Announcement

Collapse
No announcement yet.

Redox OS, MINIX, Hurd & Genode Had Their Time At FOSDEM Too

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Redox OS, MINIX, Hurd & Genode Had Their Time At FOSDEM Too

    Phoronix: Redox OS, MINIX, Hurd & Genode Had Their Time At FOSDEM Too

    While Linux is the most prominent operating system each year at the Free Open-Source Developers' European Meeting (FOSDEM), it's not a conference limited to just Linux. Once again there was a developer room dedicated to other operating systems like the Rust-written Redox OS...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Redox looks interesting, but I like the Hurd's design, and I wish they had gone this way instead (creating a small drop-in replacement for the Hurd, and compatible userspace modules, it would also have allowed some cooperation on some level).
    I also like (Stallmanite, I know) the GPL, and I wish Redox had chosen this license instead. MIT might backfire in one way or another, but that's a way to increase adoption, I guess.

    Comment


    • #3
      Bazel was pretty big this year at FOSDEM with google sending lots of people to man the bazel booth.

      There was also a presentation (https://fosdem.org/2017/schedule/event/bazel/) about it which drew quite a crowd. Let's see whether google will go for a similarly strong presence of their new build system at other events.

      Comment


      • #4
        Originally posted by M@yeulC View Post
        Redox looks interesting, but I like the Hurd's design, and I wish they had gone this way instead (creating a small drop-in replacement for the Hurd, and compatible userspace modules, it would also have allowed some cooperation on some level).
        I also like (Stallmanite, I know) the GPL, and I wish Redox had chosen this license instead. MIT might backfire in one way or another, but that's a way to increase adoption, I guess.
        We're all just taking wild guesses here, but if anything I suspect the MIT license will hurt Redox adoption unless one of the tech giants adopts the project.

        No for-profit company wants to give away engineering resources for free. But when you contribute to the Linux kernel, you accept the cost of giving away free engineering for the benefit of getting free engineering from all of the other contributors.

        The MIT license doesn't preserve that guarantee. So if Samsung adds a device driver to Redox, Qualcomm can take that driver and use it without releasing any of their own Redox drivers. So all of the companies evaluating Redox might use it, but they won't contribute back their own enhancements back.

        The only way that model works is when an ultra-wealthy sponsor backs the project. That's how WebKit and LLVM thrive despite the fact that their licenses have no GPL-style forced-to-release-changes-in-kind aspects.

        I wouldn't want to tie the success of my free software operating system to finding a rich corporate sponsor.

        Comment


        • #5
          Originally posted by Michael_S View Post

          We're all just taking wild guesses here, but if anything I suspect the MIT license will hurt Redox adoption unless one of the tech giants adopts the project.

          No for-profit company wants to give away engineering resources for free. But when you contribute to the Linux kernel, you accept the cost of giving away free engineering for the benefit of getting free engineering from all of the other contributors.

          The MIT license doesn't preserve that guarantee. So if Samsung adds a device driver to Redox, Qualcomm can take that driver and use it without releasing any of their own Redox drivers. So all of the companies evaluating Redox might use it, but they won't contribute back their own enhancements back.

          The only way that model works is when an ultra-wealthy sponsor backs the project. That's how WebKit and LLVM thrive despite the fact that their licenses have no GPL-style forced-to-release-changes-in-kind aspects.

          I wouldn't want to tie the success of my free software operating system to finding a rich corporate sponsor.
          Linux being GPLd didn't stop any of those companies from developing closed source drivers. Just look at Android. Therefore, your argument is invalid here.

          Comment


          • #6
            Originally posted by M@yeulC View Post
            Redox looks interesting, but I like the Hurd's design, and I wish they had gone this way instead (creating a small drop-in replacement for the Hurd, and compatible userspace modules, it would also have allowed some cooperation on some level).
            I also like (Stallmanite, I know) the GPL, and I wish Redox had chosen this license instead. MIT might backfire in one way or another, but that's a way to increase adoption, I guess.
            The kernel accounts for a very small portion of code here. There's basically no way to prevent anyone from having closed source additions to open source software. Even Linux has proprietary drivers even though it has been GPLd. The only reason Linux can succeed with a GPL license in a way that Hurd cannot is because Linux already received critical mass. MIT projects are far more likely to succeed these days versus GPL ones. Many companies will simply avoid anything with the GPL license entirely because of company policy. At least with a MIT license, you could get some theoretical corporate contributions over zero contributions from corporate.

            Comment


            • #7
              Originally posted by M@yeulC View Post
              Redox looks interesting, but I like the Hurd's design, and I wish they had gone this way instead (creating a small drop-in replacement for the Hurd, and compatible userspace modules, it would also have allowed some cooperation on some level).
              I also like (Stallmanite, I know) the GPL, and I wish Redox had chosen this license instead. MIT might backfire in one way or another, but that's a way to increase adoption, I guess.
              Me thinks that HURD and Redox are exciting in their own way.
              Both bring different new aproaches which i like a lot.
              I just hope both get more momentum.

              Comment


              • #8
                Originally posted by Michael_S View Post
                We're all just taking wild guesses here, but if anything I suspect the MIT license will hurt Redox adoption unless one of the tech giants adopts the project.

                No for-profit company wants to give away engineering resources for free. But when you contribute to the Linux kernel, you accept the cost of giving away free engineering for the benefit of getting free engineering from all of the other contributors.

                The MIT license doesn't preserve that guarantee. So if Samsung adds a device driver to Redox, Qualcomm can take that driver and use it without releasing any of their own Redox drivers. So all of the companies evaluating Redox might use it, but they won't contribute back their own enhancements back.

                The only way that model works is when an ultra-wealthy sponsor backs the project. That's how WebKit and LLVM thrive despite the fact that their licenses have no GPL-style forced-to-release-changes-in-kind aspects.

                I wouldn't want to tie the success of my free software operating system to finding a rich corporate sponsor.
                While everything you said is true.
                If i'm not mistaken, there's nothing preventing the owner(s) of Redox to re-license it under GPL, i think...

                Comment


                • #9
                  Originally posted by Michael_S View Post

                  No for-profit company wants to give away engineering resources for free. But when you contribute to the Linux kernel, you accept the cost of giving away free engineering for the benefit of getting free engineering from all of the other contributors.
                  In a nutshell, one might describe it as a choice between “freeloader-friendly” open-source licences and “competition-friendly” ones. Businesses may not like “competition-friendly” ones, but they do benefit the ecosystem as a whole.

                  Comment


                  • #10
                    Originally posted by mmstick View Post

                    Linux being GPLd didn't stop any of those companies from developing closed source drivers. Just look at Android. Therefore, your argument is invalid here.
                    If somebody drives into your car and you decide (not your passengers) that the damage is "so minimal" that you let them go without reporting it to the police/authorities, just polish it out so that you don't have to stand in a queue at the crappy departments and do insurance claims/pay premiums. This does not automatically mean that the laws that prevent people from damaging your property are useless. It's about getting the best out of a bad situation.

                    If your contract with your employer says you have to work at the office for 8 hours each day and you end up just being at the office for half that time, but your work is up to date and everyone is happy then that does not mean that you will get fired. We have this thing called mutual interest. Just because the law is there does not mean you have to use it at every opportunity, *looks at Oracle*.

                    It is my opinion that Google and Canonical are (indirectly) allowing bad things happen to Linux, but it is still up to the owners (not the community) to decide what they want to do and how they want to use their licensing. The owners might feel that Android/Ubuntu attracts enough people away from competitors and that they are willing to accept the new users that do not care about open source. That is totally for them to decide. I can't predict the future so I don't know who is right or wrong here.

                    Simply saying that the GPL is the problem does not sound like a valid argument either.

                    PS: If you truly care about open source use/support/advocate Linux-libre.

                    Comment

                    Working...
                    X