DMCA Notices Remove 8,268 Projects On Github In 2015 (torrentfreak.com)

An anonymous reader writes: Github's transparency report for 2015 shows that the site received many DMCA notices that removed more than 8,200 projects. "In 2015, we received significantly more takedown notices, and took down significantly more content, than we did in 2014," Github reports. For comparison, the company received only 258 DMCA notices in 2014, 17 of which responded with a counter-notice or retraction. In 2015, they received 505 takedown notices, 62 of which were the subject of counters or withdrawals. TorrentFreak reports: "Copyright holders are not limited to reporting one URL or location per DMCA notice. In fact, each notice filed can target tens, hundreds, or even thousands of allegedly infringing locations." September was a particularly active month as it took down nearly 5,834 projects. "Usually, the DMCA reports we receive are from people or organizations reporting a single potentially infringing repository. However, every now and then we receive a single notice asking us to take down many repositories," Github explains. They are called 'Mass Removals' when more than 100 repositories are asked to be removed. "In all, fewer than twenty individual notice senders requested removal of over 90% of the content GitHub took down in 2015."

Comments

Yeah....

[Anonymous Coward]

The DMCA is like the TSA:
In the words of the head of El Al, "The TSA isn't security. It's an annoyance."
The DMCA doesn't stop pirates, it just interferes with legit and acceptable things.

Re:Yeah....

[zrobotics]

Come on though, when I want the hottest new Hollywood movies or the freshest warez the first place I turn is github. Seriously, this is some weapons grade bullshit, I've never run into pirating on that site. I may have seen something that was patented, but I'm not sure that 20: print "Hello World" isn't covered by someone's patent anyway. If I sand round corners onto a block of wood, do I owe apple money?

Solution --

[Anonymous Coward]

Don't host anything in the US. Ever.

Host your projects in more liberal countries like Russia or China.

My icon...

[Anonymous Coward]

I had a 16x16 icon of a question mark that a popular project was using that had forked several thousandth times... I made my request and they were all taken down. Justice served!

Re:Definition?

[henni16]

The obvious case would be people using it like any other file(-sharing) host.
I don't think there's much if anything stopping you from adding e.g. an archive containing a movie or a bunch of ebooks into a throw-away repository and linking it from somewhere else.

Using Git might even make it easier to reupload stuff after one repo gets taken down - just add another remote to your source repo and push it.

Commercial source code, copyrighted graphcs, etc.

[Anonymous Coward]

GitHub actually them available online: https://github.com/github/dmca

Some of them are like someone using one of Adobe's .svg graphics or a commercial script file or source code file being distributed.

And some of them are abuse. Like this: https://github.com/github/dmca/blob/master/2016-01-15-AlgoTrader.md and they did license their source code under the GPL as shown here at the bottom https://web.archive.org/web/20140208002555/http://code.google.com/p/algo-trader/

So it looks like most DMCA takedown requests are legitimate, but presumably a small percent are abuse.

counter notice

[Anonymous Coward]

I was one of those counter notices. An ex business partner tried to claim ownership of my work, after I pulled my code from the project. Github disabled and then re-enabled my repository, within 1 week of sending in a counter notice. The person claiming infringement profited from my work without giving me a cent. He failed to file against me in court, as he was as broke as I am and had no merit to his claims.

Re:Definition?

[nmb3000]

On github, what constitutes being a valid target for a DMCA takedown?

Well, you can look at the full list of DMCA complaints they have received and see for yourself.

I randomly looked through them and many appear to be pretty reasonable. Many are related to simple copyright infringement, such as storing textbooks or published homework and test questions and answers. This isn't surprising, as GitHub is basically just a place you can store files.

Some say the code or data is internal or non-public and was uploaded without permission. There are also a bunch from Qualcomm complaining about firmware images and driver code. VMWare complained about internal roadmap documentation.

Some are just files with links to other materials, such as TV shows and movies. Sony and Marvel make appearances.

The line blurs some for others. There are HTML5 versions of classic games, such as ones from Nintendo and Blizzard, that got pulled down. A few companies appear to have searched GitHub for serial numbers and license keys of their products, and requested the entire repo be pulled if it contains a single file with a serial number in it. Many of these appear to be honest mistakes and have counter-notices.

In any case, it's nice to see GitHub being transparent. The DMCA requests themselves are pretty interesting, but since the vast majority of the targeted repos are no longer accessible, it's hard to gauge how justifiable most of the complains really were.

How many false claims?

[lapm]

I wonder how many of those takedowns were done under false claim. Meaning its fraudlent DMCA takedown request... In youtube they match all sort of shit for DMCA takedown making sometimes just unbelievable claims. What chances does small person or developer to really challenge those false claims. Example in my country its perfectly legal to reverse engineer software and make compatible software. Our law also docent know shit like DMCA and hopefully newer will.

Re:How many false claims?

[ATMAvatar]

DMCA full text

The DMCA contains the word perjury twice: once in relation to the person making the claim, and once in relation to the person making a counter-claim:

(vi) A statement that the information in the notification is accurate, and under penalty of perjury, that the complaining party is authorized to act on behalf of the owner of an exclusive right that is allegedly in-fringed.

You'll note that the only statement subject to perjury is that the claimant is authorized to act on behalf of a rights owner. The first half, about accuracy, is not subject to the same penalty. The prior section (v) mentions that a statement must be made that there is a good faith belief that the subject of the claim is infringing, but it makes no mention of any penalties if this is untrue.

There is some token language later on about misrepresentation - that the claimant is liable for any damages incurred by the alleged infringer should the claimant knowingly misrepresent that the material is infringing. However, proving there was intentional misrepresentation is a pretty high bar, and in most cases, the damages are low enough as to be not worth the legal fight anyways.

"Transparency" Report Features a Few Blindspots

[Kunedog]

Why only list DMCAs? Because they're a special subset of cases that sometimes makes Github look like the good guy? How about listing the repos they forced offline themselves for petty ideological reasons?

The 2014 report failed to list the takedown of the Gamergate hub, and this 2015 report doesn't mention how Github took down WebMConverter to strongarm the developer into changing its content.

Instead of using a broad, feel-good word like "transparency," they should just call it the DMCA report since that's the one specific kind of censorship it discloses.

Re:"Transparency" Report Features a Few Blindspots

[AmiMoJo]

The GamerGate repository was removed for TOS violations, specifically that it was being used to collect information used to harass and dox people. Interestingly it also rather gave the game away for GamerGate; the readme.md file didn't mention ethics in journalism until several paragraphs down. Instead it started with a rant about feminism.

As for WebM, they were asked to remove the word "retards", and did so. It really helps if you explain these things so that people can make up their own minds as to if it was justified. It's almost as if you were hoping they wouldn't bother to check.

In any case, those things were not included in the transparency report because they were widely reported (i.e. already transparent). The transparency report is a list of otherwise non-public requests from outside parties.