How to prevent screenshots in a web browser
patrick-wied.at
patrick-wied.at
Even look at music piracy. Pirates did the big labels a favor: small, independent artists can't make a living anymore because everyone just expects it to now be free. They now are almost forced to go with a big label if they actually want to make any kind of living.
Imagine you have a stock photo website. Basically honest potential customers might try right-clicking a photo once to download it, and failing that will whip out their credit card. Basically dishonest non-potential customers aren't going to pay you and see your protections as something fun to be worked around.
There just isn't a business case for heroic measures that can ultimately be trivially defeated.
You have no idea how much time I've spent on sites doing weird things to try and stop the user from downloading their content. Like online schools that don't want anyone to save the class-lesson videos. I don't even want the media, I just find it irresistible to understand and ultimately defeat their protection.
DRM can't help artists make money. It only helps them lose more.
Examples which talk about it:
* http://www.lexi-alexander.com/blog/2014/6/1/will-the-real-pi...
> And this is my favorite part: I am 100% certain that the hacking of entertainment industry's security features provides better entertainment for these kids than the entertainment we're trying to prevent them from stealing. Let that sink in for a second, then try not to bust up laughing.
* http://brutalgamer.com/2012/03/09/cd-project-red-drm-no-long...
> Every subsequent game we will never use any DRM anymore, it’s just over-complicating things…We release the game. It’s cracked in two hours, it was no time for Witcher 2. What really surprised me is that the pirates didn’t use the GOG version, which was not protected. They took the SecuROM retail version, cracked it and said ‘we cracked it’ — meanwhile there’s a non-secure version with a simultaneous release. You’d think the GOG version would be the one floating around.”
The people releasing the content do not care about the actual content, they are not fans, hell most of them probably never watched/played/listened to the movie/show/game/song they are releasing.
The Act of cracking, releasing, encoding, etc is the entertainment for them... The content is irrelevant
All the money is in touring and merchandising.
This is a point I hear often, but there seems to be no consensus. If I understand you correctly, are you saying that small, independent artists used to make a living just from CD sales?
I've only met in person poor musicians, those that make money either from street concerts or the occasional gig, which were consistent with Courtney Love's position[1] that CD sales are not really important when it comes to revenue. But I'd love to hear an argument from the other side.
I suspect that your DRM was so overly burdensome that it turned away any potential customers, you in effect screwed yourself.
If the DRM does not get in may way, or require me to do anything(i.e it is invisible to me) then I might buy that software, for example Steam Games, Spotify, etc have DRM that is invisible to the End User. I will grin and bear this level of DRM, begrudgingly. However if you go much beyond this no thank you.
If your software has Activation Keys, Installation Limits, Locks itself to my Motherboard that I change ever 6mos, or other crap like that then you would never get a single dollar from me.
If yes, then I think you agree in principle with the goal here, even if you disagree with the specific use case.
As an amateur photographer you'd be amazed how many people think nothing of taking an image and slapping it on their website, or in their print media.
People routinely use services like imgur.com to rehost images they have no permission to use, and can and will strip watermarks and be silent when asked for attribution.
I'm used to it now, but damn it's frustrating when you take that _one_ great shot and find it everywhere on the internet the next day with no credit.
If the watermark can be removed without making the image largely useless then you're not watermarking properly. Sure, its not pretty, but this isn't a pretty world. I never understood the lazy bottom-right watermark. Its trivial to remove and sites like reddit reward fraud because people like to pretend its their content as narratives like "Oh i bought a camera and took my first photo" are eaten up by low-information voters (who i imagine are the plurality of votes on reddit).
Newspapers, big online sites that should know better. They take pictures and half the time you never even know unless you use tineye, or a friend/colleague will share a link with you "Hey nice work getting included in the New York Times", etc.
I do understand hard work, and skill, and putting in the time that made you the great photographer that you are, and I understand the need to earn a living, and I understand capitalism, economic profits, and accounting profits and economic profits.
But a large number of highly trained and skilled people put a lot of hard work and effort into writing the open source software that is running the internet and the webservers that your images are on.
I understand your perspective: you'd rather buy a proprietary microsoft operating system and a proprietary microsoft webserver so you could run a proprietary Oracle database to keep track of your DRMed photos.
But understand that I wouldn't, I'd rather have a public domain web where photographers share their work like artists do and computer programmers and graphic designers and bloggers and everybody else.
I'm just not that excited by your great photo, there are a lot of great photos.
full disclosure: I took a significant news photo once and made money selling it. I enjoyed spending the money. I don't chase around the internet stopping people from duplicating it.
Most of my pictures are prints, hanging on walls. But you need a portfolio to attract (paying) clients, so images have to exist!
> I understand your perspective: you'd rather buy a proprietary microsoft operating system and a proprietary microsoft webserver so you could run a proprietary Oracle database to keep track of your DRMed photos.
So if I said I'm an ex-Debian developer who's released a pile of open-source software, for free? And does so regularly?
Seriously you're getting too personal there, and just wrong. Photography started as a hobby because I love people - in the same way that woodworking is a hobby because I like doing things with my hands that are "real".
But there are costs involved, I've probably spent close to 10,000 in camera-equipment, to have people steal your work, means that stuff doesn't pay for itself any more.
You mention proprietary software? How would you feel if you wrote open-source software and people abused the licenses? That's copyright infringement, and the same thing is a huge deal in the photography world.
Sure everybody has a camera, and fly-by-night people will take money for pictures. But that's like hiring a self-taught PHP-programmer instead of a kernel developer. There are skills, styles, niches, and similar involved that clearly differentiate photographers.
Is it big enough to start deploying the lawyers? That's what software companies have to do. If you're not doing that, then it must not be a big enough deal to you. If you CAN'T do that, then I think you should get out of the biz.
But, I bet you won't because you are probably getting something more than money out of your career. Maybe you enjoy the work. If it were all about money, you would have become a banker, right?
No sympathy here either. You have no natural right to get paid every single time one of "your" images is used. (Oh and who owned the stuff you took a picture of? Should you pay them to photograph it? Should they get paid every time the image is used?) With no natural rights, you have to use the legal system. Nobody cares if you sue some big corporation, but please...complaining about people posting stuff to imgur.com? What money was lost because your image showed up there? Did you try sending them a cease and desist? I bet they'd take it right down. If you don't do that, then you're not trying hard enough.
I have sent cease and desists in the past, in the same way that I've sent DCMA takedown notices against people who post documentation I've written with no attribution, and against people who've forked my code and pretended to have authored it.
But I'm done with you now.
That was really my main gist. Sorry for not being sympathetic. I thought I made a pretty good argument. I was not trying to be impolite, but I can see how it could be taken as such. Could you just answer that one question? I'm really interested in the fine line that is being discussed here.
1. The article is an experiment on techniques and what can be achieved.
2. At the VERY BEGINNING of the article, the author links to his own comment in a previous HN post which said "I do not advocate using any of the techniques, please don't use it, it's terrible and will give your users headaches! I'm against ressource "protection" on the web, what's on the internet should be fully consumed by users of it! The article describes just a few experiments I set up to see how far someone could go if it WAS important."
3. The article acknowledges these techniques are ultimately stupid.
What is this attitude? Would you have the same attitude with someone experimenting with XSS flaws, hacks and so on to figure out what is possible? Did you all forget than, in order to do good, you have to know what is bad first?
Regarding your points 1, 2 and 3... where does it stop? Should we always be enthusiastic about experimentation for the sake of experimentation? At what point would you say "that's stupid"?
For instance, what if the title was: How To Dump Toxins In Your Local Water Supply
Would you read the article and then come here and congratulate the author on their efforts? I bet most people would glance at the article and then come here and complain about too many toxins in the water and what a jerk the author is.
You know where it stops? It stops when you start putting people in danger. I cannot believe you're seriously asking that question (and yet, I have to believe you are seriously asking that question, because the alternative is that you're an awful troll).
I would like to encourage you to read my comment, and then read it again, and keep doing that until you understand the difference between killing people and blogging on the internet. I would then like to encourage you to read up on things such as "white hat hacking", pentesting and so on to give you an insight as to why, in the IT world, if the good guys don't push the limits first, the bad guys will.
Let's try another hypothetical title: How To Force Your Neighbors Take Out Your Trash or, How To Write A Virus That Steals Banking Information...How To Really Embarrass Your Coworkers...9 Ways To Enrage Internet Commenters...(this is fun!)
Would you applaud those experiments? I'm guessing YES since "It stops when you start putting people in danger." and nobody is being put in danger.
(BTW, I'm just doing my own experiment here - I guess I should have said that. You're not being a very good sport though ;) Seriously - I'm sitting here trying to understand your point of view and I made some POLITE arguments. You mis-represented one of my arguments and then patronized me to the extreme, so is that the kind of behavior that you're promoting?
Also this part:
> seems like there is a problem — let’s solve it for fun and profit
It's like saying: "Police state struggles with policing measures. Let's help it for fun and profit". I hope you get the idea why this can sound offending.
Ksnapshot, imagemagick (import), ffmpeg, avconv etc. etc.
> let’s assume things on the web should be protected…
Let's assume DRM is a dumb and nasty idea which always leads to very crooked practices. Period. As if we don't have enough of this EME nonsense.
What if there was a way to provide access to your medical records to people who need them, only for as long as they need them? What if even the provider of the service couldn't easily circumvent this?
DRM can not be trustworthy by its mere definition (because trust is a mutual thing, and those who deploy DRM don't trust you - the user. So you have every reason not to trust them in return). And it never can be valuable since it's not only unethical and insulting towards users (since it uses presumption of guilt and police state ideas), but it's also completely ineffective and all it does is punishing paying customers, while the vast majority of actual pirates don't deal with it. DRM should just die out for good.
> What if there was a way to provide access to your medical records to people who need them, only for as long as they need them?
Authentication and information security has nothing to do with DRM, even though both can have features like encryption. Let's not mix unrelated subjects.
Authentication and information security has nothing to do with DRM, even though both can have features like encryption. Let's not mix unrelated subjects.
What is the limiting of who can have my personal data and when, but Digital Rights Management, or DRM? One cannot stop at authentication, because to be complete, one has to prevent copying data. Without this, it becomes problematic to trace who has released information, and without some legal support, nothing at all is practical.
DRM can not be trustworthy by its mere definition
Sorry, but this is fluff. You could apply this "logic" to authentication as well. DRM has to involve some mutual trust and cooperation to the same extent that any protocol involves mutual trust and cooperation.
The key is to realize that DRM as practiced by corporations is evil and unworkable because the economics were utterly unrealistic. Really, where DRM could help individuals would be the elimination of trivial deniability on the part of corporations. Right now, when your data privacy is breached, there is no cost to "break" anything and it's very hard to pin this on a particular corporation. Add a DRM mechanism, and then there is something tangible to apply protection laws to and a labor cost with definite intent to breach an individual's privacy. In aggregate, the cost and potential legal liability becomes too high to for virtually any corporation to contemplate. We would arrive at a situation where corporations could afford to violate the privacy of a few select individuals, but could no longer do so to the public wholesale. Ordinary citizens would enjoy a measure of protection, though rich individuals and corporations would not.
Sounds pretty good to me.
Underlying everything, I'm not so much advocating DRM, as I'm calling to question emotionally charged "magical thinking" combined with naive induction. Let's not just leave things at "DRM is bad" devoid of real examination of what's happening.
EDIT: It also occurs to me that the historical factual differences between individuals and corporations should be used to modify the current formulation of "legal persons." "Legal Persons" should be considered as "persons" only in terms of a few specific factors, like the owning of property. The historical difference between actual and legal persons with regards to data privacy abuses would be an important corpus of information in support of this.
DRM can employ authentication in itself, sure. But DRM is not about ensuring that you are you. DRM is about limiting what you can do with your own system in various ways and / or spying on you (even when it already knows that you are you). Simply because DRM is preemptive policing.
I have no problem with authentication, same way I have no problems with you using a lock on your door (normal security). I have problems with DRM, same way I'd have problems with police putting their camera in someone's house (overreaching preemptive policing).
Hopefully this makes it more clear. I often see people confuse DRM with merely security / authentication / encryption etc. That's plain wrong.
> The key is to realize that DRM as practiced by corporations is evil and unworkable because the economics were utterly unrealistic.
No. It's evil because it's using police state methodology of treating everyone as potential criminal and employing presumption of guilt, which results in overreaching policing measures which violate one's privacy and security.
Isn't that what you're doing when you lock your doors though?
No. You lock your doors to protect your house from external threats. Here it's completely reversed. DRM invades your private digital space (your computer, your system, the programs you run etc.) for the sake of policing you. I.e. it's not like a lock on your house doors. It's like a police camera placed inside your house. That's exactly what makes it overreaching and unacceptable.
Policing itself is a not an evil idea in general. But it's evil when it's overreaching. Saying - "let's prevent crime" is OK. Saying "let's place police cameras in everyone's house to prevent crime" is not OK.
In the personal DRM case it would not be invading "your private digital space", you would be using it to protect your digital space.
It is not like "a police camera placed inside your house." It is the equivalent of the homeowner placing a security camera themselves to protect their property.
Protecting your own digital space is called security. Let's be clear on terms usage, otherwise time will be wasted because of misunderstanding.
DRM is always about policing others by invading their digital space.
Or about voluntarily giving some autonomy up. In the case of organizations like Facebook, they would be giving up autonomy, such that they would only run certain audited versions of certain software. Given that such companies have large numbers of computers aggregated in a relatively small number of locations, the economics of verifying these mechanisms is much more favorable than the inadvisable "traditional" use of DRM has been.
You are not going to be able to process ideas like this properly, if your only background is uneducated Internet backlash.
Protecting your own digital space is called security. Let's be clear on terms usage, otherwise time will be wasted because of misunderstanding.
Throughout, you have been insistent on an imprecise, popularized usage of terms. I will agree, however, that time has been wasted because of misunderstanding.
They created the term and they polluted it for good. Trying to whitewash it now serves no useful purpose except causing confusion. If you want to talk about concepts of protecting your personal information - just use another term, otherwise misunderstanding is guaranteed (like above).
> popularized usage of terms
No, it was you who tried to create your own interpretation of DRM which differs from what its designers put into it. That's up to you, but don't expect anyone to understand you.
Again proof that you have no familiarity with the abstract concepts or its history, prior to the popularized furor. Nor do you care or are particularly curious, or are capable of processing the logical implications of such new information.
No, it was you who tried to create your own interpretation of DRM which differs from what its designers put into it.
Let me assure you that the designers of DRM had the other interpretations in mind the whole time. You are basically arguing for your own ignorance.
don't expect anyone to understand you.
In other words, I should expect only uninformed rubes on the Internet.
Let me explain AGAIN -- I'm not talking about such a system running on your machine. I'm talking about such mechanisms running in Facebook's, Apple's, Google's, and the NSA's server farms. It's about 4 messages into this thread. Please get a clue.
Which is? Security threat for them (same as DRM which runs on your machine is a security threat for you). It doesn't sound sensible.
1) That's the same kind of half-baked wishful thinking that's been going around the Internet since before the web. 2) I'm not advocating doing what loser corps have been pursuing. I'm advocating flipping the economics around 180 degrees.
someone else breaks it shortly thereafter, whether it's DiVX, or DVD encryption, or BluRay encryption
It's wishful thinking to ascribe magical powers to anything, even hacking over the vast labor pool of the Internet. Modern content encryption based on VMs gets broken but takes labor to break, and only works as well as it does for mass-distributed media because enough people care about it. (I'm not talking about DVD encryption and earlier. Keeping a single symmetric key secret is fundamentally broken.) Flip around the economics, so equally powerful protections are applied to data only a few people care about, and the addition of a little legal support would be enough to form an effective deterrent.
It is fundamentally broken -- for corporate mass distribution. But that's not the only thing it can be used for.
What is fundamentally broken is the idea of overreaching preemptive policing (DRM). I.e. the one expressed by Sony execs in the past:
https://en.wikipedia.org/wiki/Sony_rootkit#Background
> "The industry will take whatever steps it needs to protect itself and protect its revenue streams... It will not lose that revenue stream, no matter what... Sony is going to take aggressive steps to stop this. We will develop technology that transcends the individual user. We will firewall Napster at source - we will block it at your cable company. We will block it at your phone company. We will block it at your ISP. We will firewall it at your PC... These strategies are being aggressively pursued because there is simply too much at stake."
It's as broken and crooked as it gets.
What is fundamentally broken is the idea of overreaching preemptive policing (DRM).
The record clearly shows that applied to individuals, DRM becomes "fundamentally broken overreaching preemptive policing."
The same record clearly shows that turned around 180 degrees, DRM is entirely appropriate. What's broken for protecting a single piece of information the whole Internet wants becomes viable for protecting information chiefly valuable to the individual. DRM applied in reverse against corporations is "economically viable historically supported policing."
I think you simply use the term DRM incorrectly. You think DRM means information security. It doesn't. "Turning it around" makes it not DRM anymore.
I think I've made the mistake of discussing while assuming the wrong level of information background amongst the participants.
They mean what they were designed for. And that is excessive preemptive policing. About "trusted execution" you are wrong. Original meaning of trusted execution is exactly about security, not about DRM. Actual usage of that term is more than often happens to be about DRM (which is conflicting with security actually, because DRM can be viewed as a security risk).
See https://youtube.com/watch?v=XgFbqSYdNK4
If you constantly mix up security, authentication, trusted execution with DRM, I think you should consider your own level of knowing what you are talking about. I explained everything above.
Now the only way you can steal my sweet, sweet content is by writing it out letter by letter or writing your own noble script that filters out all the junk. Checkmate.
You might say that this solution is unintuitive, massively inefficient and completely unnecessary. Well that's not the true Hacker Spirit now, is it? " ;)
That should be one line with no spaces if gets line wrapped.
I don't think there's any easy way for the obfuscation to stay ahead of JS reversal. Thanks for taking a look!
If those support HDCP then it's encrypted all the way to the display, and it would be challenging to screenshot it.
(Of course, you could take a photo of the screen, but that's a substantial degradation).
The site being slow is a better deterrent than anything written in the post. 9 Seconds to first byte[2].
[1] - http://www.patrick-wied.at/talks/image-protection/demos/cat....
1) Build a library for hiding encrypted information inside images. It should be hard to detect (indistinguishable from random) and robust (e.g. survive printing and scanning).
2) Build a web crawler, coupled with a key store. Crawl the internet for images that contain our encrypted info.
3) The copyright owners will be our paying clients. They will display images only to logged in users, use our service to embed the user information in the image, and deposit the encryption key with us.
4) Whenever the crawler finds a stolen image, we notify the copyright owners and send them the details of the user who leaked it. We don't notify the user, of course.
Right?
otherwise known as Steganography.
That may be an idea for a copyright troll, btw. Disseminate not-for-commercial-use for free, then charge dozen of thousands of dollars when infringement is found.
(You can guess how well that turned out for them.)
I've got to hand it to them, that is sneaky, but a bit underhanded. And you could sort of see something was a bit off in screenshots of the more gradiented blue skyboxes (in which it was more visible), but I'd thought that was just a poor JPEG encoding routine, until someone looked closer of course.
Shuffle an image, re-assemble with a bunch of divs with background offset and just the right amount to bring the picture back. If you download the image you get a scrambled imaged.
Doesn't protect against screenshot.
Very fun :)
2. Stop trying.
I was curious and tried his live demo URL. and I used OS X's built in screen capture while using Chrome. The screen capture worked, no special tricks needed. I didn't see any visual degradation between my screen shot and the original.
The long and short of it is his DRM did not work.
If you want to give out indelible images then go back to print.
For this to work perfectly, your monitor refresh rate would need to be at least 2x higher than what your eyes can detect. And why would anyone build hardware with such a fast refresh rate?
It's a cute idea, but in the end it ruins the image for normal uses just like the watermarks do.
EDIT: Ah, I see he uses window.requestAnimationFrame [1] to sync with the frame rate, so this is not a problem.
[1] https://developer.mozilla.org/en-US/docs/Web/API/window.requ...
Here's the relevant code from the demo:
var work = (function() {
var toggle = false;
return function tester() {
frame1.style.opacity = toggle?.5:1;
frame2.style.opacity = toggle?1:.5;
toggle=!toggle;
requestAnimationFrame(tester);
};
})();
requestAnimationFrame(work);I am associated with a... certain industry... (ahem) which frequently finds its content reposted (and it wouldn't be right or proper of me to identify any further, but I'm sure it's a moderately generalised problem).
One must be realistic about these things - if we're going to publish things, some people are going to pirate them and that's that. It leaves our control to some degree when we let it loose, regardless of what legal rights we may have. We're going to find them plastered all over Tumblr or TPB or something no matter what we (or Tumblr!) do. That is not necessarily all a negative, especially if it might drive interest and traffic, which may be worth more than any one or two pieces of content for some. If we get enough customers to support ourselves and what we do, that's fine (and we can thank our loyal repeat customers for that; many others are simply not in such a position).
"The invisible wall" might be worth it as a courtesy "please don't save this image" thing, but developer tools exist in browsers, so that really is only a courtesy thing: anything else is a fantasy and nothing else is really workable, but perhaps it presents a sort of reasonable-ish balance, if it doesn't degrade anyone's experience and if people are at least aware we'd prefer that not to happen on balance. I feel the same about any light DRM (for example, for PC gaming, I'd say Steam is fairly "courtesy DRM" compared to its peers, excepting its peers like GOG that have no DRM at all).
(I'm also a reverse-engineer of some experience, so I've found myself analysing others' solutions in other industries quite a lot over the years - it's interesting to look back and see how my own opinions on this matter have evolved from a vehement rejection of any DRM as a matter of principle, to a slightly more nuanced position that it is occasionally not devilspawn but we'd still have a better world without it and anything that may cause users any inconvenience is completely unacceptable.)
Fairly discreet and tasteful watermarking is all we've ever found to be a reasonable solution. What we see when people do repost them, as a result, is more sales - when a few people obviously like what they saw and looked it up. Not a lot, but some. Yay. Have we gained more than we've "lost" (not that we ever had those in the first place)? I don't know, I don't have a crystal ball to guess at events that have never happened, but we've gained some as a result of this, and some is always better than none.
Except those people who edit or crop out the watermarks then reblog them. Fuck those people. That gets right up my nose; there's piracy, and then there's plagiarism, and that's how I interpret that one - it feels like they're taking credit for it (and if they are, that's dangerous: people might think they know what they're doing and get themselves in dangerous situations), or at least they're definitely removing ours, which is irksome particularly when you see a comment thread asking for, as the parlance goes, "sauce" (the source of the content). Those are people who wanted more, dammit, and you cropped out the thing that could have linked them where to get more, and since it's been reposted, they probably don't even know. That's just putting barriers in people's way for no good reason. (Thank you for those few that do reverse image searches in such cases and find us that way anyway, you're heroes. ♥)
Any thoughts?
If a customer willing to pay for content is really intrigued by your image, he'll find the sauce.
[1] http://graphics.stanford.edu/papers/protected/protected.pdf
Nonetheless there are pretty smart ways to monetize content like this and any professional source will pay to use it.
SyntaxHighlighter Can't find brush for :jscript
and then the whole browser became unresponsive and needed to be restarted. On the plus side, you've helped Firefox development by discovering a new bug!
But cool write-up, I guess.
That said, i took a screenshot on windows 8.1+Firefox and it worked just fine. I took a few more screenshots.... and they were all fine too...
1) Ship your own lock-down hardware (tablet) to customers, so they can visit your website
2) Install booths around country which will show your website. No cameras or phones allowed inside!
The solution is trivial: launch a VM, make a screenshot, hook APIs such as DirectX, etc.
otherwise, on a regular screen its barely noticeable. at first glance it looks fine, but then i get the feeling something is off. if i blink or look away quickly, i can see some of the interlacing. but overall, interesting idea of treating the images more as movies rather than images
2 minute job, tops.
For most people (especially the billions in the developing world) it's a good use of their time.
Infuriating, bad design then. Still is.
Disclaimer: I only skimmed the artcle