Understanding Private vs Public · Sunday July 30, 2006 by Crosbie Fitch
John Cowan said 6751 days ago :
I think you seriously overstate the case against the GPLv3, which does not by any means require the publication of private source code, and is in fact much more careful not to do so than many free software licenses.
It merely permits someone else to impose such a restriction on modified code without making it incompatible with the GPLv3. So if Alice writes a web application with a “Push here for source code” button, and requires that this button be functional in all derivative works (which is a requirement the GPLv3 does not make), Bob can safely incorporate Alice’s work into his larger work published under the GPLv3, and Alice’s requirement is imposed on him and anyone who wishes to modify his work.
But if Bob doesn’t use Alice’s work, he need have no such button.
Crosbie Fitch said 6751 days ago :
Unlike GPLv2, GPLv3 is permitting a new class of software that has fewer freedoms. This new class of software seeks to enforce a restriction even on private derivatives.
Just that fact alone, that the GPLv3 would encroach into the private domain, is a dismaying development.
The idea of the GPL is to preserve freedom in software and in PUBLICATION of derivatives, not simply to give people the choice as to whether to do so (as the BSD does).
To even permit some users of the GPL to add clauses that restrict private derivatives is a departure from the principles in GPLv2.
I really do think the Affero license and the GPL should remain distinct as the former seeks to prevent private use or derivation, i.e. it obliges continuous publication.
Obviously license harmonisation leads to a simpler life, but if you take it too far you’ll end up with the BSD.
Affero and GPLv2 are incompatible in their respect for privacy. It is a mistake for GPLv3 to attempt to become Affero compatible.
dave crossland said 6653 days ago :
For me, GPLv3 represents fixing the ways that GPLv2 has become broken over time. Proprietary interests have worked out several ways to break the spirit of the GPL without the exact legal terms, and fixing it is very important.
One of the most basic ways that proprietary interests have broken the General Public License, that defends the public’s rights to freedom in the software they use, is to use networks to evade the distribution clauses of GPLv2.
A developer can receive GPLv2 software, modify and improve it with new powerful features, and allow the public to use their modified version, without offering the corresponding source code. How can this be? They hide in the shadow of “private” modification, despite conveying their software the public, because the software interacts with its users through a network, but runs on their private computers. This is widely known as the ‘Application Service Provider loophole’.
In my opinion it is unethical and unsustainable, just like proprietary software would be on my own computer.
In 2002 this issue first surfaced with the publication of the Affero GPL.
—8<—Date: Fri, 15 Mar 2002 09:24:38 -0700 (MST)
From: Richard Stallman rms@gnu.org
To: nelson@crynwr.com
CC: bruce@perens.com, brian@collab.net, license-discuss@opensource.org,
moglen@columbia.edu, board@opensource.org, rms@gnu.org
Subject: Re: OSD modification regarding what license can require of user
The reason we’ve decided that this ASP requirement is legitimate is
that it is a matter of requiring making the modified source code
available in a case of public use. It extends existing GPL
requirements coherently to a new scenario of usage.
It would be wrong to require publication of modified versions
that are used privately, but inviting the public to use a server
is not private use.—8<—- http://linuxmafia.com/faq/Licensing_and_Law/gplv3-plans.html
However, the FSF attitude appears to have changed:
“any attempt to interfere with the right to keep those subsequent modifications private, interferes with the right of private execution and private modification, rights that are not to be tampered with.”
– http://fsfeurope.org/projects/gplv3/barcelona-moglen-transcript.en.html(I know this is slightly out of context)
and
“We thought about putting some kind of clause like that into GPL version 3 but we decided that would be too drastic to put it on all GPL covered programs”
– http://fsfeurope.org/projects/gplv3/barcelona-rms-transcript.en.html#affero-clauseI understand that this clause remains a controversial requirement both in the individual and corporate free software communities, but I wonder why it is felt to be too drastic.
To me, the final paragraph of Richard’s email is quite straightforward, while Eben’s comment is in error, because Affero clauses ought not to intefere with the right to keep those subsequent modifications private; they should only intefere with the right to use the software publically in a way which denies the public its rights – just as the GPL has always done.
Unfortunately, the current wording of this section is flawed in this way, because “through a computer network” is too broad. As Eben has said, “the right of private execution and private modification … are not to be tampered with”. Instead it could simply be “through a public computer network”; this would clearly defend truely private modification.
While you also make this distinction, you throw the baby out with the bathwater.
Also, defining clearly what “network session” means is also important. Perhaps this could be “any of the applicable methods in section (6)”, as would ‘network session’ include a bittorrent download? New forms of distribution like Bittorrent have also broken the GPLv2, and are being fixed in GPLv3. No doubt more new forms will emerge in the future.
Perhaps “users” needs defining too – I woudl say, “a user is some one who uses the software” – it must not be able to be broken to mean a person who uses the softare on his local machine, and in some way exclude people who use the software through the network.
Summary: I am sad to see this clause as optional and not included in the main text of GPLv3. It also has some problems but these can be easily fixed.
I am saddened that the FSF no longer feels the ASP requirement is legitimate. I do not understand why its attitude has changed, why it is now felt to be ‘too drastic’. I feel it should be part of the GPLv3 as standard, or at least the reasons for the FSFs change in attitude published.
Though I’d readily accept a private explanation if it was thought to be too political to publicly announce. Might it be “too political” or “too drastic” because it would disenfranchise corporate patrons like Google? I’ve met Free Software developers who have declined to discuss this because of who their current employer is.
Crosbie Fitch said 6653 days ago :
Dave, you say “A developer can receive GPLv2 software, modify and improve it with new powerful features, and allow the public to use their modified version, without offering the corresponding source code.”
However, no software has actually been distributed to the public user. The GPL is not about ensuring users of software get copies of it, but about ensuring that people who do receive copies or derivative copies of software aren’t constrained by law, DRM or obfuscation as to what they can do with it.
In your case, the user has not been given a copy of the software that enables them to provide this interactive service on their own server.
Because they don’t have a copy, the user’s freedom is not being constrained by an artifical economic incentive or commercial privilege (copyright, patent, DMCA). They have a choice to use the service or purchase a copy. The GPL does not require copies to be given away free of charge (not free beer).
The developer is enjoying their freedom to demonstrate, exploit, or sell their private modifications for any price they can get for them. Free software as in liberty, not as in free of charge. If the private mods are so good they’re worth paying for, then buy them. You have no right to demand that the developer surrender their work free of charge, nor to prevent them exploiting the fruits of their labour.
So, if a user accepts an invitation to interact with someone else’s private computing facilities through a terminal, then they accept the constraint that private facilities are being used that remain inaccessible to them. If the user requires a copy of the software behind the facilities then they can purchase a copy, and thanks to the GPL, consequently enjoy complete freedom to exploit them in the same way, or if they really want to, publish that software free of charge.
This has been the same for any GPL program. You can take a GPL program, privately modify it, and use it say to manage accounts for many clients. None of those clients has a right to a copy of your private modifications simply because their receipts and invoices are data input, and the software’s output is given to them as a print-out.
Producing an interactive service that simulates the experience of a locally running application is nevertheless distinct from having a locally running application. Either purchase GPL software for a locally running application, or if you would like to offer a local application simulation service then purchase such a GPL based local application simulation service.
Not having access to someone’s private domain is NOT an imposition on your liberty, it’s simply the boundary of the public domain within which the GPL assures your freedom. The GPL does not grant you access to everyone’s private domain (nor should it).
The reason that SaaS (software as a service) is considered unfair only arises from a reciprocation perspective (which is not the same as emancipation). If you consider you donate your private modifications to the public free of charge on the understanding that everyone else’s private modifications will be donated to the public free of charge in turn, THEN you would indeed have a legitimate grievance with SaaS being in conflict with your expectations. However, this reciprocation is a distinct concept from liberty, it would involve a social contract, a bargain that all people agree to. Copyright is another such social contract, i.e. that we all agree not to make copies, but to buy them or borrow them.
Liberating the public from the social contracts of copyright, patents, and DMCA, is what free software is about. If you want to go further and prohibit public exploitation of private modifications, then you need a distinct license such as the APL or HPL that exploits copyright’s unethical reach into the private domain to prohibit private derivatives (forces them to be published on demand and free of charge).
drew Roberts said 6634 days ago :
“The developer is enjoying their freedom to demonstrate, exploit, or sell their private modifications for any price they can get for them.”
And I think I agree with the developer that is honestly doing this being OK. I have a problem with the developer who has not intention of selling the mods at any price but rather of making his profits from the “demonstrations.”
I am not sure I agree with the exploit word in your above thought.
If someone were to do that to GPL programs that I have released, I would feel that they were exploiting me in ways that I do not wish to be exploited. I would be happy to see this “loophole” plugged.
I am not sure I have thought it through fully, but I see private mods as properly being mods that are used privately by the entity.
How would one write a clause that would allow good faith demonstrations while not allowing the ASP option as the main plan?
“You have no right to demand that the developer surrender their work free of charge, nor to prevent them exploiting the fruits of their labour.”
Ah, yes I do if it based on my code. A legal right in any case, once I release my underlying work with a license that spells things out. If that developer want additional rights, he can negotiate with me. What right does he have to demand more from me for free?
“This has been the same for any GPL program. You can take a GPL program, privately modify it, and use it say to manage accounts for many clients.”
This is not in any way similar to the case I object to. (It is to my mind, a straw man.) Bully for you if you can do this. You are doing the work. leasing out my software for your clients to do the work on is another matter.
Look, I think that using your theory, the GPL could be totally gutted by just having an entity put GPL programs that have been provately modified on their computers, and then leasing those computers to clients with the computers to be installed on the clients premises.
“They have a choice to use the service or purchase a copy. The GPL does not require copies to be given away free of charge (not free beer).”
Wher is it a given that there will be a copy available to purchase? I do not, and rightly so, have to sell anyone a copy of my provately modified GPL programs.
I look forward to this discussion.
all the best,
drew
(+1)/10 to send email…
All private intellectual property is derived from public intellectual property. And yes, even our common DNA is public intellectual property – each person has their own private derivative, which is their own private property until they publish it.
If you privately derive something from a public work, that derivative is your private intellectual property. You are the owner of everything in your private domain. It is wholly yours until you publish it.
I have noticed twice recently, that IP reformists are failing to see the distinction between the private domain and the public domain. It seems that they perceive a problem with intellectual property itself, rather than the inequitous property-like nature of the privilege that copyright grants publishers over public property.
There is no wrong in private entities owning private intellectual property, nor in the public owning public intellectual property. Each is master of their domain. This is the natural order of things.
There are two ways in which this can go wrong:
If enforcement of such controls is attempted, it is effectively a violation of our human rights to privacy and liberty.
In the last few centuries when only a very few publishers were in a position to distribute or reproduce published works, the economic incentivisation experiment that is copyright wasn’t too widely perceived by the public as significantly impinging upon their privacy or liberty. However, now that everyone is in a position to produce copies or derivative works (privately or publicly), the fact that copyright does indeed impact our liberty and privacy is becoming ever more apparent.
Copyright not only violates our freedom in prohibiting us from publishing reproductions or derivatives of published works, it also violates our privacy by prohibiting us from preparing derivatives of published works in the privacy of our own homes.
What’s Could Possibly be Wrong with Intellectual Property?
Some people don’t think the problem is with copyright, but with the very notion of intellectual property itself. These people say that intellectual property is a figment of the imagination, a contradiction in terms.
I don’t believe this is actually the case, the contradiction only occurs because people have attempted to disassociate publication from relinquishing ownership.
The contradiction in terms occurs when a private entity seeks to assert continued ownership of intellectual property they have published. Copyright would appear to convey some aspects of such control by dint of a privileged monopoly, but without copyright, there is no logical basis upon which such control could be expected. You control what you possess, not that which you have given into the possession of another. And publication is delivering your work to the public.
So when people challenge the very concept of intellectual property, what they should be doing instead is challenging the idea that published works can remain the intellectual property of the publisher. Such works are still intellectual property, but naturally, and in the digital age, inescapably public intellectual property.
The fallacy IP maximalists nefariously propagate is to ascribe some distinct quality of ownership over artistic works in the belief that this can be maintained separately from the works themselves. It is as ridiculous as believing that one can piss in the ocean and yet still retain title to each molecule until it is relinquished. The ocean owns your water as much as the public owns your publications. If you would retain control, don’t publish – keep your product private and it remains yours alone.
Intellectual Property Does Exist
So, intellectual property does indeed exist, and it is most sensible for it to exist.
N. Stephan Kinsella in his thesis Against Intellectual Pproperty seems to miss the clear, physical, property-like characteristics of private intellectual property.
He proposes that property must necessarily be tangible, that intangibles such as ideas or representations of them on an easily copied medium, cannot be apprehended and thus cannot be controlled and consequently do not fit the requirements for property.
I disagree. The nature of property is not so much whether it is tangible or intangible, but whether one can have exclusive physical control over it. This clearly applies to tangible works. However, it also applies to intangible works whilst they remain wholly within the creator’s private demesne.
For example:
I may possess a secret, it is mine to control until I publish it. I may confide it to another, and then two of us possess it – shared ownership if you will. It nevertheless is clearly within our control, and none are clearly entitled to wrest it from us. If someone does seek to obtain it, perhaps to find it recorded within a diary, then they are as much a thief of this secret as a burglar would be of a necklace. However, if either of us choose to publish this secret, it patently transfers ownership of it to the public. What conceit could lead anyone to presume the secret remains theirs to control? How could they think that they should continue to dictate who may tell it to whom, or who may print it?
I will concede that a secret, once published and transformed into public property, cannot easily be returned to the state of private property, but that is simply due to the scale of the distribution, the consequent timescale for any reversal, and typically the reluctance of the public to relinquish its property.
Public Property Does Not Oblige Public Use
Public intellectual property is public precisely in order that it may be freely used in private. It belongs to each citizen to do with as they please. They can privately reproduce it, or make private derivations for their own use or amusement. This is because private use does not remove something from the public, it permits our creative skills free rein without public scrutiny or criticism. The choice to make our private work public is part of our human right to privacy. However much the public may be interested to see the work that is done in private, it has no right to it, irrespective of the fact that all private work is based upon public work.
Just as the creator of a private work has no right to control how the public use it once published, so the public also has no right to control how a citizen uses a public work in private.
The public may not restrict private use of its property, nor place obligations upon the private user.
GPL v3 must not be seduced into violation of privacy
GPL v3 is in danger of overreaching its remit, by exploiting copyright’s reach into the private domain.
Copyright prohibits private copies and derivatives, irrespective of how feasible the policing of this may be. How can one artist publish a work and thenceforth have not only control over it in the public domain, but also have control over it in everyone’s private domain? It is an intrinsic violation of privacy. Everyone should be free to create, duplicate, or derive any works of art in their private domain. No-one should have a right to invade or violate another’s privacy in order to police the use of their published work.
Because GPL v3 is based on copyright, and is being steered by some towards an agenda that compels public visibility at all costs, even over an individual’s freedom and privacy, it may take advantage of copyright’s intrinsic privacy violations to accomplish this.
GPL v2 was good in that it didn’t seek to control private use, it merely required that any published copy or derivative work retained the same freedoms as the original. GPL v2 effectively neutered copyright, by not only removing restrictions on free use and derivation, but also ensuring that no-one could re-instate them. GPL v2 emancipated software. This was better than the BSD license, which didn’t prevent re-instatement of copyright on derivatives.
Unfortunately, some people have decided that the GPL should mandate openness, rather than freedom. They see the strength of the GPL not in nullifying copyright, but in obliging visible, collaborative development processes. I’d suggest that visible, collaborative development processes are simply a natural consequence of copyright’s nullification, and do not need to be enforced. There is no further encouragement needed, or restriction to be removed. To actually go so far as to seek to prohibit private use and development is to lose sight of the ‘free’ in free software, and stoop to churlishness.
There are some who believe that it is inherently wrong or unfair for private entities to exploit public intellectual property. I call these people ‘gift economists’, people who believe that GPL software is necessarily created by gifted labour, and is consequently the proprietary property of the development community, and should not be commercially exploited by any private business without authorisation, e.g. purchase of a commercial use license. This is simply the proprietary mentality switched round, and just as bad. Given copyright has been with us so long, it shouldn’t be too surprising that people continue to have difficulty with the idea that they do not have a right to control how other’s use their published intellectual property. So, there are some GPL advocates who believe the goal of the GPL is to permit a public community to possess intellectual property and retain control over how it is used in private.
GPL v3 is in danger of appeasing these advocates by exploiting the restrictions copyright grants even over private use. One of these is in requiring private derivatives to preserve any mechanism that obliges on-demand delivery of source code to any user. How on earth is a restriction on private derivative use increasing the people’s freedom? Especially one that deliberately compromises their privacy.
One argument is that public use of a private derivative is unfair exploitation unless this derivative is published. These are the gift economists talking again, pursuing reciprocity rather than freedom “You can’t use our IP unless you pay in kind”.
Another argument is that public use should be included within the definition of publication. However, this doesn’t logically follow. Just as public use of a machine is not publication of the machine’s designs, so public use of software is not publication of the software (source code or binary derivatives). If the machine was ‘published’ by being reproduced and sold, then each machine could be considered to represent an instance of the design and each new owner should be entitled to it. However, if the machine remains in private hands, then neither it nor its design is being published.
Keep Free Software Free!
The freedom free software should be concerned with is that of the public. This also means respecting their privacy, and their right to freedom within it. The GPL should grant freedoms to the public, and prevent their removal from subsequent publications. It should certainly not impose restrictions on private derivation in order to serve some other agenda that pursues source code visibility or IP equity rather than freedom.
Free software belongs to the public who are free to do anything with it they want in public or private, but may not introduce restrictions on publications of their copies or derivatives (which includes withholding or obscuring the source code).