On Wednesday, August 19, 2015 9:09:59 PM Rich Freeman wrote: > On Wed, Aug 19, 2015 at 7:40 PM, Fernando Rodriguez > <frodriguez.develo...@outlook.com> wrote: > > > > The law is not clear about that. But how can it not be a derived work if it > > doesn't work without it? > > > > A is only a derived work of B if the law says it is.
The letter of the law is constantly changing. I'll grant that your argument is more inline with the letter of the law because the law wasn't written with this specific case in mind. But the FSF's argument is more inline with it's spirit. Until a court decides one way or the other it's just a grey are so there's no point in arguing. > My pot isn't a derived work of my stove. My browser isn't a derived > work of the kernel it runs on. > > Copyright law doesn't talk about interoperability when it comes to > derived works. It talks about translations, adaptations, etc. These > are derived works because they incorporate substantial portions of the > original work. MST3K incorporates substantial portions of the movies > they're parodying. Rifftrax does not. That is the difference. > > A kernel module does not incorporate substantial portions of the kernel. > > And interoperability is actually a legal defense against copyright. > If the only way to make something interoperate with something else is > to partially copy it, the court tends to view that as fair use. The GPL symbols are not necessary for interoperability. For that you need little more that access to the hardware and an interface to userspace. Most of those GPL symbols are convenience routines to enable reuse of code among different subsystems and drivers. -- Fernando Rodriguez
