Announcement

Collapse
No announcement yet.

Linux 4.9 Kernel Tacks On Over 200k Lines Of Code

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

  • #21
    Originally posted by Master5000 View Post
    This is the reason Windows 10 is dead. Because of morons like you and that includes you too. Lack of ABI stability is the reason Windows 10 won't take over windows 7.
    fixed.

    Comment


    • #22
      Originally posted by hansg View Post
      Well, that's just your opinion, isn't it? I want that. The people I work with want that. My customers want that (or at least don't care about it but would sure appreciate the benefits).
      All these people's opinion is irrelevant, even if you were right.
      Developers decide, you stfu. Welcome to opensource world.

      Having a stable ABI would allow GPU makers to release a single driver valid for all Linux kernels. Shock horror, Linux might actually get day-1 stable drivers for all GPUs! And at work I'm 'blessed' with hardware that you have never even heard about (obscure interface cards) that are supported on Linux - if you are running a very specific, ancient version of Linux, and only that version and no other. Since it lacks a modern C++ compiler, sadly I cannot run my software on it, so we just continue to run Windows instead.

      Really?
      - Quite a few companies would be happy to maintain a Linux driver if there were a stable ABI, but they aren't interested in chasing kernels.
      - As a corrolary, much of the effort spent on writing Linux drivers could be spent on more productive goals.
      - A leaner, simpler kernel would be easier and more fun to develop for.
      - An unstable driver cannot bring down the kernel if it is running in its own process.
      - Recompiling the kernel would be massively faster.
      - New hardware could be supported on older Linux releases.
      - New hardware could be supported on day 1, instead of when Linus decides to release a new kernel.
      - If a driver breaks in a new kernel release, it would be easy to roll back just that driver.
      - Instead of an all-or-nothing approach, a combination of drivers and kernel could be picked that works best for a certain task or environment.

      Which of these things is bad for developers?

      If there were a stable ABI there would be no need to 'see what is happening'. And 'adding features' should never be a reason to break that stable ABI anyway. Maybe extend it, maybe introduce an entirely new version (side by side with the old one), but not break it. That's software engineering 101.

      You may argue (as some did earlier in the thread) that Windows occasionally got the stable ABI thing wrong. That just means it is hard; it does not mean it is not worth having, or that having a completely different ABI in each release is therefore more desirable.
      You are wrong and you don't understand a damn of software development. Read this that was written by actual developers https://git.kernel.org/cgit/linux/ke...i_nonsense.txt

      Duh, you think? And here I was thinking you could just compile it with the -rewrite-to-microkernel switch... (/sarcasm)
      Then you know that what you ask is even more irrealistic than switching to stable ABI.

      Strawman...
      It was a question.

      No, most people don't know that. That's because it is not true.
      I said "most people HERE (phoronix = linux/unix forum) know that drivers are developed with kernel" and yes, it is true.

      Just because Linux does it like that, it does not somehow become a law of nature that it must be so, and in fact in a healthy ecosystem drivers would be maintained (and run) outside of the kernel.
      Considering that not even Windows has a healthy driver ecosystem, this is all wishful thinking.

      Comment


      • #23
        unapproved post for hansg above

        Comment


        • #24
          Originally posted by starshipeleven View Post
          All these people's opinion is irrelevant, even if you were right.
          Developers decide, you stfu. Welcome to opensource world.

          You are wrong and you don't understand a damn of software development.
          Based on your pleasant discussion style I'm going to guess you either haven't seen a real human being in at least a decade, or that you were born after I wrote my first line of code. _Long_ after...

          For the rest, try answering some points next time. I'll leave you alone now. Feel free to keep ranting and foaming at the mouth some more, but don't bite into the furniture please.

          Comment


          • #25
            Originally posted by hansg View Post
            Based on your pleasant discussion style I'm going to guess you either haven't seen a real human being in at least a decade, or that you were born after I wrote my first line of code. _Long_ after...
            Wrong and wrong, I work as sysadmin and techsupport on both linux and Windows stuff.
            Due to my job experience I tend to react very harshly to people that clearly don't know shit and think that Windows has stable ABI (wrong), and that a stable ABI has anything to do with good drivers (wrong again), and that we need a stable ABI in Linux too so we can have on Linux the same retarded bullshit issues that are a total PITA to fix on Windows (if fixable at all due to closed-source) during my day job.

            For the rest, try answering some points next time.
            Nice bullshit. I actually did answer all your secondary points, and the explanation in the link answers the main point you made.

            Or you want to claim that a lead kernel dev like Greg Kroah-Hartman is a liar somehow and want to show the class you are good too by pointing out where he is wrong, with sources?

            Feel free to keep ranting and foaming at the mouth some more, but don't bite into the furniture please.
            Your tears, they are delicious.

            Comment


            • #26
              Originally posted by starshipeleven View Post
              Wrong and wrong, I work as sysadmin and techsupport on both linux and Windows stuff.
              Yeah, but "hansg" is a developer, not techsupport. He has programmed a lot. And somehow, I think a developer's experiences about software development are more credible than a tech support guy?

              Originally posted by starshipeleven View Post
              Due to my job experience I tend to react very harshly to people
              Is that so? Nah, I dont believe you.

              Originally posted by starshipeleven View Post
              that clearly don't know shit and think that Windows has stable ABI (wrong),
              Wrong. You are misinformed. Windows DO have a stable ABI. And Windows does not have these Linux upgrade problems.

              Windows XP has stable ABI, as do Vista, and Win7, etc etc etc. ALL Windows OSes has stable ABIs. Sure, when you upgrade from a Windows OS to another Windows, device drivers might not work anymore because ABI might change between Windows versions. But ABI does not change _within_ Windows versions. If you upgrade from Windows XP to Windows XP Service Pack 3 - all your device drivers work. If you upgrade from WinXP to Vista, Microsoft does not promise your drivers will work anymore. But often they do.

              However, when you upgrade a minor Linux kernel version to the next (not say, from v2.4 to v3.0) all hell might break loose. Linux changes ABI within the same kernel revision. No other OS do that. All other OSes has stable ABIs within the same kernel version. Linux does not.

              You did not know this, and you are a tech support guy? Maybe you think people are a PITA, because you lack some computer skills and knowledge? If all your serviced computers break, and people complain - maybe you should work on your IT chops?

              Comment


              • #27
                Originally posted by UbuntuRulez View Post
                Yeah, but "hansg" is a developer, not techsupport. He has programmed a lot. And somehow, I think a developer's experiences about software development are more credible than a tech support guy?
                1. neither of the people here has demonstrated their claims about their job. He claimed to be a developer (I doubt it), I claimed to be sysadmin/techsupport. If you work on baseless claims you are a moron.

                2. What matters is that the linked faq, it is in the kernel's public git repo, it was written by a known and high-profile kernel developer/mantainer, and that what this developer says proves wrong any of your silly claims. https://git.kernel.org/cgit/linux/ke...i_nonsense.txt

                Ubuntu 12.04 has stable ABI, as do Ubuntu 14.04, and Ubuntu 16.04, etc etc etc. ALL Ubuntu LTS OSes has stable ABIs. Sure, when you upgrade from a Ubuntu LTS to another Ubuntu LTS, device drivers might not work anymore because ABI might change between Ubuntu LTS versions. But ABI does not change _within_ Ubuntu LTS versions. If you upgrade from Ubuntu LTS 12.04 (kernel at release) to Ubuntu LTS 12.04 (last updated kernel for 12.04 shipped as update from Canonical) - all your device drivers work.
                fixed.
                Do you understand what I did? I did this to show that also Linux has the same level of ABI stability within releases as Windows.

                Also, you know right that RHEL (and its free version CentOS) have like 10 years of ABI compatibility as they keep using the same kernel (with backported fixes/stability/security patches)? (more if you pay RedHat)

                If you upgrade from WinXP to Vista, Microsoft does not promise your drivers will work anymore. But often they do.
                XP -> Vista... not that much, but Vista -> Win7 yep. Other windows, it's a crapshoot. Depending on when they change subsystems the drivers will likely work or not (with or without some .inf file hacking).

                However, when you upgrade a minor Linux kernel version to the next (not say, from v2.4 to v3.0) all hell might break loose. Linux changes ABI within the same kernel revision.
                Tell me, you come up with this blatant, outrageous, baseless bullshit on your own or it's the voices in your head that tell you to write this?

                This is plain wrong. Linux kernel has a stable ABI within the same revision.
                All kernels in the same major version are only receving safety and stability fixes. (for a longer or shorter time depending on their LTS status)
                Any new feature (might require an ABI change or not) CANNOT be added to the kernel after the so-called "feature freeze", that happens when the kernel is released.

                You did not know this, and you are a tech support guy? Maybe you think people are a PITA, because you lack some computer skills and knowledge? If all your serviced computers break, and people complain - maybe you should work on your IT chops?
                Nice trolling, 8/10.

                Comment


                • #28
                  Originally posted by starshipeleven View Post
                  1. neither of the people here has demonstrated their claims about their job. He claimed to be a developer (I doubt it), I claimed to be sysadmin/techsupport. If you work on baseless claims you are a moron.
                  Are you saying we should not pay attention to what you say? That you might as well as lie? Nice.


                  2. What matters is that the linked faq, it is in the kernel's public git repo, it was written by a known and high-profile kernel developer/mantainer, and that what this developer says proves wrong any of your silly claims. https://git.kernel.org/cgit/linux/ke...i_nonsense.txt
                  So, the FAQ proves that Linux does not have device driver problems when upgrading the Kernel? Just because someone wrote it in a FAQ, it proves that all the threads in Linux forums about upgrade problems are invalid? Eh?

                  Do you understand what I did? I did this to show that also Linux has the same level of ABI stability within releases as Windows.

                  Also, you know right that RHEL (and its free version CentOS) have like 10 years of ABI compatibility as they keep using the same kernel (with backported fixes/stability/security patches)? (more if you pay RedHat)
                  This only applies to LTS distros. All the 99% of us that dont use LTS? Is it ok that upgrading the kernel breaks device drivers for us?


                  This is plain wrong. Linux kernel has a stable ABI within the same revision.
                  All kernels in the same major version are only receving safety and stability fixes. (for a longer or shorter time depending on their LTS status)
                  Any new feature (might require an ABI change or not) CANNOT be added to the kernel after the so-called "feature freeze", that happens when the kernel is released.

                  Nice trolling, 8/10.
                  But no one is interested in ancient LTS distros. You can not install the latest software, because it uses new libraries. So you need to backport the new library functionality to the old Linux kernel version, and who does that? No one. So you are stuck with your old ancient software. Users dont want that.

                  So if you upgrade something, say a library becuase you want a new version of the software, then that library will trigger upgrade of something else, which will trigger upgrade of something else, etc. In an avalanche you have upgraded everything, and you have left LTS. So LTS is a broken concept and only valid if you are not willing to install new software nor upgrade. LTS does not work. Linux ABIs might change at any whim, and Linux ABI model is broken. No other OS has that model. This must mean that Linus Torvalds is smarter than the rest of the world, right?

                  Comment


                  • #29
                    Originally posted by UbuntuRulez View Post
                    Are you saying we should not pay attention to what you say? That you might as well as lie? Nice.
                    I'm saying that anything without a proof is a lie, and yes I might lie as well for fun and profit.

                    But I'm providing proof here.

                    So, the FAQ proves that Linux does not have device driver problems when upgrading the Kernel?
                    No, it proves that ABI stability isn't solving the problem.

                    This only applies to LTS distros.
                    No, the Linux kernel does not break ABI whithin the same major version, LTS or not, so what you claimed before was bullshit.

                    All the 99% of us that dont use LTS? Is it ok that upgrading the kernel breaks device drivers for us?
                    Upgrading kernel does not break drivers, you get newer drivers that might have new bugs and that's not the same thing as breaking drivers.
                    Old drivers stay there with the previous kernel and you can switch back at boot time easily.

                    Closed-source drivers are another matter, if all this whine is because kernel update breaks your NVIDIA driver it's because you are a noob and didn't follow the correct update procedures (or installed an unsupported kernel)

                    But no one is interested in ancient LTS distros.
                    I'm talking of kernel LTS status, not the userspace's. Many people/distros/companies are interested in a kernel with ABI stability for like 5 years or so.

                    you need to backport the new library functionality to the old Linux kernel version, and who does that? No one.
                    Yup. Because you don't add "new library functionality" to the kernel, as it is a kernel. Libraries are in userspace, not in the kernel. Try looking with your package manager how many packages depend from the kernel version.
                    Hint: you won't find many.

                    So if you upgrade something, say a library becuase you want a new version of the software, then that library will trigger upgrade of something else, which will trigger upgrade of something else, etc. In an avalanche you have upgraded everything, and you have left LTS. So LTS is a broken concept and only valid if you are not willing to install new software nor upgrade. LTS does not work.
                    Distro (Ubuntu) issue, kernel is irrelevant here.

                    and Linux ABI model is broken. No other OS has that model.
                    Apart from Windows and OSX, and probably minor OSs too, as said above.

                    This must mean that Linus Torvalds is smarter than the rest of the world, right?
                    No it means that you are still an idiot troll.

                    Comment


                    • #30
                      Originally posted by starshipeleven View Post
                      I'm saying that anything without a proof is a lie, and yes I might lie as well for fun and profit.

                      But I'm providing proof here.

                      No, it proves that ABI stability isn't solving the problem.

                      No, the Linux kernel does not break ABI whithin the same major version, LTS or not, so what you claimed before was bullshit.

                      Upgrading kernel does not break drivers, you get newer drivers that might have new bugs and that's not the same thing as breaking drivers.
                      Old drivers stay there with the previous kernel and you can switch back at boot time easily.

                      Closed-source drivers are another matter, if all this whine is because kernel update breaks your NVIDIA driver it's because you are a noob and didn't follow the correct update procedures (or installed an unsupported kernel)

                      I'm talking of kernel LTS status, not the userspace's. Many people/distros/companies are interested in a kernel with ABI stability for like 5 years or so.

                      Yup. Because you don't add "new library functionality" to the kernel, as it is a kernel. Libraries are in userspace, not in the kernel. Try looking with your package manager how many packages depend from the kernel version.
                      Hint: you won't find many.

                      Distro (Ubuntu) issue, kernel is irrelevant here.

                      Apart from Windows and OSX, and probably minor OSs too, as said above.

                      No it means that you are still an idiot troll.
                      Ok, so it seems that you might as well as lie as telling the truth. So you admit you are a troll then. Nice.

                      Regarding the device driver issues that everybody has, it does not matter how much you link to a FAQ or talk about LTS or whatever - fact is people are having lot of device driver issues when upgrading the kernel (even within minor revision). Maybe you have missed all those forum threads in all the Linux forums. Do you want me to link to some of them? You apparently dont visit any Linux forums because you deny these threads exist.

                      And fact is, people dont have these problems with other OSes.

                      There are facts, and there are opinions. And there are trolls, that lie for fun - just to trigger a flame war. You know, triggering flame wars is what Troll does. Just what you are doing here. Denying all the facts (threads with driver problems) and lying about all other OSes share the same driver model as Linux - which is not true. So you lie, and start flame wars. The very definition of a Troll. Is it possible to ban Trolls here? I mean, you say you are lying and you start flame wars - you are a troll.


                      Comment

                      Working...
                      X