the right to make a difference

Xen list at
Thu Jun 2 12:35:56 UTC 2016

Sam Bull schreef op 02-06-2016 13:36:
> On Thu, 2016-06-02 at 12:31 +0200, Xen wrote:
>> > So I think the unfairness is very much there now. Spengler is now
>> > actually getting something for nothing, where before he was not,
>> > and
>> > the whole intention of many of the contributors to linux has been
>> > subverted.
>> Untrue. There was no before. Linux was already available to him. It
>> was given away for free.

Thank you for your response. It is very meaningful to me.

> You seem to repeatedly suggest that nothing is owed to the kernel
> developers and that it was given away with no requirements. But, the
> kernel is not simply given away, it is distributed under the GPL, and
> as already mentioned the intention of the GPL is that changes to the
> project by other developers are shared under the same conditions, so
> that the changes might be incorporated back into the original project
> or used by others. If somebody does not want to agree to these terms,
> then they are welcome to find an alternative (e.g. Apple used the BSD
> kernel).

The intention of the GPL is not really relevant.

What happens is that the authors remain to have a say about how the 
product is used, if copyright is at play (at least the idea of 

I don't think you can really embed a form of monetary compensation 
within a copyright license itself.

Normally when you sell something, it has license terms, but the terms 
themselves do not describe the contract; hence a license alone cannot 
constitute a contract for monetary exchange, or anything else. 
Explicitly, the terms do not state that the intent is for the original 
author to have copyright on your work and if they did the thing would 
probably be deemed illegal by all courts.

This is because a derivative work can in itself be copyrighted (at least 
the work you have done yourself, which must be a substantial work) and 
so a free transference of copyright to another party would never work.

Now in actual effect that is what would happen. The person doing the 
work would see his livelihood taken away by "thieves" and his added 
value would cease to be any added value; as such the whole idea and 
reason for copyright would cease to exist; it is contradictory in terms.

So what I am saying is that what the new author owns the original 
author, is that if the original author's interests are threatened, that 
person can have a say in it.

Now the question becomes what those interests are. GPL is really 
irrelevant in that sense. If you can make a case that those interests 
imply the free access to and integration of any additional work done by 
other parties free of charge, then you may have a point.

Then, when that point is made, the GPL agrees with that. If that point 
can never pass, then the GPL can never be legal.

But you MUST explicitly mention and make available the idea that this 
contract forms a valid basis for interaction, and is a valid way to 
"take back" seeing as that your structural component (the kernel) 
already exists everywhere, and that module (it is not a module) (I 
think) it almost like a car on a road. Does the owner of the road also 
own the car?

So if the original authors really have an actual interest that can be 
recognised, then yes. But this interest must first be stipulated for the 
terms themselves to have any merit. In pretty much every copyright case 
I have seen (not many now) the courts decided based on real interests, 
not simply license terms.

There would be two points:

1. derivative work
2. fair use.

> Saying that nothing is owed and that there is no agreement, would
> suggest that after buying a proprietary piece of software, I can now
> modify it and resell it to other people, despite this being against the
> intention of the author.

Nothing monetary is owed, and therefore not any amount of work either. 
After all, in any such agreement, it would not depend on whether the 
work was actually done, you cannot in advance claim possession of all 
work done by anyone.

Even if you wanted it to have some merit, you would still need to 
stipulate any exact or moderately exact agreed upon exchange in advance. 
The taking of all work for all eternity is simply a form of hostage 
taking. That is like saying you sign a license deal with a record 
company, but the license states that you can never go to another company 
ever after; you are bound to it for all eternity.

So what is saying you should have access to the work? Well everyone has, 
doesn't it?

Some implicit agreement that is in itself rather questionable to pretty 
much everyone, especially if you characterize it like this (which is the 
truth, really) is not going to be very strong if and when the product is 
being mass distributed free of charge, any sense of exclusivity is 
non-existent, and if you are only using it not to maintain copyright, or 
to maintain copyright on the premise of wanting to control your real 
immediate interests (such as those mentioned before) but only to acquire 
free work from others?????

Is not what copyright is meant for.

You can of course make distinctions between non-commercial and 
commercial use. That would imply an ongoing agreement and settlement and 
business practice of entering into real contracts with commercial 

That would also imply that your product is a rather exclusive thing, and 
cannot be used as, or substituted for, a presently omnipresently 
available core system component that everyone needs to use, or be 
without entirely. It is not an optional component here.

So no, you cannot modify it and resell it, because that would hinder the 
market for the original product. This is an extension, not a 
replacement, and that is a crucial distinction.

> The main question appears to be about a legal loophole, in whether the
> work is considered a derivative when it can be distributed as a
> separate patch. I don't know what these patches are, but I imagine if
> they are mostly a set of self-contained files (like a module) it would
> not be considered a derivative, whereas if it involves significant
> changes to the kernel, then it would be considered a derivative, and
> the GPL should be enforced.

Technical terms do not imply everything. I don't think they are modules, 
too much needs to be changed for anything to really change the core 
functioning. You could do it with a module, but you would have to 
replace a lot of system calls, and I don't know or think it would be 
feasible really. Moreover modules can be unloaded and I don't know what.

I just don't think "the GPL should be enforced" follows from what you 
say. That should or would only follow if you consider the GPL holy in 
any sense. In actual copyright law, the /more/ changes something has, 
the /more/ it is an original work, or substantial work, and the more 
substantial your changes, you have more access to a copyright of your 
own. And the more value it adds, the less it will be seen just as a 
copy, or modification. It will be seen as an addition.

The technical terms actually do not matter so much.

Then when it is an addition, the question of derivate work comes into 
play, but this license is not going to get judged on its sole behest.

There will be a comparison and a weighing of interests because fair use 
is also an important component.

Maybe if these aspects are judged in isolation.... according to the 
letter of the law.... you could win this. But the entire picture needs 
to be taken into account, I think. And, the question of commercial 
impedance will come into play, and you will fail on that.

The patch does not hinder the kernel developers. Their business does not 
suffer because of it. If their business is only defined as a return on 
investment by taking hostage every other maintainer or developer who now 
wants to build something on a system that has already become omnipresent 
and pretty much "public domain or at least common infrastructure" (which 
is almost the same) -- an unavoidable, essential part of life -- (and 
the name of the kernel is synonymous with the operating system; suppose 
you were to replace the kernel with something that is binary compatible; 
what then? Are you still going to call it Linux?) -- if that is your 
definition of business, that you create a monopoly for yourself, make 
yourself into an indispensable platform, and then monopolize this by 
claiming the work of everyone and everyone on this platform....

Then you are doing exactly what Microsoft has done, and you will be 
flagged for it.

> Another question, nobody has asked: It sounds to me, like the developer
> is possibly accepting that the patches they are selling are licensed
> under the GPL, but saying that if they use their rights to redistribute
> them, they will not continue to sell them updates.
> So, if you consider the developer is allowing anybody to exercise their
> rights for the piece of software they have received and that the
> developer can choose not to sell a new version of software for any
> reason; is the threat enough to break the conditions of the GPL?

The original poster he considers it "tort". I think tort is a very broad 
term, I am not accustomed to English legal terms yet.

Maybe you can consider it devious, but it is only so with respects to 
something else that might prompt anyone to take such measures.

I think redistributionary rights for a 2nd-degree licensee that wants to 
distribute the full product and not just personal modifications, is just 
a very weak argument.

That licensee can distribute the original product, and the original 
authors thereof can also still do it and probably maintain a sense of 
monopoly on that.

That licensee can probably distribute individual patches just fine of 
the new product (a patch of the patch, as it were) and if not, I would 
have a problem with that, but scarcely anyone would have an interest in 
doing so now.

So the only thing that licensee cannot do is give the product to someone 
else, which is what copyright was designed to prevent.

So you think you have a strong moral case, I don't know.

Does it breach the GPL? Maybe not but it is a conflict of interests. The 
GPL states that in any conflict of interests, the license is void. But 
that is what the GPL says.

In that sense, it does breach the GPL (but it is not very hard to do 

The GPL does not state anything explicitly about future versions, I 

It also does not compel anyone to distribute to anyone.

But it is a form of interference, for sure.

The author is basically saying "my product is no longer GPL. It has the 
added condition that you cannot redistribute my product". Simply because 
apparently the product itself is part of a support contract.

Again, again, again, complications.

More information about the Ubuntu-devel-discuss mailing list