On April 5th you asked Ian Clarke of FreeNet many questions about this new project, which is designed to permit almost totally anonymous Internet posting of almost any kind of material. Here are his answers.
Who is liable?
You said that this: "allows information to be published and read without fear of censorship because individual documents cannot be traced to their source..."
I'm all for an open forum for free speech, but this seems almost reckless. In most venues of speech, accountability for someone's words is fundamental. The Internet has opened up the possiblity of free speech without accountability to a small degree, and look at what has happened. Do you fear any legal repercussions to your group for creating this forum based on this fact?
Thanks for your question tcd004. The problem is that it is sometimes impossible to have free speech unless it can be delivered anonymously, since the threat of retribution can be a very effective deterrent against people stating their opinions. I would even go so far as to say that a forum which forced people to identify themselves was not permitting true free speech (This is why Slashdot allows "Anonymous Cowards" to have their say). I should further point out that there is no reason that people cannot digitally sign information they place in Freenet to indicate that they are the authors of a piece of information, but we don't force people to do that. Someone could even build up an anonymous reputation by signing all of their work with the same private key.
My personal feeling is that liable, and liable law, assumes that people will believe everything they read. This might, to an extent, be true in this time of centralized media, but my hope is that systems like Freenet (and indeed SlashDot) will encourage people to make judgments about the reliability of information themselves rather than relying on a corruptible centralized source.
Why the name FreeNet?
Maybe I'm just showing my age, but to me a "FreeNet" is a local free Picospan/shell account. Maybe it's a bad idea to take the name of an existing and quite venerable free service?
Here is part of the Detroit Freenet FAQ:
* What is a Free-Net? A Free-Net is a free, public-access community computer system. Free-Nets can serve populations of any size, from large metropolitan areas to small cities and towns. They offer a wide spectrum of on-line information services to the public, including community and government databases and worldwide electronic messaging. They don't charge for their services, so everything on them is free. Free-Nets also have an interactive aspect, in that users can dialogue with information providers. While there are many Free-Nets around the world, each Free-Net is tailored to meet the needs of the local community, so no two Free-Nets are identical.
Seems like the existing Freenet is already a very good and useful thing, and it really doesn't need the confusion.
Basically our experience is that while other systems and projects have used the term "Freenet" in the past, for the most part it had fallen out of common usage (if you don't believe me just take a look at the current contents of alt.freenet - which is primarily spam). We believe that since we are developing a "Free Network", the abbreviation of "Freenet" is too good not to use, and since it isn't really in common-usage anywhere else right now, we have resurrected it.
Poisoning the waters
by Mike Schiraldi
What protection is there against someone poisoning the system with malicious data? For example, let's say MPAASoftRIAAOL Corp. sets up a system of computers all over the place with wildly different IPs, and they feed either random or specially crafted bogus data into the system.
This is sort of analogous to renaming Barry_Manilow.mp3 to DaveMatthewsBand.mp3 and putting it on Napster. How do we prevent it? Some sort of decentralized, everyone-is-created-equal moderation system?
You raise an interesting point, and one which has created much discussion on the mailing list since these "cancer" attacks are probably the most difficult to defend against in any system, particularly if it is decentralized like Freenet. Right now, the way Freenet works limits the damage that can be done with such a node, but in the future we plan to implement mechanisms which will make such an attack even less effective. Basically Freenet avoids becoming dependent on any particular node since requests will be spread almost equally among all Freenet nodes (by the nature of the dynamic caching and mirroring). This means that even if 0.1% of the total number of Freenet nodes are corrupted, at most about 0.1-0.4% of requests will be affected by it. Right now whenever someone requests some data in Freenet, in a sense they are voting for it, and all of the nodes which were involved in retrieving it. This is less than ideal since you don't actually know what you will get until it is too late (much like voting in the real world!). We plan to implement a mechanism to address this, where you can "undo" your vote if you aren't happy with the result, and thus Freenet gets a much more accurate idea of the quality of different types of information, and the quality of the nodes used to retrieve it. This will mean that a malicious node of the type you describe will eventually be ignored by the rest of Freenet, so hopefully the threat you describe won't be an issue when we release Freenet 1.0.
Bandwidth and Piracy
In your own FAQ, you pretty much sidestep the entire issue that FreeNet would become a humongous "warez" distribution system by saying that it's merely a more efficient means of doing what others have been doing before. Ignoring the seeming subtle endorsement of piracy through the system, I'll raise an important question for adminstrators of FreeNet nodes.
In your FAQ, you say that it is very hard for FreeNet node admins to know what is on their site. With the inevitable proliferation of "warez" on the site, how will the system avoid getting bogged down with hundreds of illegal copies of popular pieces of software?
For example, when Diablo 2 finally comes out in the stores, what would prevent servers from being overloaded with:
You could literally have hundreds of 650 Mb images of games floating around jamming up everyone's nodes. With the lack of searchability, no one would know what keys hook into what files. Without this knowledge, warez people might keep uploading copies to different keys, thus flooding the system. In essence, does not the lack of protection against piracy and the seemingly intentional goal of keeping admins from controlling their system threaten to bring down the entire network under the burden of warez and junk?
The simple answer is that copyright is economic censorship (ie. restricting the free distribution of information for economic reasons), and thus Freenet will make it difficult or impossible to enforce copyright. As for whether Freenet will be "overloaded" with Warez junk, well current methods of distributing Warez work fine already, but as for Freenet - the system will contain whatever information people request. If people request Warez, then there will be Warez on Freenet, if they request pornography, then there will be pornography on Freenet, and if they request political documents then there will be political documents on Freenet. Freenet makes no distinction, and if it did it could not claim to permit true freedom of information. If you want to know more about why we just might be able to live without copyright take a look at our philosophy page.
Regarding your comment about large files clogging up Freenet, right now, inserting a huge file into Freenet probably won't work, since a node won't store a file which exceeds its disk-cache limit. We also plan to make large files need to justify themselves in-terms of the number of hits they receive, so that they don't displace loads of smaller files.
Why such bad press, what is being done to fix it?
It's pretty scary when Wired slams you with the headline, "Alternative Net Protects Pirates", which contained in the story gems such as:
"Eric Scheirer, a music technology researcher at MIT's Media Lab, said Freenet is an interesting experiment, but said it would likely be used only by a small community of pirates and "privacy nuts."
And, failing Monday's piece in the Nando Times , that's actually been the best article so far. The New Scientist is running "out of control: The Internet is about to get even harder to police" in their current issue, and ABCNews.com did a one-paragraph style summary of this article, with the lead of "An Internet system designed to guarantee anonymous free speech on the Web could be used by child pornographers and terrorists, according to New Scientist magazine," which then proceeds to all but call you and the other programmers pedophiles in a grammatical burp.
My question is, if this is to be successful (which I for one am all in favor of, I'm in close contact with Brandon and Steven, two of the FreeNet programmers, and am very much in support of the existence of this), FreeNet can't come off as a tool for criminals and miscreants, lest you attract more attention than you'd like from the Fed-types. Now, you may say that because it's open-source and already available, etc., that the Feds can't put it down, but if it is branded as an evil tool for child pornographers (like it is currently), it will never gain the popularity and user-base needed to make it sufficiently robust against machine removals.
To get something called a tool for privacy nuts by Wired is pretty bad--and the rest of the press has been worse; is there any plan to get this project out of the gutter?
Well, we actually have had more good press than you suggest, the recent Wall Street Journal article was very positive, as was an early Brave GNU World article (see our publicity page for links). As for the bad press, well I think whenever you do something really new you are always going to encounter resistance. I think most of these journalists were trying to stir up some controversy, which is what journalists do. Interestingly many of these journalists have explained that they are often given a brief by their editors before they have even spoken to me on what tone the article will have. You rapidly develop a thick skin when you are involved in this kind of project, although for the most part I have been reasonably happy with the press coverage we have received.
Re: Why such bad press, what is being done to
"Eric Scheirer, a music technology researcher at MIT's Media Lab, said Freenet is an interesting experiment, but said it would likely be used only by a small community of pirates and 'privacy nuts.'"
I stand by my quote in that article, although naturally it's a little short on context. Let me make clear that I am in favor of privacy, security, and anonymity when appropriate, and I despise the current attempts to make the WWW more corporate-controlled via both code and law. I don't think there's anything wrong with Freenet, I just don't think it will ever take off in the mainstream.
The fact is that most of the things that most people like to use the WWW for--such as e-commerce and Slashdot--cannot be built on Freenet, since it has no cookies and no memory. Given this, I can't see anything happening with Freenet except that it becomes a huge storehouse for illegal porn, pirated MP3s and 3l33t w8r3z.
It's a shame, because the potential political benefits that it raises, by allowing dissident speech in repressive countries, is great.
I guess my question for the Freenet developers would be: I am not a pirate, a privacy nut, a political dissident, or someone trying to spread illegal trade secrets. What does Freenet offer me? And are these benefits broad enough to a broad enough segment of the world population to create the momentum needed for Freenet to work sociologically as well as technically?
-- Eric Scheirer
MIT Media Laboratory
Brandon (one of the other core Freenet developers) did e-mail you twice to answer the comments you made in Wired, he still awaits your response on the matter. Ok, what does Freenet offer someone who doesn't care about anonymity? One point that many people miss is that it is actually a very efficient way to distribute information due to its dynamic caching and dynamic mirroring. Freenet will move information to where it is in-demand, and will duplicate popular information automatically so that you should never encounter "The SlashDot Effect" with Freenet. In other words, your ability to publish information is no longer limited to the Bandwidth you can pay for. Because of this it should actually be a better way to distribute information than, say, the WWW or Usenet, even ignoring the fact that the information can't be censored. TheCarp mentions this below.
I think Freenet would dovetail nicely with wireless network technology. I system of Freenet servers 1-2 km apart could blanket metropolitan areas and eliminate dependency on ISP's for network service.
What are your thoughts on this? Are any hardware people interested in looking at this problem? Building some prototypes?
I agree completely, we have actually kept the protocol packet-based for just this sort of reason. I think Freenet would be perfect for a distribute decentralized radio network, and it would be a very exciting project.
If files live longer the more they are thrashed, will this not just breed thrasher bots and crowd out data from clients with less connectivity? How about a voting system for one or more directories which does not add files easily but they are there for good. If it is that good a resource it deserves a champion to protect it.
Also, I take it you are comfortable with already having divulged the identities of the entire first wave of sysadmins of FreeNet nodes? Seems like your most vulnerable time is now.
I've long considered the value of a peer to peer system for countries underdeveloped in the areas of infrastructure and rights. Unfortunately it seems that social engineering is steadily on the side of repression. Wouldn't the best way to get FreeNet into such environments be to make it a source of economic strength? In other words, your growth metric might look much better if you include authorship, copyright, and microcashpayment management. I can't see the Declaration of Independence sticking in the current system for long.. but it is in both a good library and a good bookstore.
Basically you have built a distribution system which in its optimal configuration has no delivery time since you already have the commodity on your hard drive... make it work for business as well and it may reduce prices and take on a life of its own.
Well there is certainly more than one question here!
Firstly, the issue of "thrashing" or flooding Freenet is covered in the FAQ - I refer you to section 4.2. To summarize, the dynamic caching mechanism makes it very difficult to artificially make data more popular (since Freenet will just cache it on a node right beside you, and all of your requests will be soaked up by it).
As for divulging the addresses of the first few people to set-up Freenet nodes (I assume you refer to our "Inform.php" mechanism) that is merely a mechanism to "boot-strap" Freenet for testing purposes. Once Freenet is up and running properly we will have no need for such a mechanism, but it is useful in the early stages (and people have the option to switch it off in the config file if they really care that much). I should make it clear that Freenet now is still at an early stage of development. The project is lucky to have some very talented and hard-working developers, particularly Brandon Wiley and Oskar Sandberg who have really helped turned this from a dream into a reality, but we still have much work to do. Data modification, Content Hashed Keys, local data encryption, the list of ideas which we want to implement before a 1.0 release is long - but this is indicative of the cutting-edge nature of the project.
In terms of making it possible to sell information using Freenet, or a Freenet-like system, I suspect that might be missing the point we are making! Never say never though...
Kiddy porn, rape movies, snuff films.
No matter what zealots tell you, no freedom is absolute. Your freedoms end when they infringe on the rights of others. This includes your freedom of speech.
My thoughts when I first heard about this project were extremely positive for the first five seconds or so. I was going to set up a server, and suggest all my other bandwidth-rich friends do the same. Then I thought about what would be going to and from my server.
Anonymity has its place from time to time, but usually in the cases of an abuse by a higher power against an individual. But in the general case, I feel that freedom of speech entails the responsability of accountability.
If I'm going to say that I hate Virgos, and all Virgos should be locked up and treated as the inhuman beasts that they are, I should have the conviction to do so without a pointy hood over my head.
If I'm going to be distributing porn, I should be able to do it with a clean conscience. If I wanted to post naked pictures on a Web site, I'd be in some way traceable. And if I wasn't identifiable, there at least would be a mechanism in place (an e-mail to my upstream provider) to curb my freedom of speech if I was posting vile material.
The ideal of individual freedom falls apart in the environment of actual individuals who abuse it.
I'm not saying in any way that this should be a legal matter, or that the product should be banned, just that in the case that it turns out like I expect it to (the majority of traffic for illicit files, both violating copyright and basic human decency) I will have no respect, even a measure of contempt for the people that do run the servers. THEY will be the ones I will hold accountable for the 'free speech' being exercised on the network. And if they were to be sued off the net by the RIAA, Church of Scientology and MPAA, I can't say I'll be surprised, or all that upset.
What arguments can you make FOR free, anonymous access to kiddie porn, snuff films and rape/torture erotica? Why should *I*, a server operator, nurture these sorts of activities in an ideal environment?
You cannot have free speech without tolerating speech that you personally don't agree with. If you don't want to risk aiding the distribution of "kiddie porn" (which is *already* freely distributed on the Internet anyway), then steer clear of Freenet - it's not for you. On the other hand, if you want to help build a system which will help humanity share information, even though some of that information will be distasteful to you personally, then set up a Freenet node. See our philosophy page for more information on this.
some technical questions.
. (Score:4, Interesting)
Once something is put on freenet it cannot be removed. What does this mean? No censorship, but also misinformation stays in the system just as long as correct information, so long as it can 'trick' people into requesting it (by seeming to be relevant, for instance). This can be exploited intentionally to censor (some things are nearly unfindable on search engines because of 'key' collisions - the band 'Reload', for instance), or unintentionally - I write something, post it, and five minutes later learn that I was mistaken. Oh well! People will just have to decide for themselves what is truth. Even if I DO post a retraction, there is no way to verify that a trusted entity (such as the original author) retracted it.
As a medium for sharing artistic works (e.g., music, essays, images) this is not as important, but to carry actual facts, (e.g., hardware specs, controvertial news items, etc.) this seems a major shortcoming. Is there any solution to this problem in place or in progress? I ask because I feel that this is not adequately discussed in the FAQ.
As I mention above we are working on improving the way that people can "vote" for the validity of information - I hope that this will address some of the concerns you raise. Having said that, you can't really drownout information on Freenet in the manner you suggest. If you have the key, you can get the data unless it has died out due to never being requested.
The whole area of choosing appropriate keys for the data you wish to insert is a huge, and ripe for further research. For some things, like MP3s, or poems, choosing an appropriate key is pretty easy ("music/mp3/artist/album/track" or "poem/poet/title" for example). It merely requires a standard way to refer to these things - and since it is in everybody's interest to use the same standard, hopefully good standards will emerge quite quickly. Other things are less easy. When people created the Internet, they probably never thought it might be used the way we are using it now - and they didn't need to. Similarly, Freenet is a platform upon which we hope others will build.
Re: Why compete?
Actually....the Freenet Has a huge technical advantage over http protocols. The thing is... it's not just hard to track down who wrote it (unless they sign their name...it's only anonymous if you want it to be) and where it's stored...
It has cacheing built in. When you request something, it propagates. This means more copies exist. So if a document is REALLY popular, then no one server is bogged down with distributing it.
Imagine some really popular band that believes in mp3 distribution puts out a new mp3. Now everyone 80% of college students go to download it.
WHat happens? After the first few downloads at each college...the local university freenet server will have a copy of the mp3 and will be serving it to that university.
None of the requests for it will be going outside the local university network. Its the basic equivalent of everyone in the world being behind multiple layers of http cacheing proxy, except somewhat better (its built into the protocol)
What can I say? Cool - you get it! Just to reiterate though, Freenet is an experiment. It is quite different from projects like Linux and Mozilla, which are essentially open source re-implementations of technology which has already been proven (and, incidentally, I am a big supporter of both of these projects). There aren't really any precedents which we could follow in designing Freenet, certainly nothing that came close to what we wanted to achieve. The difficulty with Freenet is that we can only *really* test it by encouraging people to use it on a reasonably wide scale. There is much further work to be done on Freenet, it is really only at the beginning of its testing phase right now - so please don't expect it to change the world just yet - but please do try it out!
You may also be interested to note that we have released Freenet 0.1beta.
All the best,