On Sun, 2016-08-28 at 09:47 +0200, Greg KH wrote:
> 
> So I strongly believe that without the help of companies contributing to
> Linux, we wouldn't be here where we are today.  It's taken everyone to
> get here, we can't do it alone.  So let's never try to cut ourselves off
> of people we perceive as "the enemy", and instead, treat them as merely
> "oh, they don't quite get it yet", and work to educate them and bring
> them into our community[2].

Greg, this is completely unhelpful hyperbole. All your examples are
arguing against a massive straw man.

Nobody is calling companies "the enemy", or trying to say that they
have not contributed massively to the development of Linux.

Of *course* we want corporate involvement. But we just want everyone to
follow the terms of the licence.

There is a *small* minority of bad actors who simply don't do that the
licence requires of them. And it *isn't* harmless. When a closed but
"good enough" solution exists, it massively reduces the motivation to
work on a *proper* open source solution. Without OpenAFS, we'd have got
true AFS support in the kernel much sooner. Without the nVidia driver,
we'd have have much faster progress on nouveau. Without the binary-only 
Broadcom wireless drivers, we'd have had a much better platform for
OpenWRT and the like to develop on.

Sure, some GPL violations are a "victimless crime", but many aren't.
Many violators actually do complete with honest companies who are
paying the extra cost of doing the right thing, and undercutting them
by taking the shortcuts.

James said that the "solid business rationale" for contributing to
GPL'd code is based on the fact that "your competition HAS TO show you
the code they would complete with you on".

That's *my* stress on the words "HAS TO", of course — because in a
world where nobody was *ever* actually held to the terms of the
licence, and everyone got away with just treating it as if it were BSD
and contributing back only when they felt like it, the bottom would
fall out from under James's "solid business rationale".

Failing to contribute GPL'd code back when the GPL says you "have to"
is a risk. You seem to be arguing that you want to make it
unambiguously *not* a risk to break the law in that way, and you don't
seem to realise that we *need* it to be a risk because that's the
underpinning of *all* the quiet negotiations that we *hope* will always
be sufficient to resolve every violation.

-- 
dwmw2