Peer to peer websites

I been pondering about this idea, a website that isn't hosted by servers but by the users as they use it I know is not original but how come no actual websites use it?

>>6821 This is just my random speculation, but I'd guess the model could be hard to sustain for a few reasons. It would require users to be willing to trust and reliably store all of the sites data (hard for imageboards and other sites that have user-generated content) Most users probably aren't dedicated enough to a website to be willing to host it. Makes the availability of the site susceptible to day/night cycles in user computing power (assuming people shut down whatever they are hosting the site with when they sleep / are not using it). Aaand I just realized that you probably mean users host the site exclusively while they are on it, not users that like the site can choose to host a copy of it and share that copy.

>>6826 I forgot, I imagine synchronization with dynamic sites could be a nightmare, although I don't know enough to say that with any sort of authority.

There are a few projects that are trying to implement this, the few I could think of are zeronet, IPFS and dat out of which IPFS seems like the most popular

Isn't OpenBazaar like this?

>someone posts mozarella italian round pies >everyone in the swarm gets v8ed First, hypertext transport protocol wasn't designed to be distributed, nor the browsers are. There are possible hacks around in protocol level, like default Freenet behavior or optional function on top of GNUnet, or maybe TahoeLAFS over i2p. but there is no real need for all info to be distributed. Yes, there are torrents and they are fine for what they are, distributed file sharing. I guess, it's possible to seed html pages over bittorrent and view them locally, and even add some dynamic features too.

>>6821 Its called Zeronet. Its based on the BitTorrent protocol and uses BitCoin addresses (not actual money though) for adressing. On itself its 100% censorship resistant though not privacy-enhancing. But you can tack on tor for that. Its actually pretty nice and fast. Also completely open source: https://www.zeronet.io/

>>6831 >someone posts mozarella italian round pies >everyone in the swarm gets v8ed This is a really hard problem to solve in distributed file storage. Freenet tried to solve it by encrypting the file fragments that nodes keep in their cache but anyone can request any file so you can always figure out what files your fragments are part of with enough effort. It's more of a legal problem than a technical one. Will the justice system give you the benefit of the doubt that you cannot control or reasonably know what files your node is serving? Or will they just derez everyone involved?

>>6841 >>6831 no its not. 1. No one gets a vanning just for one picture. 2. we are talking about only a distributed net, not an encrypted one. Therefore it is possible to see and delete the offending files (like with Zeronet), based on that it is also trivial to implement hash-based banlists of files you do not want to see. These lists can of course be user-defined on top of getting them from publicly available places. So you wont see my dickpics either if you chose so.

>>6842 >it is possible to see and delete the offending files This depends on how the technology works. Most decentralized networks generally require nodes to move data on behalf of other users without necessarily knowing who they are or what it is. Like in the Freenet example. Even if the data isn't encrypted for some reason, you still can't manually audit every single packet going through your node, even at dailup speeds. Additionally, it is generally a point of these distributed storage systems that it is not possible to arbitrarily remove data from the network or discriminate what specific files a node is requesting and filter them. Otherwise it's not censorship resistant is it? >it is also trivial to implement hash-based banlists of files you do not want to see. Your idea is to block your peer's traffic based on a 3rd party blacklist of checksums? Nice try NSA.

>>6844 >Most decentralized networks generally require nodes to move data on behalf of other users without necessarily knowing who they are or what it is. Like in the Freenet example. No. Freenet, GnuNet and i2p are encrypted. They are not just anti-censorship apps but also privacy enhancers. Zeronet is not. Its only concern is distributing the files. Its just one click to delete any file there. >Your idea is to block your peer's traffic based on a 3rd party blacklist of checksums? Nice try NSA. User defined blacklist. Read my post again. Its only to satisfy the paranoid "muh c p" worriers.

>>6828 >>6833 I heard about zeronetbefore but because it is tied to bitcoin wallets it makes me groan on the stupidity. IPFS looks more promising but I'll have to read more about it >>6831 >>6841 >>6842 >>6844 >>6845 A decentralized webstie doesn't mean a site without moderators or administrators

>>6849 >I heard about zeronetbefore but because it is tied to bitcoin wallets it makes me groan on the stupidity. Dont worry they just use the bitcoin wallet addresses as addresses. They are nicely uniform which is good for a key based routing and as an added bonus you can send money to specific sites without any detour. But this is completely optional. It does use 90% of IPFS in the background since IPFS itself uses bittorrent as distribution protocol too.

In terms of actual adoption and what people have already made, the two modern ones I can think of are: - IPFS, used by neocities for 'ipfs archiving' (https://neocities.org/permanent-web) - Beaker Browser which uses the dat protocol (https://beakerbrowser.com/docs/apis/dat.html) The latter looks more promising.

>>6845 >No. Freenet, GnuNet and i2p are encrypted. They are not just anti-censorship apps but also privacy enhancers. Zeronet is not. Its only concern is distributing the files. Its just one click to delete any file there. >https://www.zeronet.io/ Open, free and uncensorable websites So which is it? Uncensorable or no CP? You can't have both. >User defined blacklist. Read my post again. Its only to satisfy the paranoid "muh c p" worriers. Are you literally retarded? A "user defined blacklist" is just you remembering not to click the same CP link twice. Otherwise you are getting the blacklist from a 3rd party who essentially decides what information you can consume.

>>6855 >So which is it? Uncensorable or no CP? You can't have both. yes you can. Its not censorable by a central authority. It is very much censorable by the site owner and by the peers who hold the site (well, not completely for the latter one, they can only blacklist certain files) >Are you literally retarded? A "user defined blacklist" is just you remembering not to click the same CP link twice. Otherwise you are getting the blacklist from a 3rd party who essentially decides what information you can consume. No. Its just a data structure which holds hashes of files which will not be loaded. How you fill this structure is up to you. You can blacklist the files on by one, you can use a 3rd party list or a mix of both.

>>6857 >Its not censorable by a central authority. It is very much censorable by the site owner and by the peers who hold the site From what I understand, Zeronet site owners can update files and sign them with their private key and then well behaved peers will serve the new version and drop the old version. The "censorship resistance" comes from the idea that so long as a single peer seeds a site then the site will be available. So peers pick and choose which sites they will redistribute, which appears to avoid the problem >>6841 raised. Peers are solely liable for every site they visit and seed with no plausible deniability and are very easy to track down, which may be significantly worse depending on your threat model and where you live. It also means that it would be very easy to enumerate every node who is seeding a site and DDoS them off the net. Or for an ISP to block connections to specific peers who are seeding a specific site. So this censorship resistance is very limited and misleading. >How you fill this structure is up to you. You can blacklist the files on by one, How do you get the checksum of a bad file without downloading the file first? The only thing this does is stop you from encountering the same thing twice which is almost useless, like >>6855 pointed out. >you can use a 3rd party list This means a 3rd party can stop you from consuming and sharing any file/site he wants. I really wish people like you would not participate in decentralized networks because you are not just harming yourself with stupid ideas like this. Does Zeronet have a whitepaper or any serious documentation. I cannot find anything except this marketing slide deck which doesn't really explain much. https://zeronet.io/files/ZeroNet_Presentation.pdf One thing I am interested in is these updates. If I visit a site, and now I'm seeding it. What happens when the owner pushes out an update with illegal content? Maybe I don't visit the site again and do not notice. I am then basically serving illegal files without knowing it. Law enforcement arrives and I need to explain why I am distributing CP even though there was no CP when I visited the site and hope they believe me? Not good.

>>6866 you fundamentally misunderstand a few things. >Peers are solely liable for every site they visit and seed with no plausible deniability requires actual illegal content. Requires this content to be so hidden as for the peer to not see it but so exposed for the law to see it and so much of it for them to actually do something about it (protip: even for cp, in 99% of the cases they wont). And of course this requires a situation where the law discovers a peer who does not use tor and who coincidentally is inside their area of responsibility in addition to the above. Also, yeah you might be killed by an impacting asteroid this instant. Its about as likely. > and are very easy to track down No. You cannot influence which peer will serve you the site. You certainly cannot reliably get all peers of a site just as you cannot reliably get all peers in a BitTorrent swarm. >The only thing this does is stop you from encountering the same thing twice Yes > which is almost useless no its very useful. How would it be useless? You delete the file and make sure you never see it again. Thats the very definition of useful in this context. >This means a 3rd party can stop you from consuming and sharing any file/site he wants. I really wish people like you would not participate in decentralized networks because you are not just harming yourself with stupid ideas like this. Are you 12?? Please inform yourself at least a little bit before spouting nonsense like that. We are talking about a purely theoretical situation. So first, above, you rant about how oh so bad it is to have one single "illegal" file on your disk even by accident. Now you switch around and rant about how oh so bad it is to implement one single filter mechanism to prevent such a situation. Could you please pick a side and stick to it! Realistically if, and only if, you personally decide to import a hashlist to block unwanted files. You will pick one from a serious, respectable organization with a clear purpose. Lets say the NYPD-Known-CP-Images.txt and the NYPD-Known-CP-Videos.txt (hypothetical situation). So how exactly would that "stop you from consuming and sharing any file he wants"? Much less a whole site because thats clearly impossible (which you would know if you had done just a minimum of research) because a site consists of many different files, changing even one single of them will lead to a completely different hash of the whole site. >Does Zeronet have a whitepaper or any serious documentation. Keep reading. There is much more documentation out there. I am not your Butler. >One thing I am interested in is these updates. If I visit a site, and now I'm seeding it. What happens when the owner pushes out an update with illegal content? Maybe I don't visit the site again and do not notice. Others will. If a site is served by 2000 peers today and only by 20 tomorrow you bet that there was something fishy going on. And looking at a site is literally just one single click. Pic related its the dashboard and the file view. >Law enforcement arrives No it wont. Law enforcement goes only after the big fishes. They are too understaffed and overworked to bother with that one user of some ultra-obscure Technology who had tow thubnails which may or may not display a 17 year old with her tits out. Think about it logically: If it were really like you fear it is, no one ever would spread CP. This would be really great. Unfortunately it continues to be a problem today.

>>6851 >since IPFS itself uses bittorrent as distribution protocol too. that's not entirely true, the IPFS protocol is slightly different from the bittorrent one, you couldn't just torrent an IPFS website. IPFS also has parts that aren't bittorrent-like, like IPNS.

>>6877 IPFS on itself without the added tech like IPNS is basically just one single big BT swarm. There is no such thing as a dedicated IPFS website. Only files. Some of them can be html files.

>>6821 IPFS is what you are looking for.

>>6872 >You cannot influence which peer will serve you the site I have an IP address and/or and onion domain, I can DDoS it. >You certainly cannot reliably get all peers of a site You sound more like a clueless fanboy than an expert. Explain to me how it is not possible? If you cannot find peers you cannot download the file. Therefor it must be possible to enumerate peers who are sharing a specific file. You connect directly to them and download the file. Isn't that how it works? >Could you please pick a side and stick to it! Your blacklist idea maybe solves one problem but adds a new problem so it is not a good idea. The only "side" I pick is to not use this garbagenet you are shlling. >You will pick one from a serious, respectable organization with a clear purpose. Lets say the NYPD-Known-CP-Images.txt >So how exactly would that "stop you from consuming and sharing any file he wants"? It would stop you from consuming and sharing any file he wants because he can put any file hash into the blacklist. Your solution is trust the government to hold the blacklist and to only put bad things on it? fuarrrking seriously? >Keep reading. There is much more documentation out there. I am not your Butler. I already linked you to the documentation I found, so clearly I have looked. Since you are not aware of any other sources of documentation I will assume you know even less about the project than I do now. How sad. >Others will [notice]. That's not the point retard. It's an obvious weakness in the network that can trick users unto seeding illegal content without them knowing. It doesn't matter that statistically, someone is likely to notice after the attack has happened. >Think about it logically: If it were really like you fear it is, no one ever would spread CP. This would be really great. Unfortunately it continues to be a problem today. It's not just about CP. Don't tell me what my threat model is. It is a big world with many different types of laws and governments. Your whole defense is "law enforcement only go after the big fish hurr durr". The world is significantly less fair than you think. https://www.theguardian.com/us-news/2015/sep/20/teen-prosecuted-naked-images-himself-phone-selfies

>>6872 >No. You cannot influence which peer will serve you the site. You certainly cannot reliably get all peers of a site just as you cannot reliably get all peers in a BitTorrent swarm. According to the document that >>6866 posted, whenever you connect to a site, the site gives you a list of visitors' IP addresses. Those visitors are, presumably, seeding the site. You may argue that we don't know that just because someone visited the site that they're serving the site. Fair enough. So we just have to influence which peer will serve us the site: >You cannot influence which peer will serve you the site. I bet you can. Like >>6884 pointed out, you could DDoS all of the peers except one, forcing people to download the content from that peer. If you can DDoS all the peers except for that one and still get the file, then that peer is serving the file. However, DDoS is overkill. Since you have a list of IP addresses to connect to, instead of connecting to all of them, just connect to a single one and see if they're currently serving the file. It's It's interesting that you bring up the idea of using a blacklist of hashes to avoid accidentally downloading cp. That list has to come from two sources: either LE themselves, or a person not in LE. Why would LE want to give up that list? LE uses those hashes to find who is serving those files elsewhere. If they publish their list of hashes, cp vendors are just going to check the hashlist and kill off (or, more likely, change one byte of) any files they're serving that match a hash. Suddenly LE has a lot harder of a time catching the big fish cp vendors. Not in their best interest to publish this list. If a non-LE person starts publishing this list, LE is going to wonder where that person is getting the hashes. LE deduces that the person must be coming across cp and taking the hashes, raids them, and, best case scenario, the person has his or her reputation destroyed by the local newspaper's article about it. Worst case scenario, that person goes to prison. It's not exactly in their best interests to publish this list. >>One thing I am interested in is these updates. If I visit a site, and now I'm seeding it. What happens when the owner pushes out an update with illegal content? Maybe I don't visit the site again and do not notice. >Others will. If a site is served by 2000 peers today and only by 20 tomorrow you bet that there was something fishy going on. And looking at a site is literally just one single click. Pic related its the dashboard and the file view. So you're going to check every site that you're seeding daily for any bad files? I don't want to be a janitor for every site that I visit.

>>6884 >I have an IP address and/or and onion domain, I can DDoS it. No. >Explain to me how it is not possible? Because thats a fuarrrking unsolved problem in networks and distributed systems. If we could, it would solve many other problems. Unfortunately we cannot know if the network has given us all nodes we want, short of flooding the network with requests. BT is not a flooding based Network. Gnutella is (or was). You connect to one or more but not reliably to all. >e only "side" I pick is to not use this garbagenet you are shlling. The only side you pick is that of FUD to discourage people from a legitimate technology based on nothing but idiocy and twisted facts spiced with a healthy dose of not having a fuarrrking clue. See above. This is prime crackpottery on your part. On a level of assuming that the earth is flat. >It would stop you from consuming and sharing any file he wants because he can put any file hash into the blacklist. Alternatively you do not use a fuarrrking blacklist. No one is forcing you. But if you are a paranoid glitterboy then maybe the drawbacks a blacklist offer are OK for your. >I already linked you to the documentation I found, so clearly I have looked No you did not. Its only a tiny fraction of whats available online. THere is the whole API doc for example. And of course the sourcecode itself. >That's not the point retard. It's an obvious weakness in the network that can trick users unto seeding illegal content without them knowing. No one is tricking anyone here. Its all in your mind. If you really are this deathly afraid then feel free to use the blacklist, but apparently you dont want that either. The whole argument is ridiculous because you have that very issue with the normal internet too. No one can stop a website owner from pushing cp on your PC via JS. And yet this does not seem to be a widespread issue. >It doesn't matter that statistically, someone is likely to notice after the attack has happened. Yes it matters a whole lot. Given enough people such an attack becomes impossible. Ignoring of course why anyone would start an attack like that in the first place. Again your ignorance on how and why cp spreads clearly shows. Its not some "for teh lulz" kind of attack, its spread by sick people to other sick people. >https://www.theguardian.com/us-news/2015/sep/20/teen-prosecuted-naked-images-himself-phone-selfies One incident does not a statistic make >I bet you can. Like >>6884 pointed out, you could DDoS all of the peers except one Does not work like that. You cannot DDoS a BT swarm, therefore you cannot DDoS a Zeronet site because its the same technology. And trust me, people would do that if it was possible. Think of the money and tech the MAFIAA has which it can throw at piracy prevention. But it does not work. >Why would LE want to give up that list? Right now thats exactly what they are doing. You can actually get such lists if you want. >If they publish their list of hashes, cp vendors There are no "vendors" of cp. Thats not how it works. Its a sharing "economy". >If a non-LE person starts publishing this list... Moot as your premise was wrong. See above. >So you're going to check every site that you're seeding daily for any bad files? Again, thats not how cp works. Also: Do you check all temp files after each internet browsing session just to be sure that no one maliciously included cp on his site? You guy(s) remind me of the brain-amputees from the thread about yacy search engine. Are you the same?

>>I have an IP address and/or and onion domain, I can DDoS it. >No. No what? You don't have any IP addresses, or you can't DDoS them? If you have an IP address, you can DDoS it. If you don't have any IP addresses at all, you can't connect to anyone. And, as per the image in >>6893, you do have IP addresses. >>Explain to me how it is not possible? >Because thats a fuarrrking unsolved problem in networks and distributed systems. If we could, it would solve many other problems. Unfortunately we cannot know if the network has given us all nodes we want, short of flooding the network with requests. BT is not a flooding based Network. Gnutella is (or was). >You connect to one or more but not reliably to all. Of course you can't get all of the peers of the network. That's not the issue. The issue is that you're able to get the peers who are seeding specific content just by requesting that content. That's what matters. Want to knock someone's zeronet site offline? 1. Request the IPs for the site 2. DDoS all those IPs 3. Repeat You'll eventually knock down all the discoverable peers seeding that site. The peers that aren't discoverable? They're not doing anyone any good sense they're not discoverable anyway. >The only side you pick is that of FUD to discourage people from a legitimate technology based on nothing but idiocy and twisted facts spiced with a healthy dose of not having a fuarrrking clue. See above. This is prime crackpottery on your part. On a level of assuming that the earth is flat. Argue in good faith. >No one is tricking anyone here. Its all in your mind. If you really are this deathly afraid then feel free to use the blacklist, but apparently you dont want that either. He has only pointed out that a blacklist is censorship. Not what people want from a network that claims to resist censorship. >The whole argument is ridiculous because you have that very issue with the normal internet too. No one can stop a website owner from pushing cp on your PC via JS. And yet this does not seem to be a widespread issue. People have been arrested because malware was used to host cp servers on their machines. It happens. >I bet you can. Like >>6884 pointed out, you could DDoS all of the peers except one >Does not work like that. You cannot DDoS a BT swarm, therefore you cannot DDoS a Zeronet site because its the same technology. What doesn't work like that? You have a list of IP addresses you got from requesting the site. You DDoS them. It's simple. You can finesse it further by and running a malicious tracker to more easily enumerate the participants in the swarm. >And trust me, people would do that if it was possible. Think of the money and tech the MAFIAA has which it can throw at piracy prevention. But it does not work. Do you think MAFIAA doesn't track the IPs of people who pirate things via torrents? How do you think they figure out who is torrenting their "property" and send complaints to their ISPs? >>So you're going to check every site that you're seeding daily for any bad files? >Again, thats not how cp works. I didn't say "cp." I said "bad files." Maybe I don't want to seed content calling my government fascists because my government is fascist and will persecute me for distributing such information. Maybe there's other content that I don't want to seed. >Also: Do you check all temp files after each internet browsing session just to be sure that no one maliciously included cp on his site? I don't check it, but my browser is configured to not save temporary files or any history, and my tmp folders are periodically srm'd. Your comparison is invalid. If I'm browsing some imageboard and someone posted a cp thread, but I don't see it because it's at the bottom of page 1 and I only scrolled halfway down the page and never saw the thread, then sure, the image is in the browser cache for a bit, but I haven't advertised to anyone that I've got this file. But if I start seeding a zeronet site, then a couple days after I've started automatically pushing its updates out, someone posts cp, or any other bad file, say a leak of the newest Taylor Swift single, on it while I'm sleeping, even if I wake up and immediately delete the file, my IP address was still associated with that file publicly for however long it was up before I deleted it. For me, that's bad. MAFIAA has sued people for seeding copyrighted content, and they're not going to care that I didn't realize I was seeding it because "it's just part of the zeronet protocol! I didn't see it! Promise!" >You guy(s) remind me of the brain-amputees from the thread about yacy search engine. Are you the same? No, I never posted in that thread. Argue in good faith.

>>6895 >So first, above, you rant about how oh so bad it is to have one single "illegal" file on your disk even by accident. Now you switch around and rant about how oh so bad it is to implement one single filter mechanism to prevent such a situation. Could you please pick a side and stick to it! >Doctor help my leg hurts >Cut your leg off then >But if I do that then I can't walk >Could you please pick a side and stick to it! Do you have learning difficulties or something? >The whole argument is ridiculous because you have that very issue with the normal internet too. No one can stop a website owner from pushing cp on your PC via JS. And yet this does not seem to be a widespread issue. There is a nuance you are missing in your strawman here. The problem is not accidentally downloading CP. The problem is Zeronet will *redistribute it* through the network, using your real IP address, and in the case of this attack you don't even know it happened and therefor can't stop it. Please go back to the beginning of this discussion >>6842 >we are talking about only a distributed net, not an encrypted one. Therefore it is possible to see and delete the offending files (like with Zeronet) Your argument is that users can review everything they seed on Zeronet so they can catch and delete CP. And I showed you how a site owner can inject CP after the fact without the seeders knowing what they are seeding. Now you say CP is not a big deal anyway, LE only go after "big fish" so don't worry. This is seriously not cool. >you cannot DDoS a Zeronet site because its the same technology. It's not the same technology. Zeronet doesn't split files into chunks. You get the whole file from a single peer. And this does not make it immune to denial of service anyway. >No you did not. Its only a tiny fraction of whats available online. Don't pretend you have more knowledge than you do. If you can't show me where the documentation is then you didn't read it. Simple. >You guy(s) remind me of the brain-amputees from the thread about yacy search engine. Are you the same? I saw that thread. You seem to be very good at destroying the reputation of technologies by shlling them without understanding how they work and then getting huffy at people who point out flaws. You should try to learn from your mistakes.

>>6898 >If you have an IP address, you can DDoS it. No. Read up and come back only when you understand why only an IP address is not enough to reliably DDoS anyone. >Of course you can't get all of the peers of the network. That's not the issue. And yet thats exactly the point upon which your whole argument rests. Without it its completely moot. >1. Request the IPs for the site And per your admittance aboce you do not get all IPS, thus the site is still online even if you take out some of the peers. Just like you cannot DDoS a Torrent swarm. But I repeat myself here >2. DDoS all those IPs Again not possible. See above. >Argue in good faith. Right back at you. Its about time you start doing that. >He has only pointed out that a blacklist is censorship No he (you actually, you are not fooling anyone) was talking out of both sides of his mouth and not even being subtle about it. Ignoring of course that all that is purely optional. > Not what people want from a network that claims to resist censorship. It resists censorship just fine because it cannot be censored from any central authority. Individual peers may chose to (or not) not keep certain files. Just as you as a person may chose to (or not) repeat certain arguments you heard. >People have been arrested because malware was used to host cp servers on their machines. It happens. Yes and people have also been killed by meteroids. One incident does not a statistic make. >You have a list of IP addresses you got from requesting the site. You DDoS them For the THIRD time you show blatant ignorance when it comes to BT and DDoS. Read up on what you talk about its getting embarrassing (for you). These are all completely elementary issues and you fail at them. >Do you think MAFIAA doesn't track the IPs of people who pirate things via torrents? You were talking about DDoSing them. Tracking (which they dont either, at least not on a wide scale) is something fundamentally different. >I didn't say "cp." I said "bad files. No you said cp and cp vendors. > Maybe I don't want to seed content calling my government fascists because my government is fascist and will persecute me for distributing such information. The Zeronet alone is the wrong technology for you. Use it either with Tor, or switch to Retroshare or go to Freenet, Gnunet or similar hardcore privacy-enhancing applications. Here too you commit the rookie mistake of confusing privacy, anonymity and plain anti-censorship. >I don't check it, but my browser is configured to not save temporary files or any history, and my tmp folders are periodically srm'd. Great, configure Zeronet to do the same and you are set. >but I haven't advertised to anyone that I've got this file. Does not matter (according to you) the file is on your computer and now you are going to Alcatraz for at least 200 years. >while I'm sleeping, how would that work when your computer is switched off. Also see above: If you are a paranoid (legitimate or not) then just do not safe any zeronet files, instead re-download them every time you visit. You are seeing a downgrade compared to the normal web when there is none. Anything that is safe to do in the normal web is AT LEAST as safe to do in Zeronet. And again: Anti-censorship != pro-privacy. If you need more privacy either connect to ZN using tor or switch the technology completely. But dont try to losen a nut with a chainsaw and complain to the vendor that this wrench is defective. You will only get laughed out. Just take a look at the OP: >I been pondering about this idea, a website that isn't hosted by servers but by the users as they use it >I know is not original but how come no actual websites use it? This is not about surviving in north korea or spreading bibles in Saudi arabia online, its just about not needing a central server.

>>6906 >No. >Read up and come back only when you understand why only an IP address is not enough to reliably DDoS anyone. You are being too literal and ignoring the context. It is correct that an IP address is not enough. The point is that A) you have a way to connect directly to a node (IP address or onion address) and B) it is listening on a port and will talk a known protocol (Zeronet). These two conditions ARE enough to launch a denial of service attack. >And per your admittance aboce you do not get all IPS Actually, what >>6898 said was >Of course you can't get all of the peers of the network. That's not the issue. The issue is that you're able to get the peers who are seeding specific content just by requesting that content. Don't take quotes out of context like that. >It resists censorship just fine because it cannot be censored from any central authority. Freenet is censorship resistant because the only way to stop a file from circulating is to take down the whole network, or stop users connecting to every other node in the network. Zeronet is not really censorship resistant because it is very easy to identify which nodes are serving which file and only take down those nodes or stop users connecting to those nodes. This is particularly poignant because "uncensorable" is part of their tagline (Open, free and uncensorable websites, using Bitcoin cryptography and BitTorrent network). The whole "Bitcoin cryptography" stuff also smacks of nothing but empty hype. >You are seeing a downgrade compared to the normal web when there is none. Anything that is safe to do in the normal web is AT LEAST as safe to do in Zeronet. We have already established that this is completely wrong. On the normal web, only the server and my ISP knows what files I download. In Zeronet the whole world knows what files I download. That's not "as safe", that is much much worse. On the normal web I download files and that's it. On Zeronet I download files and automatically share them with other people. If that file turns out to be illegal I am in much bigger trouble because distributing illegal content is much worse than simply downloading it.

>>6908 >These two conditions ARE enough to launch a denial of service attack. No. Seriously just stop. The internals of the internet are a bit more complex than that. >Don't take quotes out of context like that. That is the context. You cannot get all nodes reliably that seed a certain file. It seemed that you had finally realized that too but I guess you did not in the end. Does not matter, this is a fact. Feel free to dispute the relevant scientific papers about high-churn p2p networks. >s very easy to identify which nodes are serving which file and only take down those nodes or stop users connecting to those nodes. 3x wrong. - easy to identify some nodes that are serving the content (again: ffs. look up key based routing before making an even bigger fool out of yourself. I recommend "distributed systems" by Tannenbaum) But not possible to identify all. - not at all easy to take down a node. Again: look up the details on what a DDoS is and what it is not. Your crass oversimplifications or lack of understanding make your faulty arguments possible, not reality. - >stop users connecting to those nodes What the hell do you mean whith that? Thats even more impossible than taking down all nodes with DDoS. Unless you are in charge of every ISP through wich users might connect you cannot do that. >This is particularly poignant because "uncensorable" is part of their tagline And thats exactly what it is. Unfortunately you lack the knowledge to understand that. As well as (much worse) the desire to aquire that knowledge. >We have already established that this is completely wrong. No, we have established that you know jack soykaf about anything network related, nothing more. >In Zeronet the whole world knows what files I download fuarrrking wrong again. (are you trying to see how often you can be wrong in a single post or something?) No one knows the files you download in Zeronet. They would have to be connected to the same swarm as you are, and even then there is only a nonzero probability that your node would tell them that it hosts the same file too. Nothing more. >That's not "as safe", that is much much worse. In your imagination, yes. >On Zeronet I download files and automatically share them with other people See above: Only those in the same swarm are only ever (possibly) aware that you have those files. For this to be dangerous to you there has to be a whole chain of situations that are incrasingly unlikely. In the end you are fearing to get the lotto jackpot twice in a row. A few examples (I repeat myself but maybe now you learn): - The file(s) have to be so bad that LE actively wants to take down everyone of those files and the nodes that hosts them (not even cp will provoke that reaktion, everything else will provoke even less) - The files have to be accidentally on your nodes without you realizing that. - The files however have to have so many seeders that there is a reasonable likelihood that some of them are from LE. Biggest site on Zeronet has around 3k seeders. So thats a no too. - The files have to remain for a while in that state so that all (or most) nodes actually get them. So that means that they must be site-owner approved. Yet every other node must also be OK with them for some reason because apparently in our situation no one blocks or deletes those files. Again a contradiction. Lets imagine the site is nullchan and the file is a user-submitted image of cp. Now everyone, just as here and on 4chan, will immediately lose their soykaf and will point this out with instructions on how to blacklist that file. This is not hard to miss. - Back to LE: The LE node must be lucky enough to connect to another node that is coincidentally in the same region as the LE. Because of Key based routing the closeness of nodes in a DHT is uncorrelated to their closeness in geographic terms. So the LE would have to be from your country, or probably from your city even for anything to happen at all. And keep in mind that this is all grounded in statistics, there is no way to demand to be connected to one particular node meeting a set of criterias in a key based overlay network. - After the LE has connected and downloaded the offending file from your node (mind you by now the likelyhood of that happening is about one to several billions if not worse) they need to find it bad enough that they bother getting an order against your ISP and then a search order against your house. (protip: Unless the file is a new video file of some torture-cp then no one will bother). - Then they will need to prove that the file really was from you and that you willingly spread it. For that the files still has to be on your node when they seize your PC. Even less likely. - Finally the judge has to agree with them. (even less likely) After all there are enough cases where the IP address was wrong or the ISP did not keep the customer-IP mapping by the time the order came. As you can easily see this chain of events is so beyond unlikely that its quite hilarious. Right now you are turning down a great improvement over the current state of things because of your crass lack of knowledge and paranoia about impossible scenarios. While you are free to be as much of a moron as you want, you are right now also trying to turn away people from this technology who could benefit from it. And this is something I really hate. You are like those paid Microsoft sh-ills that never tire of raising one non-argument after the other when trying to discredit Linux or Open Source in general. >>6908

>>6906 >>If you have an IP address, you can DDoS it. >No. >Read up and come back only when you understand why only an IP address is not enough to reliably DDoS anyone. what is icmp flooding? what is syn flooding? what is simply making a very large amount of frequent requests for the file that the IP address is serving? If there is something here that would explicitly prevent any and all DDoS techniques, I would be gracious if you would point it out. >>1. Request the IPs for the site >And per your admittance aboce you do not get all IPS, thus the site is still online even if you take out some of the peers. However, you can get all the IPs of the discoverable peers, and you can DDoS them. Like I said, it doesn't matter if someone is still seeding the site if that person's machine isn't discoverable by anybody. >>Argue in good faith. >Right back at you. Its about time you start doing that. Nowhere in my posts (>>6898 and >>6893) have I done anything but. >>Not what people want from a network that claims to resist censorship. >It resists censorship just fine because it cannot be censored from any central authority. ISP tracks the IPs that seeds a zeronet site, then blocks connections to them. That's one potential vector of censorship. Regardless, not being censored by a central authority does not mean censorship resistant. >You were talking about DDoSing them. Tracking (which they dont either, at least not on a wide scale) is something fundamentally different. I'm talking about both. Easily being tracked is a deterrent for many people. Tracking also can lead to censorship. >>I didn't say "cp." I said "bad files. >No you said cp and cp vendors. What I said was "So you're going to check every site that you're seeding daily for any bad files?" Cp is just one example that's been used in this thread. >Here too you commit the rookie mistake of confusing privacy, anonymity and plain anti-censorship. I was not talking about censorship here. The other lainon in disagreement with you brought up concerns about seeding content from a site and then later having that site push illegal content out, that his machine would then seed. Your solution was essentially, "just look at it and delete it," but that doesn't adequately solve the issue for the reasons we've detailed. There is also the argument to be made that you can't truly be resistant to censorship without anonymity, but that is more of a philosophical argument that I feel would be a digression from the more technical-oriented discussion. On that regard, I would prefer if we could leave that point as "some people would suggest anonymity is a prerequisite to defeating censorship," leave it at that, ignore it, and continue discussing the technical aspects of Zeronet instead. >>but I haven't advertised to anyone that I've got this file. >Does not matter (according to you) the file is on your computer and now you are going to Alcatraz for at least 200 years. I have not said, nor implied, any thing of the like. >according to you https://www.youtube.com/watch?v=Pu1aQvm5MrU >>while I'm sleeping, >how would that work when your computer is switched off. I do not switch my machine off when I sleep. Most people I know do not. I would also imagine there are people with seedbox-like setups for Zeronet, and they're obviously going to have issues with this, since it's unreasonable to expect them to monitor everything 24/7. Sleeping is one example. Maybe I leave my computer for a few minutes to use the bathroom, and right as I leave, I get an update with the bad content and my machine starts serving it to people while I'm away. Or maybe, like normal people, I'm doing something else while on my computer and don't want to dedicate my time to monitoring the site. >just do not safe any zeronet files That will keep me safe, but does not mitigate any risk for the people who are involved in seeding the files. >use it with Tor This speaks to Tor's ability to be "safe" not Zeronet's. It should go without saying that there are ways to secure your opsec and do anything you want on the internet safely. That doesn't make up for shortcomings in Zeronet itself. >Just take a look at the OP: >>I been pondering about this idea, a website that isn't hosted by servers but by the users as they use it >>I know is not original but how come no actual websites use it? >This is not about surviving in north korea or spreading bibles in Saudi arabia online, its just about not needing a central server. Take another look at the OP. "how come no actual websites use it?" OP is just as much asking about what deters people from involvement in such networks. >>6908 >The point is that A) you have a way to connect directly to a node (IP address or onion address) Well, technically having an onion address doesn't give a direct connection. You can still try to DDoS hidden services by conventional means, but you'll end up DDoSing the relays and guard nodes for the hidden service as well as (or instead of) the hidden service. But there's nothing with zeronet that entails using Tor hidden services so it doesn't matter for this discussion, and even if it did, there are DoS attacks effective against hidden services.

>>6911 >what is icmp flooding? what is syn flooding? what is simply making a very large amount of frequent requests for the file that the IP address is serving? the only thing this does is kick yourself off the net because your ISP will categorize this as malicious behaviour. icmp and syn flooding is quite a bit dated. The most important point however is that you cannot DDoS one node with only your node. You need either a botnet or an amplification attack which is not automatically stopped. You might be on a laptop but many of the Zeronet nodes are full blown servers. They can soak up much more data than you and your 4 buddies can dish out. Now you do not only want to take down one node but every single one. This is something even professional internet-criminals would struggle to do. >If there is something here that would explicitly prevent any and all DDoS techniques Yes redundancy. see above. >However, you can get all the IPs of the discoverable peers, and you can DDoS them 1. IF. IF! you can get all the IPs... but you CANNOT! thats the whole problem in your argument. 2. See above even IF you had all IPs you could not take them all out unless you had spend literally years building up your botnet, And even then it wont be possible for popular sites. Otherwise (see my post a few posts above) that would be a routinely used technique by the MAFIAA to stop file sharer. >Nowhere in my posts (>>6898 and >>6893) have I done anything but. No. On the contrary. >ISP tracks the IPs that seeds a zeronet site, And they would do that because?? Also how would they be able to do that through encryption? And then add TOR in the mix. Just as impossible as that ridiculous idea of DDoSing every single IP address. Please lay off the Hollywood moves for a while. Reality is nothing like that. >then blocks connections to them. Not possible, see above. Even China does not manage to do that reliably. VPNs still get through. >Regardless, not being censored by a central authority does not mean censorship resistant. Thats what it means. Please also learn some english. >What I said was "So you're going to check every site that you're seeding daily for any bad files?" No. you ranted about cp vendors. It matters not as cp is the only thing that could possibly get you into trouble internationally. anti-communism is safe in the USA. same as pro-nazi. Blasphemy gets you in trouble in Iran and Pakistan but not in China. And so on. Only cp is universally regarded as bad everywhere. Thats the only kind of file for which you argument could possibly make traces of sense. >but that doesn't adequately solve the issue for the reasons we've detailed. yes it does. Quite adequately even. >I have not said, nor implied, any thing of the like. you did. Why else would you obsess about hypothetical scenarios where one single file gets you into trouble so much then? >I would also imagine there are people with seedbox-like setups for Zeronet, and they're obviously going to have issues with this, since it's unreasonable to expect them to monitor everything 24/7 If they do they have the same defense as operators of Tor exit nodes have. They are, depending on the country, even common carriers with nothing to fear even if the sickest cp goes through their node. >That will keep me safe, but does not mitigate any risk for the people who are involved in seeding the files. Coming back to the "normal" internet: You have the same issue for operators of servers where illegal files are located. This is not a Zeronet issue. >hat doesn't make up for shortcomings in Zeronet itself. The inability of a submarine to fly is not a shortcoming. It was never designed for that. If you want anti-censorship you take Zeronet. If you want privacy you take tor. (if you want both...) >OP is just as much asking about what deters people from involvement in such networks. Yes and your arguments from fantasy-land are not what deters people.

>>6912 >the only thing this does is kick yourself off the net because your ISP will categorize this as malicious behaviour. If that was true DDoS wouldn't exist. Don't make things up. >The most important point however is that you cannot DDoS one node with only your node. Yes you can. Reflection and slowloris are two examples. >This is something even professional internet-criminals would struggle to do. Don't project your own incompetence onto everyone else. Judging by this thread I think it's safe to assume that the majority of people are a lot smarter than you are. >>6911 >But there's nothing with zeronet that entails using Tor hidden services Hidden services seems to be how their Tor integration works. https://github.com/HelloZeroNet/ZeroNet/blob/master/src/Tor/TorManager.py#L216 See how the handshake sends an onion address instead of peer_id https://zeronet.readthedocs.io/en/latest/help_zeronet/network_protocol/#handshake >there are DoS attacks effective against hidden services. Correct.

>>6914 >If that was true DDoS wouldn't exist. Don't make things up. There it is again: Your utter and total lack of knowledge. DDoS is something else than DoS. The first D stands for distributed. Hence why I mentioned Botnets. >Yes you can. Reflection and slowloris are two examples. No, you cannot. The receiving node would have to be extremely weak. And you would have to find enough unpatched machines online that would work as amplifiers. >Don't project your own incompetence onto everyone else If it were possible there would be much more DDoSes. Right now there was not one single successfull DDoS against a BT swarm of any size. And even famous large scale attacks were rendered inefective when the sites moved to google because the infrastructure there was able to just swallow up all the traffic the botnet sent: http://www.zdnet.com/article/google-rescues-krebs-on-security-from-the-abyss/ So right now a criminal organization with a huge botnet (and not just one node as you thought would be enough) cannot take down one single other node if its supported by google infrastructure. Now you thought that it would be possible to not just take down one node but every single node (ignoring of course that you do not reliably get all nodes. But I repeat myself here) using only one node as attacker. Thats 3 total fails in one. Do you realize just how abysmal your knowledge of network technology has to be to successfully fail three times in just one single statement? And not just that. Usually people once they realize they dug themselves into a hole stop digging. They quietly walk away. No, not you. You keep on digging while insulting my lack of hole digging skills.

>>6915 there was not one single successfull DDoS against a BT swarm of any *significant* size. And even famous large scale attacks *against websites" were rendered Fixed.

>>6879 >There is no such thing as a dedicated IPFS website. Only files. Some of them can be html files. that's besides the point: you couldn't download any IPFS file using a bittorrent client, because the protocols are different. >all this talk about DDoSing bittorrent or something even taking down a single server with a DDoS for any appreciable time is not trivial, taking down thousands simultaneously would be quite a feat. Similarly with tracking bittorrent swarms: the easiest and most effective way to do this is just to seed a lot of torrents for a long time and see who connects. That's how media companies catch people torrenting.

>>6918 >Similarly with tracking bittorrent swarms: the easiest and most effective way to do this is just to seed a lot of torrents for a long time and see who connects. That's how media companies catch people torrenting. It's more effective to host a malicious tracker or DHT. Swarm participants report to the trackers, and you get their IPs. Query the other trackers involved multiple times and it's been shown to be effective in determining the participants. >>6912 >You might be on a laptop but many of the Zeronet nodes are full blown servers. They can soak up much more data than you and your 4 buddies can dish out. We're not talking about some teenage kid running LOIC and spamming 4chan with personal army requests to do the same. We're talking about a malicious actor of in-determinant resources e.g. the feds or people with sizable botnets. >1. IF. IF! you can get all the IPs... but you CANNOT! thats the whole problem in your argument. Let me reclarify: all useful IPs. Malice Alice takes her botnet, puts a portion of her zombies towards joining the Zeronet network, and then repeatedly requests IPs that serve the site and forwards those to her c&c server. With enough machines dedicated to this, Alice can reliably enumerate the peers currently involved. Or she can just host a malicious tracker or DHT and determine the swam participants that way, which has been shown to be effective in practice. She now has all the discoverable IP addresses that are currently sharing the content. Again, I will concede that she may not have every IP address that is participating, however, the IP addresses that she is unable to discover are not going to be discoverable by most normal Zeronet users, and so the undiscovered IP addresses really are not being useful. >that would be a routinely used technique by the MAFIAA to stop file sharer. The MAFIAA isn't going to risk the lawsuit for DDoSing people. That would cost them much more money than letting people pirate files. >>ISP tracks the IPs that seeds a zeronet site, >And they would do that because?? Who cares why they would do it? The point is that they could do it. Maybe they're in an oppressive regime that forces ISPs to censor negative material about the country. Maybe the ISP is just upset about the content of the site. Maybe the ISP is being paid off to censor it. Doesn't matter. All that matters is that the ISP could. >Also how would they be able to do that through encryption? An ISP can block a connection to an IP whether it's encrypted or not. And the ISPs can discover the IP addresses involved with a zeronet site by connecting to the zeronet network in the same way that you can discover the IP addresses involved to connect to them, or the same way that Malice Alice did above. >>Nowhere in my posts (>>6898 and >>6893) have I done anything but. >No. On the contrary. I would appreciate it if you could point out specific instances you believe are me ( >>6898 >>6893 >>6911 ) arguing in bad faith. >Not possible, see above. Even China does not manage to do that reliably. VPNs still get through That doesn't make Zeronet resistant. That makes VPNs good at subverting censorship. >>Regardless, not being censored by a central authority does not mean censorship resistant. >Thats what it means. Please also learn some english. If I made a theoretical service that let anyone post anything, but also let anyone delete anything, would you call that censorship resistant? No central authority is deleting anything, but anyone can run a few bots that deletes things as soon as they're posted and censor content. Therefore, simply not falling victim to censorship via a central authority does not a censorship resistant service make. >>What I said was "So you're going to check every site that you're seeding daily for any bad files?" >No. Yes. The exact quote is, "So you're going to check every site that you're seeding daily for any bad files?" And I've given other examples that some people might reasonably be worried about besides cp. Cp was only mentioned because you and the other lainon discussing this were using it as an example. >>>but I haven't advertised to anyone that I've got this file. >>>Does not matter (according to you) the file is on your computer and now you are going to Alcatraz for at least 200 years. >>I have not said, nor implied, any thing of the like. >you did. Why else would you obsess about hypothetical scenarios where one single file gets you into trouble so much then? I have not argued that simply having an certain file on your computer is going to inherently cause you legal issues or even represents much of a risk. What I've argued is that seeding certain files and publicly advertising that you have them carries risk and that it's unreasonable to expect individuals to monitor the contents of the zeronet sites they seed 24/7 for the files they deem too risky. This will deter some people from participating in the network. Even with Freenet, where there is a bigger focus on anonymity and it's not possible to determine which files you have, people have been deterred from using it due to the chance of hosting files they don't want to host. Not only is hosting certain files risky, some people may feel unethical about participating in a network where they may inadvertently help supply people with certain content they find unethical. This is another deterrent. >>I would also imagine there are people with seedbox-like setups for Zeronet, and they're obviously going to have issues with this, since it's unreasonable to expect them to monitor everything 24/7 >If they do they have the same defense as operators of Tor exit nodes have. They are, depending on the country, even common carriers with nothing to fear even if the sickest cp goes through their node. Depending on the country. People have been arrested for running Tor exit nodes. Not all countries or ISPs are friendly to this defense. Again, this is a deterrent for some people. >>That will keep me safe, but does not mitigate any risk for the people who are involved in seeding the files. >Coming back to the "normal" internet: You have the same issue for operators of servers where illegal files are located. This is not a Zeronet issue. My point. Zeronet is not any better in this regard than the normal internet. In fact, it's worse because people may be good Samaritans who are seeding zeronet sites without knowing whats on them 100% of the time, and may be investigated by LE (or random people) because of that. People who want to contribute to the network run the risk of contributing to the distribution of content they may not want to distribute. Some people don't want to do this. Hence, it's a deterrent. >And even famous large scale attacks were rendered inefective when the sites moved to google because the infrastructure there was able to just swallow up all the traffic the botnet sent: > >http://www.zdnet.com/article/google-rescues-krebs-on-security-from-the-abyss/ The provider says it could have cost them millions of dollars if it was sustained. Protection against the level of attack was offered for $150,000-$200,000 a year. That's going to destroy most people financially. >google... was able to just swallow up all the traffic Google also has 500 billion dollars and an expensive infrastructure, that's being provided to the guy for free because he's a journalist. Zeronet doesn't do anything to protect people against DDoS. Project Shield does, Cloudfare does, and the provider offering protection for $150,000 does. An arbitrary zeronet seeder isn't going to qualify for Project Shield, and DDoS protection is expensive. If you can force people to stop hosting something by making it too costly for them to do so, you can censor them. That's the point of DDoS. >Now you thought that it would be possible to not just take down one node but every single node using only one node as attacker. Where did you get the idea that anyone was suggesting using one node? We've been saying DDoS, which, as you pointed out, necessitates having more than one node. >>But there's nothing with zeronet that entails using Tor hidden services >Hidden services seems to be how their Tor integration works. >https://github.com/HelloZeroNet/ZeroNet/blob/master/src/Tor/TorManager.py#L216 >See how the handshake sends an onion address instead of peer_id >https://zeronet.readthedocs.io/en/latest/help_zeronet/network_protocol/#handshake That's pretty cool, but I reiterate that this doesn't make zeronet inherently safe, since it can be used without Tor. It only shows that Tor is safe. Anything ran through Tor is safe, sense Tor is actually censorship-resistant and anonymous.

>>6918 >that's besides the point: No thats exactly the point. >even taking down a single server with a DDoS for any appreciable time is not trivial, taking down thousands simultaneously would be quite a feat. thank you. Finally a rose among the thorns here (nohomo) >>6926 >It's more effective to host a malicious tracker or DHT. there is no such thing as a malicous dht. the D stands for Distributed. Contrary to a tracker a DHT is not "a thing" its not located somewhere. Its just an overlay network with operations defined on it. Most likely a distance metric, a get_data() and a put_data() function. Sometimes also a rating function for the trust of nodes (not on the BT DHT as far as I know) >We're talking about a malicious actor of in-determinant resources e.g. the feds or people with sizable botnets. Well clearly so far those actors have not yet done anything like that. As I mentioned in my post: Even a large criminal botnet has proven to be useless when taking down one single node. And neither have the feds or large companies (MAFIAA is pretty big) or we would know about it by now if BT suddenly became unusable for popular files. >Alice can reliably enumerate the peers currently involved. operative word here is currently. BT in the retail context (that is outside of professional tools that may use BT as one component) is a high-churn network. That is: peers join and leave constantly. Peers leave and rejoin with a completely different IP address too. Only if most or all nodes were servers with a permanent connection and thus an unchanging IP would your approach make any sense. Right now every 24h the majority of the IPs have changed because most ISPs reassign a new IP to each of their customers every 24h. That has proven to be a major "problem" for the MAFIAA who assumed that IP address = people and often have sent the cease&desist letters (or your country's equivalent) to the wrong people. This high-churn property is usually thought of as a big challenge from the network developer side of things (read the "distributed systems" book by tannenbaum thats where its explained in more details) but in this situation it works to the advantage of the swarm and against a theoretical attacker. Keep in mind please that many powerful people really hate BT and filesharing in general. If this attack against Zeronet would be effective in reality then it would also be effective against plain "standard" BT, and therefore it would be used often. True the MAFIAA and the feds may not care one bit if you share an album that is 2 years old. But they very much do care if a movie is leaked before it even hits the theaters. In this situation I have to agree with them for once: thats where lots of money can be lost. Therefore we should see this kind of attack deployed AT LEAST everytime such a leak occurs. However we do not. From that alone we can conclude that this attack is either impossible (my opinion) or not effective at all. >With enough machines dedicated to this, Alice can reliably enumerate the peers currently involved Right. Now think about how much power it took to get all nodes, or even just the majority. Now consider how much more it will take to kick each one off the net. You will need many, many bots per node because you cannot use each bot at full capacity or they risk getting disconnected either from their ISP (who automatically detects malicious flooding) or by the owner of the bot who reboots his machine every couple of minute because "Gee my PC sure is acting up today". >Or she can just host a malicious tracker or DHT Again, this DHT is probably not what you think it is. >The MAFIAA isn't going to risk the lawsuit for DDoSing people. That would cost them much more money than letting people pirate files. Well who else would if not them? See my example above with the pre-release leak. Realistically they could afford one or 20 lawsuits if it means preventing such an early leak. However they do not in the end. Thus they must have concluded that this does not work. Aside from the MAFIAA I see no other actor that would have both the money, infrastructure and motivation to even attempt a large-scale shutdown of each node in a BT swarm. >Who cares why they would do it? The point is that they could do it. Yes and also the government could just send its thugs to random houses, shoot the inhabitants and plant physical cp there. Yet they dont. Monitoring is first and foremost a huge expense. ISPs would very much love to not monitor anyone. There was a huge whining in Germany a few years back when ISPs were required to start keeping the IP adresses of their clients for 6 month. Just the IP addresses, not even the data that was transmitted between them. If you want to construct a theoretical situation where ISPs do this monitoring on a large scale you would first have to give a reason why they would spend the money on additional employees, hardware and software. The reasons you gave are basically "for teh evulz". Even china does not fit in here. They just reset all packets that contain a certain protocol ID, without even looking at the actual content. And even then (again, keep in mind on how many layers of IFs we are now, each of them unrealistic) this would only apply to a tiny part of Zeronet because thanks to key based routing, the peers are all over the world. The nodes and their closeness to each other (usually expressed as the XOR of two node IDs which are cryptographic hashes of a random number) has nothing to do with the physical reality. >An ISP can block a connection to an IP whether it's encrypted or not. Yes, but again: Zeronet and BT are high-churn networks. They block the route to all 500 nodes currently serving a site, after 24h 450 of them are reachable again because they got new IPs. >And the ISPs can discover the IP addresses involved with a zeronet site by connecting to the zeronet network in the same way that you can discover the IP addresses involved That they can does not mean that they will. What motivation would they have to do so? Not to mention that doing so would be highly illegal in most countries and obligate them to keep filtering other stuff too. If they show that they have the capacity to do this, then tomorrow the whole MAFIAA is banging on their door to do that for each of their precious songs and movies thats getting shared online. This is the last thing any (for profit) ISP would ever want. I am not talking about something like North-k-online but about real ISPs of reasonable size. This includes ISPs in terror regimes like china and russia of course. Even there they wont do this. >I would appreciate it if you could point out specific instances you believe are me ( >>6898 >>6893 >>6911 ) arguing in bad faith. Because pretty much every argument rests on impossible fantasy scenarios that are utterly unconnected to reality. I must assume that you are either inventing those out of sheer lack of knowledge (or rather the most dangerous kind of knowledge: a little) or that you are trying to deliberately keep people from using a better and safer technology. Your arguments against Zeronet rely on freak incidents like a whole family getting killed by 10 separate meteorite strikes. They may look possible on paper but they are utterly impossible in reality. Thats either bad faith or cluelessness. >That doesn't make Zeronet resistant. That makes VPNs good at subverting censorship. No it means that it is clearly impossible to block connections as you imagine. China with its billions invested in the great firewall is unable to stop people connecting to VPN servers which are clearly visible as such. How the hell then would anyone be able to block connections to Zeronet or BT nodes? when those nodes may sometimes be a zeronet node and sometimes not. When the same IP can lead to grandmas macbook and a few hours later to 1337-k1d-2005's pimped out seedbox? Keep in mind the scaling issue too. Can this be done for 1 node: sure. Maybe for 10. But what ISP or country has only 10 or even just 1000 people? (1/2)

>>6926 >>6918 (2/2) >but anyone can run a few bots that deletes things as soon as they're posted and censor content. If those bots somehow managed to delete stuff on the seeding node, or on other nodes beside themselves then yes that would be censorship. However if you delete or blacklist a file in Zeronet this is only on your OWN node that you do this. It has zero effect on anyone besides you. Therefore it is censorship resistant. >Cp was only mentioned because you and the other lainon discussing this were using it as an example. Because cp is the only thing that could possibly cause problems globally. Everything else is just a local issue. So out of all those unrealistic cases you gave, cp was the least unrealistic, and even there it turned out to not be a problem in the end. >What I've argued is that seeding certain files and publicly advertising that you have them carries risk Well you are not publicly advertising at all. Its all limited to the particular site in Zeronet. So no one ever CAN know about those files EXCEPT if they request that particular site too. And even then there is no way to influence from which node you actually get the files. Assuming the utterly unrealistic case that an overeager, overpaid and underworked LE officer (3x unrealistic assumption) bothers to actively and compulsively connect to every possible Zeronet site out there and somehow with superhuman speed and knowledge find exactly that one offending file (again: only cp could possibly be realistic as kind of offending file). And it can only be one or very few files, unless the site is called something like Pedobear's cave, but in that situation the seeders know exactly what they are doing. Now that LE has to be lucky too and hope that the node that gave him that file is also within his authority. (I explained that already in another post.) Once more: A wholly impossible scenario which you use to (I think) compulsively create a threat-scenario where there is none. You sound like the kind of mom that would never let her kid go out of the house without a hazmat suit, armed guards and tank-resistant body armor. >it's unreasonable to expect individuals to monitor the contents of the zeronet sites they seed 24/7 for the files they deem too risky. Thats unreasonable, yes indeed. And its completely unnecessary. >This will deter some people from participating in the network. There will always be people being deterred by misunderstandings and lack of knowledge. Case in point: you. >some people may feel unethical about participating in a network where they may inadvertently help supply people with certain content they find unethical. Agree. Which is why I would personally not seed a site called "Pedobear's cave" or similar. But seeding by accident one single file for only a short time before blocking it is not unethical. Again if you disagree: feel free to implement an automatic blacklist solution. >Depending on the country. People have been arrested for running Tor exit nodes. arrested but not convicted. If you absolutely want to run a tor exit node or a zeronet seedbox then you can be trusted to have done all necessary research beforehand. After all you wont be a normal user anymore. > Zeronet is not any better in this regard than the normal internet. In fact, it's worse That is wrong. It is better. See the arguments above and in my previous posts. Also keep in mind that if a server hosting illegal material is raided then the LE have access to all of its connection logs. According to you that should be much worse. >People who want to contribute to the network run the risk of contributing to the distribution of content they may not want to distribute. Not in any significant measure. I am connected to around 50 Zeronet sites, and never have I seen anything clearly objectionable. And of course: My points about the challenges LE would have trying to get to even one node as explained above and in other posts. The fact remains: Zeronet is at least as good as the normal net, and in many aspects much better. >Protection against the level of attack was offered for $150,000-$200,000 a year. That's going to destroy most people financially. You are not getting it! I used this as an example to show just how hard (or impossible) it would be to take down even one single server. Now on zeronet you have hundreds if not thousands of peers acting as servers. Do you realize how utterly impossible it would be to take down even one Zeronet site? The price akamai wanted it utterly irrelevant. Its the fact that the biggest DDoS was powerless against one single server as long as it was in a good infrastructure. And mind you, even akamai would have been able to manage this kind of traffic, they just did not want to do it for free. Now you wanted to, according to you other post, take down every single peer in a Zeronet site using only your own machine. This is why I posted this link: to show just how hilariously impossible that would be. Therefore this threat that you imagined (one peer taking down every other peer) is simply not a threat. >Zeronet doesn't do anything to protect people against DDoS. Zeronet IS the DDoS protection. Do you know how a DDoS protection (or a content delivery network for that matter) works? You assign multiple publicly reachable servers the same address. Someone wanting to connect to that address will connect to the server closest to them. This is exactly the model of Zeronet or Bittorrent. Except that its not about physical closeness. Someone requests a site, he gets a few nodes and downloads from them. If those nodes were not available in the first place (because offline, DDoSed or whatever) he would not have gotten them. >Where did you get the idea that anyone was suggesting using one node? We've been saying DDoS, which, as you pointed out, necessitates having more than one node. So you use your botnet to take down one node. How do you take down the 499 remaining nodes? Also where the hell do you get your botnet in the first place?? As we have seen powerful companies and governments are not interested in taking down all nodes (or have found out that it does not work) and criminals do not have a botnet powerful enough. Yes, IF you had the USS-Enterprise-E you could steal all the gold in Fort Knox... >Anything ran through Tor is safe, sense Tor is actually censorship-resistant and anonymous. Therefore Zeronet, if run through Tor, is safe and anonymous. It is censorship resistant even without Tor. as I have explained at length.

>>6849 they use blockchain addressing

>>6926 >It's more effective to host a malicious tracker why would people use your tracker if it were malicious?

>>6932 >Even a large criminal botnet has proven to be useless when taking down one single node. To be "proven useless" they would need to have tried and failed. What is actually the case is you are not aware of anyone trying. This is very different and not proof of anything. Moreover there is no technical reason why it is impossible. Your whole argument is "I don't know anyone has done it therefor it is impossible" is not sound. >From that alone we can conclude that this attack is either impossible (my opinion) or not effective at all. Gathering IP address of file sharing participants is perfectly possible as has been explained to you. It's the DDoS part that is not happening because it it illegal. >Now consider how much more it will take to kick each one off the net. Don't project your incompetence onto other people. You are not the attacker. Just because it is difficult for you doesn't mean it is difficult for someone who knows what they are doing and has resources. > you cannot use each bot at full capacity or they risk getting disconnected either from their ISP This has already been flagged as bullsoykaf. >Realistically they could afford one or 20 lawsuits if it means preventing such an early leak. It's not about civil lawsuits it's about prosecution. >Aside from the MAFIAA I see no other actor that would have both the money, infrastructure and motivation to even attempt a large-scale shutdown of each node in a BT swarm. Governments silencing political opposition. >Yes, but again: Zeronet and BT are high-churn networks. They block the route to all 500 nodes currently serving a site, after 24h 450 of them are reachable again because they got new IPs. What if they update the blocked IP addresses as they change? Crazy, I know... >How the hell then would anyone be able to block connections to Zeronet or BT nodes? when those nodes may sometimes be a zeronet node and sometimes not. The firewall uses DPI and identifies the protocol and uses heuristics like port number and plaintext keywords to make a decision about blocking *connections*. It is not just blocking random IP addresses. Stop projecting your own stupidity onto other people.

>>6933 >If those bots somehow managed to delete stuff on the seeding node He obviously means knock the initial seeders offline to stop new files and updates spreading through the network. >So no one ever CAN know about those files EXCEPT if they request that particular site too Wrong. https://iknowwhatyoudownload.com/en/peer/ >Assuming the utterly unrealistic case that an overeager, overpaid and underworked LE officer (3x unrealistic assumption) bothers to actively and compulsively connect to every possible Zeronet site out there Not everyone is as stupid as you. First of all you don't need every site, you only need the illegal ones. Second, there is nothing that requires human involvement. This process can be automated for free. >Now that LE has to be lucky too and hope that the node that gave him that file is also within his authority. Law enforcement do talk to each other you know. They don't just let people go because they are in another country. http://deepdot35wvmeyd5.onion/2017/12/30/luxembourg-pedophile-downloaded-3900-illegal-pictures/ >A wholly impossible scenario which you use Actually this is a retarded strawman you created. Arguing in good faith means assuming the other side have a least a micron of intelligence. >Thats unreasonable, yes indeed. And its completely unnecessary. Your first post in this thread literally said just that. If you are wrong then just admit it. >>6842 >Again if you disagree: feel free to implement an automatic blacklist solution. It has already been explained to you why this does not work. >>6855 >That is wrong. It is better. See the arguments above and in my previous posts.· No, you are wrong, it is worse. See my arguments right here: On the normal web, only the server and my ISP knows what files I download. In Zeronet the whole world knows what files I download. That's not "as safe", that is much much worse. On the normal web I download files and that's it. On Zeronet I download files and automatically share them with other people. If that file turns out to be illegal I am in much bigger trouble because distributing illegal content is much worse than simply downloading it. >The fact remains: Zeronet is at least as good as the normal net, and in many aspects much better. You are in flatout denial. Why do you have so much ego invested in this soykafty little network? >Do you realize how utterly impossible it would be to take down even one Zeronet site? Expensive does not mean impossible. You are also forgetting that DDoS was only one option. A centralized authority like a government can trivially censor any Zeronet site within their country by blocking communication to the relatively small number of peers seeding the site. >Now you wanted to, according to you other post, take down every single peer in a Zeronet site using only your own machine. >This is why I posted this link: to show just how hilariously impossible that would be. That was me not the person you were responding to and your link was about a different type of attack to what I explained in my post. You will never learn anything if you assume everyone is as ignorant as you are.

>>6939 You are just not getting it, are you? >To be "proven useless" they would need to have tried and failed. Which is precisely what happened. A large criminal group sicked its botnet(s) on one single server and that attack failed flat. Even though they shot at the target with over 600GB/s of traffic. You have the proof of the failure right here: http://www.zdnet.com/article/google-rescues-krebs-on-security-from-the-abyss/ All your other arguments resting on the possibility of such an attack are thus faulty because such an attack has failed. > Moreover there is no technical reason why it is impossible. Yes it is: astronomical hardware costs. Among many, many others. >Gathering IP address of file sharing participants is perfectly possible as has been explained to you. No, the proposed method has been found to not work reliably, just read the freaking book "distributed systems" already, or study a few academic papers on that subject. >It's the DDoS part that is not happening because it it illegal. It has happened, see the link above. And most importantly: IT FAILED! (you score another bullseye on your foot) > Just because it is difficult for you doesn't mean it is difficult for someone who knows what they are doing and has resources. And here is where your lack of knowledge prevents you from realizing your lack of knowledge. I do not base my arguments on imagined castle-in-the-sky-scenarios but on cold hard facts from academic research. This research, based upon solid mathematical models tells you that what you propose does not work. Not because I do not know about any fancy product or technology (IBM's new auto-DDoSer-9001-pro) but because math and science says its not possible. >This has already been flagged as bullsoykaf. No this is also a real scenario, contrary to yours. Feel free to try it out. It also works if you try and send too many emails from your PC, you will see that you will lose your connection very fast. Maybe even permanently. And thats just the trouble on your end. On the end of the victim node, their ISP might also decide that you or even your whole ISP is malicious and simply refuse to forward any packets from there. This, again, has happened multiple times in the past already. You would know that if you would just read up a little bit about this topic. >It's not about civil lawsuits it's about prosecution. It does not matter what law system you use. Fact is that a huge conglomerate of companies like the MAFIAA could afford to lose many lawsuits before running out of funds. Sony BMG for example has shown that they will even go so far as to infect millions of PCs with a rootkit just to try and stop the spread of copies. For this criminal act they only received a slap on the wrist from the feds. Now a rootkit is much worse than a simple denial attack. Therefore once again my argument stands correct. >Governments silencing political opposition. There are enough governments out there who would be motivated to do that, yet they dont. Proving again that this is not a valid attack. the MAFIAA at least are motivated by money, they can immediately know if an attack is worth it. China will find it much more difficult to justify such an over the top attack. Case in point: they never use it, instead prefering trojan horses to gather info instead of denying the spread of info. >What if they update the blocked IP addresses as they change? Crazy, I know... Indeed, crazy. Also useless. The attacker would have to use the infrastructure intended for the actual DDoS to keep looking for nodes all the time, ignoring once more that this is not realistically possible in the first place. >The firewall uses DPI and identifies No it does not. Thanks to this little magic spell called TLS. >heuristics like port BT can use any port you want. This too fails. >plaintext keywords nope. Not with TLS.

>>6939 >>6940 Also: Your whole sentence rests upon the false idea that there is somehow just one global ISP for everyone on this planet. Because this is clearly not the case at all, even if your proposed methods would work, they would only ever work for those nodes inside the net of the ISP doing the blocking. Everyone outside of it would not even notice that anything is wrong. > It is not just blocking random IP addresses. Technically it is. Its blocking those adresses from the random addresses the site gave you that are coincidentally inside the net of the ISP trying to block it. > Stop projecting your own stupidity onto other people. Yeah, I see clearly just how extremely annoying that is with you. Dont worry, I promise I wont do it. >He obviously means knock the initial seeders offline to stop new files and updates spreading through the network. Oh wow, another agrument that is only possible because you did not bother to do even ONE BIT OF RESEARCH! If you had you would have found out immediately that in Zeronet the original node can grant partial or total edit rights for all or for part of the site to peers. This is what makes distributed messageboards or distributed mail or chatrooms possible. If the owner is offline, the site works just fine. I tried it, it really works. On top of that: you do not know which one of the nodes is the original one, as this is only done over the cryptographic signature. Which is completely separated from the IP address (which you get if you query for nodes) >Wrong. https://iknowwhatyoudownload.com/en/peer/ Shows a few (not all) of the BT files I torrented, some files which I never torrented and NOT A SINGLE ZERONET FILE! Go one, try it out for yourself. > you only need the illegal ones. That is not what your original argument was, it was about a legitimate site accidentally containing an "offending" file. But yeah, sure. Next question: What the hell is an illegal site? And how would LE determine what is and what is not such a site? And how would they know if such sites existed and where they are located? Once more you are building your cardhouses and calling them the empire state building. >there is nothing that requires human involvement. Literally everything requires human involvement in that scenario. You would need a strong AI at least on par with humans to be able to correctly classify a site, and even then it would not be 100% certain. >This process can be automated for free. It cannot be automated, and it certainly cannot be automated for free. Shouldn't you, statistically speaking, have brought up at least one correct argument by now? How do you do that? Why not just mash the keyboard, maybe a correct argument will come out that way. >Law enforcement do talk to each other you know. Not as much as they wish. Help requests even inside the EU can take months to be acknowledged, translated and executed (or denied), You are imagining LE worldwide as one huge hivemind, This is wrong. LEs might even outright refuse to cooperate completely. For example the polish police will probably just throw away any help request from the sauidi arab police on accounts of them being wacky, treacherous muslims. (I exaggerate). > http://deepdot35wvmeyd5.onion/2017/12/30/luxembourg-pedophile-downloaded-3900-illegal-pictures/ Error 404, Server not found. Based on the site title: Yeah 3900 cp pictures are quite significant. This is where the inter-LE cooperation could become realistic (not always however). But this is not at all in the same category as what you imagined: On single accidental file seeded for only a few hours. >Your first post in this thread literally said just that. If you are wrong then just admit it. >>6842 No this was about the possibility, not the need, to delete or even block certain files on you own node. You then went and sperged out about having to do that every milisecond of the day. That is wrong, there is no need to do that. >t has already been explained to you why this does not work. >>6855 This too was based in your own ignorance. Such a solution works quite well if you really want to have such a blacklist. I for once do not want to have one. >In Zeronet the whole world knows what files I download. No. The whole world does NOT know anything at all. You did not even read my post there, I explained that: Only the nodes having or requesting a particular site can possibly ever know about those files. Zeronet does not broadcast its files, it does not use broadcast or multicast connections at all, its still very much point to point. And also encrypted in transit (see TLS above) >only the server and my ISP knows what files I download. Unless the server is raided by the feds, then they have the logs and every IP address that ever connected to that server. (not that IP addresses are particularly dangerous as I explained in my other posts) In Zeronet if they seize a node physically (very unlikely) they get access to a lot less nodes than if they seize a server. Thats because of the topological differences of a p2p net and a "normal" client-server net. >I download files and automatically share them with other people. see above why this is not the case. You are not sharing them with the whole world as you claim. >Why do you have so much ego invested in this soykafty little network? I dont like it if people are wrong or stupid. I especially hate it if people try to infect others with their stupidity. Zeronet could be an important asset in the fight against censorship. Incompetent, snotnosed, FUD-spewing scriptkiddies should not be able to scare away potential users from it. >Expensive does not mean impossible. In that case, yes it does. if cost_of(DDoS) > cost_of(lost profit) then it is impossible. >You are also forgetting that DDoS was only one option Its the only option that looks like it could ever make sense one day. >A centralized authority like a government can trivially censor any Zeronet site within their country by blocking communication to the relatively small number of peers seeding the site. Maybe they could reliably block the nodes inside their country, until those switch to Tor or vpns. But they cannot block it globally. And that is exactly what is required to completely wipe out one Zeronet site: take down every single node. A government is limited to the nodes it has authority over. A botnet is not. But a botnet is not strong enough as we have seen. That leave percisely 0 possible attacks. >trivially censor Yeah, I's like to nag about that "trivially" in particular: Its not. Its not at all trivial to reliably identify all peers within a country because of the high-churn nature of the network. >That was me not the person you were responding to and your link was about a different type of attack to what I explained in my post. And your "different attack" would be what exactly? I attached a book about distributed systems. Its not the tannenbaum one because that one is almost 40mb in size. Maybe it will help you understand why literally every single one of your arguments so far has been wrong.

We get it zeronet is trash. Can we talk about something else? What about ipfs is that good?

>>6932 >there is no such thing as a malicous dht DHT allows you to discover peers. You can exploit this to enumerate the peers, which is malicious for our purposes. In fact, in practice, DHT has been exploited to not only find peers, but trick them into DDoSing a target. The DHT exploiting DDoS part is unrelated to this discussion, aside from being another example of how someone can maliciously use DHT, but it's pretty cool and given your interest in distributed systems and bittorrent, you might find it interesting to read about. Google "DHT DDoS" and you'll get a bunch of papers about it. >Its just an overlay network with operations defined on it. Most likely a distance metric, a get_data() and a put_data() function. Sometimes also a rating function for the trust of nodes (not on the BT DHT as far as I know) For bittorrent, it's basically just each node acting as a tracker. There's some rating built-in, e.g. "good" nodes, "questionable" nodes, and "bad" nodes. See "Routing Table" section of http://www.bittorrent.org/beps/bep_0005.html >Well clearly so far those actors have not yet done anything like that. Those actors not doing it doesn't make it a theoretically impossible task. It's not cost effective for MAFIAA, so they're not going to bother. If the feds did it, I doubt we would know, especially since they'd likely not be doing it for popular torrents but rather obscurer torrents (since very illegal torrents don't tend to get popular). >As I mentioned in my post: Even a large criminal botnet has proven to be useless when taking down one single node. It was not a single node. Krebs was being hosted by Akamai, which provides (great) DDoS protection. >As the largest distributed platform operating at the edge of the Internet, Akamai provides a defensive shield built to protect your websites, mobile infrastructure, and API-driven requests. Via 24/7 monitoring, we collect and analyze terabytes of attack data, billions of bot requests, and hundreds of millions of IP addresses to solidify your defenses and keep you informed. And we never stop innovating to stay ahead of the latest threats including malware, phishing, data exfiltration, DDOS, and other advanced attacks. > >This commitment to security is why we’re trusted by the most security-conscious industries, including 18 of the largest asset managers, 12 of the top insurers, and 8 of the top financial technology companies. So you can worry less about cyberattacks, and more about growing your business. source: https://www.akamai.com/us/en/why-akamai/world-class-digital-experiences.jsp The DDoS stopped, Krebs got the boot because it cost too much money to mitigate, and Google picked him up, using their own DDoS protection service with an undoubtedly expensive and powerful infrastructure behind it. Compare this to the usual participant in a swarm. A home computer, maybe a cheap VPS acting as a seedbox. They don't have DDoS mitigation techniqures or large infrastructures protecting them. >operative word here is currently. BT in the retail context (that is outside of professional tools that may use BT as one component) is a high-churn network. That is: peers join and leave constantly. Peers leave and rejoin with a completely different IP address too. Only if most or all nodes were servers with a permanent connection and thus an unchanging IP would your approach make any sense. Right now every 24h the majority of the IPs have changed because most ISPs reassign a new IP to each of their customers every 24h. That has proven to be a major "problem" for the MAFIAA who assumed that IP address = people and often have sent the cease&desist letters (or your country's equivalent) to the wrong people. >This high-churn property is usually thought of as a big challenge from the network developer side of things (read the "distributed systems" book by tannenbaum thats where its explained in more details) but in this situation it works to the advantage of the swarm and against a theoretical attacker. True, this makes an attack harder, but not theoretically impossible. Alice can simply continually try to find the swarm participants. When new ones join, she updates her list. Additionally, assume, for the sake of argument, that Alice has been able to DDoS all of the current nodes for a swarm. An old node joins in. The node attempts to notify the swarm that it's back, but the rest of the swarm is down. Nobody knows this node is back, nobody new discovers it, and subsequently it is of no use to the network. >They may look possible on paper but they are utterly impossible in reality. I will concede that the attacks are not pragmatic except for either a) large malicious actors or b) small swarms, but I don't agree that they're completely impossible. >>And the ISPs can discover the IP addresses involved with a zeronet site by connecting to the zeronet network in the same way that you can discover the IP addresses involved >That they can does not mean that they will. >The reasons you gave are basically "for teh evulz". You're right. It does not mean that they will, but my concern rests with whether or not they can, for whatever purposes. >>but anyone can run a few bots that deletes things as soon as they're posted and censor content. >If those bots somehow managed to delete stuff on the seeding node, or on other nodes beside themselves then yes that would be censorship. However if you delete or blacklist a file in Zeronet this is only on your OWN node that you do this. It has zero effect on anyone besides you. >Therefore it is censorship resistant. That particular example was not meant to be analogous with Zeronet. I just wanted to show that it's not necessarily true that resisting censorship from a central authority entails resisting all censorship. I have to go now, but I will finish this post later. One point that I would likely to quickly make is that the book you post later defines (paraphrasing from page 494) censorship resistant networks as those that continue to function after having half the nodes removed. If that's the definition we're working with, I will concede that Zeronet meets it.

>>6959 Continuing this post. Apologies for the split post. Some irl issues came up suddenly. >Well you are not publicly advertising at all. Its all limited to the particular site in Zeronet. So no one ever CAN know about those files EXCEPT if they request that particular site too. To reclarify: in this scenario, we're seeding the site. So, like you said, if they request that particular site as well, there is a non-zero chance that our IP address will show up when they request peers for the files. To me, that's publicly advertising that we have, or at least had, the file. >And even then there is no way to influence from which node you actually get the files. From what I've seen, I believe there is. You get a list of peers which claim to have the file. You can then request the file from one particular peer. Looking through their source code, I see nothing to the contrary of that, and in fact, you could probably just use their peer.getFile() method to automate for yourself. https://github.com/HelloZeroNet/ZeroNet/blob/master/src/Peer/Peer.py Regarding the argument about the unrealisticness of LE coming after us for it, there have been instances of people arrested for seeding files. Consider this: an automated tool that uses a hashlist of "bad files" (s/bad files/cp if you want. I'm still not convinced that cp is the only thing an arbitrary person should take into consideration, but I think it's really besides the point anyway) and a Zeronet crawler to find peers seeding "bad files". The tool then uses an IP geolocation tool to determine if the peer is within jurisdiction, then contacts the appropriate authorities. Fairly trivial to make, right? I don't disagree with your belief that it's unlikely that a specific individual will be arrested for it, but I don't think the chances are as one in a trillion as you do. After all, people have been arrested for seeding things before. Also, there is more to worry about than just LE, but it's probably not worth it to get into it. >Zeronet IS the DDoS protection. Do you know how a DDoS protection (or a content delivery network for that matter) works? You assign multiple publicly reachable servers the same address. Someone wanting to connect to that address will connect to the server closest to them. This is exactly the model of Zeronet or Bittorrent. Except that its not about physical closeness. Someone requests a site, he gets a few nodes and downloads from them. If those nodes were not available in the first place (because offline, DDoSed or whatever) he would not have gotten them. DDoS protection is a little more complicated than that. Special rules, special routing, etc. >So you use your botnet to take down one node. >How do you take down the 499 remaining nodes? Most of the nodes aren't going to be implementing their own special DDoS protection, and, if the network is anything like typical bittorrent networks, most of the nodes aren't that powerful. You got a swarm of home PCs. Do you think 600Gb/s isn't gonna knock out a few hundred typical PCs? Maybe you have the one node the site is actually hosted on have some sort of high-cost DDoS protection, you're good, but in that case you're already safe from censorship since the DDoS protection keeps you online and the encryption keeps you safe from mitm attacks. >Also where the hell do you get your botnet in the first place?? Such botnets exist - hence the 600Gb/s DDoS you linked earlier. >>6935 Malicious tracker meaning tracker that records the IPs of the swarm participants that connect to it. They don't realize you're malicious because they don't see your logs, or know what you're doing with their IPs. That's malicious for our discussion (although there's some cool maliciousness possible with trackers - advertising false statistics, exploiting it for DDoS, etc. Google for it, you'll find some cool papers) >>6940 I know that we're on the same side in this argument and that the other lain has used several personal attacks, but please try to argue in good faith as well. We'll all get to a better ending if we're polite about our qualms. >>6943 >>heuristics like port >BT can use any port you want. This too fails. Zeronet has peers advertise the port they want other peers to connect to, so you can find that port and block connections specifically to it. >>Wrong. https://iknowwhatyoudownload.com/en/peer/ >Shows a few (not all) of the BT files I torrented, some files which I never torrented and NOT A SINGLE ZERONET FILE! The site lain linked finds torrents by scraping torrent sites and dht for those swarms. Of course it's not going to have zeronet files. >>http://deepdot35wvmeyd5.onion/2017/12/30/luxembourg-pedophile-downloaded-3900-illegal-pictures/ >Error 404, Server not found. Server is up. It's a hidden service. You need to access it via Tor. As I mentioned in my last post, the book you post defines (paraphrasing from page 494) censorship resistant networks as those that continue to function after having half the nodes removed. If we're going to accept that definition, I'll concede that Zeronet meets it. I do, however, take issue with this definition. To me, If a hypothetical network is vulnerable to some attack that makes all of its nodes useless, it should go without saying that it's censor-able. Likewise, if a network is vulnerable to some attack that makes it nearly impossible to access the content, it's also censor-able. >>6949 >We get it zeronet is trash. Can we talk about something else? Yes please.

>>6972 >please try to argue in good faith The tone is aggressive but I think "arguing in bad faith" implies there is something wrong with the actual content. If I use nice words with the same meaning is it still "bad faith"? https://www.deepdotweb.com/2017/12/30/luxembourg-pedophile-downloaded-3900-illegal-pictures/ >German police reached out and informed Luxembourg authorities that an IP address in Lamadeleine had accessed and downloaded a video containing child pornography. The point is it was a single download and even though German LE didn't have jurisdiction, they didn't just let him get away. >censorship resistant networks as those that continue to function after having half the nodes removed All P2P networks are "censorship resistant" by this definition. Basically there is no central server that can be shut down. >>6949 >What about ipfs is that good? ipfs has actual documentation, not just a slide deck and an autogenerated api reference. And doesn't use meaningless marketing buzzwords like "uncensorable" and "bitcoin cryptography". https://github.com/ipfs/ipfs So far so good. I'm going to start with the whitepaper. https://github.com/ipfs/ipfs/blob/master/papers/ipfs-cap2pfs/ipfs-p2p-file-system.pdf?raw=true

>>6821 Isn't that kinda what IPFS is? Or something it can achieve?

>>6959 >You can exploit this to enumerate the peers it's *very* rare that you can actually enumerate all peers, except for torrents with only a few seeders. With 1000 seeders you will never get a complete peer list. >In fact, in practice, DHT has been exploited to not only find peers, but trick them into DDoSing a target. Sounds far-fetched, and also sounds like a protocol problem rather than a technology problem. Since DHT is not a protocol but rather an idea. >Google "DHT DDoS" and you'll get a bunch of papers about it. pic related. >If the feds did it, I doubt we would know, especially since they'd likely not be doing it for popular torrents but rather obscurer torrents (since very illegal torrents don't tend to get popular). seems like it would be more effective to simply stop the torrent from getting popular. Almost all torrenting activity is through clearnet torrent-downloading websites, and we've seen in the past that nobody has a problem with taking them down at least temporarily. >For bittorrent, it's basically just each node acting as a tracker. There's some rating built-in, e.g. "good" nodes, "questionable" nodes, and "bad" nodes. See "Routing Table" section of http://www.bittorrent.org/beps/bep_0005.html IPFS has a way more advanced node-rating system by default. >>6972 >Consider this: an automated tool that uses a hashlist of "bad files" mild distortions on the file? That's already done for visual and audio media. >>6996 yeah I really don't understand why use Zeronet over IPFS, it's just got soykafty unnecessary buttcoin integration.

>>7006 >>Google "DHT DDoS" and you'll get a bunch of papers about it. >pic related. Use google results instead of duckduckgo. pic related >seems like it would be more effective to simply stop the torrent from getting popular. Almost all torrenting activity is through clearnet torrent-downloading websites, and we've seen in the past that nobody has a problem with taking them down at least temporarily. Agreed. Makes the argument that "if it were possible it would be happening" weaker imo since when there's two possible options and one is better, people go for the better one. >>Consider this: an automated tool that uses a hashlist of "bad files" >mild distortions on the file? That's already done for visual and audio media. Doesn't help much in our situation as a seeder because we're not producing the file, but just passing it along. Since Zeronet sites' content can't be edited by an arbitrary peer before being pushed to other peers as the contents are signed by the site owner's key, we couldn't flip a bit in the file without it being detected by the peers leeching from us.

>>7010 ah, so you're talking about just getting a seeder list for a particular site. I mean, if you *really* wanted to share illicit files zeronet or i2p or bittorrent is not the way to go, they're designed for speed and availability more than anonymity and anti-censorship. Freenet is best. Also >google ew

>>7011 >I mean, if you *really* wanted to share illicit files zeronet or i2p or bittorrent is not the way to go, they're designed for speed and availability more than anonymity and anti-censorship. How is i2p not anonymous?

>>7011 >Freenet is best. Freenet is censorship resistant because nodes route requests for files to other nodes and then cache the files that are returned. So files end up being replicated all across the network in proportion to how popular they are. It is not possible for an attacker to disrupt the distribution of a specific file or to gain knowledge of which nodes have a specific file and attack them in some way. Once a file has been injected into the network there is no way to delete it. Old files will eventually be dropped from the cache if no one requests them for a while, but this is not someone an attacker can control. Freenet is not anonymous though (the project does not make that claim either). There is evidence that law enforcement have at least 1 attack for deanonymizing Freenet users. http://www.hacker10.com/internet-anonymity/police-plants-own-computers-in-freenet-makes-arrest/

That's IPFS.

https://beakerbrowser.com

>>7469 How is this any different from ipfs?

Doesn't IPFS transmit things in clear still?

>>7471 You think it's dumb that people reimplement the same ideas in parallel instead of working together? Welcome to tech.

>>7471 it doesnt used a blockchain. Its peer to peer websites with encryption used by default to verify info.

>>7502 So is it more free as in freedom than ipfs?

>>7722 ipfs doesn't use a blockchain either I have no idea what >>7502 was going on about. It seems to me like beakerbrowser is more like freenet than ipfs