GitHub repository for Sedgewick's Algorithms is taken down
github.com
github.com
https://github.com/github/dmca/blob/master/2021/04/2021-04-2...
Kevin Wayne (repo owner presumably) is listed as co-author on the book that is published by Pearson, who have issued the notice.
https://www.pearson.com/us/higher-education/program/Sedgewic...
Yours sincerely,
[private]
Internet Investigator
On behalf of:
Pearson Education, Inc.
Seems like another glorious victory of the overzealous online copyright strikers. Nothing out of the ordinary, will likely get sorted out in a day or two.
Haha... that's a brilliant euphemism for "profession Google user". Or maybe we could call it "serial scrapist"?
Of all the silly titles I tried to come up with, this sure hadn't crossed my mind yet. And it has a catchy ring to it too.
"Internet Investigator" got to be somewhere up there between "Adult Software Engineer" (middle ground between "Junior" and "Senior") and "Server Captain/Commodore/Admiral".
Ahhh I see that the DMCA is still a wonderful doxing weapon :)
I assume you're missing a word there?
If it is in error, I am curious how much work it would take to get that perjury result though.
While I don’t know the relationship between this “investigator” and Pearson, the claim that Pearson is the owner or exclusive licensee of the books identified does not appear to be in dispute.
This does not appear to be a defect in the application of the penalty by the legal system but of what the penalty applies to under the relevant law.
DMCA is an awful law.
Interestingly, the README includes an e-mail exchange with Kevin, who noted the code is GPL
Found this in the link you provided. As you can see from the code there is a GPL license footer on each file.
These DMCA requests are created and acted upon without the slightest checking, with no penalty.
There should be some sort of modest cost somewhere to stop this nonsense.
Playing the Devil's Advocate here:
But wouldn't that in turn incentivise actual copyright infringement on account of it potentially being cheaper to mass publish copyrighted material as opposed to request it being taken down?
An artist finds their life’s work online, and issues a takedown notice, not realizing it contained a few songs they don’t notice? No big deal.
A label pays a takedown notice mill, that produces completely bogus / mistargeted takedown requests on a regular basis? Scorched earth. They lose the ability to enforce any copyrights on their catalog.
Judges don't scale and they are a step function, not an incentive gradient.
> An artist finds their life’s work online, and issues a takedown notice, not realizing it contained a few songs they don’t notice? No big deal.
If he had to pay a small(!) fee for mistakes in his takedown it would mean little harm to him since he protected his works with one takdown. One could even add an exemption for first-time mistakes. If he has to do it repeatedly then one would hope that he improves his search process.
Additionally you have to consider that takedown notices can be filed against middle-men targeting many different 3rd parties at once. E.g. consider URL takedowns submitted to google. Google has no dirty hands and neither do some of the mistaken targets.
Small fees would only add up if you issue bogus takedowns in bulk. They also have the advantage of not burdening the courts and providing the right incentives. The fact that the cost exists would drive people to reduce false positives which would make the penalty being required less often.
As far as fees go one could also distinguish between fair use disputes vs. wholly incorrect claims. After all in the former case the claimant has a valid case and it is just that there exists a valid defense. In the latter case the claim itself was invalid.
> A label pays a takedown notice mill, that produces completely bogus / mistargeted takedown requests on a regular basis? Scorched earth. They lose the ability to enforce any copyrights on their catalog.
That is not what is happening in reality.
How can you explain this kind of events:
> SoundCloud takes down D.J. Detweiler's 'remix' of John Cage's "4'33" for copyright infringement.
> White Noise Video on YouTube Hit By Five Copyright Claims
Not to mention that almost all classical music pieces recorded by students are under risk of being DMCA'ed for being so similar to copyrighted recordings. Classical music has many interpretations for the same piece and they are harder to tell apart than, say, pop music.
It has never been used in response to a DMCA takedown notice. It is far past time for that to happen. The precedent alone would end this crap.
With such a precedent, Pearson would be at risk of losing the copyright on the entire textbook (and probably related course materials, tests, etc.)
Presumably Kevin Wayne was given the right to license the code as GPL, but that fact didn't get included in the text of the book, and the f*cking, stupid copyright bot Pearson uses didn't get the memo.
Any public file host has to deal with the DMCA, and (for now, barring any evil brought on by Microsoft), I'd bet on GitHub siding with the developer over the lawyers. They had lots of incentives to go the other way in the youtube-dl case and they didn't, and then they put a bunch of money in a fund if someone wants needs a lawyer to get their project back online.
Now, how can we make it popular?
That's for my company, for many open source projects the damage would be more severe since they often heavily use issues and GH project management features.
It’s cheaper and less risky not to evaluate the merits of DMCA notices and just to blindly execute them, but it is erroneous to say that thet are legally required to act in that manner.
You are sharing an opinion disguised as fact. It’s a wrong one, but I get why you’d conclude it.
Source: Write DMCA policy for UGC.
The DMCA process isn’t a mandatory process, it is a process to receive safe harbor from whatever liability would otherwise exist. If there would be no liability independent of the DMCA safe harbor, there is no mandate to follow the safe harbor process.
(If you disagree with this, here’s what you need to do, identify—with citation to supporting law—the available legal remedy that can be imposed on a provider who declines to adhere to the DMCA safe harbor process where there is no underlying copyright liability.)
Service providers make a choice to follow the notice process blindly as a risk management measure, not because it is legally required to do so.
(Which is also why the counternotice process is often less fully implemented, or why, e.g., Youtube has its own hyper-aggressive policy for certain content that goes beyond the DMCA process and lacks counternotice opportunity—the counternotice process is just as much part of the DMCA process, but because providers are able to structure their relationship with users in a way which avoids having any liability for takedowns which would benefit from a safe harbor, they can safely ignore that process.)
Source: the actual text of the DMCA safe harbor provision, and the reason it is called a “safe harbor”.
> You are sharing an opinion disguised as fact
No, you are engaging in standard industry blameshifting by misrepresenting risk management strategy adopted responding to incentives created by a law with an actual legal mandate.
> Write[s] DMCA policy for UGC
Is purposefully trying to confuse your argument, or worse yet, doesn't know any of it, that may be part of the problem
The problem is that GitHub doesn't get to decide that. If they give up their option for safe harbor by not following the DMCA takedown process, the company claiming copyright can sue them. Even if they don't win, defending these things costs time and money.
I don't blame GitHub for just not wanting to deal with all that. You seem to disagree, and believe that GH should stand up and open themselves to legal liability for every random user on their site, which I don't think is practical or reasonable to expect of a for-profit corporation.
The remedy for this is simple: we need to stop relying on centralized solutions for things like this. There are a variety of options for this: Tor hidden services, IPFS, hosting outside the copyright holder's jurisdiction, etc. None of them are perfect, but relying on a for-profit corporation to host your even-remotely-legally-questionable material, you're setting yourself up to fail.
The company can sue them in either case. The DMCA safe harbor is a particular easy basis for a motion to dismiss, but other easy bases may exist in specific cases. Blindly following the DMCA safe harbor process is a sensible risk mitigation strategy, and exactly what the law is structured to promote, but it is very explicitly not a mandate.
> I don't blame GitHub for just not wanting to deal with all that. You seem to disagree
You are a mistaking a fact statement about what the law does and doesn’t require with a normative statement about what Github should do.
I imagine they do this because of their settlement with Viacom[0] where they likely bent over backwards to avoid further appeals and lawsuits. Viacom likely said "regardless of how this goes, we and other copyright holders will sue YouTube off the internet unless you build a system that proactively takes down everything even remotely potentially infringing without us having to notify you of anything". And Google caved.
Don't get me wrong, the DMCA is still overall a bad law (though, to be honest, I think the copyright takedown process is the least bad part of it), but Google is not doing what they do because of the DMCA.
[0] https://en.wikipedia.org/wiki/Viacom_International_Inc._v._Y.... -- notably, no money changed hands as a part of the settlement, but other terms were not disclosed.
And you know who can afford lawyers? Big corporate copyright holders.
archive.org doesn't have the source, but it has the list of forks ( https://web.archive.org/web/20200908104954/https://github.co... ). The forks themselves are still online. For instance, https://github.com/FGM-148/algs4 . As usual with forks, use caution.
Then there is the website of the book https://algs4.cs.princeton.edu/code/ . This might be a less useful format, though.
Did this repo contain the code or the entire book as well?
In the former case, if I remember correctly, this DMCA claim would clearly be bogus. Yet another case where sending out DMCA claims having penalties for the sender if they are fraudulent, but putting a burden on the recipient who might suffer if they don't immediately act.
Big corps keep sending automated DCMAs with little consequence and often the little guys do have the resources to fight back.
https://www.copyright.gov/legislation/dmca.pdf
“Penalties are provided for knowing material misrepresentations in either a notice or a counter notice. Any person who knowingly materially misrepresents that material is infringing, or that it was removed or blocked through mistake or misidentification, is liable for any resulting damages (including costs and attorneys’ fees) incurred by the alleged infringer, the copyright owner or its licensee, or the service provider.”
Here's the actual text: https://www.law.cornell.edu/uscode/text/17/512#f
I'm not a lawyer but my understanding is that by now the courts have confirmed that it's not just enough to say that you own the copyright on the material: you also have to confirm that the person you're sending the claim to doesn't have a right to use the material as well and that includes fair-use. If you were, say, using a Disney clip in a film studies context they would likely be liable if a takedown bot sent a DMCA claim unless they could show that you did something like posting the entire film claiming it was for “study” purposes.
In this case, that's highly relevant since this repository was apparently released under an open source license so even if Pearson held the copyright they wouldn't be able to take away the permission granted by the open source license unless they could prove that it was never legally approved for release under that license.
The winner gets the money back and the platform gets to keep the losers deposit as a fee for having a trained, certified human make the decision.
Make the fees scale up with the number of claims by an accuser.
No one would file a frivolous claim. No one would dispute an obviously correct claim. The cost of having humans decide is covered, so platforms like Google have no excuse.
I don’t know why a counterclaim should require a deposit - DMCA is already a guilty until proven innocent law, and merely counterclaiming doesn’t put your content back online.
(And if the deposit isn't too big, do you think a few dollars will deter billion-dollar companies?)
I'm not arguing that it's a perfect system, but with a few adjustments I feel it's a much better one than exists today.
The law is fine. The enforcement is iffy.
If you can't afford to loss $100, you probably can't afford to temporarily not have it either.
In this case here, github does not employ an overzealous automated system. This is a manual DMCA claim and the issuer is the single actor here, who could be to blame.
https://docs.github.com/en/github/site-policy/dmca-takedown-...
Risky to keep these exclusively in Github.
> This public repository contains the Java source code for the algorithms and clients in the textbook Algorithms, 4th Edition by Robert Sedgewick and Kevin Wayne. This is the official version—it is actively maintained and updated by the authors. [...]
We have received information that the domain listed above, which appears to be on servers under your control, is offering unlicensed copies of, or is engaged in other unauthorized activities relating to, copyrighted works published by Pearson Education, Inc.
Copyright work(s):
Copyright owner(s) or exclusive licensee:
Pearson Education, Inc.
Copyright infringing material or activity found at the following location(s):
[1] https://github.com/github/dmca/blob/master/2021/04/2021-04-2...
http://web.archive.org/web/20201016112929/https://github.com...
And they make a ton of money.
Please watch this talk for more context (even if you don't like TED talks): https://youtu.be/BnC6IABJXOI
Not just way overpriced, the few Pearson books I happen to own are of an alarmingly low quality. The pages of a phone book look and feel like premium paper compared to my copy of Blitzer's College Algebra.
Situation in private school is not better. A prominent private high school in Bay Area, made me run around like a crazy man scavenging used books from all over the place because the recommended books were out of stock. They won’t publish the exact book name until after classes begin in August and apparently varies by teacher by year. This is after paying $$$ fees equivalent to buying a new car. Next year I pulled my kid out of it.
Google AdWords Conversion <icon>; Twitter Advertising <icon>; DoubleClik; Facebook Custom Audience; LinkedIn Ads; Twitter Conversion Tracking; Google Dynamic Marketing; Drawbridge
Analytics: Google Analytics; CrazyEgg; PulseAnalytics; Facebook Connect; Pardot; LinkedIn Analytics <icon> ; PulseInsight
Can someone give some insight into how much "Surveillance Capitalism" is being done here? Facebook is a giant red-flag to me, personally..
I am glad that many CS authors nowadays make textbooks available for free. And companies like O'Reilly, Packt, and Manning provide DRM free copies.
Binding and printing: $1.50
Royalties to professor X: $$
I literally changed majors back in my school days over those shenanigans by the department head.
I would expect better from Professor X though.
(in 2011) University tells incoming students they must check the book list for their course and get books before classes start. Books are expensive.
First day of class - professor gives everyone a PDF of the physics book I just spend $200 on. Says, "the university requires that I list at least one book for the course and they can't communicate with the students beforehand.
Never bought books before the first day of class again.
Pretty sure the publishers anticipated this.
You can pick one book for the prose and another for the practice problems and neither have to be whatever the professor recommended.
The other thing is that most TED talks follow a certain formula. They do because it works, but watching many of them it sticks out and can be a bit annoying. There's probably a meta TED talk pointing out this somewhere.
much better version.
Ever had a thought that just kept popping up in your head, that you then mentally suppress because "Dear <insert religious figure here>, how horrifying."
After doing that it keeps coming back again and again? Usually after a shorter and shorter period of time, because it becomes an indexing thought for recollections of all the times you've squashed it or thought about all the ways in which it was wrong?
Ever then talkedabout itwith someone else then had a cathartic release from the sharing, and possibly even a moment of relief that someone else besides you stumbled on <horrible thought> too?
It's the same dynamic. Just writ at a societal scale. It's why no censorship is a truly good idea. Bad ideas crop up organically as we've got a constant influx of new members of society seing the same warts, the same problems, following the same dead ends each and every one of us has. How are they to recognize a false start without being exposed to them?
The part that frightens people is that previously, you never had it possible for the whole world's population of false starters to get into one place and coordinate in real-time. You had an implicit depletion of critical mass activity that ensured the feedback loop for these ideas petered out, or remained tightly constrained to just the local nuts. The world we're in means that barrier has fallen, so everybody has to be able to cope with the entire planet's equivalent of the crazies all the time realistically.
It's a learned skill. Only practice gets you better at it, and I assure you, however good you think you are, cut the estimate in half. Odds are you've got fundamental blind spots you haven't even run into the craziness for yet.
There are still good TED talks, but you have to sort through TED the same way you sort through YouTube... and at that point, YouTube starts looking a lot better.
And "How to sound smart in your TEDx Talk":
I attended one TED back when they were super exclusive, because they were super exclusive. The good lesson: super exclusive for the sake of it a waste of time.*
TED talks were pretty much of the “I am doing this cool thing and isn’t it so exciting?” That actually does sound pretty cool: the world is full of exciting things I’ve never heard of or have heard of but never appreciated.
The reality, both in the old TED and the new: the talks are almost all structured in the same mode: “you are smart for listening to this info that other people don’t appreciate.” Validating the listener for spending their time listening.
If I wanted that I’d go on a walk with my dog. Which I do instead.
* things can be exclusive just because there’s finite space and you want to have people who contribute (e.g don’t invite me to a medievalists convention; I’ll have fun but won’t add to the discussion). This works if it’s not self-congratulatory and if people pro actively mix it up over time to bring in new ppl.
I'd happily stand corrected though.
Honestly, I'm happy as long as publishing companies don't force me to use their crappy proprietary ebook readers.
The entire thing is a racket.
The other part is that Pearsons lobbies, the tests are already accepted as mandatory and then not every teacher is capable of coming up with his own test that allows you to then perform well in the Pearson test. You probably also risk a lot if you used your own test.
It's basically industry developing its own industry.
Some education board (with heavy influence from various companies) decides that you need to select from a set of curriculum to be accredited. Other universities (and some employers) won't honor credits or degrees unless they're from accredited courses.
Pearson etc offer them a time saver they don't have to pay for - they take out the effort of assigning and marking exercises, and don't add any cost to the educator. The students pay, and Pearson uses these codes to inconvenience students and force them to buy new copies.
If you think like an MBA, it's a great idea - the person in decision making power benefits from this move (exercises handled for them), is probably unaware of the issue for students, and has more time to spend on something else they need to do (raising grants, trying to write papers to keep going on the tenure treadmill). But you also get to sell a new copy of each book annually, at full price, without used sales cannibalising your returns. Given there isn't a huge growth in universities each year, this is a great way to boost revenues sustainably.
In my experience on the other side of the pond, we used 1 textbook in the entire degree. It was reasonably priced, readily available second hand, and didn't have any single use license codes. Indeed, one instructor used to upload the relevant chapters of their own book as PDF files, insisting nobody buy the book as it wasn't worth it!!
All of my courses with smaller class sizes were much more careful with their book selections; occasionally those books were still expensive (even used), but only if they were really the best tool for the job. Even with expensive books though, this kind of online subscription nonsense was never used.
And it's especially prominent in math courses for students who don't know nor want to learn math.
This is totally false. State standardized test is the natural consequence of state public education accountability system. There must be a comparable measure across all the public schools in the state as the foundation for accountability measurements. And all (>95%) the students must be tested for equity reasons. No matter how many misinformation you read about standardized tests, the simple fact is that it is the only way to measure student learning of a huge population and get meaningful / comparable results. Is it perfect? No. Is there a better alternative? No.
The whole idea that a few testing companies imposed the standardized tests on students is simply a conspiracy theory.
The point here is- they lobby to get standardized tests mandated where they were not, then they lobby snd push their tests for statewide use very hard. All school districts have to buy test questions and other infrastructure from them only. And only Pearson certified teachers can evaluate those copies. And every step of the process involves Pearson getting paid.
This is like the Mafia.
Thanks for pointing it out.
Districts do not pay for the state test or buy test questions from Pearson...
Yes, they do.
> The federal government mandates this as a requirement for federal money.
Sure, some mandates come from the federal government; others are from state governments, and still others from local governments.
Guess who lobbies legislators making those decisions, especially at the federal level, but also at the state level, and sometimes, especially for larger jurisdictions, at the local level?
Obviously, they disagree, or they wouldn’t spend money lobbying, which isn’t something companies do just to burn money. They do it because they expect RoI on the money spent lobbying.
Compared to any of the individual incremental mandates? They’ll have a lower ratio of cost to instructional time and a more efficient educational system, probably. While the requirements are often tied to consequences, they are rarely tied to the kond of systemwide analysis of results against other factors and process improvement that would make them systematically useful rather than theater. They are generally “accountability” measures that assume without any empirical basis that the results are within the control of the entity held accountable and that the consequences attached will force decisionmakers to make changes that result in optimization of the results.
Oh, and standardized testing companies will make less money. Hence the lobbying.
Doing that, they can influence a lot on the scoring, quality, content and certifications around the standardization and fund the psychometric research establishing validity/equity of the tests.
Like any other big org biz, specing the RFP is one easy way to win the contract.
A decision marker (corporate or government) is likely to choose and mandate a pearson test rather than fund the research and development of their own which best fits their needs because it is easier.
Unlike other industries testing by is nature becomes monopolistic, nobody is going to say you can either of these three tests : it is hard to compare between two tests, so winner takes all.
Pearson and others in the space, exploit this by high prices , poor content etc because that's what monopoly tend to do.
This is all because of privatization of school vendors, why can't content and tests be developed by the state as well ?. Many countries do this.
Yes. There are places where tests were developed by the government agencies, which are not professionals. Mostly their tests lack psychometric regorious, not comparable across years.
I'm not sure what you meant "poor content". A few flaws here and there?
State government is not better at doing things than big companies. Hope your solution is not to ask the government to do everything.
Many of the big tests are developed within an academic setting with university/public funding then licensed out. Almost every book is written by university professor then published by pearson or others.
State government does not outsource writing the tax code or the other regulations to companies why is tests any different ?
If reviews can be done by psychometricians, why test building cannot ?
There are many many qualified professionals affiliated with major universities- several of them owned by the state. Is it difficult to imagine them developing the tests?
Many of these same professionals consult with testing companies for building the tests, so it is not like these companies have some magic expertise no one else has.
Using an vendor test gives that companies allied products basically an Monopoly for that test/skill.
This is a fact of this industry, and testing companies will use it to their advantage because that's their business. It is in public interest not to privatize the test itself, so there is competition amongst providers of allied products and free market can prevail.
Current setup is like government tax code is set by intuit and you can only file tax via turbo tax and nothing else
Yes. University professors do consulting works. How do they have the capacity to build a product as professional companies do? ...
All tests have frameworks underlying them , do you know how long it takes to establish a framework ? How many peer reviewed papers that have to published, and time it takes to shape the opinion of the ecosystem to make it acceptable? Without a established validity and equity nobody is going to buy the test, why would they risk it to get sued?
You think companies go through the full journey of years and take all the risk? in most cases they only come in the latter stages and commercialize and adapt it.
https://www.washingtonpost.com/news/answer-sheet/wp/2015/03/...
Standardized tests can be a window in the performance and health of the overall system, or over a group of students, but it is not reasonably determinate of the knowledge level of an individual student. Which is how they are often used
In the beginning it was used to judge the performance of a teacher or school, today is it used to judge the performance of the student something it is not really equipped for.
This is similar to BMI, BMI is useful as a health metric on a population of people, but is not very useful as a metric on the health of an individual
"Ohh its just some bot" should never be an acceptable response, and people creating these bots should have massive fines attached to their false positives. Otherwise there is no incentive to actually validate or improve what these automation systems are doing, the incentive today is the flag everything and worry about it later...
That is not something society should accept
I think it is, because the core problem is not the wrongful takedown of a single repository itself, it's the outlook that the only way to get these "mistakes" reversed is the hope that it will reach the news, Youtube/Google shows us what the latestage of this nightmare looks like. Github is obviously not nearly this bad (yet?). This case will obviously be restored, it's a popular author, but what about people that don't have this reach?
I don't see how talking about this situation can be a bad thing.
YouTube/Google is bad because they use their own system instead of or in addition to the system required to get the DMCA safe harbor.
For sites that simply follow DMCA, such as GitHub, there is no need for the issue to reach the news. There is no need to be popular or have any reach. The most obscure poster with no audience can easily get their content restored. Here is what happens at such sites:
1. Someone claiming to represent the copyright owner files a take down request.
2. The site temporarily takes the content down, and notifies whoever posted the content.
3. The poster files a counter-notice saying that they have the legal right to post the content.
4. The site put the content back up, and notifies the party that filed the take down notice, telling them who filed the counter-notice.
The site is then out of the loop. The content is back up, and the party claiming copyright violation cannot sue the site over this. The site is now in the DMCA safe harbor. If the party claiming violation wants to take it farther, they need to go to court and sue the party that posted the content.
The counter-notice is trivial to fill out and file. Here's the one for this case [1] if you want to see how simple the form is.
[1] https://github.com/github/dmca/blob/master/2021/04/2021-04-2...
this is textbook harassment.
no consequences for malicious or invalid DMCA is ridiculous (1$ penalty, paid to targeted individual/company for each invalid DMCA would be a good start)
/s