Announcement

Collapse
No announcement yet.

Phoronix Test Suite Exploring GPLv2 License

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

  • #51
    Originally posted by brosis View Post
    Doesn't have to be CDs, can be any other popular disk format.
    You missed this part:

    "If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code."

    And the source you previously linked to (4.1.4) basically says that this part allows for internet-only distribution of source code under GPLv2.

    Comment


    • #52
      Originally posted by deanjo View Post
      That is about really the biggest win the GPL had. But then again, DD-WRT falling into compliance has been questioned as well. OpenWRT pionts fingers as Sveasoft for violating GPL and Sveasoft points fingers at OpenWRT for distributing non-GPL components of their software as GPL so it hasn't proven all that effective there either.
      OpenWrt is clean, and I suggest you provide documentation if you think otherwise. DD-wrt is borderline. Sveasoft is really just noise. and yes, embedded is probably the most problematic area for GPL enforcement, so it is still not perfect.
      KHTML was licensed in Lesser GPL
      Yes, a restrictive license.
      and Apple was not in violation.
      You are of course free to believe that webkit would be released without the original LGPL. I am afraid digging up the controversy between Apple and KDE developers back then is a bit over the top for me time-wise right now, so I am not going to attempt documenting violation. I see you are taking the webkit story as a victory for permissive licensing. An interesting spin.
      Yup sure do, many drivers are still not free of closed firmware.
      Indeed that is a problem. Mostly due to Linus being too pragmatic I would say.
      There is a ton of hardware that is still not supported by linux
      Sure, but less tons than on any other OS. Do you really believe we would be where we are today on drivers without copy-left?
      http://www.kernellabs.com/blog/wp-co...ssusDriver.pdf

      Try to get that or the source from them and they ask you for thousands of dollars despite it hooking into several GPL/MIT licensed code. So despite their driver being derived work from those projects any request for code gets you a polite "Go pound sand" reply. Since they are not using my code, I am powerless to do anything about it.
      If I was a copyright holder of violated code, I do believe I would enforce it. There will always be people braking the law, that doesn't mean that the law has no effect, or is not enforceable. You never came across me as the black and white kind of person before. Every day people steal and abuse without getting caught.
      I wouldn't go that far. Intel has been going all out for their drivers and they are still not the clear choice.
      No, their drivers on Windows have been superior. Their efforts are marginally better than AMD, but still not all in. In any case, we are talking gaming here. Intel was never an option for gaming, they still have some way to go there.

      I have nothing against permissive licensing, it has it's place. However, I have severe problems if people deny the victories copy-left gave us as if it would all have happened without copy-left. If any, I think the BSDs would be all but dead by now without GNU/Linux. There may be one point in the future where we do not need copy-left anymore, but I don't think we are there yet.

      Comment


      • #53
        Originally posted by dee. View Post
        Well yes, and it's kind of funny in that "GPLv2+" is the actual, original GPLv2 license as published by the FSF. It's the "GPLv2" without the plus that is the modified version of the GPLv2 proper. The original GPLv2 does include "or later version" clause.
        False. I happen to have a copy of the GPL2 shipped with GNU binutils, and it says (section 9)
        Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.
        It is clear that in the absence of the "any later version" provision (which is what "This software is licensed under the GPL2." would be), you are not given the option of following the terms and conditions of the later version.
        If you say "under the GPL", nothing more added, it's GPL 1 or later (even if you include a copy of the GPL3).

        Now, as far as real differences goes, there's one that hasn't been mentioned: GPL3 explicitly allows combining with the AGPL3. In my mind, that's a disadvantage from the perspective of "I want every publicly available derived work to be distributable under my terms." And several companies which were consulted during the drafting of GPL3 indicated that they would not use software licensed under the terms that ended up in the AGPL3. (Originally it was going to be a clause optionally appended to the GPL3, but some companies said that they wouldn't use software with the clause, and if they had to read through every copy of the GPL3 to check, they would simply not use GPL3 software.)

        Comment


        • #54
          Originally posted by Del_ View Post
          OpenWrt is clean, and I suggest you provide documentation if you think otherwise.

          The OpenWRT Project, on whose work Sveasoft's product is based, has had enough. So, in March, the project notified Sveasoft that its OpenWRT license was terminated due to GPL violations. From OpenWRT's point of view, Sveasoft no longer has any rights to be distributing OpenWRT's work in any form. Sveasoft responds that it remains in compliance with the GPL, and that OpenWRT has improperly incorporated Sveasoft code which was never meant to be licensed under the GPL - a charge that OpenWRT developers deny.
          "You rattle your sword, I'll rattle mine" and nothing changes.

          I see you are taking the webkit story as a victory for permissive licensing. An interesting spin.
          I don't know how else you could spin it. KHTML was going nowhere fast on it's own. It wasn't until Apple forked and continuously developed WebKit did it start getting some traction with it's fork being under a BSD license.

          Indeed that is a problem. Mostly due to Linus being too pragmatic I would say.
          I would say it was Linus realizing that the cost/gain out ways idealism that would be hard to enforce and police. Linus is more about the code then the politics.

          Sure, but less tons than on any other OS. Do you really believe we would be where we are today on drivers without copy-left?
          Yes I do believe that it would still come along, perhaps even more support. While it is true that linux supports more hardware by the sheer amount of legacy code out there I would not say that it is the OS that has enjoyed the most support. That title would still go to Windows. The difference being is that Windows and the manufacturers of products for windows realize there comes a point of diminishing returns. Sure it is cool that I could in theory still have a Gravis Ultrasound rocking in linux today if they even sold a motherboard with ISA slots but it isn't to practical in any means. While the theory of keep stacking on support with a sounds nice, I tend to be one that would actually favor a heavy pruning of the kernel to get rid of a lot of that stuff that is virtually no longer used. I sometimes look at the kernel and think it should be on an episode of "Hoarders".

          If I was a copyright holder of violated code, I do believe I would enforce it. There will always be people braking the law, that doesn't mean that the law has no effect, or is not enforceable. You never came across me as the black and white kind of person before. Every day people steal and abuse without getting caught.
          First of all, you have to find out and prove that it has been violated. As I mentioned above, it is really easy to hide stolen code. Then if you do somehow find out about it what happens if it is against a corporate ninja team? Do you have the financial resources to take on such a fight? The FSF won't help you unless you have some pretty damning proof and not just a suspicion

          No, their drivers on Windows have been superior. Their efforts are marginally better than AMD, but still not all in. In any case, we are talking gaming here. Intel was never an option for gaming, they still have some way to go there.
          The latest linux Intel benchmarks that have been coming out for a while actually show their Windows/Linux drivers for the most part are on par with each other. Windows win some here and Linux wins some there (and to tell you the truth the intel linux driver seems to be more stable their their Win 8 driver save for one really annoying black dot bug on IB). While it is true that I won't be doing any 7680x1440 gaming on an intel IGP anytime soon like I do on the Titans, it does work fine for more common resolutions for gaming in linux for all but the most demanding titles (I'm looking at you Metro).

          I have nothing against permissive licensing, it has it's place. However, I have severe problems if people deny the victories copy-left gave us as if it would all have happened without copy-left. If any, I think the BSDs would be all but dead by now without GNU/Linux. There may be one point in the future where we do not need copy-left anymore, but I don't think we are there yet.
          That's fine if you think that, I think otherwise from my experiences in the past. We can all speculate and cite examples both for and against what could have happened but ultimately nobody can prove anything (unless you are JJ Abrams and do a Star Trek reboot). I would attribute much of Linux's success at being able to fill the right niche at the right time in computing more then it's license and development pushed forward because it had a smart man heading up the project who ultimately still calls the shots on its direction. BSD's biggest issue is that it has had in the past poor leadership and a very narrow focused goal market. That has ultimately hurt it meanwhile Linus got the ball rolling and went for a shotgun one kernel to rule them all approach in linux's development which has been able to maintain it's momentum.

          Comment


          • #55
            Originally posted by deanjo View Post
            http://lwn.net/Articles/178550/


            "You rattle your sword, I'll rattle mine" and nothing changes.
            Well, according to your link the situation did improve. Moreover, it was pretty clear that Sveasoft tried to look compliant. There will always be people pushing the limits, that does not mean that the system is flawed. Lastly, there was no hint of OpenWrt not complying to GPL. Actually, any patches to openwrt needs to be signed off by the author as GPL.
            I don't know how else you could spin it. KHTML was going nowhere fast on it's own. It wasn't until Apple forked and continuously developed WebKit did it start getting some traction with it's fork being under a BSD license.
            It was forked in 2001 because it had come quite far according to Apple, the going nowhere part started after the forking. This data point you simply cannot spin your way out of. Since it was LGPL it could not be licensed to BSD. What happened was that minimal core components webcore and javascriptcore where released as LGPL more than a year after the fork, in 2003. Then two and a half years later, in 2005, webkit was finally released, but everything except the two core components were BSD licensed. Keeping Khtml in sync was a very challenging task to say the least. Actually, it wasn't until half year later (end of 2007) that Apple finally opened ut webkit for external contributions (but then under BSD for anything but the core components).
            I would say it was Linus realizing that the cost/gain out ways idealism that would be hard to enforce and police. Linus is more about the code then the politics.
            Don't give me that anti-FSF speech. Linus is very much about copy-left, in fact he stated that choosing GPL was the best thing he ever did. Many of us do believe in open development and put in countless hours to make it happen. Linux would not fall down if the firmwares were thrown out. I have personally been chasing bugs in wireless drivers from Intel where the firmwares where updated to address bugs to no avail, and none of us could inspect or fix the firmware except Intel employees. This is an example of why copy-left is important. For some reason you seem to think it is an argument for permissive licensing. This really puzzles me.
            While it is true that linux supports more hardware by the sheer amount of legacy code out there I would not say that it is the OS that has enjoyed the most support. That title would still go to Windows.
            I am pretty sure you are wrong on this one, but I am in no way ready to document it. One error though, you say windows, but of course you mean windows 8.1, right?
            As I mentioned above, it is really easy to hide stolen code. Then if you do somehow find out about it what happens if it is against a corporate ninja team? Do you have the financial resources to take on such a fight?
            You seem to have very little experience with corporate code. Let me put it this way, GPL code is infinitely more safe from copycats than proprietary code. Take my word for it, or meet me in person. I am not about to discuss this further in a public forum. And yes, I may very well have the means necessary to enforce the GPL license.
            The latest linux Intel benchmarks that have been coming out for a while actually show their Windows/Linux drivers for the most part are on par with each other.
            and what OpenGL level are the two at now?
            That's fine if you think that, I think otherwise from my experiences in the past. We can all speculate and cite examples both for and against what could have happened but ultimately nobody can prove anything
            Who needs proof, you only need a brain. Safari, well not really, at least webkit, is open source. As for the rest of Apple's office apps, you go and find the license: https://www.apple.com/creativity-apps/mac/

            Comment

            Working...
            X