Fedora Freedom and linux-libre

[Les Mikesell]

Alexandre Oliva wrote:
> 
>> Copyright law exists to encourage creation of works and competition
>> among them.  There is nothing wrong with that concept, although the
>> original short protected life followed by passage to public domain
>> seems much more in the public interest than the current incarnation.
> 
> Looks like we're in full agreement about that.  Are we also in
> agreement that it's (current) copyright law that imposes restrictions?
> That the GPL grants carefully delineated permissions for acts that
> would otherwise be restricted by copyright law, enough to ensure the 4
> freedoms are respected, but not enough to lift all the restrictions
> imposed by copyright law?

Yes, I agree that the initial author of a work has as much right to 
impose the harmful GPL as any other copyright holder has to choose a 
more or less restrictive license.  I'm not so sure about additional 
contributors of original work who have this choice taken away in the GPL 
case only, though.  Having the choice to contribute under the GPL or not 
at all resembles that "your money or your life" scenario you presented 
earlier.

>> if competitors were allowed to use components from Linux in
>> commercial offerings
> 
> They are. 

Not in the general case. It would be unfeasible if not technically 
impossible to ship a Linux based OS containing licensed copies of all 
the components needed to match the functionality of commercial OS versions.

> And competitors could do even more if it weren't because of
> restrictions imposed by others, in their licensing offers, which they
> can do because of (i) copyright law making certain acts prohibited by
> default, (ii) unethical imposition of restrictions, and (iii)
> acceptance and passing on of such restrictions from third parties.

Those restrictions are a given, and not necessarily a problem on their 
own. Everyone involved may be perfectly happy to meet whatever those 
other restrictions might be, yet the GPL's harmful restrictions prevents 
the useful combination from being distributed.

> Your unsatisfaction with the situation is shared by me, but your
> blaming the GPL for unfortunate (and at times unethical) choices made
> by others is misguided.

How so?  The harm isn't shared by less restrictive licenses.

>>>> You aren't a victim when you make your own choices.
> 
>>> Heh.  It's not that simple, really.
> 
>>> "Your wallet or your life."
> 
>> You are confusing the value/moral judgement of the thing itself with a
>> delivery mechanism.
> 
> No, I was just providing examples in which, in spite of being given
> choice, you're still an overpowered victim.

But the concept of victim has a preconceived notion of harm, whereas 
meeting non-GPL terms may not cause harm at all.

>>> If it's intentionally misleading, this would just make it yet another
>>> case of unethical behavior. 
> 
>> Like the way 'free' is redefined to mean restricted by the GPL?
> 
> You appear to be confusing two different topics.

Perhaps, calling restricted software free is confusing.

>> What's misguided is your conviction that there is a moral value in the
>> delivery channel at all.  The value can only be defined by the
>> recipient and the effect of the content.
> 
> Reordering a bit from your response to show the contradiction:
> 
>>> It's better to give cigarettes to kids at school than to leave them
>>> without anything to put in their mouths at lunch time? :-)
> 
>> The harm from cigarettes is due to the nature of the product, not the
>> distribution.  The product in question could as easily be milk.
> 
> See?  If my conviction you disputed above is wrong, then the person
> who decides to distribute cigarettes to the kids instead of milk would
> be behaving in accordance with moral and ethics.
> 
> Do you *really* think so?  I mean, seriously, maybe you do, but...  I
> honestly hope not.

Your reasoning requires you to know that cigarettes are harmful and 
there is a body of evidence for that, yet there is no such body of 
evidence that all software covered by non-GPL licenses is harmful, and 
it's not up to a distributor to make that kind of value judgment.  They 
must respect the recipients right to make their own choices.

>> And yet, you make no requirement that has anything to do with the
>> content.  That is, you are apparently perfectly willing to distribute
>> broken or misleading content that will do much more harm than
>> something that just works without requiring any other information.
> 
> I don't agree that incomplete content would do more harm.  Given
> information, it could be completed and do good.  In the absence of
> information, it makes no difference.  Whereas including the vendor's
> bait in a product would turn me into an accomplice, and it would
> extend the long-term harm to more people, for it would be feeding the
> monster.

You are making a value judgment there not only with no evidence, but one 
that's not yours to make.  And if there really is an evil monster in the 
picture, by helping prevent a usable alternative you drive people 
directly to the monster even if it is with the pretense that you aren't 
involved at all.

> As for misleading content, I'm against that on moral and ethical
> grounds.  If source code is sufficiently misleading or abridged, it
> may indeed disrespect freedom #1, which is why I mentioned drivers
> developed under NDA in the message that started this thread.
> 
> Ethics and morals are a lot about intentions, even when the practical
> end results are the same.  If the information is missing because the
> developer just "didn't have time to add it in some meaningful format,
> here are my notes", that's not unethical.
> 
> But if the information is missing because the developer had to sign an
> NDA to get the information, "so I can't share it with you", the
> developer is a bit of a victim and a vendor's accomplice, and the end
> user ends up hurt by the lack of ethics of the vendor in denying the
> information on how to make the developer's and end user's devices work
> to their own liking.

These out-of-context speculations don't make any sense to me.  What if 
there is only one such device and the binary driver works perfectly and 
never needs a change?

>> And you are making an uncalled-for value judgment in calling others
>> foolish who may in fact have made the correct choices for their own
>> situations.
> 
> That they chose the best available options doesn't mean they couldn't
> have avoided being limited to those options, or negotiated a better
> deal, or made a different call for balance between short and long
> term.  But regardless of the options, I don't see how trusting a
> vendor that denies technical information about their product can ever
> be a clever thing to do, even if you end up deciding to live with the
> product because it's the lesser of various evils.

I don't have any different feelings about trusting a company to build 
hardware than to supply software.  Nor do I see any reason I should.  If 
they want to give away the materials needed to duplicate either, great, 
but there is no moral difference related to the type of component.

>> Agreed - it is a bad analogy.  But so is freedom in terms of
>> restrictions on software.
> 
> How about freedom of speech?  Freedom to share? 

Yes, that's the problem.  If I write code that links to a GPL library 
and to any other library, I can't share it, even with someone who 
already has both other libraries.

> Freedom to help your
> neighbor?  Freedom to control what acts on your behalf?  Freedom to
> decide for yourself what's best for you?  Aren't these freedoms?
> Aren't they worth fighting for?  Don't you see that these are what the
> 4 essential software freedoms are about?

No, I want the one that would allow me to share that code.

>>> That's not any of the four essential freedoms.
> 
>> Which is why I think they are entirely misguided.  Reusing and
>> recombining prior work is the basis of all knowledge and progress
>> and any restrictions to inhibit that are harmful.
> 
> It's the "without restriction" that's misguided.  You acknowledged the
> value of copyright in its original version (which was all about what
> you wrote above), yet it's that very copyright that imposes the
> restrictions you oppose.

No, it is specifically the GPL's restriction that I oppose. It takes 
away your choice about obtaining each component independently or sharing 
a combined work.

>>> If the components are indeed separate independent works, copyright law
>>> won't get in your way given the permissions granted by the GPL, at
>>> least as far as the GPLed components are concerned.
> 
>> But it does.  You can't add your own work to a GPL'd part to make it
>> link against a commercial library with no GPL'd equivalent, and
>> redistribute that work of your own.
> 
> s/commercial/proprietary/
> 
> It only does when they aren't "indeed separate independent works".  If
> one is derived from the other, they're not independent works.
> Inasmuch as your work is not derived from the GPLed part, the GPL has
> no claims on it.  It only does when you combine it with the GPLed part
> forming a single derived work.

Please show an example of a non-GPL'd work where there has ever been a 
copyright issue related to another original work being combined (linking 
in the case of a library) where it was clear that every instance 
involved the end user having his own licensed copy of both parts.

>>> Unless you want to accept the unethical impositions from the copyright
>>> holders of the other work, and help them impose them on others (along
>>> with or separately from the GPLed work you'd like to combine with it),
>>> that's what you should do anyway.
> 
>> I don't accept that the copyright holders of the other works
>> necessarily make unethical impositions.
> 
> They don't necessarily make them.  They choose to.

Who does?

> It's their choice.

What choice?  How do you determine that someone is harmed?

> That, along with the fact that they're harmful, is what makes them
> unethical.

The GPL is harmful too.

> Now, ask yourself why you can't say distributed the works you
> combined?  The GPL doesn't say you can't.  It says you can, if you do
> so in an ethical manner.  On what is based your conclusion that you
> cannot distribute the combined work?  What stops you from distributing
> it the way that is permitted by the GPL?
> 
> Why are you complaining about the fix rather than about the problem?

Take the case of the original BSD license. I did/do not consider its 
requirement for attribution to be unethical.  I do consider the 
restriction of the GPL to not permit combinations with items covered by 
this license to be harmful and thus unethical. Likewise, I don't 
consider all commercial/proprietary distributions to be unethical. Some 
provide good value for their terms.

>> There's no guarantee, but I'm convinced that the restrictions in the
>> GPL that prevent reuse in competitive works have made more money for
>> Microsoft than anything Steve Balmer ever did.
> 
> You're mistaken to boot.  The restrictions are from copyright law, not
> the GPL.  They didn't stop Microsoft, and I doubt they would have had
> a negative impact on Microsoft if they didn't exist.

You are speculating that in the absence of the GPL, no freely 
redistributable code would exist.  That is rather clearly wrong, given 
the proliferation of less restricted code like apache, the *bsd's, MITs 
X, etc.  The problem is that much of this code and subsequent work has 
been hijacked by the GPL which is a one-way trip, so drivers originally 
modified from the bsd versions and included in Linux took away the 
choice of subsequent contributors to use any other copyright and now 
can't be re-used, for example in OpenSolaris.  This harms everyone.

> For starters, Microsoft's EULAs are not mere copyright licenses,
> they're contracts.  So the absence of copyright, or a more permissive
> GPL, would just enable Microsoft to milk software developers as much
> as it does computer users.
> 
> And even if the lift had enabled other competitors to surface that
> didn't surface because of their commitment/requirement/determination
> to not respect their customers' freedoms, would they be any better
> than Microsoft, given the commitment?
> 
> What good outcome would you expect from this?

Competition, lower prices, better products, more choices.

> Single standards don't have to be monopolies; in fact, the whole point
> of standardization is to avoid the formation of monopolies, to keep a
> level playing field and reduce the entrance barrier.

So how does that work in the case of something like a blu-ray player?

>> The way to prevent problems is to lower the bar to building
>> competing alternatives that leave the choices up to the recipients.
> 
> As in, more of a bad thing is a good thing? :-)

Absolutely.  You aren't going to find perfection, so your best bet is 
having choices among imperfect things.  As people make those choices, 
the bad ones go away.

> Because the GPL doesn't discourage more good things, it only
> discourages that good things be used to create bad things.

It not only discourages good things, it prohibits a near-infinite number 
of combinations of things that could have been good.

-- 
   Les Mikesell
    lesmikesell at gmail.com