Hacker News Reader

Comment by pg | Hacker News Reader

This is the reductio ad absurdum of current law.

We need to work more actively to make this sort of thing stop. There are a range of solutions. We could try to change the laws. We could develop technology that would make it harder to catch file-sharers. We could attack the labels directly by trying to pinch off the sources of their revenue. I'm not sure which would work the best. My guess is that they're better at manipulating legislators, and the way to beat them is technology.

So here's a question for everyone: What new technology would harm the labels most?

View on HN

Indirect networking much like Tor or I2P.

Translated to BitTorrent, everyone would be using just one extra hop to keep the speeds up but passing around encrypted packages for others would create plausible deniability. You can claim you don't know what the packet that you routed contained, because you can't know that. You can't be held liable for routing traffic per se, much like ISPs aren't liable for routing traffic per se.

Why Tor and I2P don't want to use one-hop connections is that someone with government-level access could deduce the participants by looking at the traffic timing patterns. But MAFIAA at least not yet isn't actively hooked to the major ISPs of the world and can't apply such a traffic analysis. A simple one-layer indirection would hide who's downloading and uploading what. All you see is routing nodes.

If you're MAFIAA and seeding your fake movies and music to this torrent network, you can only see a user's IP address uploading a block to your fake clients. But since everyone is acting on behalf of others and merely routing packets, you don't know who sent the packet and you can't possibly prove that the one who did the last route was guilty. Otherwise merely connecting to the network would mean the user is guilty. The network would of course transfer legitimate files as well, and while downloading a Ubuntu 13.04 image users would also handle these one-hop routing requests of other peers who are downloading/seeding other files.

Interestingly, you would only need to transfer a certain percentage of connections over the one-hop tunnel. You could use zero-hop for the rest. The MAFIAA can't recognize those so it still looks the same as if you acted as the router and fetched a block for another peer.

This could also be implemented as a global free-to-connect IPv6 VPN network. If the IP addresses uploading MAFIAA movies are virtual, nobody knows where they are and who operates them.

http://anomos.info

Someone already attempted to make something similar to what you're describing, but it turns out that it's very hard to do right.

Unless, of course, the authorities seed Tor with a whole bunch of nodes that log everything. No timing attacks required.

Tor relies on the fact that even if there's one or two bad apples among the bunch logging everything you're probably going to be safe because you'll go through enough routing nodes that aren't logged for it to be impossible to trace.

Logging isn't enough. You would need to be able to prove that the encrypted blob originated from the potentially infringing IP address and wasn't just merely passing through it. Because if it was the potentially infringing node couldn't have had any clue about what the packet contained.

If all you can see is routing you can't know where packets originate or are destined to. You receive a packet, open your layer of encryption, see where the packet should be going next, log it, and send it forward. You can't possibly know whether the sender created the packet or forwarded it, and you can't possibly know whether the recipient consumed the packet or forwarded it.

You would need to monitor physical network links to determine patterns how the onion style packets travel around.

But with one-hop communications (as suggested by the post I was replying to) then the routing nodes (that could be made to log everything) know both the source and destination addresses as it is the only other hop in a one-hop path.

Is anyone of you using OneSwarm? http://www.oneswarm.org/

This is intriguing. I wonder how effective it is in obscuring one's identity.

Winny was very popular in Japan. http://en.wikipedia.org/wiki/Winny

I have given this subject quite a bit of thought, and I believe the solution is to fulfill this objective: Make filesharing simple and intuitive.

The article states that the girl ultimately failed to download what she wanted, which I don't doubt is due to the complexity of navigating a torrent index and installing and running a torrent client.

For most computer users, torrenting is too sophisticated, and it doesn't help that most torrent websites seem to be focused on scamming their users (including the venerable ThePirateBay). Other protocols for file-sharing are usually no better.

What if a technologically ignorant user can find, download, and play a work of video or music in just a few mouse clicks? I think this would have huge ramifications. Simply exposing file-sharing to a wider audience will increase political support and help counter the content industry's anti-filesharing propaganda.

I really like the analogy of 4 currencies when thinking about filesharing: http://www.gamasutra.com/blogs/LarsDoucet/20120222/91144/Pir...

Improving the user experience of file-sharing will not only reduce the first three costs to near zero, but after file-sharing becomes easy and accessible to everyone, it will become more normalized in society. In turn, this will gradually decrease the last cost, integrity, to zero.

The endgame is increased tolerance and political support of file-sharing, and disrupting the revenue of the content industry.

So how can we achieve this vision? I am thinking of a high quality package manager application similar to iTunes or Steam. The biggest usability improvement would be the concept of a repository, in that there is a curated database of files rather than whatever users want to upload, which is the standard right now for most forms of file-sharing.

As a consequence, there will be a canonical version of a title, removing the choice of picking between 10 different torrents and looking at seeders/trackers. Because there is now one destination when searching for a specific title, we can leverage existing and user-generated content to include things like graphics, videos, reviews, etc. on the landing page.

There are definitely huge technical challenges involved. For example, an application like this probably can't rely on a central server. So without a reliable server, how can we maintain a repository and support an easy one-click one-download feature? Moreover, we want to protect our users and implement some sort of private information retrieval.

This is an idea I've been thinking about for quite a while, and when I have time (not enough these days it seems), I have been tinkering around and trying to implement solutions to some of the technical problems involved.

I like that strain of thought with making it more easy for users to download stuff. But the problem with repositories is, that they are requiring a specific intent of the user to do something.

Let me try to contribute to the discussion with a few unrelated thoughts with the hope that some others are willing to share their own too.

I would love to see is for example an in-browser plug in, that is based on the torrenting idea. Similar to streaming, but just that the cached data is redistributed among those who stream a particular video. It somehow should reach a level, where users have a plausible deniability to not be able to differentiate between legal/illegal anymore and at the same time an implementation that is ideally only depending on an existing widespread standard/plugin.

Another idea would be a physical device that handles torrent downloads, i.e. the VPN connection, encryption of external HDs, and has a super simple and stupid interface to use. There would be lots of opportunities to cooperate with existing VPN providers plus advertising on trackers should be relatively cheap.

I like the steganographic approach, and it could also work with a repository. Consider the following:

-Let Pc be a copyrighted package in the repository

-Let Pd be a public domain package in the repository

-Let Kc be the "key" of Pc

-Let Kd be the "key" of Pd

The keys are stored in the metadata of the packages, so we don't need to break any copyright laws to get them.

Suppose we have the following operations:

-E(Pc,Pd,Kc,Kd) = X

-D(X, Kc) = Pc

-D(X, Kd) = Pd

With this infrastructure, if we want to get Pc or Pd, we simply ask for X and then decode our desired file. This also gives plausible deniability, but I'm not sure how this will hold up in court.

This is analogous to downloading some thing after encrypting it. Or a zip file with a password.

Its a clever hack though. Because you can always argue, that File you downloaded has meaningless contents and that some algorithm at your end is re arranging it as a meaningful form.

But the copyright enforcers can always argue this is a clever way of stealing something. More like smuggling a big thing by dividing into pieces and transporting them in surrogate objects.

Therefore it is key that the legitimate vs questionable content ratio is not getting too much out of balance.

I am sure it helped a lot for Megaupload's public perception that regular people actually used the service for sending files and were screaming about the government stealing their holiday videos. The narrative of the Torrentfreak story is awesome as it humanizes the subject and creates sympathy for the cause.

In an ideal case an approach would fill a minor existing need in a way that is impossible to separate from "other" content. I was trying to do this in the comment above by suggesting some form of torrent driven YouTube alternative (i.e. video embedding witin a browser), but contravert's encryption idea could really make this even more awesome.

http://bandcamp.com/

Edit: I'll explain. Cutting the middleman is the big idea though band camp isn't enough for that. The big labels have very strong marketing capabilities that individual acts can't compete with. Even YouTube, the land of indie, is dominated by the large labels. So maybe bandcamp needs a breakthrough marketing/distribution/discovery platform that can somehow be more meritocratic.

Bandcamp is glorious. I end up buying loads of music there because it's easy, high-quality and I feel like I'm supporting the artists directly, not just lining someone else's pockets.

Agree discovery is a problem. I rely on blogs for that but it's pretty niche.

How about a lawsuit insurance scheme? Pay $X per month and if you get sued the insurance company will take the case to court and pay any damages.

This would kill the effortless €600 extortion scheme, and probably also DDOS the court system.

Possible backfire: It makes the labels even more lawsuit happy knowing that there is a pile of cash there if they win, and / or the cash reserves of the insurance are not sufficient and it goes bust fighting suits.

Actually this would likely be very effective. It's like the traffic-ticket lawyers. Even if you committed the crime, they normally severely reduce the penalty if not let you walk.

Cash piles mean nothing, going to court is extremely costly. The defence lawyers would be very good at their jobs. I know here in Canada, the Carter defence (using an expert, arguing that given bodyweight and the times of, and amount of alcohol consumption, that the machine was calibrated wrong) against DUI was made inadmissible - which is currently being fought on constitutional grounds - as it was 90% effective at giving a judge reasonable doubt when a person blew under a 0.14 (legal limit is 0.08, but suspension above 0.05). Now the lawyers are experts on breathalyzers, will get a court order providing them with all maintenance on the machines so they can argue directly that the machines are calibrated wrong. They're apparently getting the success rates up above 60% now.

I suspect that lawsuit insurance already exists, probably meant for businesses. I wonder how much convincing insurance companies would need to offer this to consumers. Maybe there is some reason they don't think it would be a very good deal for them, or maybe they simply haven't perceived a demand and haven't bothered offering/advertising it as a result. Perhaps worth looking into, if you have enough assets already that a lawsuit would present a large concern.

I doubt this would work. If you are not willing to pay for a CD,DVD, Software...etc why would you shell out on a monthly basis on the off chance that you are caught?

I, and about fifty percent of people I know, would be happy to pay a small monthly sum (say, up to nine euro) to be able to listen to and download any music we want, at any time, without risk of persecution. Ideally, this money would go to support the musicians, but that's not necessarily a requirement. Unfortunately, no such service exists right now - most of the music streaming services only work for the US, streaming's useless to many of us anyway because we listen to music in the car and need it on CD, spotify doesn't have half the musicians we like...

There's lots of grey area between "doesn't ever pay for any media, ever" and "I refuse to pay $ 12 for a single DVD with unskippable ads and warnings if I can get the movie off pirate bay in seven minutes, wtf".

Forgive me if I'm wrong, however isn't this what Kim Dotcom was going to do with his music service? Hit the labels where it'd hurt them the most (less profits) - then the raid came.

I also wonder if Microsoft is currently thinking of something like this. I recently watched 'Halo 4: Forward Unto Dawn' whilst I've never played a Halo game or anything (I've just known about it), and instead of making a movie directly to theaters they release it free on mediums people enjoy (bit-by-bit on YouTube then released it as a movie). They definitely have the distribution and technology to do more of this. I hope they do, I really liked that Halo movie, I kind of want to buy an Xbox and play the games now... :)

Focusing on harming the labels directly is probably not the best idea, but unfortunately that's what everyone seems to have focused on.

To me the most important question is why do new musicians go to the labels at all? If technology can help at all, it'll probably be by creating solutions that address that question better.

Instead of better tools for distribution and discovery of old label music, it might be better to focus on the tools for creating new music, starting from composing/jamming to recording to releasing the music to listeners.

But, regardless of how good and effective such a technology may be, it will still boil down to the benevolence of the largest successful company that produces such technology.

For example, Amazon has done a great job of dragging books into the 21st century, but its evil colours are beginning to show. Going from paper books to ebooks is a bit like going from Encyclopedia Britannica to the Encarta CD-ROM, only worse because they can be remotely erased. I have high hopes for Wattpad to take us to the Wikipedia stage, but once they're there, who knows how evil they will turn out to be?

Similarly, someone may solve this problem for music temporarily with superior technology, but in the long run, it will come down to how greedy they are.

>To me the most important question is why do new musicians go to the labels at all?

For the same reason people would rather release an app in the Apple app store than on the WWW: exposure. Someone with an insane amount of talent will make less money than a mediocre performer backed by a media blitz from the music industry. I'll leave examples up to others to come up with.

It's the whole chain. Labels decide what is played on the radio.

What's this "radio" thing you speak of? I'm 40, and I haven't listened to the radio in years. My kids (14 and 17) have hardly ever utilized the radio and have no problem keeping up with pop music.

The radio is on the internet now. There are probably a couple of dozen iApps for it.

Actually a better phrase for this situation is reductio ad malum.

I agree that the laws shouldn't be as crazy as they are now, but why make it harder to catch file-sharers? It's illegal to share stuff like this, and I don't see any justifiable reason why this shouldn't be illegal. However, it should be a small fine in my opinion, and it's a waste of time to break down the doors of these people downloading music unless they are making a ton of money by selling this pirated content.

I don't see any justifiable reason why this shouldn't be illegal

Because sharing is good and it's not our problem if the currently dominant business model has troubles with it?

Even if you think that sharing music freely should not be allowed, the fact remains that we are dealing with an industry that refuses to adapt to new technologies and new customer expectations, and engages in despicable conduct to preserve their outdated business model, as this story illustrates. In this view, accelerating the change, or doom, of this industry is rendering a service to society. Letting them go after file sharers is simply helping them resist change, and only by fighting the symptoms (we're losing some sales) rather than the causes (it's easier to "pirate" than to download legally, legal downloads are too expensive and impose unwanted restrictions).

It will obviously be a startup company that solves this problem, and technology will be involved, but it is not fundamentally a technology problem. It's primarily a problem of misunderstood economics.

Instead of trying to monetize recorded music, which has virtually no hope of maintaining the artificial scarcity (and thus salability) it enjoyed during the last century, musicians need to return to their roots and find new ways to monetize live performance and other real-time interactions. Recorded music should become a loss-leader for these kinds of transactions. You may not get a $100 or even a two-drink minimum for "tickets" to a live stream, but you might also fill a few orders of magnitude more "seats".

The ultimate unit of marketable scarcity is immediacy. There is only one "now". Once it's gone, you can never, ever have it again. Barring the mass-availability of time travel, "now" isn't going to be commoditized, so it ought to be a fairly safe thing to base your business on. For example, "we'll post the recording for free next month, but if you want to see it happen right now, you pay (directly or indirectly) for access." Whatever drives the last nail in the coffin of the labels, I'll gladly wager 5 years of my life in a startup that it's going to involve this principle.

Second, there's a problem of social conditioning. Not so much on the demand side, but on the supply side. Musicians are taught to worship blindly at the alter of intellectual property. When Napster first appeared, I was a student at Berklee College of Music. That was pretty much ground zero and I remember it feeling like armageddon. Selling IP was all that is good and holy. To most of my musician friends and acquaintances, it still is.

This bias has to be overcome by helping musicians make more money from non-IP sources than from IP. That is the only thing that will convince them to throw away everything they've ever learned about their own industry, and throw off the yokes of their voluntary servitude.

With a nod to Etienne de la Boetie: "I do not ask that you place hands upon the tyrant to topple him over, but simply that you support him no longer; then you will behold him, like a great Colossus whose pedestal has been pulled away, fall of his own weight and break into pieces."

As an aside, my startup is out to make the music business a better place for musicians and their fans. We're looking for a UX/graphic designer who's also passionate about the profession of music. We can offer room, board and equity for the right person. Email me if you're interested (address in profile).

It strikes me that whoever solves the App Store Discovery Problem can apply the same solution to music. Right now all the biggest gaming companies are solving this problem by signing up (paying) companies with high DAU games to join a sort of gaming social network, sometimes unbeknownst to the user. The network then lowers the cost of user acquisition for all participants. Makegameswithus seems genius in the sense that making it extremely developer friendly will help them win the mindshare of developers and in time, will produce quality high-DAU games for their network.

Labels are essentially big gaming companies, pouring an immense amount of resources for those couple home-run hits. So maybe in the future Makegameswithus will spin off a Makesongswithus and make it easier for any amateur artist to promote their music on iTunes but with all the producer work outsourced like it currently does for art.

I think we need a decentralized, anonymous filesharing protocol, something like i2p or tor, but optimised / adapted for bittorrent.

It would have to be something that anyone can use and is a no brainer to get working. Could even use something like paypal / google checkout / bitcoin to allow artists to accept payments for their uploads.

Apple and Google have large enough platforms where if they publicized a "direct-to-artist" payment system, you could see it getting a lot of use.

Something like that would obviously not be healthy for their relationship with the record labels and would never happen. But you need a large enough platform to make a difference. What about Facebook?

Facebook won't be any good because the labels can just bribe them to turn you in after you use their system to download a song.

Distribution is solved. Marketing is not.

Every work of art has its audience. The focus now should be on helping the art and the audience find each other.

Reply on news.ycombinator.com