> Even if they did, it's under a proprietary license and it's not clear that proprietary kernel modules are necessarily derivative works, they are commonplace.
True, proprietary kernel modules are commonplace, and not necessarily derivative works. But you very rarely see proprietary drivers "in-tree" (that is where the source code file is placed into the Linux kernel source code tree and built as part of the kernel build process). I remember reading an argument by Torvalds that any driver built in that way is a derivative work of the kernel (even if it is built into a module) but unfortunately I can't find it now.
> Even if they are derivative that may make Samsung in violation of the GPL but not actually make their code covered (these are be separate issues).
Mmm... I'm not entirely sure... Normally I'm a "separate issues" kind of guy, but here I'm hesitating...
To make this interesting let's say Samsung has distributed to rxrz in binary form and that the driver is a derivative work. Then the driver is part of the Program distributed to rxrz under the GPL and Samsung has granted rxrz this right (as per section 1 of the GPLv2):
"You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program."
Isn't that exactly what rxrz has done? :) Why would rxrz not be allowed to argue that in court? I agree that normally the issues are separate, but in my mind the GPL ties them together in this case. I'm not saying that rxrz is right, but it's not entirely clear-cut to me.
True, proprietary kernel modules are commonplace, and not necessarily derivative works. But you very rarely see proprietary drivers "in-tree" (that is where the source code file is placed into the Linux kernel source code tree and built as part of the kernel build process). I remember reading an argument by Torvalds that any driver built in that way is a derivative work of the kernel (even if it is built into a module) but unfortunately I can't find it now.
> Even if they are derivative that may make Samsung in violation of the GPL but not actually make their code covered (these are be separate issues).
Mmm... I'm not entirely sure... Normally I'm a "separate issues" kind of guy, but here I'm hesitating...
To make this interesting let's say Samsung has distributed to rxrz in binary form and that the driver is a derivative work. Then the driver is part of the Program distributed to rxrz under the GPL and Samsung has granted rxrz this right (as per section 1 of the GPLv2):
"You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program."
Isn't that exactly what rxrz has done? :) Why would rxrz not be allowed to argue that in court? I agree that normally the issues are separate, but in my mind the GPL ties them together in this case. I'm not saying that rxrz is right, but it's not entirely clear-cut to me.