On 05 May 2005 18:00:12 -0400
"Nathan J. Williams" <nathanw@wasabisystems.com> wrote:

> This is simply wrong. To make a somewhat extreme example, many drivers
> in NetBSD were written or enhanced by looking at how Linux drivers for
> the same hardware worked (and even say so in the CVS logs). If your
> theory of derived works applies, then the entire NetBSD kernel is now
> a derived work of GPL-licensed code and as such is subject to the
> GPL. Do you believe this to be the case?

Yes.  Sounds like some people really screwed up and should be more
careful about how they write drivers.  If NetBSD ever develops deep
pockets, it won't go well.

> > There may be no contamination from looking at BSD licensed code, but
> > APSL is not BSD.  Besides, if you like Darwin code that much, why
> > not run the operating system that is based on it?
> 
> I don't at all see what you're getting at here. Their code does
> something that ours doesn't, involving a weird bit of hardware (the
> locking-down caps lock on ADB keyboards) that is specific to the same
> company that wrote the code. It seems entirely natural to see how they
> handle this situation when considering how to handle it in our code.

In other words, you have no way of obtaining the knowledge for that
hardware without looking at the Darwin code.  How can you argue that the
code you then write was not derived from the Darwin code?  Did the
knowledge magically pop into your head?  This is precisely my point. 
The hardware is proprietary to Apple and they have not released any
documentation for the hardware except for the code.  Any code written
based on knowledge obtained from Darwin code is then subject to APSL.  

tim