GitHub Files Court Brief Criticizing 'Vague Infringement Allegations' (github.blog) 24
"One project going dark — due to a DMCA takedown or otherwise — can impact thousands of developers," GitHub warns in a blog post this week:
We saw that firsthand with both leftpad and mimemagic. That's why GitHub's designed its DMCA process to follow the law in requiring takedown requests to identify specific content. We want developers on our platform and elsewhere to have a clear opportunity to remove infringing code yet keep non-infringing code up for others to use, modify, and learn from.
Ensuring that software copyright allegations are specific and actionable benefits the entire developer ecosystem. That's why GitHub submitted a "friend of the court" brief in the SAS Institute, Inc. v. World Programming Ltd. case before a Federal Court of Appeals.
This case is the most recent in a ten-year litigation spanning both the UK and the US. SAS Institute has brought copyright and non-copyright claims against World Programming's software that runs code written in the SAS language, and the copyright claims drew comparison to the recent Google v. Oracle Supreme Court case. But this case is different from Google v. Oracle because here the alleged copyright infringement is based on a claim of "nonliteral" infringement. That means there is no allegation that specific lines of code were literally copied, but only that other aspects, like the code's overall structure and organization, were used. In nonliteral infringement claims, the questions arise: what aspects of the "nonliteral" features were taken and are they actually protected by copyright...?
GitHub believes that for claims involving nonliteral copying of software, it is critical that a copyright owner provide — as early as possible — examples that would allow a developer, a court, or a software collaboration platform like GitHub to identify what was claimed to be copied. Our brief helps educate the court why specificity is especially important for developers.... We urged the court to think about efficiency in dispute resolution to avoid FUD (fear, uncertainty, and doubt). The sooner infringement allegations can be made specific and clear, the sooner infringing code can be changed and non-infringing code can stay up. That should be the result for both federal lawsuits, as well as DMCA infringement notices.
Ensuring that software copyright allegations are specific and actionable benefits the entire developer ecosystem. That's why GitHub submitted a "friend of the court" brief in the SAS Institute, Inc. v. World Programming Ltd. case before a Federal Court of Appeals.
This case is the most recent in a ten-year litigation spanning both the UK and the US. SAS Institute has brought copyright and non-copyright claims against World Programming's software that runs code written in the SAS language, and the copyright claims drew comparison to the recent Google v. Oracle Supreme Court case. But this case is different from Google v. Oracle because here the alleged copyright infringement is based on a claim of "nonliteral" infringement. That means there is no allegation that specific lines of code were literally copied, but only that other aspects, like the code's overall structure and organization, were used. In nonliteral infringement claims, the questions arise: what aspects of the "nonliteral" features were taken and are they actually protected by copyright...?
GitHub believes that for claims involving nonliteral copying of software, it is critical that a copyright owner provide — as early as possible — examples that would allow a developer, a court, or a software collaboration platform like GitHub to identify what was claimed to be copied. Our brief helps educate the court why specificity is especially important for developers.... We urged the court to think about efficiency in dispute resolution to avoid FUD (fear, uncertainty, and doubt). The sooner infringement allegations can be made specific and clear, the sooner infringing code can be changed and non-infringing code can stay up. That should be the result for both federal lawsuits, as well as DMCA infringement notices.
Do Judges really understand the complexity? (Score:5, Interesting)
Most DMCA takedown requests are so vague that they're the equivalent of Ford asking a judge to force GMC to stop making vehicles because they infringe one of their 50K patents.
Re: (Score:3)
The problem is that the law -- and thus reporting forms -- only require identifying the supposedly infringing URL and the original work. Sometimes the target will be something very broad like a site's front page. And the takedown process doesn't allow the target to argue merits. Takedown is automatic, as is reinstating the content, as long as certain procedural steps are taken. This is unfortunately complicated by the additional requirement to have something like a "three strikes" policy to ban repeat o
Re: (Score:2)
Pointing to someone's front page and claiming infringement fails to identify the original work unless the claim is actually for the entire front page.
Re: (Score:2)
That doesn't stop it from happening. In some cases, perhaps there was a story or link on the front page that is the alleged infringement. Lack of actionable specificity is a weakness of the overall scheme.
Re: (Score:2)
In theory, that should void the complaint until they meet requirements. In fact, the kopyrite kops have people running scared everytime someone says boo.
The fundamental problem is that getting sued is just too expensive no matter how right you are in the eyes of the law. The notified host doesn't want to risk it, and their customer probably can't even afford enough legal help to get through discovery if they choose to contest the notice.
Re:Do Judges really understand the complexity? (Score:4, Informative)
Indeed. I'm on the board of an orchestra, and one of the Warner companies sent us a vague DMCA notice that literally says that a few unidentified seconds out of a multi-hour concert infringes their copyright. They failed to identify the original work, the portion of the concert that infringed, or whether the claim is based on a recording copyright or a publishing copyright. And they filed the claim only in third-world countries, ensuring that nobody in their right mind will bother disputing the claim, presumably because that lets them run ads on our recordings or pages for free or something.
F*** all of them. The way the process works gives big media companies the right to steal the hard work and advertising dollars from small artists and ensembles, and this is very much by design. The only way it's going to get better is if we educate the public about what a scam the current copyright regime is and get a huge groundswell of popular support for congresspeople who have some common sense.
Re: (Score:2)
Where did that come from? That's not in DMCA. Is this just a GitHub policy, or something else?
Re: (Score:2)
Fuller’s work was not only taken down, he was also given a strike under YouTube’s “Three Strikes and You’re Out” program.
Re: (Score:2)
It came from the DMCA. Title 17 of the U. S. Code, section 512, paragraph (i)(1)(A). A provider only enjoys any protection under the Act if that provider:
In practice, counter-notic
Re: Do Judges really understand the complexity? (Score:2)
They don't understand the complexity, that's why this case is being brought up in the first place. And this has huge ramifications for languages that require definition order for compilation and it won't be long before we see claims like "int b before char a is just like our int c before int b"! This is just dumb but maybe it'll finally end the debate of curly braces on new lines vs in line.
Should be a significant penalty (Score:3)
for filing "false" allegations.
If there isn't, why not start filing millions of takedown requests for certain corporations IP.
Re: (Score:2)
The penalty is that if you're found to not have a good-faith belief that the work infringes the work you named (and courts have ruled that "good-faith belief that it constitutes unauthorized use of your work" must allow for fair use) then you're fully liable for any damages resulting from sending the note, including all associated legal fees.
So the reason it's a terrible idea to carpet bomb Megacorps with fake notices, but that Megacorps are free to do so without concern, is that the penalty is proportio
Re: (Score:2)
"Or, if they host it themselves, your note goes in the shredder because the DMCA doesn't even apply."
Please explain this further. I seem to not be able to find a context for this statement that would make it true.
Re: (Score:2)
The context would be that their content isn't hosted on a cloud/social platform like youtube but on their own streaming service like disney+. That doesn't really shield them from responsibility, as you can still strike their hosting company.
Chances are their hosting provider and/or ISP partners aren't going to be particularly reactive when doing so involves losing million dollar contracts.
If anyone tries that BS "non-literal copying" (Score:2)
Just don't use github (Score:1)
DMCA equals guilty until proven innocent (Score:4, Insightful)
The whole concept of the DMCA is harmful to everyone. It basically boils down to guilty until proven innocent which is the fundamentally incorrect legal principle to begin with.
Add to that the fact that most companies have no choice but to comply with such a request and it is waiting for an innocent bystander to get caught up in it or for another company to start abusing the DMCA takedown system for their own benefit.
The judge should take this brief seriously (Score:5, Interesting)
GitHub's owner, Microsoft, is an authoritative voice on this topic, given that it spent a good chunk of cash helping to bankroll TSCOG's vague infringement allegations against Linux. They know what they're talking about.
“Look & Feel” revisited? (Score:3)
This sounds suspiciously like an attempt to resurrect the failed “look & feel” lawsuits of the 80s. Lotus, Xerox & Apple all took their shots (and lost in court).
Ref:
https://enwp.org/Look_and_feel... [enwp.org]
In passing, this “nonliteral infringement” claim, like those earlier “look & feel” claims, appear to be rooted in what is called “trade dress”, not “copyright” per se.
For example, Sun/Oracle suing Google over Java APIs is a borderline copyright infringement issue (since it involves actual code, viz. the interface design & code snippets of the Sun/Oracle Java APIs).
By contrast, the earlier “look & feel” lawsuits by Lotus (spreadsheets: Lotus 1-2-3 -v- Borland Quattro), Xerox (GUI design: Xerox Alto -v- Apple Macintosh), and Apple (point-and-click OS design: Apple Macintosh OS -v- Microsoft Windows) were more an issue of digital design “trade dress”. The alleged infringement involved issues like mimicking visual layout, menu sequences, GUI widgets (both visual & behavioral aspects, hence “look” & “feel”), action sequences, basic capabilities & core functionality, technical terms (e.g., UI labels), and the like.
Critics derisively ridiculed those efforts as analogous to trying to leverage a Model-T Ford automobile patent to claim exclusive licensing rights to the steering wheel interface (as opposed to knobs & “joysticks” & levers to steer, as in a crane or helicopter or snowboard/toboggan or amusement park fanciful contraptions).
It should be noted that the principal effort then with “look & feel” lawsuits was to establish in the courts a precedent for “software patents”, not to extend software copyright.
The core claim was that software, as an active process that is “executed” on a computer, is more akin to business process patents and/or manufacturing processes than to mere copyright of printed text (“source code”) and visual images (“GUI widgets”). That didn't fly in court. They should have claimed trade dress instead, or even artistic performance rights, rather than “computer process as patentable”. The SAS suit mentioned in the article seems to be aiming in this direction.
To wit, the “look & feel” critics were quick to point out that, as in chemistry, chemical engineering, and pharmaceuticals, one can patent the process for manufacturing a product but not the chemical equation for the end product per se. That would be akin, they argued, to patenting a mathematical formula or theorem or algorithm. One can receive a Nobel Prize or Turing Award for such discoveries, but not a patent for the discovery itself. Proponents of the claim, however, tried to argue that computer algorithms are more than just formulae & equations & scientific / mathematical facts: they are active, animated processes. The courts were not convinced. The problem is that an algorithm is in essence a mathematical concept of how to proceed in calculating something. Any specific implementation of said algorithm is protected by copyright, but the algorithm itself, as a mathematical concept, is not patentable.
Another analogy raised was that of cooking recipes as copyrightable but not patentable. One cannot force licensing of the use of a recipe in a commercial setting, but one can sue for copyright infringement if said recipe is published. Hence the preponderance of “secret recipes” in the culinary arts industry: they are trade secrets, not published material. Trade secrets can be licensed (typically under strict NDA) but publicly disclosed (published) recipes cannot.
Finally, to clarify, “trade dress” is traditionally the domain of protecting textures, shapes & color schemes (like the d
I'm probably in trouble now... (Score:2)
... because I tend to organize source code by keeping the majority of any subroutines/procedure/methods/etc in alphabetical order (to simplify finding bits of code while scrolling through a source file). I'm sure that's going to run afoul of someone's idea of something that can be copyrighted.