weak arguments to begin with.