Announcement

Collapse
No announcement yet.

Some Linux/Open-Source Letdowns Last For Years

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

  • brad0
    replied
    Originally posted by pal666 View Post
    and complain that they need to do more work to get different behavior
    That doesn't even make sense, but so typical of you.

    Leave a comment:


  • brad0
    replied
    Originally posted by pal666 View Post
    it was useable enough to be default fs for enterprise distro for many years
    That's not saying much. Still would not use it.

    Leave a comment:


  • pal666
    replied
    Originally posted by aht0 View Post
    Funny statement, considering the fact that most Linux distros are increasingly slaves to the whims or RH and IBM.
    you live in funny imaginary world where "most linux distros" repackage redhat distro and complain that they need to do more work to get different behavior

    Leave a comment:


  • pal666
    replied
    Originally posted by brad0 View Post
    It might actually be useable 10 years from now.
    it was useable enough to be default fs for enterprise distro for many years

    Leave a comment:


  • pal666
    replied
    Originally posted by geearf View Post
    btrfsf has been mainlined over a decade ago, how is that next-gen? Is it not current-gen at best?
    ext4 is current-gen(and you can mainline old-gen code later btw)

    Leave a comment:


  • arQon
    replied
    Originally posted by f0rmat View Post
    No one will really do much - [s]until one day[/s]even after Google is hacked and personal date - I am sorry - "anonymized data" is stolen.
    FTFY.

    This is MURRICA. It doesn't matter how utterly you screw up, how much you lie, or even how many laws you break: as long as you're the one with the money and/or the friends in high places, you get off scot-free. And if you aren't the one with the money, you can go pound sand.

    Leave a comment:


  • oiaohm
    replied
    Originally posted by skeevy420 View Post
    I did a lot of reading last night. What's kind of funny is that, technically (especially when reading Stallman and using his POV), all that BSD and MIT code in the kernel should be GPL'd due to the GPLv2's viral nature. They're just granted exceptions so it's all good. Exceptions so it's all good. Heh.
    They are not granted exception at all. BSD/MIT licensed code will dual license with GPLv2/V3 perfectly fine. The original license grants of the code don't go away because it becomes part of the Linux kernel. This has been tested in court. Stallam is right that all BSD and MIT code when shipped as part of the Linux kernel source is GPL thing he missed is the original license is not removed so its now a dual license item. CDDL straight up conflicts so cannot come a proper dual license item instead a either or license item and what is nightmare with ZFS is you are depending on CDDL for the patent grant so.

    Do note attempting to remove even BSD/MIT parts from the Linux kernel you do have to be careful as the patent grant in different areas can be GPL only. Yes that one when the Linux kernel flags particular internal function as GPL only is a really check here carefully there can be patent dragons hiding here.

    Originally posted by skeevy420 View Post
    Here's the thing though -- if OpenZFS or any other module is considered a derived work of the kernel, how come Alien Isolation and every Linux native game on Steam isn't a derived work of Glibc? System library exceptions; that's why. It wouldn't be possible to run a modern Linux system without system library exceptions. The GPL is simply too viral.
    Glibc is under LGPL so has a linking exception.

    Its worst than that. Linus exception of the Linux kernel https://github.com/torvalds/linux/bl...x-syscall-note with out this Linux kernel would reach into userspace as well. The exceptions are important.

    Originally posted by skeevy420 View Post
    I love how "system library exceptions" are given with Glibc in regards to linking. But how is that any different than a kernel module?
    There are differences in the licenses and those are important and there are operational reasons not to allow it.

    Originally posted by skeevy420 View Post
    How is a kernel module any different? In the simplest of terms, a bit Layman perhaps (but also how you'd explain it to a jury), a kernel module is a kernel's library. A module is to the kernel what a library is to libc. Like a library is to a system, it isn't necessary for the kernel to function and is only there if the user puts it there and wants it ran. But if libraries are allowed special linking exceptions, how come kernel modules aren't granted one either?
    Except a linux kernel module is not like a library dynamic library in a very important way the lack of stable ABI so its now like a static library. Static library built into application has to be of compatible license to the license of the application. It does not stop there. Linux kernel is monolithic memory model with a full mathematically sound memory model as long as you are using what comes in the kernel this is a security feature this is why the taint flag gets set when you load a non GPLv2 compatible module you have loaded something that has undermined the security of the OS..

    Originally posted by skeevy420 View Post
    Does that mean modules should be granted a de-facto special exception?
    The answer is strictly no. The Linux kernel developers saying that your source code of your module need to be under GPLv2 compadible is the same as Microsoft saying your drivers to load in windows have to be signed as well because they need it under a particular license so it can be peer reviewed effectively.


    Originally posted by skeevy420 View Post
    Are Nvidia and AMD not being sued ever for a closed source kernel modules the case studies we need to allow a special module exception in an updated GPLv2? AMD had Catalyst and still has AMDGPU-Pro. Nvidia has their driver. They both clearly violate the spirit and meaning of open source by using closed source modules linked against a GPLv2'd kernel....but they've never once been challenged as far as I know.
    This has a big error. AMDGPU-Pro does not have a closed source kernel module it uses the same amdgpu kernel module as the open source driver only thing that is closed is the user space that is protected by the Linus exception of the Linux kernel. This usage of the same kernel driver by AMD allows you to have 1 applicaiton using the mesa stack and 1 applicaiton using the AMDGPU-Pro stack on the same video card and same screen at the same time. Yes you cannot use the closed source and open source stack at the same time under Linux with Nvidia because they don't have a single kernel module.

    You are ingoing how many times Nvidia binary driver has been borken by the Linux kernel.

    Originally posted by skeevy420 View Post
    I think that the GPLv2 needs to be updated to address libraries, modules, and to clearly define what special exceptions mean and where they're given. Either that or I've shown that closed sourced and/or GPL-incompatible licensed modules are acceptable in Linux by Linus and the LFS.
    The answer is strictly no with the Linux kernel. Remember the kernel is a core program that your complete system depends on. To make it as secure as possible black boxes in kernel space is not a good thing. Microsoft demanding drivers be signed and validated is attempting to address the same problem.

    GPLv2 is not getting updated. Linux kernel does not have the or later clause so to add a new exception would require doing the nightmare process of totally re-licensing the project and getting every involved party to agree to the new license. Yes you would need over 90% agreement this is happened with nothing in Linux kernel history and I do me in history there is not even 100% agreement over the name Linux from day one. OpenZFS re-licensing has a better chance than the Linux kernel by a huge margin.

    The is the syscall exception this covers file systems using fuse and other userspace driver solution.

    skeevy420 its really simple to miss that if you want a stable OS the closer you get to the kernel the stricter the rules that need to be followed to put code there.

    Other thing skeevy420 there are very few closed source Linux kernel modules left. Reason why Linux developers have not bothered with court is up until now the API/ABI changes of the Linux kernel ruining their day have got most of them to release their drivers open source. Also look at the SCO vs IBM and Oracle vs Google cases if you go to court over this stuff you have be ready for a 20-30 year drag out fight.

    skeevy420 you also do need to have a look at the Nvidia blob some time. Nvidia does what they call a wrapper so the ABI of the blob is not meant to be a Linux one. There was a attempt back in history to make a shared kernel binary drivers between Linux, freebsd and Unix that would have not been the Linux kernel ABI so would not have been a derivative work of the Linux kernel.

    Please note Nvidia issues with there driver is the fact you have a binary blob serving up what they use under windows then shoe horning it so it works with freebsd and Linux. Yes we see zfs for Linux doing the same thing with solaris API on Linux. Both are running into the problem the Linux kernel mainline is not going to accept their code and will be totally willing to change stuff that totally breaks them.

    Something to remember here the Linux kernel developers you are right are unlikely to go after you directly for binary drivers or non gplv2 compatible instead are going to come after you indirectly with API/ABI changes. But there are parties who will come after your hide those with patents and they will wait for when they think there is a big enough install base to make the profit they want. ZFS patent issue is quite a major issue.
    Last edited by oiaohm; 02 January 2021, 08:48 PM.

    Leave a comment:


  • brad0
    replied
    Originally posted by pal666 View Post
    there's only one next-gen linux fs - btrfs
    It might actually be useable 10 years from now.

    Leave a comment:


  • aht0
    replied
    Originally posted by Volta View Post
    skeevy420

    There's nothing more free than Linux. Others are either slaves or whores. You can always use this crap with 'open source' slowlaris if you want. Or blame Oracle for choosing incompatible license.
    Funny statement, considering the fact that most Linux distros are increasingly slaves to the whims or RH and IBM.

    Leave a comment:


  • f0rmat
    replied
    Originally posted by skeevy420 View Post

    That's starting to change, thankfully. Look at Massachusetts. They just expanded their right to repair law because Tesla was pulling shenanigans in regards to non-Tesla repairs...all because some guy on YouTube wanted to fix a broke down Tesla and they jerked him around. Hopefully more states pass and/or expand laws like that.

    Frankly, that John Deere crap is no different than AT&T and Verizon blocking custom firmware and rooting on phones because someone might hack something to get free service when in reality most people just want to unblock crappy tethering restrictions and tweak frequency tables for either more battery life or more performance. Why does Phone to PC over BT cost nothing while Phone to PC over USB cost money? Being nickeled and dimed be those corporations sucks. If the AT&T and Verizon networks are so vulnerable that nerds running AOKP with overclocked GPUs bring it down, they're doing something wrong.

    My dream is to figure out something so I can live off the patent royalties. I'm the second guy.
    Do not get me started on cell phones and the bloated crapware from the providers. I buy cell phones now from anyone but the service providers. BS on their stupid stuff and their lame *ss apps. I use Android because I have to - otherwise I would use a flip phone that is old school style. Samsung is also just as bad with their crapware. They make good hardware, but I give a rat's *ss about their "helpful" software. Let me just get a vanilla cell phone and then spend a couple of hours trying to tame Google's BS tracking. Especially since you have to go to several locations to turn things off. I love all these large corporations talk of "privacy" and my "rights." They certainly have bought hook, line, and sinker into "telemetry to improve the user experience" and "customizations" only to make money off of advertising. No one will really do much - until one day Google is hacked and personal date - I am sorry - "anonymized data" is stolen.

    You live in the US, if Massachusetts did something, you might question who is behind it - maybe Tesla did not contribute to the "appropriate" cause.

    Leave a comment:

Working...
X