Intel Pulls Mir Support From Their Xorg Driver

intel-banner

A commit on the Intel Xorg driver branch yesterday showed a very damning comment from “Intel Management” informing the world that Mir will no longer be supported by them.

The commit came yesterday morning with the following message within the news section:

We do not condone or support Canonical in the course of action they have chosen, and will not carry XMir patches upstream.
  -The Management

Bad news for Canonical

The turn around comes after a commit that was implemented by Canonical’s Christopher Halse Rogers a few days ago was later reverted. The comment above is vague to say the very least, and one has to wonder why Intel decided to change their minds like this. Canonical’s Jono Bacon had this to say on Google+:

This is very disappointing from Intel. Good contributions to Open Source projects deserve fair and reasonable feedback, particularly when accepted as a technically worthwhile contribution and then reverted just a few days later. Given that “the management” ordered this, I presume the revert is due to either strategic or political reasons as opposed to technical.

Whilst I agree with Jono to some degree, I can see the other side of the coin. Mark Shuttleworth did commit to supporting Wayland only to do an about-turn a couple of years later. So in my personal opinion, although this is disappointing for Cannonical and Ubuntu users, Mir was a decision that was taken by Canonical, so unfortunately they will have to live with consequences of their actions.

What Does This Mean For Me, An Ubuntu User?

It’s too soon to tell at the moment, but I would assume that Ubuntu will simply need to take on the work that Intel were going to do and patch the Intel Xorg driver themselves. This means more work for the Ubuntu development team, but some would say that they’ve made their bed, so now they should lie in it.

Intel do already have a vested interest in Wayland, and have a couple of full time developers working on the project I believe. So some could say that this is a strategic move by Intel. But you have to ask the question, why the hell did Canonical decide to start the Mir project in the first place? Wouldn’t it have made more sense to commit upstream to Wayland instead?

What do you guys think? Do Canonical deserve this, is this the first nail in the coffin for Mir, or will Canonical simply roll with the punches, knuckle down and get on with it? Feel free to tell us your thoughts in the comments section below.

Intel Pulls Mir Support From Their Xorg Driver
User Rating: 0 (0 votes)
  • Fewt

    I think: http://i.imgur.com/zh7a7zZ.gif

    Canonical is its own worst enemy and their company is driven by so much arrogance that they will never admit fault. You can smell the arrogance from all of their employees calling this bad on Intel on their blogs and feeds, but really they are getting exactly what Canonical wanted – after all they elected to develop Mir in a secret bubble. As you said, they have made their own bed and now they can lie in it. :)

  • http://www.mhall119.com/ Michael Hall

    “Wouldn’t it have made more sense to commit upstream to Wayland instead?”

    Commit what? People seem to assume that the Wayland project wants the things we built into Mir, and we’re just not giving it to them. Either we’d have tried to force unwanted things into Wayland, or we’d have had to settle for unwanted things from Wayland. Neither of which is “the open source way”.

    • http://www.refugeeks.com/ Kev Quirk

      At a hight level, to me and many other members of the community it seems that it would be better to have the maximum amount of people focussing their efforts on one product (ie Wayland OR Mir) than this disjointed effort that results in being two display servers, one of which may not be needed.

      I’m sure that Wayland doesn’t tick all the boxes for you guys, but surely it’s the same thing with X at the moment, as it has been for god knows how many years. To me personally, the maximum amount of people possible contributing to a piece of software is more “the open source way” than going off on your own and duplicating what has already been done in some fashion.

      Wayland, Mir, or X. I really couldn’t care less to be honest with you. It just seems to me that there is a lot of wasted effort here when there doesn’t necessarily need to be. And I think that’s what most of the community think.

      • http://www.mhall119.com/ Michael Hall

        We have a dozen programming languages, half a dozen widget toolkits, tens of desktop environments an windows managers and hundreds of distributions. It’s kind of weird to complain about duplication and fragmentation now.

        “Scratching your own itch” is an essential part of what has made open source successful. Linux didn’t just contribute to GNU or Minix. Miguel didn’t just contribute to KDE. Ian didn’t just contribute to Slackware SLS.

      • http://www.refugeeks.com/ Kev Quirk

        I agree with that, open source software is definitely fragmented, no argument there. However, I think of the fragmentation in OSS as a kind of family tree (bare with me here ha ha), with Linux at the top, then a handful of core packages and binaries, after which come applications, after which come the many, many distributions. They all have the same beating heart basically, as you go deeper and deeper into the system there is less fragmentation.

        But working our way back up that tree, there is less and less fragmentation the closer to the top of the tree you go. I would consider the display server to be pretty high up in that tree, and therefore shouldn’t really be fragmented that much at all, if at all.

        Couple that with the fact that Mark openly said Canonical was supporting Wayland, then all of a sudden that’s not on the cards and Mir comes along. Like I said, to me personally it just feels like an unnecessary waste of resources.

      • http://www.mhall119.com/ Michael Hall

        Less fragmentation perhaps, but even without Mir you still had X11, Wayland, Android has SurfaceFlinger, I think Enlightenment had their own at one point, and Mozilla has yet another used in Firefox OS.

        When Mark said Canonical was supporting Wayland it was supporting wayland. That was 3 years ago, IIRC. Mir is only about a year old, as far as I can tell. A lot of things can change in 2 years, it’s not fair to expect people to not change their plans when their situations change.

      • http://www.refugeeks.com/ Kev Quirk

        That’s true. People are free to change their minds. I just don’t want to see Ubuntu suffer for a bad decision.

        On the other hand, it might pay off and work out. If so, more power too you guys! :-)

      • https://twitter.com/xarinatan Alexander ypema

        ” It’s kind of weird to complain about duplication and fragmentation now.”
        I think it’s actually more reason to complain about fragmentation right now, if there weren’t any fragmentation there wouldn’t be anything to complain about.

        And regarding the ” “Scratching your own itch” is an essential part of what has made open source successful.”: That’s true, but not because it made things more fragmented; It’s because you can contribute your scratching right to your favorite project. You just opened up the offending bits, scratched all you pleased, committed the patch and bam, everyone enjoys your patch. THAT’S the power of OSS, that everyone can contribute, whereas your version sounds more like what you get if everything were closed source, since you can’t submit your patches anywhere, or even see where the problem is located in the first place.

        I’m not saying it’s not good to start projects of your own, if there’s absolutely no match with your vision anywhere and you think your vision is something many people could benefit from then by all means really, some choice is nice, but there’s a thin line between choice and fragmentation and it’s to be guarded carefully.

        Personally, that used to be one big great feat about Ubuntu back in the 6-7-8.xx era; They used widely used packages and committed the patches upstream. Ubuntu was easy to use and almost every bit was familiar because the packages were widely used. Canonical has been doing the exact opposite, scratching an itch meant starting a completely new project. This actually works because both Ubuntu and Canonical are so large internal projects do get ground. But the Linux and OSS community is bigger than Ubuntu, and there are many projects out there that might be doing the exact thing you’re trying to achieve with a new project. How awesome it would be if the large Canonical and Ubuntu committed all their work Upstream, where everyone could benefit from it, and not just Ubuntu and Canonical themselves.

        I do understand your vision, as someone with many tiny projects of my own for personal use that do more or less the same as some big projects out there, I know it’s really tempting to just start out with a clean slate when your ideas do not 100% coincide with that of the bigger projects already out there, especially if you have the resources. I just don’t think it’s the right way to do this for large community projects that involve the opinions of tens of thousands if not more people.

      • http://www.mhall119.com/ Michael Hall

        “It’s because you can contribute your scratching right to your favorite
        project. You just opened up the offending bits, scratched all you
        pleased, committed the patch and bam, everyone enjoys your patch. THAT’S
        the power of OSS”

        That’s great, as long as your patches are actually accepted by upstream. If they’re not, the only you and the people who get the patches from you get to enjoy it.

        “How awesome it would be if the large Canonical and Ubuntu committed all
        their work Upstream, where everyone could benefit from it, and not just
        Ubuntu and Canonical themselves.”

        You do realize that this article is about Canonical and Ubuntu committing our work upstream, and then upstream rejecting it (after accepting it), and then Canonical and Ubuntu being told that we shouldn’t be submitting those things upstream.

      • https://twitter.com/xarinatan Alexander ypema

        Not sure how far upstream developing your own displayserver goes versus developing on the existing ones. Also I can’t judge how much difference there is between Wayland and Mir and whether or not either has obvious advantages over the other, but considering the rather negative response towards Mir and positive towards Wayland I’d personally pick the latter to work on. But that’s just my two cents.

    • Leon

      No, people just assumed that when Canonical announced their intentions to support Wayland they would have put some effort into it and actually contributed some code. Their intentions started becoming pretty evident looking at their Wayland related blueprints, especially when the question of how to implement the $XDG_RUNTIME_DIR went from tossing around a couple half-serious ideas to idling for months and apparently waiting for someone else to solve the problem.