A successful defense against a copyright troll
At the 2018 Legal and Licensing Workshop (LLW), which is a yearly gathering of lawyers and technical folks organized by the Free Software Foundation Europe (FSFE), attendees got more details on a recent hearing in a German GPL enforcement case. Marcus von Welser is a lawyer who represented the defendant, Geniatech, in a case that was brought by Patrick McHardy. In the presentation, von Welser was joined by Armijn Hemel, who helped Geniatech in its compliance efforts. The hearing was of interest for a number of reasons, not least because McHardy withdrew his request for an injunction once it became clear that the judge was leaning in favor of the defendants—effectively stopping this case dead in its tracks.
Copyright trolling
![Marcus von Welser [Marcus von Welser]](https://static.lwn.net/images/2018/llw-von-welser-sm.jpg)
Von Welser began by noting some of the reasons that Germany is an attractive venue for copyright trolls. It is inexpensive to file cases and the process is relatively quick. The practice of the courts in Germany is to issue an interim injunction without an oral hearing. In order to avoid that interim injunction, defendants must sign a cease-and-desist declaration that then subjects them to contractual fines. There is also a lack of transparency in the German court system, which allows trolls to operate largely anonymously.
The tactics used by McHardy are to first notify the company of the GPL violation and ask for a cease-and-desist declaration that would subject the company to a flexible contractual fine. If that is signed, it is followed up with another letter, pointing to a different GPL violation, that asks for a second cease-and-desist declaration with a fixed contractual fine. After that is signed, further violations are alleged, each of which comes with a request for the fines (which can be a five-figure value per infraction), along with further cease-and-desist declarations with even higher fines.
A timeline of the McHardy versus Geniatech case was then presented. The company received a letter from McHardy on July 17, 2017 requesting the first cease-and-desist declaration in order to avoid an interim injunction. Instead of signing it, the company replied a week later questioning some of McHardy's claims, which appeared to anger him. On July 26, he replied by terminating Geniatech's GPLv2 license on McHardy's netfilter contributions in the kernel. Hemel noted that there were two elements here that were quite different from most of McHardy's cases: the defendant did not sign the first warning and McHardy explicitly terminated his license to the code for the defendant.
In his filings, McHardy referred to the Linux kernel in general, not to a specific version or one that was tied to the firmware in Geniatech's product. In its reply to his first letter, Geniatech raised questions about McHardy's contributions to the kernel, noting that his contributions were not integral to the kernel itself. The regional court in Cologne did not understand, however, and granted an interim injunction on August 23 that stopped Geniatech from distributing (or even hyperlinking to) any version of the kernel.
Arguments
That injunction was appealed and there was an oral hearing at the higher regional court of Cologne on March 7, 2018, where McHardy eventually withdrew his application for an injunction. The arguments that Geniatech made at that hearing seemed to resonate with the judge, according to von Welser and Hemel.
McHardy claimed joint authorship in the kernel, which gave him the right to enforce the copyright of the whole, but Geniatech argued that McHardy's contributions were simply adaptations. The kernel was authored by Linus Torvalds and released under the GPLv2, which explicitly grants the right to modify the work. Modifying the work under that clause does not give rise to joint authorship, but were instead adaptations under German law. If it is not a work of joint authorship, McHardy could only enforce the copyrights of his changes, not those of the whole Linux kernel.
Beyond that, McHardy's arguments that his modifications are protected by copyright and were infringed by the defendant were not proven. He alleged that he contributed 50,000 lines of code over the years and provided a CD with the changelogs corresponding to those changes, but did not show that those modifications fulfilled the requirements for copyright protection. In addition, he did not show what parts of the modifications he made were actually used by the defendant.
![Armijn Hemel [Armijn Hemel]](https://static.lwn.net/images/2018/llw-hemel-sm.jpg)
Another argument that Geniatech used was that McHardy was abusing his rights by approaching multiple companies with the intent of gaining monetarily. This hurts more than just the companies involved; it harms the entire Linux ecosystem by making companies insecure in their use of the operating system, it argued. Hemel said that the judge had heard from colleagues about other companies that had suffered at the hands of McHardy, which helped this argument.
The final element of the abuse-of-rights argument concerned the revocation of the GPL terms for Geniatech by McHardy. This was done to raise pressure on the defendant. In his efforts to enforce the GPL, McHardy is not following community norms, specifically "The Principles of Community-Oriented GPL Enforcement", which prioritizes compliance and not financial gain. He is also not entitled to terminate his GPL grant to Geniatech as it does not comply with German law or the GPLv2 itself, von Welser said.
The final argument, which von Welser thought was "quite striking", was that that the scope of the interim injunction was too broad and not well defined. It enjoined Geniatech from distributing the Linux kernel, not some specific version or versions of the Geniatech firmware. There are more than 100 officially released versions of the kernel that do not include any contributions from McHardy, von Welser said. By ordering Geniatech not to distribute any version of the kernel, the court was covering kernels that were not even part of the dispute with McHardy.
After those arguments had been presented, McHardy and his lawyer discussed it over the telephone (McHardy was not present, according to Harald Welte's account that was linked above). At that point, the request to continue the injunction was withdrawn. McHardy is required to cover the costs of the proceedings, including court fees and lawyer costs for both sides.
Recommendations
The recommendations for companies contacted by McHardy "are quite obvious", von Welser said. They should refuse to sign the cease-and-desist declaration. They should then ensure GPL compliance for their products and prepare a legal defense strategy. That may be easier to do now, since there is more information available from this case. Those that sign the declaration generally do so in the hopes of avoiding a conflict altogether, but it is important to recognize that it likely will not succeed in doing that. Hemel added that if companies do end up signing something with McHardy, they should try to get rid of the confidentiality clause that he demands; shining more light on his tactics will help others defend against them.
In answer to a question from the audience, von Welser said that since there is no judgment, there is nothing to translate to English for others to study. He also noted that it sets no precedent since there was no ruling, but he thinks the arguments make logical sense and will be successful elsewhere.
James Bottomley noted that it was good to have fended off McHardy in this case, but wondered how legitimate (i.e. community-oriented) enforcement would be able to proceed in Germany. Von Welser said that McHardy was trying to make wide claims to make things simpler for himself. If he had showed which parts of the kernel he modified, demonstrated that these modifications are copyrightable (which is not difficult in Germany), and showed how those modifications were used by the defendant, he might have found a different reception by the court. Bottomley said that the goal should be to stop trolls like McHardy, but to still allow legitimate GPL enforcement for the kernel. Von Welser indicated that should still be possible.
[I would like to thank the LLW Platinum sponsors,
Intel, the Linux Foundation, and Red Hat, for their travel assistance
support to Barcelona for the conference.]
Index entries for this article | |
---|---|
Conference | Free Software Legal & Licensing Workshop/2018 |
Posted Apr 24, 2018 0:52 UTC (Tue)
by pturmel (guest, #95781)
[Link] (4 responses)
Posted Apr 24, 2018 5:48 UTC (Tue)
by epa (subscriber, #39769)
[Link] (3 responses)
Posted Apr 24, 2018 11:54 UTC (Tue)
by khim (subscriber, #9252)
[Link] (2 responses)
That was outlined before. No details, but you can guess: chinese supplier which does not care about GPL compliance at all sends goods, German company sued... only this time German company was daughter company of that same supplier... which meant it was just matter of asking mother company to publish source for kernel. 9 times out of 10 (if not 99 times out of 100) chinese companies don't comply with Linux GPL license not because they maliciously want to hide something, but because they just don't care to comply. It's cheaper for them that way: do nothing in the beginning, dig out sources when caught. Since many products would never be sold wide enough to catch attention of GPL enforcers... that saves them money!
Posted Apr 24, 2018 12:46 UTC (Tue)
by armijn (subscriber, #3653)
[Link]
What we did on the engineering part in this case was actually a bit more than just publishing the source. We made sure that the source code could be rebuilt and that the built kernel matched the one in the firmware. We also provided decent build instructions.
Posted Apr 24, 2018 12:59 UTC (Tue)
by aggelos (subscriber, #41752)
[Link]
Almost everything in the article has been reported before, in considerable detail (see e.g. Harald Welte's blog post linked to in the article). The main contribution of this presentation (and its reproduction here) seems to be the one-sided concern for the plight of commercial-scale copyleft violators. I find it telling that it was James Bottomley who (according to the coverage) introduced concerns about actually enforcing the GPL (as, IIRC, he has in the past been vocal in advocating in favor of minimal (if indeed any at all) GPL enforcement).
Posted Apr 24, 2018 9:56 UTC (Tue)
by jwilk (subscriber, #63328)
[Link] (1 responses)
Does he mean very old versions that were released before McHardy started contribution (i.e. < v2.6.13 AFAICS), or something else?
Posted Apr 24, 2018 12:42 UTC (Tue)
by armijn (subscriber, #3653)
[Link]
Posted Apr 24, 2018 11:59 UTC (Tue)
by bengen (guest, #14957)
[Link] (1 responses)
As I understand it, the copyright trolling business case here is built upon offering the defendant/victim a (costly) way to keep things out of court while threatening to seek an injunction that would cause greater damage to the defendant's business. Of course, this is non-transparent because thus far no court has been involved. Is there anything *within* the court system that ought to be more transparent?
Posted Apr 24, 2018 12:51 UTC (Tue)
by armijn (subscriber, #3653)
[Link]
Posted Apr 24, 2018 17:52 UTC (Tue)
by Cyberax (✭ supporter ✭, #52523)
[Link] (19 responses)
Wonderful.
Posted Apr 24, 2018 17:59 UTC (Tue)
by armijn (subscriber, #3653)
[Link] (14 responses)
Posted Apr 24, 2018 18:23 UTC (Tue)
by Cyberax (✭ supporter ✭, #52523)
[Link] (13 responses)
> Beyond that, McHardy's arguments that his modifications are protected by copyright and were infringed by the defendant were not proven. He alleged that he contributed 50,000 lines of code over the years and provided a CD with the changelogs corresponding to those changes, but did not show that those modifications fulfilled the requirements for copyright protection.
means that pretty much nobody can enforce the copyright on Linux contributions.
Posted Apr 24, 2018 18:28 UTC (Tue)
by armijn (subscriber, #3653)
[Link] (11 responses)
Posted Apr 24, 2018 18:47 UTC (Tue)
by mdolan (subscriber, #104340)
[Link] (10 responses)
Posted Apr 24, 2018 18:58 UTC (Tue)
by armijn (subscriber, #3653)
[Link]
This had indeed nothing to do with compliance, as Geniatech wanted to be compliant and put in the effort to become compliant as well. Compliance is absolutely necessary in case you want to fight in court. Also, if a company doesn't want to come into compliance I don't help them.
Posted Apr 28, 2018 16:27 UTC (Sat)
by paulj (subscriber, #341)
[Link] (8 responses)
When are _you_ going into compliance with the GPL?
Posted Apr 29, 2018 20:17 UTC (Sun)
by armijn (subscriber, #3653)
[Link] (7 responses)
Posted Apr 29, 2018 20:28 UTC (Sun)
by paulj (subscriber, #341)
[Link]
Over to Mike.
Posted Apr 30, 2018 19:54 UTC (Mon)
by paulj (subscriber, #341)
[Link]
Posted May 4, 2018 7:50 UTC (Fri)
by paulj (subscriber, #341)
[Link] (4 responses)
The question remains: If a work A is made to directly and explicitly use and rely upon abstractions and functionality provided by a GPL work; e.g. by work A including header files of the GPL work, so as to make function calls into the GPL work to avail of its functionality, and to register callbacks to conform to abstractions so that the GPL work can orchestrate functionality on behalf of work A, along with other technical mechanisms that cause work A to explicitly depend on the GPL work for critical functionality; would the Linux Foundation consider work A to be a derived work of the GPL work?
Strange that it's difficult for the Linux Foundation to give a public answer to that...
Posted May 7, 2018 22:42 UTC (Mon)
by jospoortvliet (guest, #33164)
[Link]
Posted May 8, 2018 9:12 UTC (Tue)
by lyda (subscriber, #7429)
[Link]
Posted May 16, 2018 3:18 UTC (Wed)
by fest3er (guest, #60379)
[Link]
Perhaps the question that should be asked is, "Are included header files (that don't contain 'code') non-factual code? Or are they merely facts that do nothing by themselves?" If the latter, then there should be no questions and no issues. (Refer to the timezone data debacle, where some court clearly determined that facts cannot be copyrighted.) When something cannot be copyrighted, there can be no one who can grant or deny license to use it.
Posted Jun 2, 2018 12:03 UTC (Sat)
by paulj (subscriber, #341)
[Link]
Background: The Linux Foundation are hosting a project who vehemently claim that source code that is intimately built on GPL code (like work A) is /in no way/ subject to the GPL licence. This is a deliberate, long-running campaign to dilute and undermine the GPL licence (on at least the code-base concerned, which includes code of mine), backed by a number of corporate bodies.
The Linux Foundation appears to be A-OK with this.
Posted Apr 24, 2018 19:52 UTC (Tue)
by rgmoore (✭ supporter ✭, #75)
[Link]
The critical point you are missing is that "were not proven" is not the same as "cannot be proven". As the article says:
So somebody trying to enforce their copyrights in the kernel can do it; they just can't cut corners with their legal case.
Posted Apr 24, 2018 22:42 UTC (Tue)
by mageta (subscriber, #89696)
[Link]
Posted Apr 28, 2018 14:24 UTC (Sat)
by auerswal (subscriber, #119876)
[Link] (2 responses)
IANAL etc.
Posted Apr 29, 2018 20:14 UTC (Sun)
by armijn (subscriber, #3653)
[Link]
Posted May 3, 2018 10:42 UTC (Thu)
by Wol (subscriber, #4433)
[Link]
McHardy *failed* *to* *show* that he actually held *any* copyrights. If you can't prove that the plaintiff has copyrights, you can't punish the defendant for violating them. Simples!
Cheers,
Posted Apr 26, 2018 2:29 UTC (Thu)
by marcH (subscriber, #57642)
[Link] (7 responses)
Well, the German legal system is better than the US one for at least that one thing
https://en.wikipedia.org/wiki/American_rule_(attorney%27s_fees)
Posted Apr 26, 2018 8:46 UTC (Thu)
by jwilk (subscriber, #63328)
[Link] (6 responses)
Posted Apr 29, 2018 22:21 UTC (Sun)
by giraffedata (guest, #1954)
[Link] (5 responses)
The US stands alone among major countries in requiring parties (usually) to pay their own costs of prosecution or defense.
But loser-pays isn't all that great either. When you look at someone exploiting the legal system to gain money you don't think he's really entitled to by filing or defending a lawsuit he probably won't win, it sounds great. When you look at a person who is not rich being damaged by a rich person and not having access to the courts because there's a small chance the court won't see it his way and the person will have to pay the exorbitant legal costs the other party chose to incur and courts consider reasonable, well, it doesn't sound so great.
I have another idea: if you lose, you pay the other side's costs up to the amount you paid for your own side. That way, if I decide to go pro se (not pay a lawyer), I don't have to finance the other side's hiring of an expert trial lawyer to beat me to a pulp.
(The US does, incidentally, have exceptions for vexatious litigants - people who file lawsuits they know they won't win just to make the other side incur legal costs).
Posted Apr 29, 2018 23:32 UTC (Sun)
by marcH (subscriber, #57642)
[Link] (3 responses)
So you're considering the case of rich guy hiring expensive lawyers to sue poor guy on questionable grounds to extract money from poor guy. That doesn't really sound like the common case (that's an understatement) Let's also assume rich guy wins. Even then the courts would have the decency not to ask poor guy to pay in full the legal bills of rich guy if they're extravagant.
> The US stands alone among major countries in requiring parties (usually) to pay their own costs of prosecution or defense.
.. and I can't help but suspect the real cause is US lawyers keeping the US a "nation of lawyers" - as seen not just in the number of lawyers per capita but also in the tech industry which is made of (legal) immigrants for the most part.
> I have another idea: if you lose, you pay the other side's costs up to the amount you paid for your own side.
OK, even better but which fix stands the better chance to be actually implemented some day?
Posted Apr 30, 2018 0:03 UTC (Mon)
by giraffedata (guest, #1954)
[Link] (2 responses)
No, I'm thinking of the case of a rich guy damaging a poor guy and refusing to pay for the damage without being sued, then hiring expensive lawyers to defend that lawsuit based on a legitimate possibility that the rich guy doesn't actually owe the money.
If you have trouble picturing this, make the rich guy a large corporation and the poor guy a consumer of the corporation's products.
In this case, the poor guy can't really sue even if he's most likely owed the money, because even if he can afford a bare-bones prosecution, he can't afford to risk losing and having to pay for the gold-plated (but legally reasonable) defense.
I can see that increasing access to the courts for gray-area cases does expand the need for lawyers, but it also seems that loser-pays could be in the attorneys' best interest. What pitch would get a plaintiff's attorney more business? 1) You have a good case, but note that even if we win, you'll only get 2/3 of your money back because 1/3 will go to me; or 2) you have a good case, and it won't cost you a dime because when we win, the other guy will pay me.
Posted Apr 30, 2018 7:48 UTC (Mon)
by mpr22 (subscriber, #60784)
[Link] (1 responses)
For many such cases, the sum at issue will be less than ten thousand pounds sterling, and thus in the UK would be handled via the small claims procedure, where costs are not awarded and it's standard practice for private plaintiffs to appear pro se. For cases above ten thousand pounds, I would be inclined to take it to the court of public opinion before taking it to a court of law anyway.
Posted Apr 30, 2018 9:34 UTC (Mon)
by farnz (subscriber, #17727)
[Link]
Also in the UK, the rule of thumb (consult a real lawyer, not a software developer if you want the gory details) is that you must first make an offer to settle the case before you are eligible for a costs award; costs are then limited to "reasonable"[1] costs incurred after the settlement offer was rejected, and are only normally awarded if the court makes a more generous award than the settlement offer represented. Further, we permit so-called "conditional fee arrangements" (CFAs), where the lawyers agree a nominal fee to represent you, but get their fees paid if they win (from a costs award, if there is one, or from your award if there is no costs award). CFAs have limits (generally, that they cannot recover from your award if there is a costs award, and cannot recover more than a fixed percentage of any award), so that you are never in the situation where the lawyers are the only winners.
This, in turn, means that our rules of civil procedure require both parties to start out by making settlement offers; fail to do so, and all costs incurred before that point are not recoverable. For the case given, of a rich guy hiring lawyers on the chance of a surprise success, you hit two issues:
Also, don't forget that a single case in civil court is not a straight win/lose proposition. Let's say (for the sake of argument) that the case is about an insecure device leaking information; the plaintiff (poor guy) wants a full refund on the device (not fit for purpose), and compensation for the data leak (GDPR, soon). You can make a settlement offer that consists of "full refund, legal fees to this point, plus a written apology for the data leak", and thus block off any arguments over the data leak part of the award; if the court awards a full refund for the product (the bit that's not grey), then legal fees cannot be awarded to rich guy - he could have settled at the point the settlement was offered, and chose to litigate anyway.
[1] Reasonable is defined in terms of market rates, and in terms of the reasons given for rejection; in general, you are expected to bring all your reasons for rejecting a given settlement offer before the judge, as they should all be grey areas that need the court to rule on the case, and not things that you already know. The fewer reasons you litigate in court having rejected a settlement, the higher the definition of "reasonable" on the part of your opponent's lawyers.
Posted Apr 30, 2018 4:33 UTC (Mon)
by Cyberax (✭ supporter ✭, #52523)
[Link]
A successful defense against a copyright troll
A successful defense against a copyright troll
A successful defense against a copyright troll
A successful defense against a copyright troll
A successful defense against a copyright troll
That was outlined before.
A successful defense against a copyright troll
A successful defense against a copyright troll
transparency
transparency
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
Von Welser said that McHardy was trying to make wide claims to make things simpler for himself. If he had showed which parts of the kernel he modified, demonstrated that these modifications are copyrightable (which is not difficult in Germany), and showed how those modifications were used by the defendant, he might have found a different reception by the court.
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
TLDR; GPL is the new BSD
Wol
A successful defense against a copyright troll
https://www.clements.com/sites/default/files/resources/Th...
A successful defense against a copyright troll
A successful defense against a copyright troll - loser pays
A successful defense against a copyright troll - loser pays
A successful defense against a copyright troll - loser pays
So you're considering the case of rich guy hiring expensive lawyers to sue poor guy on questionable grounds
suspect the real cause is US lawyers keeping the US a "nation of lawyers"
A successful defense against a copyright troll - loser pays
A successful defense against a copyright troll - loser pays
A successful defense against a copyright troll - loser pays