youtube-dl has Received a DMCA Takedown from the RIAA
I am a translator!
All the more reason to not use Github or Youtube.
Exactly.
The more users keep relying on these predatory service "providers" and the more they put themselves at risk of these terminations of service.
The takedown notice was from the RIAA, not Google/YouTube, and unlike a normal DMCA takedown notice, against which the devs could file a counter notice and Microsoft/Github might legally be able to restore access to the repo, since yt-dl is accused of being illegal DRM-circumvention software, this was Github's only legal course of action and anyone else would probably have done the same.
By all means, we should avoid Microsoft and Google in general, but that misses the significance of what is happening here. The RIAA is saying that using alternative software to determine a video URI instead of executing YouTube's non-free JS in the browser constitutes DRM circumvention, and that by having that capability the software code is illegal, regardless of how you use it (youtube-dl has many legal uses, and despite the name works on many other sites[1] as well as YouTube). If the RIAA halts yt-dl development, either via the law or via intimidation, it will be a great loss to society, horrible for software freedom, and set a frightening precedent that could affect similar projects. Depending on how far the RIAA goes with this, GNU/Linux distriutions that package yt-dl could be in legal jeopardy as well. Don't be surprised if Trisquel eventually does exactly what Github did. This is not a "no-brainer". This will not be solved by having opinions about particular companies. We should be worried, not smug, right now.
I am a translator!
since yt-dl is accused of being illegal DRM-circumvention software, this was Github's only legal course of action and anyone else would probably have done the same.
Couldn't Github ask youtube-dl developers to remove the offending code instead of disabling the repo especially since youtube-dl supports downloading from websites that provide the video link (e.g. the mp4 video or m3u8 playlist) in the source of the HTML page without running any JS or bypassing any copy protection scheme?
> Couldn't Github ask youtube-dl developers to remove the offending code instead of disabling the repo
I am not a lawyer, but I doubt it. If distributing the software is illegal, which is what the RIAA is claiming, then Github continuing to distribute the software would be illegal. Even if Github were to donate legal resources to help the yt-dl devs come up with a legally safe version of the software, they would still need to take down the repo in the meantime in order to comply. The repo may be back up eventually, but there will probably have to be a lot of discussion with lawyers that we won't be privy to before that happens.
> We should be worried, not smug, right now.
I do not see anyone being smug here.
Being worried is fine but is no different from having opinions in terms of problem solving efficiency.
Being condescending to people expressing opinions is not helping either. Especially given that what you are suggesting is, in fact, to wait and see.
Maybe "smug" was the wrong word, but some of the comments here came across to me as something like "Well, that's what you get for using Github and YouTube. Just use something else and problem solved." I want people to understand that the problem here is something else.
> what you are suggesting is, in fact, to wait and see.
For the time being I'm not sure what else to do that would be helpful, apart from communicating why it will be important to support yt-dl if/when they ask for support. That could at some point take the form of contributing to a legal fund, or making public comments about ways in which we have used yt-dl lawfully. As far as I know they have not yet made any statements other than acknowledging the takedown. Hopefully that is because they are seeking legal guidance before deciding on a course of action. So yes, wait and see, keep following this story, and if/when there is something we can do understand the importance of doing it. Just don't dismiss it as a simple cautionary tale against using certain services that we already knew were bad. Maybe no one is doing that, in which case I regret overreacting, but that's what it sounded like to me, which is why I said something.
I think the best case scenario is that RIAA hasn't really thought this through or is bluffing, that Google has left yt-dl alone because their lawyers think that there isn't a strong case against it (Google *has* shut down similar projects that use YouTube's API, so maybe they agree with the theory that because things like yt-dl and Invidious don't use their API they are not bound by YouTube's TOS), and that yt-dl can come out of this relatively unscathed.
Your efforts to be pedagogical and wary of confusions and knee-jerk reactions are commendable.
I cannot speak for other posters but I do not think that the recommendations to avoid the "usual suspects" has been made here only because they are the usual suspects.
We all know that the list of sources available through youtube-dl is a very long one, but there is no mystery why it is called youtube-dl and not videos-from-a-long-list-of-many-sources-dl. To the best of my understanding, the DMCA takedown notice specifically refers to copyrighted content available on Youtube. If I am not interested in downloading coyrighted material and if I am not using Youtube, I do not see how I should be concerned about that notice, to put it short.
If the xxx-dl developers used the same approach, their repository would arguably not have been taken down. If there is no way to design such a tool without allowing its users to "circumvent technological protection measures on copyrighted videos without permission of the rightholder"[1], which I very much doubt, then indeed the tool simply cannot exist within the current legal framework.
Of course it is abusive that such takedowns can happen before anything has been clarified and yes, I would be ready to support any effort aiming at limiting the probably of such bullying behavior happening. For the time being, though, the only course of action available is to stay clear from the abusive actors and keep advising accordingly.
[1] from Julia Redas's viewpoint as quoted in https://blog.codeberg.org/on-the-youtube-dl-dmca-takedown.html
> We all know that the list of sources available through youtube-dl is a very long one, but there is no mystery why it is called youtube-dl and not videos-from-a-long-list-of-many-sources-dl
I think the reason is historical. yt-dl started out as a tool for downloading YouTube videos, and the scope expanded as such a tool was found to be useful for other sites as well. Changing a project name after it is already popular is a great inconvenience.
> To the best of my understanding, the DMCA takedown notice specifically refers to copyrighted content available on Youtube.
Yes, it does refer to specific copyrighted content on YouTube, but the way I read it it does not sound like the accusation merely is that yt-dl facilitates downloading that specific content. Rather, it sounds to me like the use of that content in unit tests is being used as *evidence* that the *intent* of the software is circumvention for the purpose of copyright infringement. Even if yt-dl changes the URIs in these unit tests, the RIAA may still try to argue that neither the design nor the intent of the program has fundamentally changed. Will they go that far? I'm not sure. They may be bluffing and not actually think they have a shot in court, and won't go any further than sending a takedown notice to a company they know will comply to play it safe. Or they may have a goal of setting a dangerous precedent against software similar to yt-dl. We probably won't know until yt-dl makes a move and RIAA either makes a second move or does nothing.
> If I am not interested in downloading coyrighted material and if I am not using Youtube, I do not see how I should be concerned about that notice, to put it short.
Here are some reasons I'm worried:
* I am working on a project related to my dissertation that involves a large number of clips mostly from online sources. Some of these clips are from YouTube, but many are from other sites supported by yt-dl. An interesting example is C-Span, whose videos of US Congress sessions are public domain, but do not provide a direct link to the video file, requiring a yt-dl like tool to download it. Even where I am downloading copyrighted videos, the way in which I am using them I believe to be covered by the US's current exemptions[1] to Section 1201 that allow circumventing DRM for the purpose of several kinds of fair use. So even if it is determined that yt-dl is a DRM-circumvention tool, it would be legal for me to use it in the way that I am. However, the exemptions only protect *use*, not *distribution*, so the development of yt-dl could be fragmented or suppressed, hindering its ability to keep up with supporting all of these sites.
* After this thread[2] I began working (slowly, it's like a fourth-degree side project at this point) on an experimental web browser based on Readability (the library that powers Firefox's Reader View) and youtube-dl, with the goal of using tools like these to separate out content from crap in web pages and applying browser-side navigation and styling in order to create a not only usable but actually polished and pleasant JS-free and CSS-free experience for a large subset of the web. I may or may not have to give up on this idea, or limit its scope to a much smaller subset of the web, depending on how things play out with yt-dl and the RIAA.
> Changing a project name after it is already popular is a great inconvenience.
True. Although in the light of the events discussed here, now might be the right time. Anyway, I have only seen links to de-googled youtube videos on this forum, which I would take as an indication that yt-dl is mostly used on these, which was my point.
> Even where I am downloading copyrighted videos, the way in which I am using them I believe to be covered by the US's current exemptions [...] that allow [...] fair use [...] However, the exemptions only protect *use*, not *distribution*, so the development of yt-dl could be fragmented or suppressed, hindering its ability to keep up with supporting all of these sites.
In that case "fair use" would indeed fall in the category of wishful regulation: the law stating that you are allowed to do something while also saying that the means you would need in order to exercize this right are forbidden. In many legal systems, such a law could not exist. This case might be a good occasion to determine where the law actually stands in the US.
But is it not that anything that shows up on your video or sound card can be grabbed there, whatever uri obfuscation tools have been used upstream? I have no idea how painfully inconvenient such a method would prove for the specific usage case you are mentioning but I do not see how RIAA or any similarly repressive organization would argue their case against such a tool. Would they also try to take down all repositories for, say, screenshot tools? Or Audacity?
I realize that DRM have diverging consequences depending on whether the matter at stake is being able to make legal use of something you have legally acquired or of something you have somehow legal access to but have not been granted by default any further explicit right to copy, store or distribute. In the former case, it is arguably akin to robbery. In the latter case there might be room for a more nuanced discussion. Maybe I am wrong and it is always a bad idea, though.
> an experimental web browser based on Readability [...] and youtube-dl, with the goal of using tools like these to separate out content from crap
Now I am worried, because that could be a massive hit. It would still be interesting to see what a takedown-proof version of such a browser could do. If most embedded videos are linking to obfuscated Youtube url then again, the core problem is our dependency on Google or similar power grabbing entities. Could the crap not be removed locally, while streaming, when no other option is available?
I might somehow be playing devil's advocate but this discussion is showing up on a regular basis, most probably each time such bullying move is made, so I agree it is important to make sure where we stand, and why.
> It would still be interesting to see what a takedown-proof version of such a browser could do. If most embedded videos are linking to obfuscated Youtube url then again, the core problem is our dependency on Google or similar power grabbing entities. Could the crap not be removed locally, while streaming, when no other option is available?
To clarify, I am planning to put yt-dl to much broader use than YouTube. Not only does yt-dl have a long list of supported sites, but it even works on many sites that are not specifically supported but have a video uri somewhere in the page source. When I am trying to watch a video online without JS, I don't even check the list of supported sites. I just try the web page URL with mpv (which uses yt-dl as a backend for streaming) and most of the time the video plays. With yt-dl, videos on the web go from mostly inaccessible to mostly accessible without JS. Any page suspected of featuring a video would be run through yt-dl. As a result, YouTube will be just one of many sites that work better than they would on a traditional browser with JS disabled.
That said, yes we should also move away from YouTube, because centralization is bad. This is a separate issue from JS though, which is even a problem for many smaller websites.
> Could the crap not be removed locally, while streaming, when no other option is available?
I was planning an "archive" feature, which would be kind of like a bookmark, but would save a copy of the archive page to disk and index the description and (in the case of articles) content for use with a local search engine that searches your archived documents. (The same functionality can be used to expose a public search engine.) But this option could for example be disabled for YouTube videos, and yt-dl could just be used to retreive the video URI for playback via Gstreamer. We'll see. By the time I have a chance to dive in and make some progress on this we will likely have more information about the yt-dl/RIAA situation.
Codeberg, a freedom-focused alternative to Github, challenges the legitimacy of RIAA's claim, but acknowledges that they would have done the same as Github. They call for the FSF and EFF to get involved[1]:
"First of all, Codeberg e.V. and the Codeberg platform was founded to support Free and Open Source (FOSS) code and content development. We support the open source community and aim to make the Free and Open Source movement stronger. We see ourselves on the good side and we do not support illegal activities on our platform.
"However, if we for example host a legitimate open source tool and we would receive a similar notice, then we most likely would have to disable the repository until the matter is resolved by court ruling if such is fought through by the project owners. This could include the code, issues, and documentation, which would immediately threaten/weaken the development community around that project.
"We express severe doubts that the RIAA reference to the above mentioned Hamburg Court ruling is legitimate. As a large portion of content hosted on video streaming platforms is of educational nature created and posted by owners under permissive licenses such as Creative Commons with the clear intent to be freely shared for the common good, rendering a general-purpose tool like youtube-dl illegal is not justified.
We sincerely hope that organizations with established standing and funding resources much greater than ours, the Free Software Foundation FSF and the Electronic Frontier Foundation EFF just two among the most well-known, will join forces and prevent this from becoming a threatening precedence potentially eroding the most basic human rights on free access to education, knowledge, and communication."
[1] https://blog.codeberg.org/on-the-youtube-dl-dmca-takedown.html
Interesting that Google would make this very monopolistic move in the week that they were sued for their monopolistic business practices. Not really good timing, from a legal perspective.
EDIT: it is amazing that it took so long for any sort of antitrust action to be initiated against Google.
This is the RIAA, not Google, though they are using Google's TOS and alleged DRM (it's debatable whether YouTube's on-the-fly URL calculation constitutes DRM) as justification, so if Google had wanted to make this case they could have with the same evidence. Hopefully the reason that they have not come after yt-dl is that their lawyers think that the RIAA's argument is as flimsy as I (as a non lawyer who can only guess) hope it is. There could be other reasons though, such as not wanting to appear more monpolistic than they already do, or not wanting to set a precendent against using certain web scraping techniques, given that their search engine data is acquired via web scraping.
Some are already fighting back:
Sourcehut, another freedom-respecting Github replacement has commented on the situation[1]:
"As a US entity, SourceHut is obliged to comply with DMCA notices. In this event, our next step would likely be to coach the affected project through the counter-notice process, and contribute to their legal costs if we believe that they’re in the right. We know that the DMCA is a constantly abused force for censorship, and there are no friends of the RIAA here. They represent much that our mission statement — to support and improve the free- and open-source software ecosystem — stands in opposition to."
They also note that by using a decentralized, email-based workflow, Sourcehut is more resistant against censorship than a centralized model like that of Github and its imitators:
"Even beyond our principles, however, the mailing list based workflow we espouse is resistant to this kind of censorship. Git repositories are easily re-hosted, of course. However, if a lot of your project’s value requires rapid updates and the ongoing support of its development community, a centralized, pull-request-style system is vulnerable to censorship.
"If you use mailing lists, you might not even immediately notice that something was wrong. That contributions pass through a centralized mailing list is often only a formality — the project maintainers and others likely to have comments will usually be Cc’d on the emails, and they’ll be delivered directly to them without the list’s help. Recipients can provide feedback by replying to the email, sending their comments directly to the contributor, and bypassing the mailing list entirely. The list faithfully records these emails and forwards them to interested parties, but it’s not actually required for this to work."
[1] https://sourcehut.org/blog/2020-10-29-how-mailing-lists-prevent-censorship/
"Asking Microsoft to resign from the RIAA over youtube-dl takedown demand"
https://sfconservancy.org/blog/2020/oct/26/microsoft-github-riaa-youtube-dl/
New git instance is up on git.rip website with updated download instructions and new commits in the past week: https://git.rip/lucifer/youtube-dl
Clearly development is still occuring somewhere, as a new release (2020.11.01) is available on yt-dl's website.[1] However, the latest tag[2] of the git.rip repo is 2020.09.20. Are you sure that this is an official repo and not a mirror or someone's personal branch?
I don't know, but its instructions show it using curl and wget to pull installs and updates from the yt-dl.org site, so if you followed the git.rip instructions you'll still end up with the latest.
Some good news: Github has reversed[1] the DMCA taketown, citing a letter[2] from the EFF.
[1] https://github.com/github/dmca/blob/master/2020/11/2020-11-16-RIAA-reversal.md
[2] https://github.com/github/dmca/raw/master/2020/11/2020-11-16-RIAA-reversal-effletter.pdf
"To borrow an analogy from literature, travelers come upon a door that has writing in a foreign language. When translated, the writing says “say ‘friend’ and enter.” The travelers say “friend” and the door opens. As with the writing on that door, YouTube presents instructions on accessing video streams to everyone who comes asking for it." -- RIAA takedown reversal EFF letter.
Trolling the RIAA must be something.
I love how GitHub, in their blog, paints "doing the bare legal minumum" as "standing up for developers."
> "doing the bare legal minumum"
I think you meant "minimum".
All these steps seem to be going in the right direction anyway:
https://github.blog/2020-11-16-standing-up-for-developers-youtube-dl-is-back/#what-were-changing
Yes, looking at takedown requests, only processing ones that GitHub/Microsoft deem to be valid, etc. that's the bare legal minimum I refer to. It's hardly "standing up for developers." It's PR spin and my prediction is that nothing much will change.
For one example: "The claim will also be carefully scrutinized by legal experts to ensure that unwarranted claims or claims that extend beyond the boundaries of the DMCA are rejected." Considering GitHub/Microsoft's "safe harbor" protections under the DMCA evaporate if they don't follow this and that the potential penalties to GitHub/Microsoft for continuing to host allegedly infringing material after having received a take down notice is substantial, it's safe to assume that those "legal experts" will council to be more conservative and favor GitHub/Microsoft's interests over developers and take the material down over not doing so, especially in cases that go near the DMCA boundaries or are somehow questionable or otherwise untested cases like with youtube-dl.
In fact, all that's happened in this is case of youtube-dl is the normal next step in the DMCA process: The counternotice has been made. Now it's up to the RIAA about what to do so this isn't necessarily over just yet.
> this isn't necessarily over just yet.
I think we all agree about that.
I am trying to see some light in this. Everybody does PR but the EFF reversal letter could also have been trashed, or processed without much further reaction from GitHub.
I judge on facts, not on predictions or assumptions.
For instance, I see that PureOS still appears to be endorsed by the FSF. Or is it the GNU project? The announcement was published on the fsf.org website but the FSDG and the list of endorsed distros appear on gnu.org. I am not sure who is actually in charge.
Anyway, it says that "we only ask that a distribution's developers make a good faith effort to avoid including nonfree software, and commit themselves to removing such programs if any are discovered." Was it the case with PureOS in the three years that passed since endorsement?
> Anyway, it says that "we only ask that a distribution's developers make a good faith effort to avoid including nonfree software, and commit themselves to removing such programs if any are discovered." Was it the case with PureOS in the three years that passed since endorsement?
Based on my cursory examination of PureOS in this thread,[1] no, not at all. And since writing those posts, I have come across Purism's repository[2]for the Librem 5's non-free WiFi firmware, which is an even more blatant FSDG violation. However, I don't see what this has to do with youtube-dl or Github.
> I don't see what this has to do with youtube-dl or Github.
Sorry, I realized the relative off-topicness of it only after you had answered, so I could not edit my post.
It was supposed to be an example of judging on facts, and not on assumptions, guesses or predictions. Facts seem to tell that the FSF/GNU do not enforce their own endorsement rules. I cannot guess what anyone else would have assumed or predicted about this case.
> I love how GitHub, in their blog, paints "doing the bare legal minumum" as "standing up for developers."
Yeah, any and all credit here goes to the EFF.
> all credit here goes to the EFF.
Absolutely. Though lore has it that Meriadoc Brandybuck had a minor contribution in it.
I noticed that youtube-dl is updating normally again from the command line. So I guess our long national crisis is over?
That remains to be seen. This was more like crossing the mines of Moria, Mount Doom is still hundreds of miles away and now Gandalf is gone, according to goblin media at least.
The RIAA should still be trolled to death for these never-ending three weeks in the dark.
The ball is back in the RIAA's court, I guess. Hopefully at this point they'll decide that this fight is not worth it and move onto whatever their next scheme is after this,[1] perhaps a worldwide ban on any human ears with the illicit capability to listen to copyrighted music?
[1] https://www.technadu.com/riaa-mpa-target-domain-name-hosting-providers/223938/
They'll have to go back to their favorite scam - suing deceased grandmothers for file sharing: https://arstechnica.com/uncategorized/2005/02/4587-2/