Elastic promises "open"—delivers proprietary [LWN.net]

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 0:09 UTC (Thu) by kemitchell (subscriber, #124442)
In reply to: Elastic promises "open"—delivers proprietary by pizza
Parent article: Elastic promises "open"—delivers proprietary

I believe you're mixing two disparate concepts.

One is the idea that copyright licenses—setting aside the dubious hard distinction between licenses and contracts—have to regulate conduct covered by copyright. That's true. In the USA, terms license—or permit—others to do what would otherwise be prohibited for anyone but the copyright holder under 17 U.S.C. 106.

To put that more concretely, if you're going to sue someone for copyright infringement, you need to show the court that they did something with your work that involved one of the exclusive rights of copyright owners, without permission. Rights like reproducing, distributing to the public, preparing derivative works, or publicly displaying or performing. This is usually very easy to do with software, since running it usually entails copying, sharing distribution, and building on it preparing derivative works.

The other is the idea of what license terms can say, and what rules they can impose, in return for permission to do what copyright would otherwise prohibit. In other words, what you get in exchange for granting a license. In commercial context, that usually includes some money. But for either open or closed software, it almost always involves some obligations and boundaries, too.

For example, setting copyleft entirely aside, nearly all common open source licenses require passing along notices of license terms with copies of the software. That's not a part of copyright law, or specifically called out as something only copyright owners can choose not to do. It's implemented as a condition to being able to reproduce and distribute to the public.

There are some limits on what you can demand in exchange for copyright permission. The usual prohibitions on illegal terms, terms against public policy, potential competition law violations, and so on. There are also some limits specific to copyright, like the doctrine of "copyright misuse". These come up rarely, and tend to narrow in time. Copyright misuse came about largely by analogy to patent doctrine, but is currently much disfavored by the courts.

I'm afraid I don't follow your point on discrimination and copyleft. If it helps, I'm a practicing California attorney advising primarily software companies and developers. I've written, read, and negotiated countless licenses.

to post comments

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 1:00 UTC (Thu) by pizza (subscriber, #46) [Link] (8 responses)

> For example, setting copyleft entirely aside, nearly all common open source licenses require passing along notices of license terms with copies of the software. That's not a part of copyright law, or specifically called out as something only copyright owners can choose not to do. It's implemented as a condition to being able to reproduce and distribute to the public.

Yes. But the key here is "with copies of the software" -- If you do not make a copy of the software (or create some derivative work) then those conditions/restrictions cannot not apply to you, as all other activities (including running/using the software!) are out of scope.

(and since the ephemeral copies/adaptations made by installing/loading/compiling/etc is "an essential step in the utilization of the computer program" they can't be used as a hook to impose field-of-use restrictions)

> I'm afraid I don't follow your point on discrimination and copyleft. If it helps, I'm a practicing California attorney advising primarily software companies and developers. I've written, read, and negotiated countless licenses.

(Were these licenses you negotiated stand-alone copyright licenses or were they part of larger contracts?)

My point about discrimination is that the way you were saying it should be interpreted is so broad that it it is effectively nonsensical; any restriction at all becomes "discrimination" of some form or another. Especially copyright itself.

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 1:58 UTC (Thu) by kemitchell (subscriber, #124442) [Link] (7 responses)

On license conditions: There are "public licenses" outside the open source canon that impose enforceable restrictions on manner and purpose of use. For example, Creative Commons Non-Commercial. We also see use and territory limitations in other public grants, like unilateral technology pledges. The recent Open COVID Pledge comes to mind. As far as I know, there's no rule of law that says you have to license exclusive rights of copyright holders entire, without limits. You can grant less, and draw the line in your own terms.

I believe your comment on negotiated contracts is going for a distinction between license and contract. I'd invite you to have a second look at that, from primary sources rather than commentaries. It was part of the FSF catechism, especially via Eben Moglen, for many years. Largely, I think, as a kind of rearguard action to try to influence copyright law policy, which came to naught, but also a reflection of the fact that it just hand't come up is US court yet. I'd argue subsequent legal developments went decided the other way, even setting aside countries like France where the distinction never made any sense.

Plaintiffs suing for license violations tend to make both copyright and contract claims. See, for example, all the lawsuits Artifex brought about Ghostscript. And it's unclear how a US court would interpret and apply license terms, other than by applying contract-law rules. There are some legal questions about preemption—legal double dipping—which might prevent winning both a contract claim and an infringement claim for the same conduct. But that matters largely when it comes to what you can get the court to order if you win—legal "remedies"—not whether you have the right to set rules in the first place.

Sounds like we actually agree on "nondiscrimination". I don't argue that the nondiscrimination criteria, as written, should preclude copyleft. Others try.

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 3:00 UTC (Thu) by pizza (subscriber, #46) [Link] (6 responses)

> On license conditions: There are "public licenses" outside the open source canon that impose enforceable restrictions on manner and purpose of use. For example, Creative Commons Non-Commercial.

Of course! But they can still only impose conditions tied to activities that fall under the purview of copyright. That is to say, making copies or derivatives. If you don't engage in those activities (eg by making further copies), then those conditions simply do not apply.

(If I am wrong here, I would appreciate a proper explanation...)

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 4:06 UTC (Thu) by kemitchell (subscriber, #124442) [Link] (5 responses)

I'm not quite sure I know where you're coming from, but as a stab in the dark: Are you familiar with 17 USC 117, the way it differed from the CONTU proposal, the licensed-versus-sold distinction, and Vernor v. Autodesk?

So it doesn't go unsaid: I appreciate your comments. It's clear you've put time and thought into this. And I'm not any kind of guru with all the answers. Just another schmuck who's put too much time and thought in to quit.

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 12:02 UTC (Thu) by pizza (subscriber, #46) [Link] (4 responses)

> Are you familiar with 17 USC 117, the way it differed from the CONTU proposal, the licensed-versus-sold distinction, and Vernor v. Autodesk?

Yes, Yes, Yes, and... somewhat. :D

That unattributed "Rightful Possessor" vs "Owner" change is still being argued over in many venues, with the "we agree this decision sucks, but we're bound by precedent, and it's up to Congress can fix this" punt in Vernor v. Autodesk appeal. Meanwhile, that "licensed vs sold" hand-wavery attempt to avoid transferring "ownership rights" (while simultaneously saying things like "own your copy today" in advertising copy) is one of the reasons laypeople have strong opinions about the legal profession.

But there can still be an effective "transfer of title" that occurs when the copy is handed over, which means the recipient is a "owner" of the copy. (see the UMG vs Augusto case, where the 9th circuit ignored their Vernor precedent. That scenario does seem more directly applicable to F/OSS than Vernor..)

What a mess, eh?

> Just another schmuck who's put too much time and thought in to quit.

Don't sell yourself short; unlike nearly everyone else here, you actually _are_ a lawyer, and have direct experience with this stuff.

So, um, thank you for sticking through this too!

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 15:15 UTC (Thu) by kpfleming (subscriber, #23250) [Link]

I know it's a divergence, but 'while simultaneously saying things like "own your copy today" in advertising copy' is such an important point.

At the same time that the major US content producers leverage the DMCA as a cudgel, and push the RIAA to act on their behalf, those same content producers loudly proclaim that you can 'own' a movie by purchasing a copy. The content delivery networks will even advertise that you can 'own' a movie by purchasing an indefinite-term access token to be able to play the movie from their service. Of course these are both BS, there is no 'ownership' involved in either transaction, and the license being granted in the transaction is revocable at any time and terminable at the will of the content provider/producer with no notice requirement and no recourse available to the licensee.

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 18:52 UTC (Thu) by kemitchell (subscriber, #124442) [Link] (2 responses)

Augusto is the right citation. Thanks for poking me to read it again.

You could cite the case as weakening the conceptual case for the rule from Vernor. But it's also empirical proof positive that the courts will disparate treatment for software. They can and do treat it as a special case. It kind of is a special case.

Particularly with regard to computer software, we have recognized that copyright owners may create licensing arrangements so that users acquire only a license to the particular copy of software and do not acquire title that permits further transfer or sale of that copy without the permission of the copyright owner.

The court mentions that there were more factors to the rule in Vernor than just one. But it doesn't actually run through them systematically.

Go looking, and you can find distinctions it does make that maybe wouldn't go the same way for open software, rather than commercially licensed software. The absence of a "prior arrangement" for delivery of the copy. No attempt to keep track of specific copies. No response indicating agreement, though I'd argue open licensing is still very much "consensual". But the court's just piling on. They have the special statute about what happens when you mail unsolicited merchandise to people.

As I'm told Moglen used to say: "Don't learn copyright law from free software people." It's important to remember that we don't see copyright like other users of the system. Our policy preferences are often the opposite of prevailing and established public policy.

Software might have been covered by its own, special IP statute. We saw that for a time in Japan, for example. But it was put under copyright instead. Some of the established expectations and rules for older media haven't translated over to software. The general trend is making room for industry to monetize rights in software, as it figures out how to do that. Especially in light of how quickly and often distribution and marketing have changed.

If you mail a game critic a boxed copy of your next title and it ends up on eBay, I wouldn't be surprised to see a decision applying Augusto for the same result. But I suspect that would have as much to do with honoring long-held expectations about physical products as with any refinement of the First Sale Doctrine.

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 20:31 UTC (Thu) by Wol (subscriber, #4433) [Link] (1 responses)

> As I'm told Moglen used to say: "Don't learn copyright law from free software people." It's important to remember that we don't see copyright like other users of the system. Our policy preferences are often the opposite of prevailing and established public policy.

May I suggest you take a look at groklaw.net? Sadly, it's a cobweb site now, but a lot of us cut our "lawyerly" teeth there, and the site owner was a paralegal who made us do our homework! (The Library of Congress picked it out explicitly for archiving.)

While I'm not aware of the Nazgul actually contributing to the site, I'm pretty certain they would have been monitoring it for the SCO vs IBM case ...

Cheers,
Wol

Elastic promises "open"—delivers proprietary

Posted Jan 28, 2021 20:34 UTC (Thu) by kemitchell (subscriber, #124442) [Link]

I read groklaw back in the day. One of many influences that lured me off the straight and narrow path of programmer to the dark side...