GNOME to drop support for BSD, Solaris, Unix?

Take this one with a pinch of hearty pinch of salt for now, but, in a post to the GNOME Developer Mailing List, Jon McCann – a tour de force in the GNOME world and pioneer of GNOME Shell itself – has urged that GNOME not only become an OS, but forgo keeping support for other non-Linux operating systems such as BSD, Solaris and Unix in the process.

Systemdwut?

What’s the beef? First let’s get up-to-speed on the issue courtesy of Redditor deebeeoh: -

“[GNOME] are proposing systemd as a dependency for GNOME Shell to function. systemd has made it quite clear that no port to other OS’s would be supported (or even easy due to technical details).

In short, by depending on systemd, they force the OS to pick a particular init system which is only supported in Linux. This is all proposed for GNOME 3.2, so very soon.”

‘The future of GNOME is an OS”

Today, in reply to Debian’s Josselin Mouette on the issue of whether adopting the Linux-only init replacement daemon ‘Systemd’ was an agreeable step to take, McCann wrote: -

“The future of GNOME is as a Linux based OS. It is harmful to pretend that you are writing the OS core to work on any number of different kernels, user space subsystem combinations, and core libraries. That said, there may be value in defining an application development platform or SDK that exposes higher level, more consistent, and coherent API. But that is a separate issue from how we write core GNOME components like the System Settings.

It is free software and people are free to port GNOME to any other architecture or try to exchange kernels or whatever. But that is silly for us to worry about.

Kernels just aren’t that interesting. Linux isn’t an OS. Now it is our job to try to build one – finally. Let’s do it.

I think the time has come for GNOME to embrace Linux a bit more boldly.”

This bold, rather rousing, read is completely understandable in the context of GNOME becoming the best that it can.

The good news is that the ‘systemd saga’ is on-going and many further discussions are yet to be had before any commitment to a decision is made either way.

GNOME Devel mailing list via Reddit

No related posts.

Post a comment or leave a trackback: Trackback URL.
  • https://launchpad.net/~andycandet Andy C.

    Oh, goody. Another minority Linux distro coming up.

    • Anonymous

      dont you mean another minority “gnome-shell” distro?

      because clearly gnome, gtk and its programs will still be part of ubuntu.

      i think is a good idea for them to worry only about linux, that way they have more people/time to work out the bugs faster (i assume).

      and bsd can still use awesome kde meanwhile (and i believe most of them were using it anyway as default).

  • http://www.tux-crazy.com Tux Crazy

    :§)  Oh look! Smiley has a moustache.

    • http://omgubuntu.co.uk/ d0od

      I like that 

    • http://omgubuntu.co.uk/ d0od

      I like that 

    • The Negative Shape

      Smiley is angry.  >;§{ (

  • triplezone3 triplezone3

     Ubuntu hasn’t committed to systemd yet, so this could potentially exclude Ubuntu Linux as well.

    Bear in mind this is still just a discussion on a mailing list, not a decision.

    • http://twitter.com/om26er Omer Akram

       >>Bear in mind this is still just a discussion on a mailing list, not a decision.

      thats why i thought there was a question mark in the title

    • http://twitter.com/om26er Omer Akram

       >>Bear in mind this is still just a discussion on a mailing list, not a decision.

      thats why i thought there was a question mark in the title

    • Anonymous

       This is a bit more than that, actually. The discussion to make GNOME an OS has been going on for a little while, and it’s a bit divisive.

  • http://geojorg.tumblr.com Jorge Guerrero

    If accepted it would be difficult for Ubuntu to ship Gnome 3.2 ? Ubuntu, Debian and Arch are not planning to move to systemd in the near future.

    • https://launchpad.net/~jassmith Jason Smith

      I can see where that might cause issues. It might mean Ubuntu would have to switch init systems to get GNOME Shell (depending on the hardness of the dep) or ship a degraded GNOME Shell. Switching init systems is something no distro takes lightly. 

    • https://launchpad.net/~jassmith Jason Smith

      I can see where that might cause issues. It might mean Ubuntu would have to switch init systems to get GNOME Shell (depending on the hardness of the dep) or ship a degraded GNOME Shell. Switching init systems is something no distro takes lightly. 

      • http://twitter.com/jspaleta Jef Spaleta

        Its important to read the entire discussion. For example.

        http://mail.gnome.org/archives/desktop-devel-list/2011-May/msg00433.html
        “What I am willing to support is builds of systemd that consist only ofthe tiny mechanism daemons, and leave the core of it outside. That wayfolks can install these mechanisms and stick with their old init systemsfor a while.”

        My reading of that, is for linux distros who can’t turn the corner quickly and can’t use systemd as an init system in the short/near term can still use the evolving systemd session session management bits needed for GNOME and use something else for init.

        It really helps to read up on the systemd roadmap document the systemd upstream developers floated recently which lay out the vision for systemd for session management to get context.

        There are very real and very powerful technical and maintenance wins for every linux distribution.  I don’t think there is much in the way of disagreement on that, even from Ubuntu devs I’ve seen chime in.  There is some concern over timing and pace of development and I think that is recognized and the systemd devs are trying to be accomedating for linux distributors on that score in good faith. But you can’t wait forever. If this is the best technical path forward, then there are real market reasons to get it done as fast as possible instead of dragging it out for 2 years.

        If the bsd or solaris kernel development can’t keep up, then how relevant are they in the marketplace really?  Is Shuttleworth going to get to 200 million Ubuntu users by slowing down needed system integration work so BSD and Solaris kernel developers can play catch up? No way.

        Welcome to the land of hard choices and time sensitive prioritization of desired goals. Systemd is the way forward, its just matter of how long everyone is willing to wait to bring its technical benefits to the table.

        -jef

        • https://launchpad.net/~davidnielsen David Nielsen

          For once Jef and I are in complete agreement. People are not reading all the material and overstate the long term consequences while understating the benefits of committing to such moves.

          More such thoughts:
          http://davidnielsen.wordpress.com/2011/05/18/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means/

          • http://twitter.com/jspaleta Jef Spaleta

             I believe your agreement with me is one of the signs that herald the coming of the ends of days. So right on schedule for a rapturous May 21st.

          • https://launchpad.net/~davidnielsen David Nielsen

             I believe you and I agree on a lot more than you think Jef. Regardless it is always nice to see a herculean sagacious freedom warrior speak sense.

      • Anonymous

        Are/were you deebeeoh/DBO? 

    • Dmitry Shachnev
      • http://geojorg.tumblr.com Jorge Guerrero

        Systemd is proposed for Gnome 3.2 wich should be the default in Ubuntu 11.10. So maybe ubuntu would ship with 3.0 and not with 3.2 to avoid systemd problem

        • http://profiles.google.com/lilianftp Moraru Lilian

           Ubuntu 11.10 will not have Systemd, that’s 100%. They said that before releasing an LTS(12.04) they will not make this kind of switches, they stay with a known, stable “configuration”. 11.10 is kind of a “Getting ready for 12.04, no major changes”.

    • Anonymous

       are you sure about arch? systemd is already in the repos

      • http://geojorg.tumblr.com Jorge Guerrero

        It does, but it is not used as default 

        • https://launchpad.net/~isantop Ian

          The only thing that is standard and used by default in arch is a Kernel. 

          • Yi Sun-sin

             systemd is in [community], and the init scripts are in [core].
            If it’s not even in [extra], it means it’s really not the standard way.

    • http://twitter.com/jspaleta Jef Spaleta

       Debian uses sysvinit by default right?  If that is the case then Ubuntu has already shown that its not a problem to diverge from Debian with regard to default init system with the adoption of upstart.

      Debian may have an issue with making systemd default due to their support of non-linux kernels as a release target, but systemd existing in Debian as a non-default init is certainly doable. Debian’s default init choice is a matter of Debian’s choice to be bigger than just linux. Perhaps its instructive for you to review why Debian doesn’t use upstart by default either.

      http://lists.debian.org/debian-bsd/2009/07/msg00122.html

      Think very hard about the implications of that thread from nearly 2 years ago with regard to the _same_ issues concerning the portability of upstart.

      Ubuntu doesn’t have a bsd kernel yet so there is no equivalent barrier for Ubuntu picking up systemd as the default. There is literally nothing new here that was not at issue with the introduction of the linux-specific upstart codebase.

      -jef

  • http://geojorg.tumblr.com Jorge Guerrero

    If accepted it would be difficult for Ubuntu to ship Gnome 3.2 ? Ubuntu, Debian and Arch are not planning to move to systemd in the near future.

  • Anonymous

     Gnome as an OS! HA!

    • http://twitter.com/the_madman Marcus Harrison

       We’ve already got that. It’s called MacOS.

      Yeah, I went there.

      • http://pulse.yahoo.com/_KFF2PURLMSEKI3ALERT2PGJHTY NathanS

        No, because Mac at least lets you open a terminal in BSD.

        Actually, you guys won’t believe it, but I already have a mock-up of GNOME-OS, codename “Cultural Revolution”. So far it requires quad-cross-fire graphics cards with 8GB of DDR5, another 8 GB of RAM and a hex-core CPU, and features anti-aliasing so smooth it makes the baby Jesus cry. the final RC is going to consist entirely of the following menu, with your eyes being read by the machine to move the cursor, will only display through 3D holograph projection, and will feature six amazing apps, all coded in Mono:

        “Check your Facebook” “Post to Twitter” “Read some ‘blogs’!” “Delete word processor, PDF viewer and money management applications” “Reinstall Windows” “Buy a Mac”

        GNOME has announced that they will fill the GNOMEnix kernel with various cripple-functions if users attempt modifications either through the GUI, java, CSS, or the command line, because “from a design standpoint free and open source software isn’t about choice, it’s about Stalinism.”

        In reality, the good thing is that by the time GNOME-OS is in
        development, in about three years this dude will be too “old-fashioned”
        and not “innovative” enough for the GNOME team.

        • http://profiles.google.com/bwat47 Brandon Watkins

          Are you trolling? Please tell me this ridiculous post isn’t serious… 

          • http://alaukik.myopenid.com/ Alaukik

            surely trolling .

          • http://pulse.yahoo.com/_KFF2PURLMSEKI3ALERT2PGJHTY NathanS

             Okay, a Mono reference, the suggestion to buy a mac (or System76 for that matter), two jokes about communism, one about buzzwords and about people over 25 not being ‘hip’ enough, a list of non-existent hardware… can this be about anything but the lulz?

  • Anonymous

     Gnome as an OS! HA!

  • Akshat Jain

    This is indeed a very bad news for distros that are unlikely to switch to systemd like Ubuntu(Upstart), Arch and Slackware(BSD style), Gentoo (custom), Debian(System V) etc.

  • Akshat Jain

    This is indeed a very bad news for distros that are unlikely to switch to systemd like Ubuntu(Upstart), Arch and Slackware(BSD style), Gentoo (custom), Debian(System V) etc.

    • http://carlaalvarez.mp/ Carla Alvarez

      Well, at least in Arch systemd is already in the official repos, and it can be installed side by side with the traditional init scripts.

    • Martin Stitz

       i’m using systemd on my arch notebook without any problems

    • https://launchpad.net/~davidnielsen David Nielsen

      Ubuntu switching seems to be more a question of before or after  the next LTS. They’ve gone far down that path and switching will be painful but if they keep it post LTS they are left holding that codebase alone. Every major distribution has committed to systemd, that is a lot of shared labor and much needed standardization to win. 

      I believe even more in the light of this excellent proposal to leverage systemd’s capabilities in the desktop, I think sanity speaks to not discounting switching to systemd in Oneiric.

      I am sure we can crowd source a lot of the tasks, such as reviewing the 1000 or so patches Ubuntu carries to integrate with Upstart. List them, review them – sounds downright papercut like :) 1000 papercuts we stand to lose, a nice maintance burden to share potentially.

      Systemd is a fact. Like it or not it has won a lot of hearts and brings a lot of great technology for a wide range of users (desktop, server, embedded, admin). It is not a political statement, it is a technical win and a big play for cross distribution cooperation and foundational standardization.

      • https://launchpad.net/~mpt mpt

        “Every major distribution has committed to systemd”

        That is not true. Ubuntu has more users than Red Hat, Fedora, CentOS, Arch, Debian,Gentoo, Suse, Mandriva, and Mint combined. It is the only “major distribution” now.

        • http://www.facebook.com/john.t.folden John T. Folden

          Well, someone is buying the marketing…

        • http://castrojo.tumblr.com Jorge Castro

          … and RHEL6 and SLES all ship Upstart as their default.  

          • https://launchpad.net/~davidnielsen David Nielsen

            Besides the point. Red Hat even invested development time in Upstart once upon a time. They use the old init system in older still supported pre-RHEL6 releases and with their base Fedora moving to Systemd this is overwhemingly likely to be the choice for RHEL7.

            Same thing with SLES. openSUSE is switching, likely the next SLES release will as well.

            So that is entirely a null argument.

          • https://launchpad.net/~davidnielsen David Nielsen

            Besides the point. Red Hat even invested development time in Upstart once upon a time. They use the old init system in older still supported pre-RHEL6 releases and with their base Fedora moving to Systemd this is overwhemingly likely to be the choice for RHEL7.

            Same thing with SLES. openSUSE is switching, likely the next SLES release will as well.

            So that is entirely a null argument.

        • http://twitter.com/jspaleta Jef Spaleta

           Is this statement based on the wikimedia stats?

          There continues to be a discrepancy in the wikimedia tables they release with regard to handling of total counts “linux” counts. They don’t add up when you try to double check for self-consistency across tables.  Which is an eyebrow raiser.

        • https://launchpad.net/~davidnielsen David Nielsen

          1) You have provided zero data to back the number of users.

          2) You then use that assertion to effectively state that absolute numbers of users is all that matters for the determination of cross distribution collaboration efforts, not that a given piece of technology has won over every major group of developers relevant. Ubuntu in this sense is not it’s user base however big it is, it is one distribution out of multiple which all work to make up Linux.

          I don’t think argumentation means what you think it does.

          • http://twitter.com/jspaleta Jef Spaleta

            Whoa, did you just question a userbase estimate not backed by publicly stated methodology?

            I’d be interested in hearing what you consider adequate public disclosure with regard to these sort of estimate. I’m keen on establishing some reasonable best practices that all distros could choose to adopt so we can get a better overall view of the size of our collective ecosystem of users. This is an historically difficult problem, and the lack of methodology disclosure and best practices isn’t helping.

            -jef

  • Yi Sun-sin

    “ Microsoft ended up with a patchy, pithy OS due to having support multitudes of hardware”
    WTF do you mean ?
    Windows support less hardware than Linux. The desktop edition of Windows support only x86 and x64.

    • Anonymous

       I was wondering too what on earth they are talking about…

    • Anonymous

      Sure if you take his statement only to mean CPU architectures which it clearly wasn’t.

      • Yi Sun-sin

        Well, what was it ? Do you think it was because they tried to support too many printers ? Or maybe because they supported too many different keyboards ? Or too many webcams ? Too many graphic cards ? Too many monitors ?
        I just don’t get your point.
        Windows desktop supports only personals computers. That’s definitely not was I would call “a multitude hardware”, especially when comparing to Linux, which goes in super-computers, in VCR, in phones, in tablets,in routers, …

      • Yi Sun-sin

        Well, what was it ? Do you think it was because they tried to support too many printers ? Or maybe because they supported too many different keyboards ? Or too many webcams ? Too many graphic cards ? Too many monitors ?
        I just don’t get your point.
        Windows desktop supports only personals computers. That’s definitely not was I would call “a multitude hardware”, especially when comparing to Linux, which goes in super-computers, in VCR, in phones, in tablets,in routers, …

    • Anonymous

      Sure if you take his statement only to mean CPU architectures which it clearly wasn’t.

  • http://twitter.com/azwarain Mohd Azwar

    gnome for linux? Wow… anyway most of BSD & unix use kde & xfce except ghostbsd, solaris alike & etc.

  • http://twitter.com/luisdavim Luis Davim

     And what about gnome (gtk) based apps??

    • https://launchpad.net/~shnatsel Shnatsel

       GTK will remain cross-platform for sure.

  • http://twitter.com/JamesDenholm James Denholm

    Am I hearing an OOo/LO move? I think I’m hearing an OOo/LO move…

    • http://www.facebook.com/trombonecricket Dillon ‘Queso’ Connell

      That is one possibility. Also, seeing as Unity is bound to get much better (Yes I am a supporter of Unity, I just think it need a lot of work), maybe Canonical should consider dropping GNOME and begin a project of building Unity from scratch. It would take a lot of time and energy, but it would not be beneficial for GNOME’s plans to force Ubuntu into directions they don’t want to go. I think that would happen if GNOME followed such a closed-minded path. One way or another, I feel like GNOME asking other kernels to suffer so that they can put out an OS based on their own narrow views is going against everything the linux community stands for. As of now, I still prefer GNOME to KDE, but this may force me to take a look at other environments. 

      • http://pulse.yahoo.com/_KFF2PURLMSEKI3ALERT2PGJHTY NathanS

         Canonical doesn’t do upstream coding, they do marketing. Which is fine, but your hopes are misplaced.

    • http://www.facebook.com/trombonecricket Dillon ‘Queso’ Connell

      That is one possibility. Also, seeing as Unity is bound to get much better (Yes I am a supporter of Unity, I just think it need a lot of work), maybe Canonical should consider dropping GNOME and begin a project of building Unity from scratch. It would take a lot of time and energy, but it would not be beneficial for GNOME’s plans to force Ubuntu into directions they don’t want to go. I think that would happen if GNOME followed such a closed-minded path. One way or another, I feel like GNOME asking other kernels to suffer so that they can put out an OS based on their own narrow views is going against everything the linux community stands for. As of now, I still prefer GNOME to KDE, but this may force me to take a look at other environments. 

  • Anonymous

    Why would GNOME Shell care what init script is used? To satisfy one man’s need to troll the world?

    • Mr Dandy

       Because this RedHat peoples. Like developer of systemd – Lennart Poettering, also from RedHat. Its commercial company. It does not care about other distributions. Linux kernel must cease to be – it should be OS, by the work of RedHat. This will ensure the influx of customers.

      • https://launchpad.net/~cscarney ~cscarney

         Be careful what you say about Red Hat.  RH employees wrote a *lot* of the code in Ubuntu.

        • http://twitter.com/kmetamorphosis Bertel King, Jr.

          More than Ubuntu developer’s, I’d say, since a large chunk of GNOME is maintained by them.

          I’d say Mr Dandy’s comment is entirely uncalled for, since Canonical is also a commercial company, one with a far greater history of not caring about other distributions.

          • Anonymous

             Red Hat is to thank for a lot of the technical innovation you see in various Linux distros. I moved away from Red Hat related distros when Fedora Core three was released, but you cannot deny their influence over all things Linux. Canonical is not yet profitable, maybe when they are we will start to see their contributions. I will admit they have put alot into usability and marketing, but I’m not sure what direction they are heading towards with the Unity interface. Having said that I’m not sure what the Gnome developers are thinking either. I’m currently running openSUSE 11.4 with KDE. I haven’t run KDE since Mandrake  9.2, but I’m really enjoying what  they’ve done with it. Linux is about choice, use what you like and support the community.

        • Anonymous

          you could argue that their influence over holding on to ugly and old apps (e.g. network-manager) and not making way for the new stuff written by others, has held gnome2 back.

  • http://profiles.google.com/lilianftp Moraru Lilian

     I’m actually happy to hear this news…
    That means more time for developers to make a really good product for only one platform “Linux”, where it is really used…

  • Polly

     I think it is a great idea and they should go for it.

  • https://launchpad.net/~davidnielsen David Nielsen

     It just means dropping support from non-systemd platforms. It is no different than any other technical requirements. This will help make GNOME awesome. If a platform provides the Systemd API and functionality then it is supportable. Just like say for gtk+

    There is no reason why *BSD e.g. could not work to provide the functionality needed and expose it in a compatible fashion. Sure it sucks, I like FreeBSD, but I am sure they will compete if they so desire, they have done before. I would not discount the alternatives just yet.

    In the meantime we gain a lot of desirable functionality and much needed standardization. Each of which alone are worth this.

    • http://pulse.yahoo.com/_3YATP2IRIAY3Z5F5IJATPVDD3E Juan

      I don’t think any BSD will push systemd just because GNOME shell depends on it. They will use it only if it is the best thing to do for BSD, if it’s not, there are tons of other Desktop Environments.

      BSD in general is about having a rock-solid OS, they can use any other DE (KDE, XFCE, LXDE, etc…) and that’s it.

      • http://profiles.google.com/sprkv5 Subhashish Pradhan

        ” I don’t think any BSD will push systemd just because GNOME shell depends on it.”

        I think PC-BSD (version 9) may push for systemd because it looks forward to officially support main DE’s .

      • https://launchpad.net/~davidnielsen David Nielsen

        They don’t have to, they just have to support the same interfaces. Lennart has stated that he is more than willing to work towards separating out the core of systemd so say a FreeBSD backend could be dropped in instead of the Linux one currently in place. That is just a matter of technicalities not unwillingness to make this work. 

        This is about the interfaces, not systemd as one whole blob. Fear not. Lennarts initial email and the entire thread is very informing on the subject.

        https://mail.gnome.org/archives/desktop-devel-list/2011-May/msg00427.html
         

    • Anonymous

       bsd will be using kde i suppose.

      but you are right, if they want gnome, they could work on it, nothing stopping them.

      I’ve seen much less popular and hard to port projects get ported…

    • http://alaukik.myopenid.com/ Alaukik

       so does that exclude ubuntu also ?

  • https://launchpad.net/~shnatsel Shnatsel

    systemd is not a good thing to depend on. systemd is a good idea but there are several places where it really sucks. Upstart is a better idea because it’s crazily flexible. For example, you can make it work mostly like systemd but use a different approach in the places where systemd fails. http://netsplit.com/2010/05/27/dependency-based-event-based-init-daemons-and-launchd/
    I wonder why they’re not happy with D-bus…
    And GNOME as a distro is… crazy. No popular distro ever shipped vanilla GNOME. AFAIK the only distro that did that is Foresight Linux, which is a clear outsider.

    • Anonymous

      It’s not like I really understand init systems, but just reading a thinking about it made me realize a few things:

      1. The on-demand approach is really a subset of the event-based approach.

      2. There are some things that are better expressed as events, such as “when music CD is inserted, start Rhythmbox”. I simply can’t see that expressed as a dependency without getting really convoluted and doing things in a hackish way.

      3. There are things better expressed as dependencies. For example, to specify a dependency as an event you might say something like:
      “when NEPOMUK starts, start Virtuoso first”
      “when Virtuoso stops, if NEPOMUK is running, start Virtuoso”
      “when NEPOMUK stops, stop Virtuoso”

      Or you could simply say:
      “while NEPOMUK runs, run Virtuoso”

      It describes what conditions should exist throughout a given situation, which is really the same as defining a dependency. 

      4. Lennart is a douche. He went on the blog of what seems to be the creator of Upstart and said something along the lines of “Upstart is a pile of crap”.

      I’m not surprised with what Lennart said about the various distros that use Upstart only using it in init compatibility mode, given that it should be easier to specify dependencies using an API that is designed for that rather than an API designed for describing events, and most of what they’d want to do with an init system is describe dependencies. I do wish that Upstart would add API for describing dependencies as part of it’s “native” API.  it could be nothing more than a shorthand for a bunch of lines of event-driven code, but would make the user’s work far easier. Really, dependencies and events are complimentary, not contradictory to each other.

      It would be nice if more states than just “is running” and “is not running” would be accessible. For instance, programs might want to tell Upstart what their state and provide handles for triggering behavior other than just starting or stopping the program. For example:
      “when udev.musicCD.isInserted Rhythmbox.Play(udev.musicCD)”
      Or something like that.

      It would also be nice if there was a way to use abstract events and dependencies and trigger abstract actions, but I can’t really think of any concrete example.

      There was a comment about blocking other services from accessing certain files or folders until they undergo some sort of setup procedure, but that sounds to me like file lock, which should be implemented in the kernel if it isn’t already. If it gives programs an error when trying to access a locked file, perhaps there should be an option to block the program’s execution until the file becomes available, and only return an error after a decent timeout.

  • Anonymous

     I can’t help myself thinking this is the answer to Mark Shuttelworth’s criticism of percieved lack of leadership in GNOME community. But not the answer MS would like to see.

    • http://twitter.com/humphreybc Benjamin Humphrey

      And if it is, from what I’ve seen, McCann is not a suitable leader.

  • http://profiles.google.com/h.venhorst hugo venhorst

    would it be as simple as getting gnome 3.2, removing gnome-shell, removing systemd, stick unity on it, and have no problems? or is systemd tied into more of gnome?

    it just sounds backwards to have a ui layer, which would be at the top of the stack, demand changes on something all the way at the bottom of the stack, causing incompatibility with multiple kernels.

    you must REALLY want that UI  0_o

    besides, isn’t Gnome-shell running already? what would systemd do to it to make it so necessary?

  • http://profiles.google.com/h.venhorst hugo venhorst

    would it be as simple as getting gnome 3.2, removing gnome-shell, removing systemd, stick unity on it, and have no problems? or is systemd tied into more of gnome?

    it just sounds backwards to have a ui layer, which would be at the top of the stack, demand changes on something all the way at the bottom of the stack, causing incompatibility with multiple kernels.

    you must REALLY want that UI  0_o

    besides, isn’t Gnome-shell running already? what would systemd do to it to make it so necessary?

  • http://www.go2linux.org/ Guillermo Garron

     Does this possible new direction has something to do with the decision of Ubuntu of letting Gnome for Unity?

    • Anonymous

      what you mean by “gnome”?

      as you can see ubuntu is still using all gnome, except gnome-shell.

  • http://twitter.com/Surlent777 Sabaku no Surlent

    KDE is looking so much nicer every day. It doesn’t neuter you, it doesn’t think you’re an idiot, and the dev team isn’t trying to take over/troll the world. You want to use screensavers? You have screensavers. You want to minimize? Man, they’ve got you covered. You want to work on more than one application at a time without feeling silly about it? Can do. You want your default policy to be that unplugging your laptop doesn’t decrease your screen brightness?  Not a problem. You want to change your theme? No third-party apps or patches needed dude, you go right ahead. Seriously, KDE has been a godsend this release. I used to just dabble with it, but I’ve made it my primary DE on my laptop now and will very likely move over from GNOME 2 on my desktop shortly, especially given that Classic is supposed to disappear next release.

    • http://profiles.google.com/sprkv5 Subhashish Pradhan

      That’s it, I am switching to KDE. Screw Unity and Gnome Shell. 

      • http://www.facebook.com/people/Jeremy-Newton/512458865 Jeremy Newton

        Unlike lol, I’d rather go with whatever mint is doing before KDE. I think they may be using the fall-safe session in GNOME 3 (looks like gnome 2), modified with compiz or whatever.

        • http://twitter.com/yannanth Γιάννης

          Nuh-uh! They’ll be using vanilla GNOME 2 instead as porting things to GNOME 3 is a pain in the butt and results in many bugs! So yeah, good luck GNOMEfag!

      • http://pulse.yahoo.com/_A7FFFBS2QYWWY3J5Y3YYVSAAYQ thiyagi

        Guys..seriously Unity need lots of debugging and fixes. Its like, it is still in alpha version. 

        I already changed to KDE and i am loving it…

      • http://pulse.yahoo.com/_A7FFFBS2QYWWY3J5Y3YYVSAAYQ thiyagi

        Guys..seriously Unity need lots of debugging and fixes. Its like, it is still in alpha version. 

        I already changed to KDE and i am loving it…

    • http://profiles.google.com/sprkv5 Subhashish Pradhan

      That’s it, I am switching to KDE. Screw Unity and Gnome Shell. 

    • http://mark-y-a.myopenid.com/ Mark

       I wish they’d (Ubuntu) give as much attention to KDE. No, Kubuntu is not
      an excuse.  Bling-wise, KDE is pretty much up there! I’ve always had
      issues with Gnome appearance (I’m not saying it’s fugly) and the only
      distro to make me switch to Gnome is Ubuntu. Damn, I was a KDE user for
      several years.

      Ok that’s it I’m moving to Arch!

      • https://launchpad.net/~andycandet Andy C.

        Exactly my thoughts. KDE on Ubuntu downright sucks, and openSUSE just doesn’t cut it for me. I want my PPAs, man.

        • Alex Abbott

           You know that you can get anything from build.opensuse.org that you can from PPAs right?

    • http://www.facebook.com/profile.php?id=100000528522053 Ladislav Ezr

       one big – …. its German……

    • http://www.facebook.com/profile.php?id=100000528522053 Ladislav Ezr

       one big – …. its German……

      • https://launchpad.net/~isantop Ian

         Und ihr Punkt?

      • Anonymous

         What’s your problem with that?

        • http://www.facebook.com/profile.php?id=100000528522053 Ladislav Ezr

           I don’t like them at all O_o (Im from EU BTW…)

    • http://www.facebook.com/profile.php?id=100000528522053 Ladislav Ezr

       one big – …. its German……

    • Anonymous

      On the other hand, GNOME is more streamlined, simpler, more usable, less distracting,  easier to manage, and less resouce-intensive. There are benefits to both. I personally go the third route and choose XFCE (and LXDE, if it ever becomes stable enough) — really don’t want the extra cruft that comes with GNOME and KDE, plus I prefer a more responsive desktop.

      • http://pulse.yahoo.com/_4LXTMAC7KVMD7JBKWBCCBU4HSU inner_turbulence

        when was the last time you have used kde?

      • http://pulse.yahoo.com/_4LXTMAC7KVMD7JBKWBCCBU4HSU inner_turbulence

        when was the last time you have used kde?

      • http://identi.ca/gerlos gerlos

         Obviously, you didn’t tried KDE recently. It runs fine on my 8 years old computer, as on my recent laptop, and after a day or two of use and configuration, it can be made really clean and tidy, fitting tighly your needs.

        You don’t need any more a developer team to tell you how your desktop should be to be productive. They can’t know your tastes and your needs as you. KDE doesn’t assume to know everything about you, and leave you free to adapt the tool (the DE) to you.

        • http://twitter.com/SarcasticSloth Steven Garza

           ”after a day or two of use and configuration”
          That’s the problem, nobody wants to spend a full day configuring KDE.

          • Nowardev-1 NowardevTeam

             you have not to spend a day…
            just an example

            http://nowardev.files.wordpress.com/2011/04/gnome-defaltt-panel-kde.gif

          • Anonymous

             I hate to say, but that mac theme looks awful (even by OS X standards). I’ve been trying KDE on my openSUSE install. I have no qualms with it so far, I’m certainly more accustomed to GNOME and Ubuntu, but I’m going to keep experimenting with it.

      • Jean-Philippe Green

         http://apachelog.wordpress.com/2011/05/17/kubuntu-11-10-sneak-peak/

        As you can see on that link, there will be a low fat Kubuntu configuration in 11.10. Sure GNOME is less distracting, especially since GNOME-shell focuses on 1 application at the time, but since Ubuntu uses Unity there’s not much difference in distraction between Ubuntu and Kubuntu.

        They will also improve the accessibility in kubuntu 11.10. That’s a thing that GNOME always has been better at, but now it’s about to become just as good on KDE.

        Ubuntu has taken some small steps away from GNOME lately. Unity, Wayland, Qt etc. I don’t see why they don’t go with KDE as a base instead, now that it’s so mature. I’m sure it’s possible to do some sort of KDE-unity hybrid.

    • http://twitter.com/MarcCoquand Marc Coquand

      “You want to change your theme?” GL finding a decent looking complete icon theme that isn’t blue or faenza and then finding a good theme to fit it.

      And lastly GL installing it.

    • Anonymous

       KDE looks childish.  It’s “bubbly” and too fuzzy/glossy.  I tried using it, and it isn’t bad, but it just sits ill under my skin like Win7 does.  To each his/her own for sure.  I have no problems using Gnome.  I’ve yet to try to do something I couldn’t.  I’ve seen no issue with changing screen brightness when unplugging my laptop.  I have no issue with screensavers – built right in, no app needed.  No problem changing themes either.  I have plenty I’ve switched between.  I’m not sure how you feel silly in Gnome working on more than one app at a time.  I have no issues switching between them or having more than one on the same screen.  Maybe you used a very old version of Gnome?  2.32 seems just fine so far.  I’m not sure how I feel about 3.0 yet.  It won’t run on my system. (Ubuntu 10.04)  Maybe I’ll try it when I upgrade to 12.04.

      I’m sure some people like it.  But there’s no need to bash Gnome, especially for things it does just fine.  (at least for me)

      • http://twitter.com/kmetamorphosis Bertel King, Jr.

        The post was referring strictly to GNOME 3.0, and the issues raised were valid. I still love the new version of GNOME far more than GNOME 2. For me, it’s a welcome change.

        At the same time, I still also love KDE. Both desktop environments are radically innovative in completely different ways.

      • http://twitter.com/kmetamorphosis Bertel King, Jr.

        The post was referring strictly to GNOME 3.0, and the issues raised were valid. I still love the new version of GNOME far more than GNOME 2. For me, it’s a welcome change.

        At the same time, I still also love KDE. Both desktop environments are radically innovative in completely different ways.

    • http://profiles.google.com/bwat47 Brandon Watkins

      Saying gnome is trying to “troll/take over the world” is just ridiculous.

      I think gnome 3 is great, and love that they have the boldness to try and do their own thing, even if it can be unpopular with some.

      Its linux, you are always free to use a different alternative.

    • Anonymous

      Just because it’s always my inclination, I happened to do a crap-ton of research when I first installed Linux (first distro being Ubuntu, of course :) ) .  At that time, most of the BS “GOME vs. KDE” crap I read made a point of saying that KDE was, essentially, the power-user’s DE: it’s meant for people who want to have all the tools there, preconfigured and all  (so as to avoid spending a great deal of time setting up *Box, Awesome or whatever) while still offereing an insane amount of control to the user.  I love KDE, even though my laptop can no longer handle it, and I’m looking at an absence of a couple years because of that.  I immediately became enamored of Dolphin, Kate, and Konsole (and of course Konqueror, for those of you who dig the “all-in-one” thing) Those folk have it down and, unlike the GNOME devs, they inted KDE to be a tool to use one’s computer with, rather than a full-blown DE that drags the user along with it regardless of change (the one exception I would note, though, is KWin; those folk have made their intentions clear. )  KDE is also available for Linux, BSD, Mac and Windows, ‘cuz hey, why screw people out of somehting they might find useful?
      This deal with GNOME OS has been their goal for a long time, and the fact is it’s not gonna go anywhere worthwhile.  One of GNOME’s strengths has been its ability to insert itself into mutliple systems, so its simplicity can appeal to a wide audience; that also happens to be its only source of external influence.  They survive solely on the good-will and monetary donations of those who care bout them.  And as any historian/biologist/chemist/sociologist/antrhopologist/physicist/engineer/person-who-spends-a little-time-looking-into-it  will tell you:  cut a system off from external influence–be the system physical or metaphysical–and it WILL stagnate and die.  

    • http://pulse.yahoo.com/_QGUOKPDPIPMOPJK7RCCXIALHRI Ene Dene

       KDE looks very nice why I don’t use it is all that Ksoftware, I have no idea what what is, it’s too much installed by default. I don’t know how it interacts with gtk applications.
      And of course the primary reason why I don’t use it is because I’m satisfied with Gnome and I’m used to it.

    • http://pulse.yahoo.com/_QGUOKPDPIPMOPJK7RCCXIALHRI Ene Dene

       KDE looks very nice why I don’t use it is all that Ksoftware, I have no idea what what is, it’s too much installed by default. I don’t know how it interacts with gtk applications.
      And of course the primary reason why I don’t use it is because I’m satisfied with Gnome and I’m used to it.

    • http://www.facebook.com/profile.php?id=100000055513961 Nikita Nikishin

      I despise GNOME3 and Unity, especially their “the user has ADD, is an idiot, and has no reason to configure our perfect environment, Therefore, we won’t let them.” philosophy. It’s ridiculous that the developers even expect users to function properly with those types of setups.

      I think the developers forgot usability when they were on their slidy-fady-spinny effect spree. The communist Apple philosophy seems to be seeping over into Linux, which saddens me quite a lot.I use Arch Linux, but it doesn’t make a difference. I stopped updating my system after GNOME3 infected the repositories, and now I’m quite sad.

      What has GNOME and Ubuntu become? :[

  • http://pulse.yahoo.com/_3YATP2IRIAY3Z5F5IJATPVDD3E Juan

    And which package manager is this “GNOME OS” going to use? I sense a big deb vs rpm flamewar in the future :P

    • Anonymous

       RPM of course, Red Hat support GNOME team and GNOME so..

    • Yi Sun-sin

       Well, this :
      http://packagekit.org/

      • http://pulse.yahoo.com/_3YATP2IRIAY3Z5F5IJATPVDD3E Juan

        From that page:

        “The actual nuts-and-bolts distro tool (yum, apt, conary, etc) is used by PackageKit using compiled and scripted helpers. PackageKit isn’t meant to replace these tools, instead providing a common set of abstractions that can be used by standard GUI and text mode package managers.” 
        So, from what I understand, you would still need to decide for an existing package format (rpm, deb or whatever) right?

        I should have used “package format” instead of “package manager” in my previous comment :P

        And that’s not the only decision they would need to make if they want to ship a full OS, except for the desktop environment of course ;)

        • Yi Sun-sin

           They don’t plan to make a distribution. They plan to make an operating system. That’s not the same.

          • http://pulse.yahoo.com/_3YATP2IRIAY3Z5F5IJATPVDD3E Juan

            Of course, but this smells like a distro to me. 

          • Yi Sun-sin

            Well, the new article on OMG clearly states that they want to promote PackageKit, and not promote any underlying backend.

          • Yi Sun-sin

            Well, the new article on OMG clearly states that they want to promote PackageKit, and not promote any underlying backend.

    • https://launchpad.net/~davidnielsen David Nielsen

      Not at all that is where PackageKit comes in. Most people misunderstand it and dislike it just for having seen it’s current default UX. However the purpose of PackageKit is to give GNOME and other developers a supported single way to interact with the Package Manager no matter which one so long as it implements a PackageKit backend.

      So there will be no such flamewar, and as for the UX. Most of the tools people know and love in Ubuntu are powered by aptdemon which is a kind of specialized and incompatible PackageKit. The AppStream project is moving to power the Software Center to PackageKit under GSoC this summer.

      Being able to rely on PackageKit when running under GNOME will allow very specific integration for e.g. installation of say a plugin in Banshee. Support for the format when an unknown format is encountered (typical .mp3, .rar, etc.).

      There are still things that are suboptimal in PackageKit but it specifically allows us to evolve a superb e.g. upgrade, update story. Solving specific problems in an elegant way regardless of the platform underneath. Rely on the specified dbus interface instead. 

      Anyone not currently supported are I am sure welcome to submit patches. I am using it on Oneiric regularly to update my system and it works well.

      So no format wars, instead a standard for using formats uniformly for the purposes a desktop cares about.

      • Anonymous

        It’s interesting that you mention AppStream and how Ubuntu intends to integrate PackageKit in the Software Center. Just yesterday a blog post by a Kubuntu dev mentioned how they’d be moving away from PackageKit to Muon, the main reason being: “As it is using APT directly it enables Muon to have a much tighter integration into Debian-like systems (such as Kubuntu) as well as expose specific functionality of APT/DPKG more directly.”

        EDIT: the link to the blog post: http://apachelog.wordpress.com/2011/05/17/kubuntu-11-10-sneak-peak/

        • https://launchpad.net/~davidnielsen David Nielsen

           I mostly agree that short term there are some performance and feature advantages of tight integration, but I would propose that those are bugs and can be fixed.

          I trust Richard Hughes to deliver and I value the benefits of the assumptions PackageKit gives us to provide a superb experience. more thoughts on the matter:http://davidnielsen.wordpress.com/2011/05/18/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means/

          • http://www.manishsinha.net Manish Sinha

            I haven’t checked PackageKit for a long time, but does it allow
            * Running transactions are allowed to be paused?
            * Debconf support?
            * Resolving config file conflicts?

            “those are bugs and can be fixed” is not an excuse. Fix them and then come back toshow your face is what every applications should do. Esp when the application you intend to replace doesn’t have those bugs.

          • https://launchpad.net/~davidnielsen David Nielsen

            there is debconf support and I believe pausing as well. Resolving config files you’d have to ask the PK developers about.

            Richard solved a number of concerns from Debian/Ubuntu already.

            Some references can be found here:

            http://davidnielsen.wordpress.com/2011/05/18/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means/

          • http://www.manishsinha.net Manish Sinha

            @openid-16777:disqus 

            I heard debconf support was added, but didn’t get any links for it.
            I hope you know why aptdaemon was started. If PackageKit does not support everything which aptdaemon supports, I am against replacement.

        • https://launchpad.net/~mpt mpt

          No, he did not say “Ubuntu intends to integrate PackageKit in the Software Center”. AppStream is not part of Ubuntu.

          PackageKit is an extra layer of abstraction over apt — and therefore inevitably slower, less featureful, and less precise than using apt directly. It’s all very well to say “those are bugs and can be fixed”, but the same amount of work put into improving aptdaemon would help several times as many users.

    • Anonymous

      It would probably be RPM, since it’s less distro-dependent (from what I hear, DEB is really only for Debian-based distros).

    • Anonymous

      It would probably be RPM, since it’s less distro-dependent (from what I hear, DEB is really only for Debian-based distros).

      • http://identi.ca/gerlos gerlos

         But there’s a HUGE number of debian-based or apt/dpkg/deb-based distro out there! A lot more than rpm-based ones.
        And if it’s not enough, there are some problems with next versions of RPM… unfortunately, different vendors are working on different, incompatible, versions of RPM :-(

  • http://www.facebook.com/profile.php?id=100000528522053 Ladislav Ezr

    Srsly, this is a joke? O_o

  • Anonymous

    “Jon McCann – a tour de force in the GNOME world (…)”

    One can make just about anything sound smarter with the use of French expressions, but it’s often a good idea to double check that one actually knows what they mean :-P

    Sorry, but I just can’t help myself. Content-wise, though, I found the article both informative and interesting!

    • Yi Sun-sin

       French people tend to use English expression when they want to make something sound smarter. Just like Christini in Dobermann (best French movie of the last decades). But actually, it’s really fun to see all those “en français dans le texte” in book translated into French :D.

  • Anonymous

    “Jon McCann – a tour de force in the GNOME world (…)”

    One can make just about anything sound smarter with the use of French expressions, but it’s often a good idea to double check that one actually knows what they mean :-P

    Sorry, but I just can’t help myself. Content-wise, though, I found the article both informative and interesting!

  • http://pulse.yahoo.com/_DID3BVH23GHHZDFQITQDT7EPZY Sean

    Gnome just keeps getting worse

    Gnome 3 (and the shell) are good in a lot of ways, but my biggest complaints include lack of customisability (hopefully this will get better with time)

    I mean getting rid of the minimize button . . . really?

    I can imagine putting my entire family in front of Gnome Shell and Unity, they’d understand Unity and then then when I’d show them Gnome shell they’d be like “WTH? Where are the windows, where is the minimize button?” One of my family already complains about it frequently when using my PC

    And now dropping support for other OSes and turning Gnome into ANOTHER linux distro (yeah 400 isn’t enough apparently)

    Why Gnome? I thought your goal was to offer a consistent DE across as many environments as possible?

    • http://twitter.com/riccosobers Ricco Sobers

      google chrome has a minimize button :p  I think middle mouse click on the top bar of the window is enough :D

    • Anonymous

      The minimize button is gone because a) there is no panel a window could minimize to, so, for the user, the window just disappears, and b) that functionality has been shifted to workspaces.
      But if you really need it, you can add it back.

    • Anonymous

       You have apparently not been keeping up. Google for “Gnome-tweak-tool” and “Extensions”. That will show you that you are currently wrong. There are plenty of customizing to do if you like.

      You have also apparently missed the point of Gnome OS. (If that ever see the light).

  • Anonymous

     This is worrying on so many levels:
    1- Does this dependency on systemd apply to the whole of GNOME or just GNOME-Shell? If the former, it will be a big problem for Ubuntu because it’s unlikely to just abandon Upstart.
    2- Not every distribution will like changing the init system just to use a specific DE. Why would a DE impose low-level things like the init system, anyway?
    3- IIRC, GNOME is a “GNU package”. You can bet Stallman wouldn’t be too happy about it being Linux-specific. When the Hurd is released in 2073, it won’t be compatible with GNOME.
    4- BSD users already hate the many linuxisms found in lots of programs. I can imagine how they will feel about this.
    5- GNOME OS? Oh dear, that’s a whole different can of worms. I can list another 5 points about why that’s a terrible idea.

    (Obligatory joke: That’s it! I’m moving to GNOME OS)

    • http://twitter.com/humphreybc Benjamin Humphrey

      This is interesting, because only two weeks ago I was reading some articles by Canonical and Red Hat developers about init systems.

      Long story short, the Red Hat developer was insistent on systemd and listed a bunch of reasons why it’s better than anything else (targeting upstart.)

      He also questioned why certain distros (hyperlinked to ubuntu.com) weren’t changing to systemd straight away.

      In reply, Robbie Williamson from Canonical explained why they’re sticking with upstart due to an upcoming LTS early next year, among other reasons.

      We know that Red Hat has an influence in GNOME. My question would be, is this dependency something pressured by certain people just so they can get their way?

      • Anonymous

         If that is so then Canonical is the one who started and is to be blamed by this war.

        Technically systemd is the better choice. If Ubuntu can’t make that transition then i feel sorrow. If they could switch to Unity then they should not have problem to switch to systemd.

        • Anonymous

          I can’t believe you’re blaming Canonical. They have been developing Upstart for 5 years (while everyone else was happy with their SysV and BSD style init scripts). Suddenly, Lennart starts to develop systemd and practically demands  everyone switch to it immediately (He doesn’t say it directly, but his tone says so), even before it’s one year old.

          Personally, I wouldn’t complain if Canonical decided to switch to systemd, but it’s Canonical’s decision, not Lennart’s, and not GNOME developers’.

          • Anonymous

             Couldn’t agree less with you. I was simply just responding to Benjamins last sentence.

            Starting up a blame game would not end up in Canonicals favour. This i am saying as a user of Ubuntu since 6 years back.

            I do not care what Lennart have said or not said. I want systemd as soon as possible. The quicker they start implement it the better.

            Ubuntu are losing ground due of the decision to use Unity and now they are potentially loosing more ground by sticking around with Upstart. Upstart was indeed an improvement and does partly what systemd will do fully. By waiting they will get overrun by the other dists.

          • http://twitter.com/jspaleta Jef Spaleta

            It’s too early to know if the decision to build Unity is losing ground or not.  There’s very little in the way of measurable technical deliverables. The real measure of Unity will be OEM market penetration Canonical can achieve with it. 

            If Canonical is able to re-introduce Ubuntu Light based on the strength of the new Unity UI and get OEMs to pay for it as a pre-install dual-boot with windows on intel hardware as Canonical originally envisioned, that’ll be a significant win.

            Or if Unity-2d is an effective hedge against Android dominance in the ARM OEM space, and we see retail ARM based products coming out with Unity-2d pre-installed that would also be a significant win.

            But it’s way too early to pass judgment on the Unity strategy.

            -jef

          • http://twitter.com/jspaleta Jef Spaleta

            It’s too early to know if the decision to build Unity is losing ground or not.  There’s very little in the way of measurable technical deliverables. The real measure of Unity will be OEM market penetration Canonical can achieve with it. 

            If Canonical is able to re-introduce Ubuntu Light based on the strength of the new Unity UI and get OEMs to pay for it as a pre-install dual-boot with windows on intel hardware as Canonical originally envisioned, that’ll be a significant win.

            Or if Unity-2d is an effective hedge against Android dominance in the ARM OEM space, and we see retail ARM based products coming out with Unity-2d pre-installed that would also be a significant win.

            But it’s way too early to pass judgment on the Unity strategy.

            -jef

          • Anonymous

             @twitter-55957641:disqus  Yes, you are totally right about that. I was however talking about the current situation. Where we can see a lot of people fleeing. But that would not mean that this could turn into a success story eventually. ;)

        • http://www.manishsinha.net Manish Sinha

          Yes. Canonical is responsible for next ice age too or the upcoming apocalypse.  

          • Anonymous

            You are pretty much missing the point i was trying to make. I simply made an answered to Benjamin as he was implying foul play by Fedora and Gnome.

            I have nothing against Canonical.. actually i have loved that company for 6 years. Even though they now have been taking some strange decisions that is still the case. ;)

            But to be fair Canonical did start this whole mess.

          • http://www.manishsinha.net Manish Sinha

             I have no love-hate equations with any company, but blaming Canonical for everything is what I call stupid

            Anyway you failed to tell how is Canonical to blame for this?

          • http://www.manishsinha.net Manish Sinha

             I have no love-hate equations with any company, but blaming Canonical for everything is what I call stupid

            Anyway you failed to tell how is Canonical to blame for this?

          • Anonymous

            (response to the below comment)

            The only one stupid here is you who doesn’t seam to comprehend what are happening here. I am far from blaming Canonical for anything. I was just giving a warning of taking that road. If you have followed the Unity vs Gnome situation you would understand what i am pointing at. This things seldom end up good.

          • Anonymous

             Almehdin: “If that is so then Canonical is the one who started and is to be blamed by this war.”
            Almehdin: “I am far from blaming Canonical for anything.”

            Either way, this whole discussion is moot now. See http://www.omgubuntu.co.uk/2011/05/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means

          • Anonymous

             Almehdin: “If that is so then Canonical is the one who started and is to be blamed by this war.”
            Almehdin: “I am far from blaming Canonical for anything.”

            Either way, this whole discussion is moot now. See http://www.omgubuntu.co.uk/2011/05/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means

          • Anonymous

             You are taking that out of context @imgx64:disqus .  The first was a response to Benjamin. And meant as a reminder from where all this mess started. The second a response to Manisha as a clarification. I hope you think about that next time you decide to cite someone wrongly.

        • John Penland

           from the comments, it sounds like Cannonical doesn’t want to change too much until after their LTS version is released. After that, they would consider switching. I don’t know that for a fact, but it is what I gathered from reading the article/comments.

          • Anonymous

             Yes, me too and they are fully allowed to make such a decision. I do not agree it is wise however. Also.. They have a full year to implement it which i am sure is more than enough. Other have done it in less time. They would even get a chance to test it in the wild when 11.10 comes.

        • http://twitter.com/jspaleta Jef Spaleta

          Careful, its not completely non-trivial to move from upstart to systemd once you have native upstart jobs. systemd was designed for compatibility with sysvinit. There is definitely an additional burden to port native upstart jobs into systemd units. 

          There’s no benefit in throwing blame around at all. 

          Upstart was a legitimate effort at solving a subset of the problem space.  It could even be described as a successful effort considering the history. I remember very well the heated discussion inside Fedora about the need to move to a better init system for multiple releases. Initng was a real contender for adoption in Fedora when upstart came into being.  The ecosystem was ripe for a change, and upstart was a timely effort which dragged us further along the path, which believe it or not actually had less political baggage compared to other alternative init systems already available. Some of us were dragged kicking and screaming…but dragged along none-the-less.

          Successful things can be obsoleted by better things, without their being any need to blame anyone.  This could be one of those times.

          -jef

          • Anonymous

             Yes, that was my intent and in response to Benjamin. I could of course choosed a more wise way to warn about it. I have clarified what i meant in my other comments here though.

            Although non-trivial they have a year to do it. Something that should be enough. I have also seen some ideas of running them side by side at first. To keep compatibility to Upstart while still benefit of Systemd.

          • http://twitter.com/jspaleta Jef Spaleta

             I expect RHEL6 to offer to run them side by side for customers as well. I’ve yet to see a RHEL6 statement with regard to a full switch over.

            This is one of those things where Canonical’s chosen release model approach which mixes LTS targets and rapid development targets into the same product roadmap makes it harder for them to turn the corner on technology enhancements.

            Red Hat has the same constraints, but because they have very specifically disconnect their target audience for RHEL from Fedora it gives Fedora more breathing room to turn the corner on a technology, and break it in without compromising enterprise customers.

            So don’t bregrudge the reluctance here on Ubuntu’s part. This is just a trade-off in their release model.

            But if you really think systemd brings benefits to the table and you really want those benefits soon as an Ubuntu user, do everything you can help Canonical make it possible to land it as an option.  The best way to do that might actually be to help it stabilize as an option in Debian so Ubuntu can merged it in from there.

            Or if you are one of the rare species of Canonical customer or partner, reach in through your customer/partner channels and impress on Canonical how important this one technology is to you to see integrated by next LTS. Flash some cash. Maybe tuck a few dollar bills in Mark’s wasteband at a conference or something equally as subtle.

            -jef

      • https://launchpad.net/~andycandet Andy C.

        Do you happen to have links to those articles? I’m curious about this whole systemd argument (and why the hell does a DE depend on an init system).

      • http://twitter.com/jspaleta Jef Spaleta

        Uhm… you know in the long long history of effort to modernize the linux desktop experience….tighter integration with the init system is really not out of character. Udev/HAL introduction was just as disruptive because it really changed how hotplug was handled and distros had to adapt.  I know your pretty young, so you might not really have a feel for the bad old days before udev so you might not appreciate how far and how much has been done historically to craft the modern desktop linux experience. Work done by some of the very same people you imply have some hidden agenda. It’s far from hidden.

        And of course, when HAL was deprecated (2009?) in favor of pushing a lot of what HAL did into linux-specific udev, that decision took away portability to other non-linux kernel based systems and those other kernel developers had to play catch up to re-implement matching interfaces in their udev-like daemons. Everything old is new again.  For you to suddenly take notice, especially when you are very much enjoying the fruits of previous rounds of modernization embedded in the Ubuntu system stack, is a little naive.

        This is an iterative process in modernizing the concept of the linux operating system. A process which really started way back with Project Utopia.What were you like 12 when that was formed?  Systemd is just the next step a long the road. A road easy to see in hindsight.

        Now excuse me I have to go brush in some Just for Men into my neck beard, and cover up the patches of grey created by having to relate ancient history to an angst filled youth culture.  Kids these days.

        -jef

        • Anonymous

           Couldn’t have said it better.

      • http://twitter.com/jspaleta Jef Spaleta

        Uhm… you know in the long long history of effort to modernize the linux desktop experience….tighter integration with the init system is really not out of character. Udev/HAL introduction was just as disruptive because it really changed how hotplug was handled and distros had to adapt.  I know your pretty young, so you might not really have a feel for the bad old days before udev so you might not appreciate how far and how much has been done historically to craft the modern desktop linux experience. Work done by some of the very same people you imply have some hidden agenda. It’s far from hidden.

        And of course, when HAL was deprecated (2009?) in favor of pushing a lot of what HAL did into linux-specific udev, that decision took away portability to other non-linux kernel based systems and those other kernel developers had to play catch up to re-implement matching interfaces in their udev-like daemons. Everything old is new again.  For you to suddenly take notice, especially when you are very much enjoying the fruits of previous rounds of modernization embedded in the Ubuntu system stack, is a little naive.

        This is an iterative process in modernizing the concept of the linux operating system. A process which really started way back with Project Utopia.What were you like 12 when that was formed?  Systemd is just the next step a long the road. A road easy to see in hindsight.

        Now excuse me I have to go brush in some Just for Men into my neck beard, and cover up the patches of grey created by having to relate ancient history to an angst filled youth culture.  Kids these days.

        -jef

    • http://twitter.com/humphreybc Benjamin Humphrey

      This is interesting, because only two weeks ago I was reading some articles by Canonical and Red Hat developers about init systems.

      Long story short, the Red Hat developer was insistent on systemd and listed a bunch of reasons why it’s better than anything else (targeting upstart.)

      He also questioned why certain distros (hyperlinked to ubuntu.com) weren’t changing to systemd straight away.

      In reply, Robbie Williamson from Canonical explained why they’re sticking with upstart due to an upcoming LTS early next year, among other reasons.

      We know that Red Hat has an influence in GNOME. My question would be, is this dependency something pressured by certain people just so they can get their way?

  • http://twitter.com/foxoman foxoman

    A complete OS will need a package management , so next time they will force us to use only RPM  :)

    Never mind I ‘am waiting for canonical/Ubuntu official response .

  • http://twitter.com/foxoman foxoman

    A complete OS will need a package management , so next time they will force us to use only RPM  :)

    Never mind I ‘am waiting for canonical/Ubuntu official response .

  • http://twitter.com/conorsulli Conor O Sullivan

     Well it would not really matter if it was just a shell thing and not system wide.. in that case unity could just be used.. Im fairly sure it has more users than gnome-shell anyway.. the unborn child that people forgot about.

    >> ready to get flamed on so many levels :-D

  • daas88

    Those gnome developer really don’t want people to use their DE. First they make it a counter-productive mess, then they make it exclusive for linux and systemd.
    Next thing is make it exclusive for gnome club members. Not that I care, I’ll probably stick with unity or switch to kde in the next releases

    • https://launchpad.net/~davidnielsen David Nielsen

      I wrote some thoughts on why this isn’t about creating an exclusive club but about defining needed interfaces to provide the user experiences GNOME3 strives to deliver.

      http://davidnielsen.wordpress.com/2011/05/18/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means/

      • http://twitter.com/kmetamorphosis Bertel King, Jr.

        Well worth the read. Oh, and “GNOME as an OS is (partly) about interfaces, not defining a Linux only desktop that runs only on Thursdays if the window is open” is a great subhead.

    • Anonymous

       Unity runs on gnome anyway.

      • daas88

        Yes but if I understood correctly, ubuntu devs will have to do extra work to make unity work on top of gnome if gnome devs make systemd a dependency.

        • John Penland

           hhmm, one could hope they get lazy(in this particular case then? )

        • Anonymous

           I don’t think so, they will just have to provide the systemd stuff, which has nothing to do with Unity afaik and they were apparently planning on doing that anyway I think.

  • Code Dash

    Gnome: We want to build a Gnome OS so we can create a stable, standard api os to make life easier for users and developers.

    Ubufans: Nooooo we want mess and stuff from different upstream sources that are about to break at any time.

    Me: LOL keep drinking the ubucola. Gnome FTW

    • http://profiles.google.com/daengbo Daniel Bo

       I agree with your basic sentiment, though I’ll probably stay with Ubuntu.

      I support GNOME going full-stack. Make everything work, from top to bottom! Keep it LGPL, and people can fork and port to whatever kernel they like. Gnome apps have for too long been worried about getting crap running on Windows and OS X.

      Have a primary target, damnit!

      • http://profiles.google.com/bwat47 Brandon Watkins

        I agree, having a more focused development can really make gnome great :). The biggest gnome distro, ubuntu is already splitting off and doing its own thing, and thats fine, and I think this is an opportunity for gnome.

  • http://wiki.go-docky.com/index.php?title=Main_Page PsyberS

     ”Redditer user deebeeoh”? Please.  Try DBO, aka Jason Smith, aka Unity developer.  Please take anything a Unity dev says (that has a negative connotation) about GNOME/GNOME Shell with a grain of salt.

    Don’t get me wrong, I still love you Jason. ;-)

    • http://profiles.google.com/downwardspindle David Walker

      He’s a Unity developer? Really? It’s a shame that he deleted his original Reddit post. I was having a great time watching him try to justify all the claims he was making, only to have just about everyone disagree with him (while citing actual quotes from the actual mailing list proving he was either failing at basic English comprehension or making stuff up.) But this guy is also a Unity developer? Man, I assumed he was just some random guy who was relatively new to linux and the English language.

    • http://profiles.google.com/downwardspindle David Walker

      He’s a Unity developer? Really? It’s a shame that he deleted his original Reddit post. I was having a great time watching him try to justify all the claims he was making, only to have just about everyone disagree with him (while citing actual quotes from the actual mailing list proving he was either failing at basic English comprehension or making stuff up.) But this guy is also a Unity developer? Man, I assumed he was just some random guy who was relatively new to linux and the English language.

  • christoph411

    Yay! More fragmentation…

  • https://launchpad.net/~isantop Ian

     Two things:

    1. Making a DE depend on a specific Init system is just plain poor design. 

    2. If they really *have* to make it depend on an Init system, why not choose Upstart? Fedora, RHEL, and Ubuntu currently use Upstart, openSUSE and Debian have plans to, and that takes care of the overwhelming majority of linux users out there.

    Of course, Fedora is planning on switching from Upstart for FC15, so really what the GNOME project is saying is that they want to make Fedora a requirement to use GNOME.

    • https://launchpad.net/~davidnielsen David Nielsen

      They did switch to systemd in F15 already and the F15 stable release is as I recall days away. OpenSUSE has committed to systemd, Mandriva has as well if I remember correctly. MeeGo is on board as well. There is also work in Debian to support systemd but thoughts of making it default I don’t know, support will suffice to provide GNOME3.

      It is really just a matter of switching before or after the next LTS. Being left alone maintaining an init system for 6 years (time till LTS release plus 5 years support as it is on the Server product). The lead maintainer left for Google which is famed for swallowing contributors leaving their participation to mostly memories.

      A saner and saner case I think is building to support systemd in Oneiric. It is going to be a lot of work but we have a year till LTS which is when it really counts. Upstream is active and documents everything very well so we should be able to rely on that and focus on migration documentation. Every distribution is basically working on systemd transitions right now. Meaning we probably can share a big burden and unload some of the thousands of un-upstreamed Upstart integration patches.

      Outside of a range of documented technical and end user benefits from systemd is presents a big chance to standardize some really basic things like how to store the hostname. Things that might sound simple but present problems making things work uniformly on all platforms.

      I think it is a worthy effort to undertake even if it is going to be painful. I think LTS will be better for it.

      More thoughts:
      http://davidnielsen.wordpress.com/2011/05/18/systemd-in-gnome-packagekit-and-what-gnome-as-an-os-really-means/

      • Anonymous

         I am totally agreeing to this. Systemd is the more sane choice and i think every dist but debian have realized this. Upstart is not as competent as Systemd. Plane and simple. I have seen works on this for Ubuntu so there might be something coming.

        I support Gnome choosing one init system but not to go full blown OS. Although if they do i would probably use that OS.

      • https://launchpad.net/~mpt mpt

         You made a typo. “The lead maintainer left for Google which is famed for swallowing contributors leaving their participation to mostly memories” should read “The lead maintainer left for Google to help them integrate Upstart into Chrome OS”.

        • http://twitter.com/jspaleta Jef Spaleta

          His recent posts to the systemd devel list seem to imply some interest in systemd as possibly suitable for the ChromeOS target needs. I don’t think he’d be wasting time making the sort of inquiries in the last couple of weeks if upstart was the clear winner on technical merits.

          I guess time will tell. It would be ironic indeed if ChromeOS bit the bullet before Ubuntu in systemd adoption given the staffing moves.

          -jef

  • http://linuxblog.darkduck.com/ darkduck

     Don’t like this idea in general. If GNOME and KDE move into their own OSes, what will other distros do?

    • http://twitter.com/jspaleta Jef Spaleta

       Distros will adapt.

      Speaking as someone who primarily works at the distribution level I can definitely say I see a transition coming in the role of the distribution. It’s coming, its been coming for a long while now..ever since Nokia experimented with Maemo on post-desktop formfactor N800 devices. 

      Markets evolve, distribution mechanisms evolve.  Holding the current ecosystem of hundreds of linux distributions as sacrosanct is like me expecting the local Little Professor bookstore from my childhood to stay in business in the face of big box stores and ebooks. Things change, the role of the distro is changing and there is nothing you or I can do to stop it from happening.

      The distros who adapt quickly and can ride the tiger of change will be the champions of the future. The ones who don’t will get frustrated and will self-destruct.  New distributors will rise up…there’s always a new linux distribution just over the horizon.   And some will just keep on keeping on. 

      -jef

  • http://profiles.google.com/l33ts0n Arron Washington

     Hell yes.

    If Gnome chooses to do their own thing, cool; but if they’re gonna make that decision, they need to go for the gold. Make some heavy Linux-only integrations to improve speed and performance, while removing the millions of backends they might have to support.

    Use all the Linux-specific kernel features that provide speed / stability boosts.

    Use all the Linux-specific technologies that are available without having to worry about portability.

    The world still has KDE, XFCE, LXDE, enlightenment, and Unity. There’s no reason Gnome shouldn’t just go balls out and use all the sweet offerings the kernel and the technologies built on it has to offer.

  • Anonymous

    I seem to be in a minority, but I think it’s an interesting decision and I’m looking forward to it. I’m really curious what the Gnome devs can do when they design the whole OS and not just the desktop environment. I think they did a great job with Gnome 3, so hopefully the Gnome OS (if it becomes reality and not just an idea) can be really good. And if it fails, there are still other distros and desktop environments, they aren’t going to disappear :)

  • aperson

     That’s right GNOME, just keep shooting yourself in the foot.

  • http://twitter.com/SarcasticSloth Steven Garza

    “Gnome drops support for BSD and Solaris.”
    And nothing of value was lost. At least for the Linux desktop.

  • http://pulse.yahoo.com/_LLFBBO3CDQUMGQXOQNGXDDDBPM Ladymecha

    Told you Gnome is full of idiots.

  • Anonymous

     Its painfully obvious that most of the ‘sky is falling’ commenters here have not read the thread. Its also disappointing that the author of the article took the CNN approach to journalism and lazily quoted a random reddittor instead of being a tech journalist and actually trying to understand what is going on.

    What falls out of the thread on d-d-l is going to be the requirement of a handful or so DBus interfaces whose initial implementation will be provided by systemd.

    Don’t confuse interface with implementation.

    The first implementation will be linux only (likely) because it is in systemd. There is talk on how to build these services without using systemd and that will likely be the approach of some linux distros (and perhaps Ubuntu).

  • Jouni Rajala

     I dont like kde 4+, i dont like Gnome either. I use KDE 3.5.4 version and i am very happy :)

  • http://pulse.yahoo.com/_XTPJUQH2CMGBW3BZSQVPWFVQNY Frost

    This whole story is a non-event. It was caused by a comment from one person, which was quickly shot down by others in the same thread.

    Why does everyone has to become so dramatic about it?

  • http://pulse.yahoo.com/_XTPJUQH2CMGBW3BZSQVPWFVQNY Frost

    This whole story is a non-event. It was caused by a comment from one person, which was quickly shot down by others in the same thread.

    Why does everyone has to become so dramatic about it?

  • Shane Quigley

    BSD should be able to run it any way running the Linuxulator right? That’s how i get flash on my BSD box.

  • Anonymous

    When ubuntu switches over to gnome 3 will we still get the option to use ubuntu classic? 

  • Pingback: GNOME Project suffering the NIH disease | rojtberg.net