the right to make a difference

Xen list at
Thu Jun 2 18:39:09 UTC 2016

Sam Bull schreef op 02-06-2016 16:33:
> On Thu, 2016-06-02 at 14:35 +0200, Xen wrote:
>> 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 
>> copyright).
> Yes, and the authors stated they require you to allow redistribution of
> any modifications under the same conditions of the GPL.

That is not any form of usage. That is a statement about ownership, 

That goes against copyright. You do not own the modifications made by 
the other person, even if the next thing you say applies:

> This exactly has happened many times. If you take a proprietary piece
> of software, make changes and resell it (breaking the license agreement
> you received it under), then the original authors are perfectly in
> their rights to take away their livelihood, and this is no different
> with an GPL'd piece of software.

This is because that other person would not have the rights to the 
original material. It is not because the original authors would not have 
the rights to the new material.

In copyright you do not obtain rights to modified products, or at least, 
not to the work that is itself a substantial work eligible for 
copyright. So perhaps you may take away the right of another person to 
distribute; in that sense it is the same. However, in this case you do 
so because that other person is not willing to hand over ownership to 

> Nobody is saying any amount of work is owed. But, any work done, is
> subject to the restrictions of the license.

That is exactly the same thing. It comes down to ownership, dressed up 
in a different jacket.

So there is not actually anything owed, you say, so then how can you (or 
anyone else) say there is to be a form of payment. And if there is not 
to be a form of payment, how can you then take possession of derivative 

There either has to be payment, or there has not to be payment. You 
cannot let payment in actual honesty depend on whether this "asset" ever 

Some people do this thing, but they do this in complete liberty. They 
say "Pay me when you have the money" but that is dependent on trust and 
agreement and love, not on any contract. You cannot codify love, and 
willingness, in a contract.

If people in GPL software actively started saying to people "You can use 
this software, but pay me back if you ever do any work on it" that would 
sound very weird. Suddenly a form of payment is required, this was never 
stated before.

So what do you really want? And if you are clear on that, why not become 
clear on it in writing as well? State your intentions expressly, so that 
people can also object to it if they want.

If you started saying that, people might think twice before actually 
using it or becoming a user of it. Suddenly it would become clear that 
the product is not free, that the person is now actually in debt, and 
that that person might at some later time be held hostage to it if he or 
she does decide to do anything creative with the work.

Stated in such clear terms, you would receive a backlash, and if you had 
done so from the beginning, people would probably have started looking 
from alternatives. Such a new thought yes, I know. You would not have 
gained the traction you have now, if you actually had been clear about 
everything, and not just in technical license terms, but also in human 
terms and what it really means.

>> 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.
> No it's like saying you sign a deal with a record company, who take
> copyright of the songs you produce while there. After leaving the
> record company, you can't re-publish your old songs, or use them as the
> basis to create new songs, without permission from the record company.
> That is how copyright works.

But now the record company is the original GPL project, and there are 
(in the GPL world) no none-GPL projects. And you have an agreement with 
every copyright holder for the lifetime of your person, or the lifetime 
of that project, which is really the same thing.

So "while being there" is for all eternity as long as you keep using 
that software.

>> So what is saying you should have access to the work? Well everyone
>> has, doesn't it?
> No, anyone who has accepted the license and chosen to download the code
> has access. It's not like you picked up an apple from a public tree.

Incorrect. Everyone has access. There does need to be done anything 
difficult, strange, or license-accepting in order to download a copy. 
That is an illusion you make up. You can download a copy without 
accepting the license.

>> 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.
> Just because you have copyright on your own changes, doesn't change the
> conditions of the original work, unless it is such a small amount that
> it can be considered fair use.

But those conditions basically indicate that your changes now also 
belong to them. So it is not /about/ the copyright on the original work, 
it is about the copyright on /your/ work. What I said meant that I might 
consider an "added module" in that sense, and a judge might perhaps too, 
not a derivative, and even if it was judged a derivative, that would not 
be the sole basis for a judgement on this issue, because this GPL you so 
adore, apparently, also lays claim on the copyright of the other person. 
That is why it sets itself apart from regular copyright law (or claims).

In regular copyright there might not be an issue of what belongs to 
whom, there is only an issue of whether the original rightsholder's 
rights are being threatened or not. And even though a modifier might not 
be allowed to distribute, that would be the end of it. In this case, the 
originator tells him not to distribute, says she cannot distribute, 
unless the copyright of the new work, basically transfers to him. That 
cannot necessarily be judged as a threatening of interests, just because 
that original rightsholder says so. Add to that the fact mentioned 
previously, which is that in actual effect, ALL of these works are 
available to ANYONE even though you SAY they have "accepted" a license, 
but you never talked to them, you never really explained what you want 
from them, and it might be considered that the practical matter of the 
issue is simply that given the availability of the software, and the 
broad use of it, the license will be described void and the actual 
interests of the parties will be considered.

And at that point it might be considered that expecting ownership of new 
derivatives might not be considered fair or acceptable business practice 
-- yes, we are talking business here -- and that given the broad 
availability of the software, and its source, the originator is not 
really harmed if some "addon" product is devised and distributed or even 
sold, because it does not really take away, in this case, from the 
business opportunities of the original product, seeing as that they are 
not actually harmed with some addon product, and what you want could 
just be considered petty or petulance, since you really, in the face of 
it, don't actually have a real interest you are defending.

This is why I said that "the GPL should be enforced" does not really 
follow, because the GPL will not be the only thing that will be 

>> and pretty much "public domain or at least common infrastructure"
> So, anybody that has access to Windows or Word, they are allowed to
> redistribute and resell copies? Because, they must be also be big
> enough to be considered "public domain or at least common
> infrastructure".

For the most part we have not been talking about the redistribution and 
reselling of copies here. The topic at hand was some add-on product, or 
a set of patches, or a derivative, that does not replace the original 
product. You are meaning to imply that, in the face of not being 
replaced, the "derivative product" clause will be equally as strong, and 
also, currently, that what you say here applies to the discussion we've 

So no, I do not say here that Windows or Word should be allowed to be 
redistributed, because I never said or thought that GRsecurity would be 
a supplier of kernels, supplanting the original supplier.

Nevertheless I have often thought that way of Windows, and have often 
acted that way in practice, because for teenagers without money, you 
really can't acquire the product in any other way. But presently, this 
is just about addon products I want to say, and Windows actually cannot 
determine whether I will release a product for the Microsoft Operating 
System in that sense here now, and that should answer your questions 
about roads and cars, at least the way I see it.

Under the US constitution by the way, everyone has the right to use 
roads, and you only lose that right by signing an agreement on your 
driver's license, in actual form, because at that point you enter into 
agreement with the state. Without that agreement, you really had access 
all along. In practice, maybe not. But in theory; yes, this is the law.

Also in general terms those requirements for being on the road do not 
come down to the cars becoming property of the road owner. And that was 
really the comparison. So yes, the road owner may say stuff about the 
cars that run on it, but he or she may not say that the cars now belong 
to them.

And I think that is a critical distinction. That difference, those 
claims that the GPL makes, would not be accepted by anyone.

It is not about being allowed to make requirements. It is about the kind 
of requirements you would be allowed to make.

And also for what reason you would make them. That are two things that 
set it apart from regular copyright law.

> Ultimately, it seems to me, that most of your arguments are arguing
> that copyright law doesn't exist at all. It seems you are telling me,
> that as soon as someone has a copy of any piece of work, they are free
> to do anything they want with it.

Doesn't that seem to be the premise of the GPL? And then follows a 
clause or conditions: provided these and these conditions are met.

You might say that those clauses are secondary in nature, since they are 
not the first thing being said. They are not most prominent. It is about 
the modification that everyone is allowed to do.

Just showing you the nature here. We were only talking about GPL here.

You claim that GPL is something that inherently restricts, I say that 
GPL is something that inherently allows.... first , and then only 

Copyright is foremost an instrument that says what you can do with an 
"informational" piece of work, something that embodies content that can 
be described as information. It also has to be an artistic expression of 
some kind, something that is not generic and adds to the pool of 
available artistic creations or expressions.

Something that adds value on its own right in a not-pure-functional 

Then the first issue is redistribution of a work, but GRsecurity is not 
doing that.

So what is your point really?

Why do we have to go into the broad scope of whatever copyright covers, 
when GRsecurity is only doing something that covers redistribution of a 
derivative? He is not doing something that may be considered copying an 
original, as you may have mentioned above.

"Anything you want" only covers "modifying a copy" and "redistributing a 
diffset" here. GPL explicitly seems to grant the first without any 
restrictions whatsoever. So the only topic at hand is the second.

And in law we have see, I guess, that judges have decided in various 
fashions about the topic of derivatives. Sometimes a derivative was 
judged to be in violation, sometimes not.

The law article that was cited said in conclusion, of one of its first 

"Whether or not we find the analysis in these cases convincing, they do 
teach us something about the way courts address the derivative work 
problem. First, a derivative work must incorporate the protected work in 
some manner. Second, courts will fall back on an analysis that resembles 
the substantial similarity test familiar under the reproduction right. 
Third, courts will look at the market impact of the work. An alleged 
derivative that does not supplant demand for the original is less likely 
to be held infringing. Finally, courts recognize that an over-broad 
definition of derivative work may chill the market for follow-on 

The topic falls back, at present, to the issue of derivative works, that 
do not even distribute the full product, but are only a shallow subset 
of it as it were, a diffset that only contains modifications based on 
it. So they are derivative works that do not "incorporate the protected 
work in some manner". This is a very small subset of what copyright 
ultimately covers.

Then the question is mostly about whether this derivative work, if it is 
considered to be one, by a court, would really compel a judge to 
consider that the license statements made by GPL, are actually worth 
considering considering that the market impact of the derivative work 
might in fact be positive, and not negative, for the originating work.

So it might be considered that the claims you make would actually follow 
to be irrelevant because they do not cover any real interests you have, 
that can be recognised by a third party. This covers the "chilling 
effect on the market". If anything anyone should do, would need to be 
transfered to the originating author, that would indeed have a "chilling 
effect on the market", precisely what the GPL in essence seeks to avoid.

This is why the author concludes with:

"To effectuate the goals of the free software movement, the drafters of 
the GPL urge a generally expansive definition of derivative work. The 
great irony is, of course, that such an expansive definition would have 
second order consequences that are exactly counter to the goals of the 
proponents of Free Software. A broad definition of derivative would give 
code authors /less/ freedom to create software that they can truly call 
their own and do with as they please."

> So, what do you think copyright is?

Please don't embarrass me like this. You know full well that I know what 
copyright is. You don't really have to be like that here.

> p.s. Please try to keep replies a little shorter than an entire essay,
> otherwise they will never be read in full.

Some things can never be explained in short sentences, when you and I 
disagree on perspective, but your perspective is already covered in 
countless documents. So for you it is very easy to be short, you only 
have to ask questions or make short statements. For me that is not 
possible, because I am really saying something new here, that might not 
have been covered before.

So what can I do? But sometimes I am ill, and I write longer than I 
want. I apologize for that.


More information about the Ubuntu-devel-discuss mailing list