On Friday 15 June 2007 05:17:44 David Woodhouse wrote:
Hrm... Perhaps I misread your post originally. Let me read it again and see if
I didn't encounter a parsing error somewhere... Nope. Error of omission. The
text you cut changes the meaning of the passage in its entirety.
Here, I'll quote it, in it's entirety:
These requirements apply to the modified work as a whole. If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works. But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it.
In other words, it applies to *SECTIONS* of the code, not to individual object
code files. This is why kernel modules can have their own, separate license
from the kernel. It isn't until the code is shipped as a *standard* part of
the kernel that it has to be GPLv2. (Dynamic Linking, being a totally
mechanical process, cannot create a derivative work under US copyright law,
so please, don't try that old saw)
What this means is that it doesn't matter that a non-GPL module is shipped,
in "object code" form with the "object code" form of the linux kernel it is
designed to interface with - it *still* doesn't become automatically covered
by the GPL.
Nope. Not confused at all.
But you obviously are. After all, what does this have to do with whether the
GPLv2 can "magically" change the law?
Again, what does this have to do with your apparent belief that me putting a
binary of a kernel module that isn't GPL'd on a disc with the Linux kernel
causes that module to become covered by the GPL?
And it also says:
In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.
In other words, even though I've built a program that isn't GPL'd, I can still
put it on the same "volume of storage or distribution medium" as a GPL'd work
and not have to put it under the GPL. Imagine that.
And again, I have to point out that, no matter what RMS and the rest of the
people that wrote the GPLv2 may believe, it can't change the law on which it
is based. If the law says "mechanical translation or processes cannot produce
a new work", then no matter what the GPLv2 says, I can write a parser
skeleton in the "YACC" language, run it through Bison and *NOT* need
the "exception" that the FSF "thoughtfully" provides. Why? Because the
*skeleton* I wrote carries the copyright - not the output of Bison, despite
what the GPL *might* say on this in Section 0. A license *cannot* change the
law, because it gets all its power from the law.
Hrm... I see, you've included the section unmolested here - but you still seem
unable to read it correctly. Perhaps I'm wrong though.
And is this what the process of making a module does? Because that *IS* what I
had mentioned. If the mail I'm responding to was a response to that mail then
you are sadly confused as to what I was talking about.
Yet it still doesn't make them a "combined work". If it did then the simple
act of me installing a copy of the ATI or NVidia modules makes them GPL'd.
But it doesn't - if it did I'm sure that somebody would have filed a lawsuit
over this already.
DRH
--
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
-
