[linux-elitists] biggest binary blob EVAH? was: So, Microsoft spent that good will already.

D. Joe Anderson deejoe at etrumeus.com
Fri Jul 24 09:48:31 PDT 2009

On Thu, Jul 23, 2009 at 09:26:24PM -0700, Greg KH wrote:
> On Thu, Jul 23, 2009 at 07:31:56PM -0700, James Sparenberg wrote:
> > In other words they aren't willing to get caught as SCO did,
> > in a situation where they can't claim foul because of their
> > own dirty behavior.
> Which is great that they have come clean, what's the problem
> with that?

> Lots of companies have done this in the past (come clean with
> the code), and that's great, it is exactly what the large
> majority of the Linux kernel developers want to see happen. 
> Including myself.  And it is by doing this, we have brought
> more companies into the Linux kernel development community,
> and in the end, made Linux much better for both us as
> developers, and everyone else as users.
> You, yes you, have directly [benefited] from this strategy over
> the past 15 years, whether you realized it or not.

The kind of driver code release we tend to benefit from the most
are drivers for *real* hardware.  There's a dramatic qualitative
difference between the release of code that enables on-the-card
3D video rendering or that makes a network card run, for
instance, and this code that helps Linux work on somebody else's
proprietary virtual machine.

> What's that phrase about catching more things with honey than
> vinegar...

OK, so you're all about not spooking the herd wandering
around outside while you gently lead another newly-broken one
into the corral.

We get that you're using honey to catch flies; that you're
wrapping copyright's unwieldy, tortious shithammer* in velvet,
to direct and soften the blow for better effect; that you're
using the carrot, rather than the stick; that the incremental
approach is the way to boil the frog.  Fine.

It's important work, you're doing well, thank you.  We can be
glad you're making some sort of incremental progress.  At least,
some of us are open to seeing this bit as progress, rather than
as a big noisy distraction from getting more drivers for real

> I'm tired of this whole thing now.  If people aren't happy
> about what just happened, the fact that a major software
> company just validated the license of the largest and fastest
> growing software project in the history of computing, and also
> validated the statement that this project's developers made
> about the license under which all drivers that interact with
> their project, then they need to seriously rethink their true
> motives and see if it has anything at all in common with the
> people who are actually creating this project.

Pragmatists can concede that it's probably not a big deal.  But
to be *happy* about it?  Elitists, in particular?

Tell you what:  Why don't you decide whether this is something
we should be positively happy about and grateful for, as you
aver above, or whether it is no big deal, as you and others have
made out elsewhere.  Then, get back to us.

That way, we can make a decision.  We can try to decide if we're
valuing the worth of the release of this code based on whether
or not some big bad voodoo daddy software company gives our
project attention and "validation", no matter how miniscule and
no matter how offset by that company's other recent efforts.  

Or we can decide whether our only useful metrics are the extent to
which the released code does what we want and whether we're free
to take that code to use, to study, to improve, and to share.

There's not much code in this release, compared either to
Microsoft's total code base, or to the aggregate code base of
free software, and it's utility is limited.  That supports a "no
big deal" line on the release.  If there's any fuss to be made
of this release, it's symbolic.

If you want people to be "happy" about this move by a "major"
entity, it seems you're arguing a symbolic value line, which
plays right into all the pissing and moaning and wringing of
hands.  If it's symbolic, then its meaning is ever the more open
to interpretation (eg, spin), and people's perceptions and
emotional reactions to it matter a great deal, for good or ill.

For example, one might say that this effort is doing nothing so
much as turning a proprietary OS into Linux's biggest binary
blob ever.  In this interpretation, they haven't released a
driver (as they claim and which claim you "validate" by
repeating) so much as they've released just a tiny piece of glue
code for a much larger driver, the bulk of which remains
proprietary and beyond the ameliorating touch of the Linux
kernel team.

In that light, maybe it'd be best to stick to the "no big deal"

In the meantime, remember, the point of using honey to catch
bugs is to remove them as a nuisance.  The point isn't to eat
yet another bug, whether for love or money.

* props to Mr. Bad
man screen | grep -A2 weird
  A weird imagination is most useful to gain full advantage of
  all the features.

More information about the linux-elitists mailing list