1. 15
  1.  

  2. 5

    We have a definition problem. Android and Chrome OS are both very successful and running on the Linux kernel. Yet we are still waiting for the Year of the Linux Desktop. By definition a kernel cannot be a Desktop. It’s not possible. How can we win if the goals are not even clearly defined?

    The year of the Gnome Desktop would be more accurate and attainable for example.

    1. 8

      I’ve not (yet) been able to watch the video - no transcript is available for me and I’m not in a situation where I can listen (a11y people take note).

      If “fragmentation” and “commercial power plays” aren’t strong contenders I’ll be very disappointed, and Canonical are one of the major offenders here. There was no need to push Upstart when the rest of the world was leaning into systemd (be that right or wrong), likewise Mir vs. wayland, bzr vs. git, etc.

      Canonical has a remarkable desire, it seems to me, to want to be like redhat - build a big userbase, control the software they use and peel it away from the traditional foss consensus, and then dominate the ability to provide services for that software. I’m not sure it’s healthy for users, nor the concept of “linux on the desktop”.

      1. 14

        There was no need to push Upstart when the rest of the world was leaning into systemd (be that right or wrong), likewise Mir vs. wayland, bzr vs. git, etc.

        Upstart came along quite a bit before systemd, 4 years I think it was in fact. As I recall, the early systemd blog posts even referenced upstart regarding lessons learned (good and bad).

        As for Canonical pushing upstart – for a while it also wasn’t assured that systemd pickup would be as quick or as pervasive as it ended up being – Redhat pushed it pretty hard, with Fedora being the first major distro to adopt it.

        That said, Canonical certainly /did/ drag their feet converting, but then again.. look how long it took debian to change! It was something like a year after ubuntu did? (EDIT: I read the wrong date here. See here for update)

        Not that I defend upstart /at all/. I found it to be pretty darn buggy in fact.

        1. 9

          Upstart came along quite a bit before systemd, 4 years I think it was in fact. As I recall, the early systemd blog posts even referenced upstart regarding lessons learned (good and bad).

          To add to this, Lennart states: “Before we began working on systemd we were pushing for Canonical’s Upstart to be widely adopted (and Fedora/RHEL used it too for a while). However, we eventually came to the conclusion that its design was inherently flawed at its core…”

          That said, Canonical certainly /did/ drag their feet converting, but then again.. look how long it took debian to change! It was something like a year after ubuntu did?

          Not sure where you’re getting that from, but Mark Shuttleworth announced that Ubuntu would adopt systemd the same week as the Debian decision.

          1. 4

            Not sure where you’re getting that from, but Mark Shuttleworth announced that Ubuntu would adopt systemd the same week as the Debian decision.

            Ah. I couldn’t remember the timeline, and looked at the wikipedia ubuntu version history. It looks like I accidentally, and certainly erroneously, used the announcement date instead of the actual release date.

            Thanks for noticing and correcting!

        2. 6

          I doubt I’ll ever watch this video, since I don’t really care for videos. But I’d happily read a transcript.

          Mostly I’m curious about what Mark’s definition of “success” would be. I don’t have stats handy, but in my limited experience Linux desktop usage seems pretty strong in certain technical and professional settings. Meanwhile desktop OS usage of any variety has declined in relative terms, thanks to the rise of the mobile platforms. If he means success as a consumer OS, it’s not clear to me that any players besides Canonical were ever tilting at that particular windmill.

          1. 8

            So much this. People keep making the “Year of the Linux Desktop” joke mostly for historical reasons, I think. So far as I can tell GNU/Linux based Desktop and Laptop systems have been very good for quite some time, very usable (and used) by non enthusiasts, and also Desktop as a target is in strong decline.

            1. 5

              Aside from hardware issues, the one thing that’s bad about them is that they keep making changes that break fundamental parts of the platform users rely on. That or just not QA testing enough on those. This is easy for them to avoid like the proprietary ones do with their stronger assurances of backward compatibility. I mean, sure Microsoft tried doing something similar with Windows 8 but look how that went. The Linux desktops should make sure basic functionality always works and is consistent over time.

              Recent example that just happened is I can’t open PDF’s with Firefox on Ubuntu. The JS reader always clobbers the abstract texts I copy and paste in ways the native apps don’t. So, if I want to use the text, I’ll re-open the PDF in native reader from within the JS reader with open/save button or Firefox with ask feature. Suddenly, I can’t do that. It’s also suggesting opening it with a shell script, “env,” or finding the specific executable in Linux filesystem (what Windows/Mac user would…?). I’ll debug this new problem later. Meanwhile, yet another critical part of workflow has broken for no justifiable reason if any QA is getting done.

              I can’t remember that kind of stuff happening on Windows (NT onward) until Vista’s issues with hardware. Aside from bloat, it worked fine with some apps needing WinXP compatibility mode. Simple fix. I’m likewise not knocking Linux on hardware issues: just the one’s developers can avoid easily. Wireless suddenly stops working, can’t open PDF’s, weird interactions between three ways of managing packages I need, and so on. The best proof is probably that several, small distros did fix some of these problems despite not having millions of dollars.

            2. 2

              If he means success as a consumer OS, it’s not clear to me that any players besides Canonical were ever tilting at that particular windmill.

              Strictly speaking, Chrome OS put a Linux kernel (albeit a somewhat non-standard one) on all sorts of consumer machines.

              1. 9

                While loads of people talk about ChromeOS and Android in these discussions, I think “Linux on the desktop” is more about “free software on the desktop”, and they don’t really hit the mark, though Android has absolutely been a pretty important step.

            3. 1

              A proper transcript would be much nicer, but youtube does do a fairly decent job at automatically captioning the video so you can turn that on and watch it silently if you really want…

            4. 4

              I mean, the idea of a Linux desktop is nonsensical. What does it mean? A Unix desktop? Well, that’s here – it’s just called “macOS”. Linux as the foundational technology for consumers? That’s here, too, in the form of Android. But there is no such thing as a “Linux desktop”, and there never has been.

              I’m sure Ubuntu is perfectly usable for 75% of what 75% of people do with their computers, but I’ll never know, as it’s entirely unfit for my purposes.

              1. 3

                A Linux desktop is a usable alternative to Windows and Mac in their common, use cases that runs on Linux with Linux apps (not Wine). Ubuntu, SUSE, and Mint are probably best examples. If adding Windows compatibility, the best example of how I’d define the phrase was probably Linspire that aimed squarely at Windows users that might not want to learn Linux internals or lose their apps. Turns out they’re still around, too, having just done a release in May. Hmm.

                1. 5

                  Sure, but there is no single “Linux desktop”, the way there is a Mac or Windows desktop. There are many different ways to do the above.

                  1. 1

                    That’s true. It’s even inherent in Linux philosophy. I think what a “desktop” is was defined by prior systems. They had installers, GUI’s, apps for common things, etc. Linux desktop, if aiming for mass appeal, would have to be consistent with that definition. If not aiming for that and consistent with Linux uses, then it’s going to be meaningless like you say due to fragmentation.

              2. 3

                Really not sure how much this is clouded by my personal experience but hardly any of the people who traditionally used Linux actually used Ubuntu. Sure, I put in on an eeePC or on any device where I quickly needed “a working Linux” - but only in this job it’s on my main machine. Nearly every Linux “power user” I know uses something different. Or switched to Mac after using Linux on the desktop.

                Oh, and why? I’m just not a big fan of how they do things. For me it’s a less-functioning Debian (where it counts) and some added nicety of having more packages available, but I could always remedy that by building myself or using nix-pkg. Also never used the default window manager, so all the work put into that was lost for me anyway.

                So what I’m trying to say I don’t know if this mattered, but I surely wasn’t a lot of help for people approaching me with Ubuntu-specific problems, as I had to research them myself. Also I didn’t have a vested interested in converting people from Windows, so I would have given them Ubuntu, but there was no real occasion. So the fragmentation was and is still real.

                1. 2

                  I’ve been a ‘traditional Linux power user’ user since before Ubuntu came along, but I did use it as my primary OS for many years. I switched over from Debian (and Fedora/RedHat before that) around 2006, since it offered a nice “Debian with GUI defaults”; incidentally, I think they gave Debian they kick they needed to make installation a bit more newbie-friendly. IIRC Debian only had a text installer, with no live CD, and it only installed a base system, leaving the user to apt-get X/gnome/kde/etc. “manually”.

                  Ubuntu certainly feels different now, and less “power-user-friendly”. I switched back to Debian around 2010, probably to avoid Gnome3/Unity/Mir/etc. (can’t remember too clearly; I used to use Gnome+E16, then switched to KDE4). The past 5 years I’ve been on NixOS, although I still use Debian on “less serious” machines (e.g. for media playing, raspberry pi’s, my OpenMoko, etc.)

                  1. 1

                    Ha! I was going to make almost this exact comment.

                  2. 1

                    This interviewer keeps approaching the situation as if the Linux desktop has failed, while the other person is trying to talk about improving the success of it and doesn’t acknowledge any idea of it “failing” (which it isn’t)