The advent of superworms creating peer-to-peer networks which allow for easy propagation of commands to the entire network by their creator was anticipated in this Linux Journal interview and in the design for the Curious Yellow Peer-to-Peer Superworm.
This superworm for Linux is just a first attempt at an entire genre of zero-day exploit worms which create ad-hoc peer-to-peer networks as they spread.
The question which most often comes up when JXTA is mentioned is always the same, "What is JXTA for?" The answer is simple but unintuitive: JXTA is so that Sun can break into the P2P mindshare. It can't do this with Jini because Jini has already ridden the hype rollercoaster, so Sun had to come up with a new product. They took their bag of buzzwords and came up with a buzzword compliant P2P product.
Now, seeing as how are you probably a developer who does not work for Sun, what is the point of JXTA for *you*? What does it give you that you will find useful. Very little. If you happen to want to write a P2P app that uses the JXTA network model (basically Gnutella, but let's not quibble about that particular point) then JXTA has some code that you can use as a base for your application. You can also hack up an appropriately licensed Gnutella, Freenet, or MojoNation client if you like their arhictecture.
Chances are, however, you want a custom network architecture. If you read Sun's marketing hype, they will tell you that JXTA can be any architecture and it's open source, so it's up to you to add the components to make JXTA-Freenet or JXTA-MojoNation or JXTA-your-custom-network-architecture-here. This is technically true because JXTA doesn't actually do anything. It's basically a bunch of webservers that open connections to each other. The only thing that JXTA adds over a webserver with CGI or some form of RPC such as XML-RPC or SOAP is that it provides a few services that create a Gnutella-like network. If you're not going to use the JXTA services then JXTA adds nothing over an XML-RPC module in Apache or any of the small XML-RPC servers such as Helma.
People ask me for advice on their P2P applications all the time and never have I found it prudent to recommend JXTA for a project. However, JXTA does have one benefit, which is that if you use it in your project then Sun will give you free advertising announcing "yet another" JXTA-based project.
TCP/IP is a better choice for Chess if you want to place at a reasonable speed. Freenet is a better choice if you don't want a centralized game server or you're afraid of evil people inserting false moves into your game or you want to play totally anonymously.
I don't know if anyone actually has these goals. I think most people would prefer to play at a reasonable speed. But the actual point was to teach people how to write applications for Freenet. I figured it was more fun to show how to write chess than to show how to write, say, an Enterprise Application Server.
Re:Has anybody found Freenet useful yet?
on
P2P Piracy? Piffle!
·
· Score: 1
Yes, Freenet is useful. I use it quite a lot for sharing files with my friends. Rather than putting it on my website or on FTP and sending them my IP/URL, I put it in Freenet and say "The key is steve.pdf." It's very convenient. A while back we were even doing mail and newsgroups over Freenet (defunct now, but we're upgrading it). I recommend any users of the Freenet web interface to visit "webpages/gj_jump0" on Freenet. It has links to a lot of Freenet resources including legally obtained MP3s from the Tropus project.
You point out two problems in Freenet, unreliable key lists and no searching, adding up to an inability to find files. Luckily, these won't be problems for much longer as I am personally working on solutions to just these problems. These are exactly what is next on my list. If you want more reliable key lists on the key index servers then you should write to the maintainer of the Key Index CGI scripts and bug him to put in dead key culling. The people working on then Freenet implementation don't write the key index code.
If you think of Freenet as another Napster, we are certainly way behind any reasonable schedule. I mean really, how long can it take to write Napster? Luckily, we're not trying to make Napster. Every other P2P project is way behind Freenet in terms of actually implementing a secure, anonymous system. Not to knock the other projects, I wish them the best of luck.
This is an analysis of the keys listed on key index servers, analyzed by name. The key index servers are publically accessible web sites where people can enter keys for things which they have placed in Freenet.
Therefore, this analysis doesn't take into account certain facts, such as that names do not always accurately reflect content, some listed keys were never inserted into Freenet, some listed keys have since fallen out of Freenet, and some information in Freenet is not listed.
This is not so much an analysis of what is on Freenet as in what people think they should tell the world is in Freenet.
As a Freenet developer, I feel compelled to correct some of the inaccuracies being presented by commentors as fact.
"Freenet is an attempt to replace the web." - This is more true than saying that Freenet is a replacement for Napster, but it's still not true. Freenet is better than the web in a couple of ways, mainly anonymity and decentralization. If you don't need these features, then by all means use the web.
"You can't create Slashdot on Freenet because Freenet doesn't have dynamic content." - Sure you can. A web forum was already created, but is currently being overhauled. We already have a web frontend and newsgroups, mail, and hyperlinked documents in Freenet. A web forum is just an HTML frontend to a newsgroup with some bells and whistles. The reason that they use dynamically generated pages is because they use RDMS backends so that the servers can handle the load. Since the load in Freenet is distributed, this isn't necessary. Sometimes you really do need dynamically generated content, but in the case of web forums it's mostly just a performance enhancement.
"Popular == worthless. Freenet will be filled with worthless stuff." - Popularity is local, not global. If you connect to your friends instead of random strangers then the local network will be filled with items of shared interest.
"The problem with Freenet is that unpopular items are dropped." - Popularity is local, not global. You want items that no one in your local network is requesting to disappear. Files go to where they are wanted and disappear from where they are not wanted.
"I can't trust the information that I get out of Freenet." - We have tamper-proof keys that rely on digital signatures and content hashes. If you are worried about authenticity, then use those.
"Freenet must track what people request because it knows what is popular. That leaves an audit trail that compromises anonymity." - Popularity is local, not global. Your node discards items that have not been requested in a while. There is no global rating or tracking of any kind.
"Freenet requires a high-speed connection" - No, but it would certainly be nice.
This is an ideal use for Freenet. Freenet saves you the trouble of having to run multiple mirrors. All you have to do is run a Freenet node, just like lots of people do. When someone tries to shut down a web site that you value, insert the web site into Freenet. Anyone else attempting to insert the web site will be told that it has aleady been inserting.
It's perfect for this sort of thing. The point, after all, is to defeat censorship.
The new version of Freenet has a web proxy interface to allow the easy viewing of web pages in Freenet. Also, there exist convenient scripts to mirror a site into Freenet.
Following Judge Kaplan's ruling that code should not be protected as free speech because it can harm society, the Internet community was in an uproar.
However, one group, the Norwegian hacker collective known as PoMoRu say they saw this coming.
"We've known that code can destroy society for a long time. That's why we learned to code. Most programmers start coding in middle school, when they hate the world and want to destroy society, and that angst never leaves them."
When asked how exactly code could harm society, they responded with a perplexing answer.
"Well, we've got a perl script to kill the president. It's still pre-alpha, but we plan to release a beta at Defcon 9. We wanted to generalize it to kill everyone that ever wronged us, but we're developing under windows and so fork doesn't work in perl. We plan on switching to Linux and achieving genocide as soon as there are better GUI configuration tools."
It is a tool for the propagation of free speech! FREE SPEECH! It has nothing to do with violating copyright! AaRGH!
Why do you need strong encryption to pirate MP3s? You don't even need anonymity! People have been doing it on IRC for years. Napster hasn't even been shut down yet.
Your definition is most certainly not the definition being used in architectural whitepapers.
The definition of popularity in Freenet is simply what's being requested and it's local, not global. When a file is requested, it gets bumped up in the list. When there is not enough room, something on the bottom is kicked off. It's more complicated than this to avoid abuses like flooding the system with a bunch of junk so as to knock every useful off. But that's the basic idea.
But you are right that the obscure, unheard-of, and unknown gems will not last long.
HOWEVER, cult phenomenons which only a small community like will proliferate in that small community, just not on the rest of the network. And so it should be! If there is ever a boom in the desire for such an obscure text, it will still live in the archives of this small community and instantly proliferate out onto the network.
This doesn't solve all problems. (like I said, freehaven.net). But it's better than the web. It protects hot spots of interest. *Something* needs to protect hot spots of interest, after all. They are the most under attack. That which no one cares about or even knows about is an entirely different problem solved in an entirely different way. Join the two and you're doing pretty good.
It's interesting that you say that. I haven't seen any new features added to Napster or Gnutella since the first Freenet release. I'm not knocking Gnutella. It's a very cool project. I just don't understand where this idea of relative advancement is coming from when there aren't any new features being added.
I think the reason that things are progressing so slowly is because what we're doing is REALLY HARD.
Anonymous, efficient, non-abusable searching and updating is a problem no one has solved yet. I lot of people think they have solved in, in which case I invite them to the mailing list for some good intellectual jousting.
There is a hypertext distributed message board by Joe Soelbrig. There are also Freenet newsgroups that you can access via the fnclient Freenet client.
I am also working on an NNTP gateway to the Freenet newsgroups so that you can read them in your favorite news reader.
This isn't true at all. The time it takes to start downloading the file is the time it takes to contact the 5 or 10 servers and then get a response. So it's the cumulative latency between the servers, just like with TCP/IP, except that in our case it might take longer to find the file (you have to search the network for it). But the file start downloading immediately once it's found.
And if anyone in your area has already requested it, then it will only be a few hops away.
The clustering is not arbitrary. It is somewhat arbitrary. You tell your node what nodes to talk to. So if you only want to talk to nodes in Europe, you can configure it as such and the geographical clustering which Ian talks about will happen. If you want to only talk to nodes run by former Scientologists, you can do that too and then scientological clustering will occur. Or you can only talk to people you meet on alt.fan.jwz and then there will be clustering of pictures of jwz.
Freenet was not started to end copyright. Freenet was started to stop censorship. Ian Clarke is not Freenet. Ian Clarke thinks copyright is bad. Many people are attracted to the Freenet development lists because they think that's what Freenet is about and they think copyright is bad.
But Freenet originally had nothing to do with copyright and for me, it still doesn't.
Freenet makes censorship more difficult and it ends the need to know *where* something is and it helps fix the problem of overloading websites.
This whole copyright thing is a product of the press attention that Napster has been getting. Music piracy has been around for a long time. We're attempting to do something much more interesting.
The actual problem here is with downloading untrusted information. That's what causes the problem with spam, the "Pat Boone" attack, and trojans. The solution to this is a trust system. We're working on a trust system based on digital signatures. It's all worked out, it's just a matter of implementing it along with all of the other hundreds of things we need to implement.
That unpopular information is dropped is one of the most common and most bizarre criticism of Freenet. You have to have a way to manage an endless supply of information and a finite amount of space.
First recognize that popularity is local, not global. So the network won't be filled with Britney Spears MP3s like everyone seems to think. The area of the network which I form by connecting to my friends will more commonly have cool stuff that we as a group like than stuff which "the people" think is popular.
But, if you don't like the throw out unpopular stuff strategy of resource usage regulation, you can check out freehaven.net (I give you a meg, you give me a meg, permanent storage based on trading) or mojonation.net (trading via virtual currency) or the ever popular Gnutella (I keep what I want, you keep what you want). You can choose amongst the available options whatever suits your tastes in resource management.
We believed that we've fixed this problem in Freenet. We only send the hash of the key ("filename", "search string", whatever) into the network, but the file is signed with the unhashed key. Results that are not properly signed are automatically rejected. The result is that a node can only return a file (and have it accepted) if it actually has a file inserted under the same key. It can't fake it.
Of course, we don't have searching. This only stops spamming when you actually request a file. Searches have to be transmitted in the clear. To make that harder to spam you need a web of trust system with reputations and ratings so that you can find good, valid content.
You can walk into any 7-11 right now and get (for free) an Internet Shopping Card. You give the cashier cash and it is put on your card. You can use the card anywhere online that accepts American Express. Neither 7-11 or American express has your name or address. All they have is a bunch of card numbers mapped onto dollar amounts. Combined with SSL and a web anonymizer, totally anonymous e-cash without any complicated patented cryptographic techniques. And you can do it right now.
> and the third generation will probably bring > metadata tagging facilities, more powerful > searching and search path optimization.
That's exactly what I'm working on in Freenet. We've already implemented metadata. Now I'm working on metadata query based searching as well as the ability to sign metadata to limit the search range. We already have path optimization and better anonymity.
I think what you mean when you say that an Open Source project could not have produced something like BeOS you really mean that a poorly funded, volunteer-run, design-by-committee project couldn't have made something like BeOS.
There are well funded, tyranically controlled Open Source projects, too.
Slashdot Mirror
This superworm for Linux is just a first attempt at an entire genre of zero-day exploit worms which create ad-hoc peer-to-peer networks as they spread.
The question which most often comes up when JXTA is mentioned is always the same, "What is JXTA for?" The answer is simple but unintuitive: JXTA is so that Sun can break into the P2P mindshare. It can't do this with Jini because Jini has already ridden the hype rollercoaster, so Sun had to come up with a new product. They took their bag of buzzwords and came up with a buzzword compliant P2P product.
Now, seeing as how are you probably a developer who does not work for Sun, what is the point of JXTA for *you*? What does it give you that you will find useful. Very little. If you happen to want to write a P2P app that uses the JXTA network model (basically Gnutella, but let's not quibble about that particular point) then JXTA has some code that you can use as a base for your application. You can also hack up an appropriately licensed Gnutella, Freenet, or MojoNation client if you like their arhictecture.
Chances are, however, you want a custom network architecture. If you read Sun's marketing hype, they will tell you that JXTA can be any architecture and it's open source, so it's up to you to add the components to make JXTA-Freenet or JXTA-MojoNation or JXTA-your-custom-network-architecture-here. This is technically true because JXTA doesn't actually do anything. It's basically a bunch of webservers that open connections to each other. The only thing that JXTA adds over a webserver with CGI or some form of RPC such as XML-RPC or SOAP is that it provides a few services that create a Gnutella-like network. If you're not going to use the JXTA services then JXTA adds nothing over an XML-RPC module in Apache or any of the small XML-RPC servers such as Helma.
People ask me for advice on their P2P applications all the time and never have I found it prudent to recommend JXTA for a project. However, JXTA does have one benefit, which is that if you use it in your project then Sun will give you free advertising announcing "yet another" JXTA-based project.
TCP/IP is a better choice for Chess if you want to place at a reasonable speed. Freenet is a better choice if you don't want a centralized game server or you're afraid of evil people inserting false moves into your game or you want to play totally anonymously.
I don't know if anyone actually has these goals. I think most people would prefer to play at a reasonable speed. But the actual point was to teach people how to write applications for Freenet. I figured it was more fun to show how to write chess than to show how to write, say, an Enterprise Application Server.
You point out two problems in Freenet, unreliable key lists and no searching, adding up to an inability to find files. Luckily, these won't be problems for much longer as I am personally working on solutions to just these problems. These are exactly what is next on my list. If you want more reliable key lists on the key index servers then you should write to the maintainer of the Key Index CGI scripts and bug him to put in dead key culling. The people working on then Freenet implementation don't write the key index code.
If you think of Freenet as another Napster, we are certainly way behind any reasonable schedule. I mean really, how long can it take to write Napster? Luckily, we're not trying to make Napster. Every other P2P project is way behind Freenet in terms of actually implementing a secure, anonymous system. Not to knock the other projects, I wish them the best of luck.
This is an analysis of the keys listed on key index servers, analyzed by name. The key index servers are publically accessible web sites where people can enter keys for things which they have placed in Freenet.
Therefore, this analysis doesn't take into account certain facts, such as that names do not always accurately reflect content, some listed keys were never inserted into Freenet, some listed keys have since fallen out of Freenet, and some information in Freenet is not listed.
This is not so much an analysis of what is on Freenet as in what people think they should tell the world is in Freenet.
Where are the screenshots?
As a Freenet developer, I feel compelled to correct some of the inaccuracies being presented by commentors as fact.
"Freenet is an attempt to replace the web." - This is more true than saying that Freenet is a replacement for Napster, but it's still not true. Freenet is better than the web in a couple of ways, mainly anonymity and decentralization. If you don't need these features, then by all means use the web.
"You can't create Slashdot on Freenet because Freenet doesn't have dynamic content." - Sure you can. A web forum was already created, but is currently being overhauled. We already have a web frontend and newsgroups, mail, and hyperlinked documents in Freenet. A web forum is just an HTML frontend to a newsgroup with some bells and whistles. The reason that they use dynamically generated pages is because they use RDMS backends so that the servers can handle the load. Since the load in Freenet is distributed, this isn't necessary. Sometimes you really do need dynamically generated content, but in the case of web forums it's mostly just a performance enhancement.
"Popular == worthless. Freenet will be filled with worthless stuff." - Popularity is local, not global. If you connect to your friends instead of random strangers then the local network will be filled with items of shared interest.
"The problem with Freenet is that unpopular items are dropped." - Popularity is local, not global. You want items that no one in your local network is requesting to disappear. Files go to where they are wanted and disappear from where they are not wanted.
"I can't trust the information that I get out of Freenet." - We have tamper-proof keys that rely on digital signatures and content hashes. If you are worried about authenticity, then use those.
"Freenet must track what people request because it knows what is popular. That leaves an audit trail that compromises anonymity." - Popularity is local, not global. Your node discards items that have not been requested in a while. There is no global rating or tracking of any kind.
"Freenet requires a high-speed connection" - No, but it would certainly be nice.
This is an ideal use for Freenet. Freenet saves you the trouble of having to run multiple mirrors. All you have to do is run a Freenet node, just like lots of people do. When someone tries to shut down a web site that you value, insert the web site into Freenet. Anyone else attempting to insert the web site will be told that it has aleady been inserting.
It's perfect for this sort of thing. The point, after all, is to defeat censorship.
The new version of Freenet has a web proxy interface to allow the easy viewing of web pages in Freenet. Also, there exist convenient scripts to mirror a site into Freenet.
Following Judge Kaplan's ruling that code should not be protected as free speech because it can harm society, the Internet community was in an uproar.
However, one group, the Norwegian hacker collective known as PoMoRu say they saw this coming.
"We've known that code can destroy society for a long time. That's why we learned to code. Most programmers start coding in middle school, when they hate the world and want to destroy society, and that angst never leaves them."
When asked how exactly code could harm society, they responded with a perplexing answer.
"Well, we've got a perl script to kill the president. It's still pre-alpha, but we plan to release a beta at Defcon 9. We wanted to generalize it to kill everyone that ever wronged us, but we're developing under windows and so fork doesn't work in perl. We plan on switching to Linux and achieving genocide as soon as there are better GUI configuration tools."
It is a tool for the propagation of free speech! FREE SPEECH! It has nothing to do with violating copyright! AaRGH!
Why do you need strong encryption to pirate MP3s? You don't even need anonymity! People have been doing it on IRC for years. Napster hasn't even been shut down yet.
Your definition is most certainly not the definition being used in architectural whitepapers.
The definition of popularity in Freenet is simply what's being requested and it's local, not global. When a file is requested, it gets bumped up in the list. When there is not enough room, something on the bottom is kicked off. It's more complicated than this to avoid abuses like flooding the system with a bunch of junk so as to knock every useful off. But that's the basic idea.
But you are right that the obscure, unheard-of, and unknown gems will not last long.
HOWEVER, cult phenomenons which only a small community like will proliferate in that small community, just not on the rest of the network. And so it should be! If there is ever a boom in the desire for such an obscure text, it will still live in the archives of this small community and instantly proliferate out onto the network.
This doesn't solve all problems. (like I said, freehaven.net). But it's better than the web. It protects hot spots of interest. *Something* needs to protect hot spots of interest, after all. They are the most under attack. That which no one cares about or even knows about is an entirely different problem solved in an entirely different way. Join the two and you're doing pretty good.
It's interesting that you say that. I haven't seen any new features added to Napster or Gnutella since the first Freenet release. I'm not knocking Gnutella. It's a very cool project. I just don't understand where this idea of relative advancement is coming from when there aren't any new features being added.
I think the reason that things are progressing so slowly is because what we're doing is REALLY HARD.
Anonymous, efficient, non-abusable searching and updating is a problem no one has solved yet. I lot of people think they have solved in, in which case I invite them to the mailing list for some good intellectual jousting.
What you are looking for is an Eternity service. The only implementation of this that I think has a chance is FreeHaven at freehaven.net.
What would be particularly useful is a
FreeHaven Freenet gateway. Permanent information storage and fast, unslashdottable access at once.
There is a hypertext distributed message board by Joe Soelbrig. There are also Freenet newsgroups that you can access via the fnclient Freenet client. I am also working on an NNTP gateway to the Freenet newsgroups so that you can read them in your favorite news reader.
This isn't true at all. The time it takes to start downloading the file is the time it takes to contact the 5 or 10 servers and then get a response. So it's the cumulative latency between the servers, just like with TCP/IP, except that in our case it might take longer to find the file (you have to search the network for it). But the file start downloading immediately once it's found. And if anyone in your area has already requested it, then it will only be a few hops away. The clustering is not arbitrary. It is somewhat arbitrary. You tell your node what nodes to talk to. So if you only want to talk to nodes in Europe, you can configure it as such and the geographical clustering which Ian talks about will happen. If you want to only talk to nodes run by former Scientologists, you can do that too and then scientological clustering will occur. Or you can only talk to people you meet on alt.fan.jwz and then there will be clustering of pictures of jwz.
Freenet was not started to end copyright. Freenet was started to stop censorship. Ian Clarke is not Freenet. Ian Clarke thinks copyright is bad. Many people are attracted to the Freenet development lists because they think that's what Freenet is about and they think copyright is bad.
But Freenet originally had nothing to do with copyright and for me, it still doesn't.
Freenet makes censorship more difficult and it ends the need to know *where* something is and it helps fix the problem of overloading websites.
This whole copyright thing is a product of the press attention that Napster has been getting. Music piracy has been around for a long time. We're attempting to do something much more interesting.
The actual problem here is with downloading untrusted information. That's what causes the problem with spam, the "Pat Boone" attack, and trojans. The solution to this is a trust system. We're working on a trust system based on digital signatures. It's all worked out, it's just a matter of implementing it along with all of the other hundreds of things we need to implement.
That unpopular information is dropped is one of the most common and most bizarre criticism of Freenet. You have to have a way to manage an endless supply of information and a finite amount of space. First recognize that popularity is local, not global. So the network won't be filled with Britney Spears MP3s like everyone seems to think. The area of the network which I form by connecting to my friends will more commonly have cool stuff that we as a group like than stuff which "the people" think is popular. But, if you don't like the throw out unpopular stuff strategy of resource usage regulation, you can check out freehaven.net (I give you a meg, you give me a meg, permanent storage based on trading) or mojonation.net (trading via virtual currency) or the ever popular Gnutella (I keep what I want, you keep what you want). You can choose amongst the available options whatever suits your tastes in resource management.
We believed that we've fixed this problem in Freenet. We only send the hash of the key ("filename", "search string", whatever) into the network, but the file is signed with the unhashed key. Results that are not properly signed are automatically rejected. The result is that a node can only return a file (and have it accepted) if it actually has a file inserted under the same key. It can't fake it. Of course, we don't have searching. This only stops spamming when you actually request a file. Searches have to be transmitted in the clear. To make that harder to spam you need a web of trust system with reputations and ratings so that you can find good, valid content.
I was deeply saddened when OLGA shut down.
You can walk into any 7-11 right now and get (for free) an Internet Shopping Card. You give the cashier cash and it is put on your card. You can use the card anywhere online that accepts American Express. Neither 7-11 or American express has your name or address. All they have is a bunch of card numbers mapped onto dollar amounts. Combined with SSL and a web anonymizer, totally anonymous e-cash without any complicated patented cryptographic techniques. And you can do it right now.
I maintain a comprehensive list of related projects here.
> and the third generation will probably bring > metadata tagging facilities, more powerful > searching and search path optimization.
That's exactly what I'm working on in Freenet. We've already implemented metadata. Now I'm working on metadata query based searching as well as the ability to sign metadata to limit the search range. We already have path optimization and better anonymity.
Check us out at http://freenet.sourceforge.net
I think what you mean when you say that an Open Source project could not have produced something like BeOS you really mean that a poorly funded, volunteer-run, design-by-committee project couldn't have made something like BeOS.
There are well funded, tyranically controlled Open Source projects, too.
Freenet was tested on a mac recently and has been modified to run just fine.