OSNews

> looking for the link again

I want to see this post.

Linus cannot make comments in place of others. Linux is not Linus' work. Linus is neither the only contributor nor the contributor of the majority of the code.

If Linus wants to make an exception for NVidia, he needs permission from all copyright holders. Linux is distributed under the GPL, contributors contributed their work under the GPL without extra clauses unless explicitly otherwise stated.

And if Linus really makes an exception for NVidia, other companies will ask for exceptions as well, and in the end everyone will ask for ways how to bypass the GPL most efficiently.

If you really want to go this way, the result will be that Linux becomes proprietary in the long term. Why would companies want to contribute their drivers under the GPL if they don't have to? Keeping them proprietary is much more comfortable, no expensive quality review, no expensive security audits etc.

In the end, Linux will become unportable: Porting it to a new architecture will take years instead of months, like it did with Windows, because the OSS community will have to wait for blobs from the vendors. *sigh*

"Linus cannot make comments in place of others. Linux is not Linus' work. Linus is neither the only contributor nor the contributor of the majority of the code."

No, it is not that Linus spoke about the code. He spoke in terms of the GPL...am seriously looking for the link. He said Nvidia does not violate the GPL as it does not directly link to the kernel. It was a discussion on the kernel mailing lists, as well as another place. If I find it you will see what I am talking about. It was nothing making an exception for code that links to the kernel, just a statement on how Nvidia does not link to the kernel, making it not in violation of the GPL.

I want to see this post.

http://kerneltrap.org/node/1735
What Linus said is:

And in fact, when it comes to modules, the GPL issue is exactly the same. The kernel _is_ GPL. No ifs, buts and maybe's about it. As a result, anything that is a derived work has to be GPL'd. It's that simple.

Now, the "derived work" issue in copyright law is the only thing that leads to any gray areas. There are areas that are not gray at all: user space is clearly not a derived work, while kernel patches clearly _are_ derived works.

But one gray area in particular is something like a driver that was originally written for another operating system (ie clearly not a derived work of Linux in origin). At exactly what point does it become a derived work of the kernel (and thus fall under the GPL)?

THAT is a gray area, and _that_ is the area where I personally believe that some modules may be considered to not be derived works simply because they weren't designed for Linux and don't depend on any special Linux behaviour.

He goes on to say later in the LKML thread that:

In contrast, these days it would be hard to argue that a new driver or filesystem was developed without any thought of Linux. I think the NVidia people can probably reasonably honestly say that the code they ported had _no_ Linux origin. But quite frankly, I'd be less inclined to believe that for some other projects out there..

Basically he's saying that the derived work clause is fuzzy. In nVidia's case specifically, the binary "blob" is OS-agnostic, they use the same binary for Win and BSD. The linux piece is simply the kernel wrapper, which is in fact GPL. So if the binary itself isn't designed specifically for linux, and the piece that is dependent on linux is GPL, then the nVidia driver is not a clear violation of the GPL.

And I think this is the core of the issue, no matter how much people want it to be black and white, it isn't. What defines a derived work? If someone uses a clean room type reverse engineering of the linux API, would that driver still infringe?

I also think Linus' point is credible because if you read the whole thread, he is admant and unforgiving about the GPL nature of the kernel, but he's willing to admit that loopholes exist. He doesn't endorse them, but he doesn't pretend otherwise. It's an uncharacteristcly pragmatic viewpoint for him.

Clearly there are cases where binary blobs compiled against a kernel and distributed as such would violate the GPL. Yet it is also evident that the manufacturers will continue to resist opening up their API's for reasons that the community refuses to accept. It doesn't matter who is right or wrong, and linux hasn't attained the saturation point yet where it can force manufacturers to comply, they'll simply walk away.

So the inevitable middle ground is going to be either a pure GPL kernel that loses commercial momentum due to the impracticality of supporting certain hardware, or we wind up with a bunch of kernel wrappers around generic drivers, or worse ndiswrapper type solutions around windows drivers.

It might make some kind of ignorant pleb, but I'll settle for nVidia's approach until something better comes along. It's not ideal, but I've never had a problem with mixing and matching driver and kernel version when necessary, since the source is provided and it's a simple step to re-compile.