Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
×

GNU Hurd Begins Supporting Sound, Still Working On 64-bit & USB Support (phoronix.com) 312

Posted by timothy from the pretty-soon-big-and-fancy-like-gnu dept.
An anonymous reader writes: GNU developer Samuel Thibault presented at this weekend's FOSDEM conference about the current state of GNU Hurd. He shared that over the past year they've started working on experimental sound support as their big new feature. They also have x86 64-bit support to the point that the kernel can boot, but not much beyond that stage yet. USB and other functionality remains a work-in-progress. Those curious about this GNU kernel project can find more details via the presentation media.
This discussion has been archived. No new comments can be posted.

GNU Hurd Begins Supporting Sound, Still Working On 64-bit & USB Support More

GNU Hurd Begins Supporting Sound, Still Working On 64-bit & USB Support

Comments Filter:

  • In future news (Score:4, Funny)

    by c++ ( 25427 ) on Sunday January 31, 2016 @11:27AM (#51408957)

    We're proud to announce that GNU Hurd can now save and load files.

  • What's the point (Score:5, Interesting)

    by mwvdlee ( 775178 ) on Sunday January 31, 2016 @11:27AM (#51408961) Homepage

    What's the point of continuing with Hurd?
    I mean, apart from making make laugh whenever they have "news".

    • Hurd probably does make make laugh!

    • Re:What's the point (Score:4, Interesting)

      by thegarbz ( 1787294 ) on Sunday January 31, 2016 @11:33AM (#51408989)

      This is an interesting question. It's one thing if the HURD was making progress but based on this kind of news it would seem that technology is actually being developed faster than the kernel.

      • Seriously, if you google "GNU Guix", you'll see that system startup scripts will be written in Lisp, the package manager will use Lisp to describe packages. Also, I note that the microkernel architecture will allow code that is traditionally part of the kernel to run in user mode and be written in Lisp.

        It looks to me that they are building a new system that combines the best aspects of Unix and the legendary Lisp Machine. Which would be kind of cool.

        • Can't they rewrite systemd in lisp, put it under emacs and then have their entire OS? Speaking of which, why are they bothering w/ 64-bit support at all? Since the only app that would run in HURD would be emacs, they might as well just make emacs the front end of the OS, instead of bothering about bash/csh/ksh/ et al

    • Re: (Score:3, Insightful)

      by Anonymous Coward
      The microkernel architecture makes it quite cool.

      • The microkernel architecture makes it quite cool.

        For that, why not go to Minix, which is FOSS under a BSD license, and is well documented, since AST's book is the documentation of the OS. It's a good learning platform, and has some neat features, like a 'reincarnation server' that lowers the priority of hung drivers until they are effectively killed, and restart them again.

        Last I hurd, HURD is still based on Mach 3, which was a first generation microkernel, but a lot of developments have happened in microkernel concepts that have not made it to Ma

    • Its like trying to talk someone out of a cult they are following.

    • Re:What's the point (Score:5, Informative)

      by short ( 66530 ) on Sunday January 31, 2016 @11:38AM (#51409013) Homepage
      Thanks to the microkernel architecture you will no longer have to reboot system just to get rid of that stale lock on an accidentally removed USB disk or unmountable --bind mount in /proc/mounts due to non-existing user/usecount or due to some crashed driver locking up your PCI device etc. I could transparently restart crashed ntfs.sys emulated under Linux in 2003 while Linux kernel still can't do that with its native filesystems.

      • Re: (Score:2)

        by rssrss ( 686344 )

        "you will no longer have to reboot system just to get rid of that stale lock on an accidentally removed USB disk or unmountable --bind mount in /proc/mounts due to non-existing user/usecount or due to some crashed driver locking up your PCI device etc."

        Before or after my 115th birthday?

      • Ok. So by which century will Hurd be usable enough that I can take advantage of the features? And don't say "now" because not having sound support or full workinf x86_64 support does not make a usable kernel.

        • Re: (Score:2, Interesting)

          by short ( 66530 )
          The sooner you will write that the sooner you will get it. That's all what Free software guarantees you and I find it superior to anything else.

          • Re: (Score:2, Insightful)

            by jones_supa ( 887896 )
            FOSS is not a garden party or cake sale where anyone can volunteer just like that. Features that seem relatively simple to the end user can hide tens or hundreds of thousands of lines of code behind them.

          • Re:What's the point (Score:4, Insightful)

            by Chris Mattern ( 191822 ) on Sunday January 31, 2016 @01:40PM (#51409601)

            Odd. Linux has sound, full 64-bit support, is free (as in both speech and beer), and I didn't have to write any of it.

            I'm having trouble seeing HURD as being superior to that.

            • Re: (Score:2, Insightful)

              by short ( 66530 )
              HURD can become a non-crashing OS one day, Linux with its current architecture cannot. Everything has its pros and cons, pick your poison.

              • That's great and all, but if one can't actually use it it's rather useless to try to impress us with theoretical capabilities that will take countless more decades to actually happen.

              • You think we'll still be using linux a few centuries from now, when HURD might be usable? Even Linus foresees the day something else replaces linux.

          • Ah, yes the old, tried and true "Fix it yourself, bum" response.

          • Re: (Score:3)

            by Nemyst ( 1383049 )

            The sooner you will write that the sooner you will get it. That's all what Free software guarantees you and I find it superior to anything else.

            So because it fulfills a need that's not actually there, never.

      • We know what microkernels are good for. What we don't know is what the Hurd is good for. Development is moving so slow that it is forever trying to keep up with the calendar.

      • Re:What's the point (Score:5, Insightful)

        by ooloorie ( 4394035 ) on Sunday January 31, 2016 @03:20PM (#51410063)

        Thanks to the microkernel architecture you will no longer have to reboot system just to get rid of that stale lock on an accidentally removed USB disk or unmountable

        There is an implicit false dichotomy there, namely "ancient design monolithic kernel" and "ancient-style microkernel". There are many other choices.

        I could transparently restart crashed ntfs.sys emulated under Linux in 2003 while Linux kernel still can't do that with its native filesystems.

        Actually, since NTFS under Linux runs in user space, yes you can. In fact, for many kernel services (USB, file systems, networking, etc.), the kernel can call upon separate servers to handle those services. And that's another problem with microkernels: their design focuses not on what users need and the question of how to best provide that, but rather on a mechanism.

        • There is an implicit false dichotomy there, namely "ancient design monolithic kernel" and "ancient-style microkernel".

          But in the choice between Linux or BSD on the one hand, and Hurd on the other, that is the choice we're being asked to accept. Well, unless the L4 or Viengoos variants of Hurd come good.

      • Re: (Score:3)

        by batkiwi ( 137781 )

        Adding to your comments, you won't have this issue in HURD due to the lack of USB removable device support, so that stale lock wouldn't exist to begin with.

    • What's the point of continuing with Hurd?

      A long time ago you could've asked the same question about Linux. Just because it is not useful right now (or might never be...) doesn't mean it is not worth working on.

      I'd much love to have a production-ready, open source microkernel OS to toy with.

    • "news"

      Properly called "gnus".

    • Re:What's the point (Score:5, Insightful)

      by dissy ( 172727 ) on Sunday January 31, 2016 @12:08PM (#51409151)

      What's the point of continuing with Hurd?

      For the same reason anyone does something they enjoy for fun and recreation, namely so we don't become hollow and joyless, reserved to asking on forums why other people do things they enjoy :P

      I note you both read slashdot and posted to slashdot today, as well as aren't out working to do something "useful".

      Don't you think it a tad off to spend your free time doing things you enjoy at the same time as questioning other people doing the same?

      • Re: (Score:2)

        by mwvdlee ( 775178 )

        Today I worked on an addon for a popular open source javascript-based code editor, added some minor features to one of my open source projects and added a bunch of much needed unittests to another of my open source projects.

        I also took a few minutes to read some Slashdot posts and make a few comments.

        Amazingly, both can be done in a single day!

        • Re:What's the point (Score:4, Interesting)

          by dissy ( 172727 ) on Sunday January 31, 2016 @02:13PM (#51409759)

          Today I worked on an addon for a popular open source javascript-based code editor, added some minor features to one of my open source projects and added a bunch of much needed unittests to another of my open source projects.

          I also took a few minutes to read some Slashdot posts and make a few comments.

          Amazingly, both can be done in a single day!

          Indeed! Just as the Hurd team can play on Hurd and contribute to other more useful projects :}

          And I apologize for the accusation as well, it's just that the vast majority of people who question others free time activities have a high likelihood of both demanding productivity from others while not living to the same standard themselves.

          I suppose it was mostly the fact I quite literally formed the thought "I wonder which of the top three posts will ask 'what is the point?'" as I clicked the article to open the comments, and there this was right at the top in spot 1 with that exact phrase and already modded up to max.

          But I am pleasantly surprised for you shattering that expectation.

      • This is the best answer to the question. In fact, it is probably the only good answer to this particular question.

    • Why not? People are making it for free. There's actually a wiki devoted to OSes [osdev.org], and there are quite a lot of them [osdev.org]. Working on a kernel is fun.

      These are the advantages [gnu.org] and challenges [gnu.org] of the GNU kernel. If you want to understand why people like the Hurd kernel, I would suggest reading that.

    • Re: (Score:2)

      by drolli ( 522659 )

      I don't work on it, but if i would work in researching micro kernel OSes, I could imagine that a OS which is running (forget about USB or sound, for most purposes in OS research it is fine to run in an emulator) is a good base to start.

    • What's the point of continuing with Hurd?

      Well the FSF doesn't really have an operating system kernel representative of the Free Software ideology. The closest thing is the Linux kernel which Linus has clearly said [lkml.org] simply uses the GPLv2 for tit-for-tat contributions because it is a good license, not because Linux is a free software project. Of course it also doesn't do copyright assignment to the FSF and also is not GPLv3 and is unlikely to migrate to further revisions of the GPL in future.

      As the FSF evolves the Linux kernel's position on free soft

      • While this is true, RMS had abandoned HURD a long time ago, and instead took Linux, stripped it of any 'non-free' components, branded it 'libre-Linux' and promotes THAT. There are a few distros that the FSF does endorse - like Trisquel, gNewSense, and a few others that seem primarily localized for Latin America.

        If I'm not mistaken, they wasted a lot of time in determining an ideal microkernel. What I don't get is why they didn't just fork Minix, which was there, put their fork under GPL3, and then make

  • Open software (Score:4, Insightful)

    by 110010001000 ( 697113 ) on Sunday January 31, 2016 @11:44AM (#51409037) Homepage Journal
    These days Open/Free software is more important than ever. With closed kernels and binary blobs you have no idea what kind of code is running on your system. It would be nice to have a true Open kernel running on true Open hardware.

    • Re: (Score:3)

      by short ( 66530 )
      You can run Linux-libre [wikipedia.org] for the OS part, that is much easier target than GNU Hurd.

    • Re: (Score:2)

      by wjcofkc ( 964165 )
      Open kernel. Absolutely. Open hardware...? Dunno. I honestly think for that to happen we will need to print our own hardware. A team of Open Source engineers specializing in CPU design, designing and printing CPUs. Mass storage so on an so forth. Then we buy and sell to each other, perhaps even at cost. Meanwhile we as many people as are willing will through money into funds that can then be disbursed. I really do believe open hardware would need such a scheme. Of course, we will also need to print our prin

  • Many kinds of freedom (Score:3, Insightful)

    by jones_supa ( 887896 ) on Sunday January 31, 2016 @11:51AM (#51409077)
    In a way Hurd is not free software because it does not allow me to freely unleash all potential of my PC. I am restricted with digital handcuffs.

  • Hopefully one of the days we'll support that Internet thing I hear is all the rage with the kids these days.

  • He shared that over the past year they've started working on experimental sound support as their big new feature.

    Truly, we live in the future.

  • by the time they have support for my hardware it will be obsolete, worn out and thrown in to the recycle bin

  • By the time Hurd hits the streets, we're going to need a 128-bit version. I wonder if there'll be a Duke Nukem port?

  • Sound? (Score:2)

    by PPH ( 736903 )

    There's a guy that can help you with sound support. While he's at it, he could replace your Hurd kernel with his new init system.

  • Just an observation... (Score:3)

    by roc97007 ( 608802 ) on Sunday January 31, 2016 @08:04PM (#51411283) Journal

    Scanning through the discussion, it looks like conversations relating to HURD get ugly fast. Obviously there are strong feelings at work here.

    I view the whole thing somewhat with nostalgia, as I was babysitting Vaxen running BSD when I first heard of HURD. Regardless of its merits or lack of same, it seems to be on track for the world's record for slowest development of any currently developed OS. Kinda the Duke Nukem Forever of operating systems.

Slashdot Top Deals

An age is called Dark not because the light fails to shine, but because people refuse to see it. -- James Michener, "Space"

Close