Hacker News Reader

Comment by bitcharmer | Hacker News Reader

Can someone knowledgeable explain to me how a tracker source code violates DMCA?

What's next, hard disk manufacturers? Internet providers?

> What's next, hard disk manufacturers? Internet providers?

You're joking, but in some countries there are taxes on blank recording media: https://en.wikipedia.org/wiki/Private_copying_levy

And in the Netherlands those used to make sense. As downloading copyrighted material was legal for personal use

https://mashable.com/2014/04/10/downloading-pirated-content-...

Introduce levy to cover for people copying stuff, then introduce copy protection and anti-circumvention laws so copying becomes illegal. Continue to collect levy. Ingenious.

I think the basis is that you're still allowed to make private copies of CDs/DVDs/etc pretty much throughout EU.

Not when you'd need to circumvent DRM, thanks to EU Copyright Directive Article 6.

Some countries including the USA, for those who haven't bothered reading the link.

This is why decades ago you could buy blank CDR media designated for "music" at a slightly higher price. They are identical to "computer" CDRs, but the seller pays a fee to the copyright office.

In the Czech Republic you need to pay this levy on blank media (see point number 1 & 6 on https://www.mkcr.cz/frequently-asked-questions-953.html?lang...) to a shadowy state mandated cartel of "designated rightholders" where the money preaty much disappears from the face of the Earth.

AFAIK there is no way around it, even in professional environment - we build a couple hard drives for a an internal company server and this levy was clearly marked there.

Same thing for all the other extortion money they extract from music festivals, pubs, hotels, radio stations, etc. - they do use some of it to bribe the most prominent and vocal old school artists to basically do advertising for their cause but the rest safely hits their coffers.

This reminded me of the Russian/ROMS/AllOfMp3 saga. That was a service way ahead of its time... Even now it is difficult to find such a good catalogue of music in lossless format.

It shouldn't, and we can't know what was in the offending files because they've been removed.

> For the avoidance of any doubt, we are also providing you with the attached file tilted “GitHub_code_Nyaa” which shows code hosted on GitHub that provides all of the source code, templates, and utility API tools needed to host a copy of the Nyaa.si Bittorrent site, used to access infringing copies of motion pictures and television shows for which Nyaa.si provides torrent files and magnet links for the infringing content that users are looking for. The identified files and code are preconfigured to find and provide infringing copies of our Members’ film and tv content to Nyaa.si users in violation of copyright law.

That's the real issue, but again it's hard to discuss since the files have been removed.

It really shouldn't. their argument is probably that infringement is the main use case of a torrent tracker.

edit: I don't have access to the taken down repo but maybe not: https://news.ycombinator.com/item?id=25836800

In which case it should be easy to show the non-infringing uses. Many linux distributions have torrent downloads just as one example.

Agree. Whether or not the boomers in congress and lobbyists in the MPAA/RIAA care, though, is another question

The courts have a different standard though, and so it is much easier to get them to see. DMCA also is submitted under penalty of perjury, so if there is no infringement, then the courts can take criminal action against the accuser.

I'm not clear on exactly what is being claimed so I don't know how real it is in this case.

> DMCA also is submitted under penalty of perjury

No, it's not, except for the claim to copyright ownership of the allegedly infringed work. Particularly, the claim of infringement is not made under penalty of perjury (which makes some sense, as it is a legal and not a factual claim, but you don't have to allege the factual basis for the claim of infringement under penalty of perjury, either.)

The issue is that (supposedly, we can't know because the repo has been taken down) the code was pre-configured to host copyrighted material. From the complaint:

> The identified files and code are preconfigured to find and provide infringing copies of our Members’ film and tv content to Nyaa.si users in violation of copyright law.

I don't know if they mean that the code is preconfigured to connect to a torrent network which is known to host copyrighted material, or if there was a host_copyrighted_contet.cfg file, etc. Its hard to know how much they are stretching the truth because the code is gone.

Agreed, that's what I just commented before reading this reply. They are very loose with their words in takedowns. I bet it's probably not as bad as they say, though. Maybe it has a movie/TV API that it uses, thus is "preconfigured to find and provide infringing copies of our Members’ film and tv content".

I never saw the repo, but to play devil's advocate it's possible that there's a unit test or something that downloads copyrighted material. That's what got youtube-dl taken down.

Doubt, there would be no reason for it to be copyrighted material, it would just be a torrent test

Couldn't the same thing have been said for the problematic youtube-dl test case?

Those directly downloaded copyrighted youtube videos. It would instead be just a torrent tracker connection check, no reason to even transfer a file, since the tracker just matches user with user and doesn't touch any files.

The issue with the youtube-dl case was that certain copyrighted video pages did have a different structure, so they needed to test they could parse that page correctly.

There's no technical difference between a torrent containing ArchLinux.iso and TheBeatlesDiscography.zip however.

Not exactly. They had a test that saves a file from YouTube that was "protected" with some tech a lawyer might quality as DRM. YouTube only uses that thing for copyrighted materials.

Torrents don't have anything like that. You don't have to deal with anything copyrighted to test any related stuff, be it client, tracker, or web server.

You clearly didn’t read the DMCA

I'm just saying that if a unit test existed, that's likely not what would cause the DMCA. seems that the reason is that it's preconfigured to share copyrighted content, so like I said, not the unit tests.

Torrent trackers are responsible to introduce Peers to each other who are downloading/uploading the same files[torrent]. Its basically the backbone of torrent.

say for example, you have a torrent file or a magnet link. And you start downloading the file using client application, it will go to known trackers and ask them the address of peers who are uploading the file based on unique hash of it. Torrent breaks files in the chunks with size of few MB referred as pieces, Tracker will get the current status of downloaded and remaining chunks from you and give you the list of peers with same file available, at same time it will register your address for the other new peers to connect. now connecting peers have knowledge of one anther's address they will share pieces to others and download remaining from other peers to complete the download.

so, basically peer is backbone of system, its a tool which can have multiple use legal or not, one cannot simply call its source code violating the IPs, that's just an absurd argument.

Spoiler alert: it doesn't.

And neither does it violate any similar law (because not many people know, but the DMCA is not an idea from the US, just barely the US implementation of the 1996 WIPO treaty), in the European Union or anywhere else.

So it really, isn't illegal anywhere (well except obviously authoritarian regimes).

I don't think it does - I think this is just a grab to see what microsoft caves on.

Technically, the code is alleged to violate copyright, not DMCA. A DMCA notice is just the process that copyright owners use to tell service providers about allegedly infringing content.

The argument is for contributory copyright infringement via inducement, ala MGM v Grokster 545 U.S. 913 (2005) or Columbia Pictures Industries, Inc. v. Fung 710 F.3d 1020 (9th Cir. 2013), the isoHunt case. Inducement occurs where someone distributes a device with the intent of promoting infringement and actual infringement occurs. Here, the argument is that nyaa intends to promote copyright infringement because it provides a tool that is tailored for copyright infringement. The notice letter cites various references in the code to nyaa.si. Exhibit A of the letter comprises "a series of screenshots taken from the Nyaa.si website for the Project that include images of copyrighted works available through the Project." The implication here is that the devs here are encouraging similar use of the code for infringing copyright.

I'm not saying I agree with it, but that's their argument.

Based on the way certain files are listed in the complaint I suspect this is something along the lines of what happened with youtube-dl where there's some part of the test suite, some example, or even some comment in the code that refers to a copyrighted work.

With the repo down I obviously can't verify this hypothesis but it fits.

>by virtue of the operation and further development of the Bittorrent website Nyaa.si’s “nyaa” repository (the “Project”)

Basically the fact that the world's most predominant installation of the site (Nyaa.si) is serving a whole bunch of copyrighted materials through torrents makes the code itself violate DMCA.

I guess most user use Chrome to access that site, does it making Chrome violate DMCA?

> Can someone knowledgeable explain to me how a tracker source code violates DMCA?

> What's next, hard disk manufacturers? Internet providers?

Hard drives are already taxed in France because of piracy (well not exactly, but that's the real reason)

Next is mpv I'm guessing

By name it wouldn't but they link to several files where the "offending code" resides, so if they're using examples where someone is downloading a movie then that would prompt for a DMCA.

It is stated in the complaint:

> Specifically, at the URL, the Repository hosts and offers for download the Project, which, when downloaded, provides the downloader everything necessary to launch and host a “clone” infringing website identical to Nyaa.si (and, thus, engage in massive infringement of copyrighted motion pictures and television shows).

This is not a generic BitTorrent tracker. This is source code for duplicating the nyaa.si website (which in the github source code, they state is "A BitTorrent community focused on Eastern Asian media including anime, manga, music, and more" [1]).

I think it would be a stretch to say you are interested in this source code so you could then rip out large pieces of it to host your Linux ISO's on...

I agree that some DMCA crosses the line. But in this case, I don't see a legitimate use for this source code other than enabling copyright infringement.

[1] https://github.com/seco/nyaa-1/blob/master/nyaa/templates/ho...

It's a torrent index with search, categories, user management and so on. I don't see how I can't use this for hosting Linux ISOs by just renaming the categories.

You could use it, but would it be the most natural thing for you? Or are there better tools for that task? The nyaa source code includes features and organization specifically for the anime/manga scene.

For example, there is torrent classification to detect whether the torrent contains an entire series, which subtitles it contains, whether it is a reencode/remux/scene release, etc [1]. Would that be useful for you to host legitimate content (which typically is not reencoded/remuxed/fansubbed)?

I am all for criticizing the chilling effects if the DMCA was targeting the DMCA protocol in any way. But in this case, it is targeting the source code for a CLEARLY illegal website (illegal, in this case, based on the jurisdiction in which github operates and is bound to)

[1] https://github.com/seco/nyaa-1/blob/master/nyaa/templates/he...

all of those topics can have legitimate, non-infringing content; such as fan fiction released under a permissive license.

Reply on news.ycombinator.com