############### BEGINNING OF GUIDE ##########################
In this thread I document and revise how to set up a tor hidden service email server, you may substitute the servers that you are most comfortable with.
STEP 0) Collect the relevent files from The Endware Hidden Service
Set up a tor mail server using postfix or OpenSMTPd, with dovecot for imap or pop.
STEP 3) Make ssl self signed certificates for postfix and dovecot And place these in the appropriate directory
This might require entropy so you might need to run haveged first
STEP 6) Select a strong password for a new user account
# passgen --bytes 33
Write this down in a passbook and add a few random numbers and letters from your mind in here as well.
Alternatively store your keys in a gpg encrypted file on an airgapped computer with a memorizable password to open the file.
STEP 7) Create a new user with your anonymous handle /name
and edit /etc/postfix/main.cf and /etc/dovecot/dovecot.conf to reflect the new hidden service name.
edit /etc/postfix/recipient_access and change the recipient name to anon142boo
# postmap recipient_access
# systemctl reload postfix
# systemctl restart postfix
There might be errors or omissions in the above but I think that's the general process. Now you have a tor hidden service mail server, that uses a selfsigned certificate and you have a gpg public key. you use tor and ssl and gpg to secure your email communications, and you are known by your handle and hidden service .onion address.
################# END OF GUIDE ####################################
To add an extra layer of security we will be using A:\ drive floppy disks to shuttle the encrypted.asc messages to and from a dedicated airgapped encryption station preferably running openBSD on a non-intel architechture, SPARC, PowerPC, Alpha, etc. Do not use USB as a substitute for this step (STUXNET).
0. Your decryption station will have full disk encryption and be powered off when not in use
1. Generate your keys on the air gap
2. Export your public key,change the file permisions to read only and save it onto a floppy disk A:\
3. Take the floppy disk and sneaker net it to your transmision computer which has the hidden service and postfix on it.
4. Publish your anonymous user name and public key as well as the hidden service onion name using tor and icecat, links, or endcurl or however on your tor hidden service website or on a message board forum like endchan.xyz.
5. Recieve the hidden service onion name and public key of your correspondent ( by reading a published name,address and public key on a forum or other communication method or by recieving it in your inbox by postfix after publishing yours)
6. Save the public key of your correspondent onto a floppy disk A:\ change permisions to read only write a sha256sum checksum for the file and shuttle it to the decryption/encryption station.
7. check the file againts the checksum, and then gpg import the public key to your key ring
8. Type a message for your recipient in plain text on the air gapped encryption station and encrypt it to encrypted.asc. Delete the plain text file if unnecessary to archive especially if it is incriminating.
9. Write the encrypted message encrypted.asc to the floppy disk and change permisions to read only also write the sha256 sum of the file to the floppy if you have space.
10. Shuttle the message by floppy disk A:\ to the transmission computer and send to your recipient using endmail
11. You can also write the sha256 sum of the file or sha512sum of the file onto the disk before sneaker netting it.
12. You can include this sha256 checksum as another attachment to transmit to the recipient or as a second followup email
13. Recieve an encrypted message response from postfix, change its file permissions to read only ( chmod ugo-xw encrypted.asc ) and copy it onto a floppy disk A:\
14. take the sha256sum of the file and also write it onto the floppy disk with ( sha256sum encrypted.asc >> sha256sum.txt ; chmod ugo-xw sha256sum.txt ; cp sha256sum.txt /dev/fd0 or however that is mounted like /mnt/floppy)
13. Shuttle this by sneaker net to the decryption station check the sha256 sum against the file value and decrypt wtih gpg, delete the plaintext response, and either write a response and repeat or reboot the computer.
14. Power off the decryption station airgap after communications have ended to protect your station.
This step protects against keylogging or other malware revealling your message during compositon before encryption.
Maximum Security Encrypted Message Communication Protocol (MSEMCP)
The great thing about this system is that it is opensource, distributed, anonymous, transport layer encrypted, and the message is RSA 4096 encrypted by gpg. Currently you do not have to have a working server to submit mail to a server. So if you were part of a 10 man spy ring with a central node the field agents only need to learn how to use gpg to encrypt and to install swaks, tor, torsocks and use a variant of endmail.sh to report back to central command. Of course it can be used for server to server anonymous communications as well. Since it uses gpg several recipients can be addressed to one anon@hiddenserver.onion and then central command can distribute the messages to their intended recipients after the first decryption. So you could encrypt with a general wrapper that contains the meta data (to: field ) that central command uses to distribute the message to the person it is intended for, and then put the from: and subject: in the targeted encryption for that person that central command/dispatch can't read. So you'd have two gpg keys one public key for central command and then a public key for the intended recipient. The field agent would perform 2 encryptions the first for the recipient with the subject: and from: field in the plain text, and then the second encryption with the to: field in the plain text of of the encrypted file.
Maximum Security Encrypted Message Communication Protocol (MSEMCP)
Author: Endwall from the Endware Development Team
Creation Date: August 8, 2017
Copyright: The Endware Development Team (c) 2017
License: You are Free to Study, Reproduce, Copy, Modify, Implement, Test, and Use this protocol as described below, in the spirit of the Endware End User License Version 1.18.
Description:
This protocol MSEMCP provides:
1. Message security through physical isolation of the encryption station by a read only Floppy Disk Message Sneaker Net (FDMSN)
2. Strong Public Key Cryptography using RSA 4096 bit.
3. Transmission and reception anonymity through Tor with 12 hops.
4. Trust and Verification through TLS 1.2 with RSA 4096 bit, and Self Signed Server Certificates to compare repeated server connections.
All of which can be implimented with 4 tools:
1) A Base install of a *nix BSD or GNU operating system in text mode.
2) GNU privacy guard (gpg) on the airgapped encryption decryption system.
3) TLS 1.2 using RSA 4096 certificates from OpenSSL used in a Postfix Mail server to allow for certificate comparison to build trust and continuity.
4) Anonymity durring message transmission and recption using the Tor network with 12 hops,(and weak encryption SHA1 RSA 1024).
Minimal instalation:
Computer 1 (Transmision computer)
1. Base install of *Nix in TEXT MODE no GUI / or use a GUI (whatever)
2. Postfix (or OpenSMTPD ), Tor, Torsocks, Swaks, OpenSSL ( or LibreSSL or GnuTLS)
3. endget.sh, endmail.sh, endfix.cf (for postfix)
4. A 3.5" Floppy drive with 1.44MB HD IBM format Floppy Disks
Computer 2 (Decryption/Encryption Station)
1. Base install of *Nix + Full disk encryption in TEXT MODE no GUI
2. Gnu Privacy Guard gpg
3. Air Gap : Unplug the computer from internet permenantly post installation
4. A 3.5" floppy drive with 1.44MB HD IBM format Floppy Disks
Encrypt and Decrypt messages on Computer 2 with gpg and Send messages with endmail.sh and Receive messages with Postfix with endfix.cf on Computer 1
Move the keys and messages by read only 3.5" floppy disk files. Do not substitute USB for floppy disk read only files (STUXNET).
I have no idea why no body recommends this as the go to solution. The messages pass through Tor. The bad guys should never be able to see any of it. it also uses Self Signed TLS certificates, Which allows you to compare if you're talking to the same server every time, and adds a layer of protection. Finally the message is encrypted using PGP RSA 4096 bit or stronger.
If all of the message encryption is done on a battery backup powered airgap on OpenBSD on a Sun Sparc or DEC Alpha, and decrypted on an airgap by your recipient, there should be no way that your messages are read by anyone other than the intended party.
And for regular business email, I'm sure that a competent person could use the endfix.cf and modify it for use on a clearnet mail server, on a different computer.
There really shouldn't be much more use for services such as protonmail, tutanota, cock.li for private communications anymore. This should fill most of that need. The only need for that kind of free public service is to communicate with the public through regular email but with psuedo anonymous properties.
Peer to peer, everyone who is serious should do it this way. It takes 30mins to set up with all of the files and instructions I've provided.
Mods go ahead and delete the wall of text in the first post but keep the picture. I tried copying from links2 into the textbox but it seems to have not copied the /n newline and carriage return. Posting from tor browser worked better. Sorry about that. I hope the information is useful.Thanks.
>>2976 Just because I used Freddie Mercury in one of the /os/ banner's doesn't mean I like that kind of stuff. I guess I did use the word Homosexual in the Endware License...no homo.
>>2975 But yeah this board has real promise/potential. I can post on here without captcha, or javascript using links browser. Doesn't get much better than that... Kudos. Good growth potential.
>>2977 You don't get it do you? You are a namefag. Not only is it not appreciated, there's no way you can prove your identity to anyone (except for using tripcodes but that just makes you an even biggr faggot).
Your identity is not relevant and announcing it merely indicates that you are a nu-phag.
>>2977 >>2978 Being a hidden service board, anonymity is held in high regard. Some users have even frowned upon the admin for using a capcode. Please lurk more before posting again and drop the cringy name.
I'll concur with the criticism about namefagging, but I'll still thank OP for taking the time to post this information. You've inspired me to host my own secure email rather than continuing to fuck around with assorted services.
Also you should do the 6 hop mod
and compile and link tor from source
and I recommend using endtorrc or a modified version of it for your torrc-defaults file.
>>2970 Also after writing the encrypted message onto the floppy disk, eject it and then flip the write protect tab on the floppy. So that you only read encrypted.asc and sha256sum.txt from the floppy on your transmission work station when sending the file. In general you can improve this protocol by having 2 airgapped computers, that share your private key. One for decrypting messages, the other for encrypting messages. That way information never flows back and forth between your airgap statioons and the transmission computer.
READ ONLY
[TRANSMISSION] <-------------- [ENCRYPT STATION]
READ ONLY
[TRANSMISSION] --------------> [DECRYPT STATION]
The transmission station has the tor mail server (postfix) with tor and endmail. The Encrypt and Decrypt station are separate computers that share your private key for pgp. Both are airgapped / full disk encrypted running OpenBSD, and floating on a battery backup (off the power grid) bassically unplug your APC from the wall for 15-20 mins while you do this operation, shutdown the computer then plug the APC back in to recharge the batteries.
If malware flows in to the decrypt station, it can't return to the decrypt station. Nothing ever flows in to the Encrypt station only out.
If both message counter-parties did this protocol, it would be pretty secure. If your messages were leaked at that point and you discovered it by some means, you could safely assume that the person your are communicating with is a rat,snitch,traitor, etc., who forwarded the decrypts along to your adversaries / enemies.
Its not unbeatable, for instance someone could install a camera in your room...and point it at your decrypt / encrypt station monitors,or keyboard, or switched your keyboard for your encrypt station with a keyboard with RF signalling, But if they've done that then you probably have more serious issues at hand.
>>2991 they would say that, though. they state they don't like people doing it because it puts more strain on the network. however, if you read to the end it's clear they're not even sure if it has a negative impact on anonymity. they make two assumptions:
1. anyone making more than the standard 3 hops is noticeable
2. an attacker controlling the final node(s)/relay(s) can statistically infer who you are anyways regardless of how many hops you configured
i think these are founded assumptions, but they're not certainties and tor devs are kikes.
>>2993 APC UPS
I have an APC UPS i meant UPS. Float on your UPS battery backup while decrypting. Or have the encrypt and decrypt stations as two laptops instead of desktops.
>>2993 If malware flows in to the decrypt station, it can't return to the encrypt station. Nothing ever flows in to the Encrypt station only out. Nothing flows out of the decrypt station only in.
>>2987 No need to listen to those fags. They are just gate-keeping nigger faggots. You have some interesting things to say, so I don't give a fuck if you use a nickname or not. Learn to ignore haters.
>>2998 I guess you have to get the public keys for your recipients onto the encryption stations, you can make an exception for that, or do it durring installation if you have consistant people that you communicate with. Or just risk brining in a new public key on floppy disk. It's a very small attack surface. 1.44MB is enough to have a book's worth of text. should be enough for passing messages. The point is to assume that the transmission computer can or will get compromised by a keylogger, or by other malware, as it is internet connected, and to shield the messages by encypting on the air gap. You might need to jump new public keys onto the encryption station but aside from that nothing comes in, only messages go out on it. With the decryption station nothing comes out, only encrypted messages go in.
the king james bible is 4.3MB uncompressed 1.3MB compressed with gzip. So 2 floppy disks should be sufficient to shuttle messages back and forth for a couple of years.
>>2985 >>2986 The "SCPL page" was made of AIDS and fail when it was developed by LARPing retards on endchan, and it's made of AIDS and fail now. What a joke. Just shameful.
Also pad your messages with random noise. You can pad the begining and end with random ASCII from /dev/urandom, or you can pad each paragraph with a paragraph of noise. passgen.sh should help to accomplish this.
>>3003 Nobody has contacted me with an update of the webpage. I wasn't involved in it's creation or the content but I offered to host it, mainly because the discussion regarding finding hosting for it was painful. I offered to host it to demonstrate how simple it would be to do so. I'll leave it going for now. It doesn't get much traffic and has no marginal cost.
>>3005 It was originally pitched as a list similar to privacytools.io, with better suggestions and tips. It didn't really achieve that goal. But if there are updates, I'll post the new updated html. It sounded like a good idea, but the originators / creators that were discussing it apperantly abandoned the project or couldn't find further input from other interested contributors. Sounded like a good idea, but it hasn't been updated since 2017, or revised or improved. If it turns into a how to guide or tutorial on each of those topics, (similar to this thread), then it would be worthwhile. Right now its just a random list of recommedations with some clearnet links.
A better approach would be define the problem and offer a solution : ( this is a problem, click link to learn more about this problem, with a well written page about the topic -> this is a potential solution to that problem with a link to a new page with detailed instructions on implimenting the solution step by step instructions ).
>>3006 I just read/skimmed through the page again. It has some interesting links and recommendations. It would be nice if it would specify why those reccomendations are superior, and why their near substitutes are inferior naming them and pointing out some of their defects. It's a list of links, some of them are interesting.
>>3006 don't pay any attention to that nigger. the SCPL page is good and has plenty of very useful advice. maybe some of it is outdated but i didn't see anything that sticks out as larping. it's at least a good entry to an enormous and complex topic.
>>3008 >the SCPL page is good
It is absolutely not.
>and has plenty of very useful advice.
It has a bunch of random links and some very dangerous advice. The torrc configuration given is absurd, and anyone dumb enough to use it is doing the digital equivalent of wearing a bright red sign, ringing a bell, and screaming "LOOK AT ME! LOOK AT ME! I'M HIDING! LOOK AT ME!"
>maybe some of it is outdated but i didn't see anything that sticks out as larping.
lol, that says more about you than it does about the list. I'm not surprised. Ever since that shitshow debuted, I've said that nobody who knows anything would take it seriously, and anyone who takes it seriously is too dumb to be made to understand what's wrong with it.
>it's at least a good entry to an enormous and complex topic.
And what is that topic? It purports to be about "Secure Computing Practices Links," but it never defines what "secure" means. It's clear from the list, however, that it's the neophyte's understanding of "security" which is really a kind of vague, mystical confluence of security, anonymity, and privacy that utterly ignores the distinction between these things. This is unsurprising, given that the list was composed almost solely by an autistic ESL LARPer on endchan who not only routinely handed out ridiculous advice, but who repeatedly made the same erroneous claims about basic cryptography concepts. When called out on it, his brilliant response was that he never claimed to be an expert. Never mind that he nonetheless felt qualified to routinely make recommendations and to be largely responsible for a "secure computing practices" document.
Nobody who knows what he's doing needs that page, and anyone who doesn't know what he's doing will derive absolutely no benefit from an almost entirely context-free list of links to a hodgepodge of hardware, software, dead projects, and random shit that caught the fancy of some autist from endchan. Here's what should have been the biggest clue that the whole thing is a LARP: the concept of threat modeling appears nowhere in the document. It's garbage. It doesn't need to be updated. It doesn't need to be edited. It needs to be fucking deleted.
>>3010 i think you are being overly critical of the guy, though you are right it is not really a great intro and i wouldn't blindly take the advice of some autist, especially with the torrc shit. as for endware and the SCPL page i never used any of it, but i did save them figuring they might come in handy for project ideas or a topical reference some day.
>lol, that says more about you than it does about the list. I'm not surprised.
:(
>The torrc configuration given is absurd, and anyone dumb enough to use it is doing the digital equivalent of wearing a bright red sign, ringing a bell, and screaming "LOOK AT ME! LOOK AT ME! I'M HIDING! LOOK AT ME!"
Please Elaborate. What's wrong with excluding countries from your node path selections? It's statistically likely to choose such a path. And if thoese nodes are also freely connecting other traffic to other areas, why would it single you out?
I dont know who wrote that but why are so many nodes excluded? the amount of nodes on the Tor network is already low. You're severely restricting the pool of possible nodes without proof that it will improve the ratio bad actors/anonymous nodes.
It is currently set for 9eyes country avoidance, + middle east Tor node avoidance. You can modify it to your heart's content.
I currently enter through France or Germany and exit from anywhere in Europe or related destinations. It seems to be working for me. I also use the 6 hop mod. Works for me...
Modify it to less restrictive (5eyes exclussion) or by pass. Make multiple coppies with different settings and call them like this
$ tor_stable -f ~/bin/endtorrc.1 &
etc 1,2,3,4, .us, .bypass, .ru
to for entry and exit through specific locations. Its like changing your identity but you have more control. The idea is to avoid CIA, FBI and NSA nodes, so that they can't do cross correlation on timing and packet rate , which can deanonymize you with 70% accuracy if they have both of the entry and exit nodes. I would be dumbfounded if they weren't routinely doing this, as it is a standard signal processing tool used to detect signal in noise, and finding periodicity in noisey data (radio signals, radar, sonar etc...) .
>>3502 >>3331 >Foolish
Not foolish at all
The security of TOR depends on different not cooperating nodes.
Uncovering of TOR user is via correlation attacks of two or more nodes.
There are some very high bandwidth nodes in EU (recently in crypto hostile France) and USA. Those nodes bend the routing of TOR clients like black hole do to light.
How high is the probability this nodes are cooperating? Eventually all running on the same snooped on hardware?
>Picking your entry and exit in different countries is not a good defence, because it only defends against adversaries that are unable to rent servers in other countries.
Obvious false advice.
You ideally want your nodes in non cooperating countries. Say US, Russia, China, India
Security and insecurity of TOR depend on routing.
If your TOR client uses always the same 5 eyes, EU nodes. If TOR project gives false advice. Makes one think.
Will be more difficult to get both the entry and exit to be correlated or owned by NSA. The individual countries intelligence services are probably doing cross correlation or power spectrum techniques on 3 hop tor. My view point is that Using Tor Browser entering and exiting from USA is a waste of time. You are the fool. Good try though.
The correlation in the USA could be done from their backbone listening posts. I just want to avoid that all together. I'm not buying this appeal to authority post. I don't trust the Tor Project, their technology seems feasible but some of their design choices and some of the things that they do are questionable. I don't trust them or their public recomendation given about "3 hops is the best, so don't change it"....Yeah right.
>>3507>>3502 You're wrong. And you smell like CIA. Save the ad hominum attacks
( You a re a fool and here is a picture of a joker that will convince the readers to question your design choices he he he) Do you think people are that stupid. (...They are admitedly)
You just don't like that you were beat by a kid living in his mom's bassement, just by using my mind, versus you and your big bad millitary budget, that you use to terorize and harass American civilians. And to download celebrity nudes from their iphones....Hey check out J-law's snatch, she took this extreme selfy this morning, he he he. Better get a FISA warrant for this batch, I'm gonna need to collect it all!
>>3502 " Please write a research paper that tells us what to do."
I just did and I posted it on the internet. Here let me tell you what to do Mr. Tor Project. Change the number 3 to the number 6 and enforce it for everyone. Done. Now everyone is doing 6 hops and your argument is now irrelevant. The argument that the attack will somehow be able to guess your path length, which also makes no fucking sense. The nodes shouldn't know anything but where the last packet came from and where the next one is going. Onion routing remember!!
My argument:
The more mixing, the more variable the time delays between servers, the more variety in distance between hops that changes every circuit, the more dificult it becomes to do successful cross correlation, even if both entry and exit ends are owned. The longer the path length , the more statistical variance, the the less clear the path becomes. Makes perfect sense.
Your argument:
Tor Project says so. You're dumb. Heres a picture to go with that. He he he.
I simply disagree with them, and I disagree with you. And anyone who thinks about it for a bit, and isn't on drugs or retarded, will probably see it my way. And this is all explained with statistics.
1000 C 3 vs 1000 C 6 which is a bigger number for the denominator for the sample space? Use your calculator and tell us the number.
My torrc file endtorc switches circuits every 30 seconds, so the delay times are variable and changing every 30 seconds. This would really fuck up their cross correlation analysis tools. All of the benefits are obvious. Large sample space (intractable), variable time delays that are non stationary, (time variant variances and expected values), and entry and exit from separate countries far far away from USA.
But won't other goverments just do that too??? Probably, but unlike the US governement they aren't harrassing me, which is why I had to build all of this stuff to begin with. So even if Germany is able to figure out that I'm the guy who exited from Turkey, they aren't bugging me about it. Unlike YOU!
I'm not going to loose my job, be blackballed and be stuck living in poverty because I searched someone's name in SEARX, if Germany finds out it was me who did the search. On the other hand in America, that is exactly what's fucking going on.
Oh look he searched for cancer in Wikipedia, he must have cancer, deny him health insurance, beh haw.
"You should change path selection to avoid entering and exiting from the same country. It is better to not manually change the path.Why?This could have unforeseen consquences and you'll probably screw it uplike what? , we don't understand it very well either.
There are many attacks and adversaries that Tor is trying to defend against at once, and constraining paths has surprising trickle-down effects on the other attacks (e.g. if I see where you exit then I know where you *didn't* enter, thus reducing your entropy, sometimes by a surprising amount depending on what path constraints are choosen). In general, changing Tor's path selection makes your client look different from other clients. Picking your entry and exit in different countries is not a good defence, because it only defends against adversaries that are unable to rent servers in other countries.
Me:
This is bullshit. And if you don't understand it very well either then why am I listening to you at all? I thought that you were the fucking expert....remember appeal to authority....I don't understand it either duhhhhhhh. What?!!
I just don't want any nodes in 9 eyes. But then I might change that up some times, with different torrc files, you never know....gotta keep em guessing ( big sample space and all that).
"Well even if you avoid 14 eyes countries they can still run nodes in every country on earth by renting server space so why bother. Just run your nodes in USA all of them and only use 3 hops. Its better Tor Project says so."
Analysis:
Do those countries have their own intelligence services, and do they also run Tor nodes? Yes ok, so what fraction of Tor nodes in say Romania are run by the US and then by independent Romanians, and then by Romanian Governement Authorities.
My guess is (US,Foriegn GOV, Foriegn Independent) = ( .10, .80 , 0.10 )
These numbers are just made up go grab some real stats to bolster your argument( thought experiment) so your chances of avoiding USA nodes in Romania are 90%. Now extrapolate this to the rest of the countries. Even if the breakdown is (0.3,0.6,0.1) your still better off in a foriegn country, than by going into the hornets nest. This argument ( They can rent so don't bother) is weak and lame and discounts the statistical approach. Which foriegn country ( with an active intelligence service and modern economy ) is going to have a breakdown that looks like (0.8,.15,.05)?
Probably not many countries at all.
Your chances are better to avoid USA nodes by going abroad. Makes sense if you think about it.
Next suppose that the probabilities of each different country are independent and have a simillar distribution (USA, Local GOV, Local Independent) = (0.1 , 0.8, 0.1)
Now by going through 6 hops the probability of all of those being USA nodes becomes (0.1)^6= 0.000001
Or 0.001%. Now calculate the probability of the two important nodes (entry and exit) being USA only. P(ends being USA)(ignore the middle nodes) = (0.1)^2 =0.01 1% probability of getting both ends owned by USA if the selection is RANDOM. Is the selection Random?
So if randomly selected path with distribution as given then you should have a 99% chance of avoiding having both ends being owned by USA by having a foriegn only path traversal. Regardless of the path length.
The real problem isn't corrupt nodes. The real problem is the listening posts at AT&T and Verizon and Time Warner etc and on the fiber optic backbone, and international cable lines. If you avoid that by not starting and ending in USA then you avoid most of the problems to begin with. Those are the points that watch the entire network and allow for the passive surviellance and real time cross correlation and deanonymization. Just go around it. First hop is out of the country.
To get that kind of telecom wire tapping they need legal cooperation from the companies involved. Its not likely that Germany will have the legal authority to Compel Romanian telecoms to allow them to install a black box tap. But it is reasonable to expect that Romanian Governement Authorities can.
USA can't have ultra vision on Germany, Germany can't have Ultra vision on Romania, Romania Can't have ultra vision on Russia. But Russia can have ultra vision on Russia. So just bounce around between them, and if you don't live in any of those locations and don't do crime through the connections, no body is going to harrass you. Because they can't.
Well well NSA will just Hack The telecom company and gain control and... and so don't even try it, we're the best you can't stop us.
Wow, I really touched a nerve there. I made my post and went off to enjoy my evening drinking brewskis and playing ping-pong with my platoon in the Tor Project Rec Room located here in the military wing of CIA headquarters--you accused me in various of your schizophrenic ramblings of being part of the Tor Project, part of the CIA, and part of the military, so I'm trying to accommodate your fantasy here--and here you were, posting the first insane shit that came to mind, then pacing in an autistic rage, biting your fingernails, your anxiety rising, until you thought of something else to say, upon which moment you rushed back to the computer to ejaculate yet another bizarrely formatted and incoherent rambling onto this thread. And you did this several times, apparently. Top kek, as they say, according to our imageboard analysis division.
>>3509 >It's also more likely that since they don't control the infrastructure in Hungary
I really don't know who the "they" you're talking about is--that's as poorly defined as your perception of fantasy and reality--but if you mean the U.S., you might be interested to know that the U.S. and Hungary just signed a defence cooperation agreement this month.
>>3510 >I don't trust the Tor Project
>I don't trust them
Yet you use their code, which I guarantee you cannot read or understand any significant portion of. I'll attach a picture I took of you through your webcam from my workstation here at NSA headquarters upon you realizing how retarded what you just wrote was.
>You just don't like that you were beat by a kid living in his mom's bassement, just by using my mind, versus you and your big bad millitary budget, that you use to terorize and harass American civilians.
It's true. You won. Despite all of the resources of the military-industrial complex that we enjoy here at Tor Project/CIA/military headquarters, you beat me by making your Tor traffic stick out like a sore thumb. We're sending a CIA drone to your mom's "bassement" right now to kill you before you get out any more information about this. Drat! No good! I can't locate you because you're using SIX HOPS! On behalf of the United States Government, I, Lieutenant I.P. Freely, USCOM/TORPROJ, CIA Division, hereby surrender.
>And to download celebrity nudes from their iphones
Perquisite.
>>3512 >Here let me tell you what to do Mr. Tor Project. Change the number 3 to the number 6 and enforce it for everyone.
Good idea. I'll petition Lieutenant Colonel Dingledine to implement this in the code for the next release.
>>3518 >These numbers are just made up
Everything you're saying is made up.
Also, Romania is a NATO member. But I'm sure they don't share anything with the U.S. Because you believe it, it must be true.
>I really don't know who the "they" you're talking about is
Yes I meant the US, the CIA, NSA, as the people I mentioned. US Military intelligence, the people placing capture and storage splitting devices at all the major telecoms in the US by lawful authority and compliance of those companies with the Law.
Being a NATO member doesn't give the US cart blanche to install surviellence gear in their telecoms in order to perform country wide surveillance on their population. That is nonsensical and it's what you're implying. Although that would be NSA's goal. NATO is an Anti-Soviet mutual defense organization, not a come in and violate our soveriegnty and our citizens privacy pact. Being in NATO doesn't mean come in and take over my country and do what you want. That's point number 1.
Point number 2 is that what I say is backed up by combinatorics, and basic probability and statistics.
>Here let me tell you what to do Mr. Tor Project.
I was responding to the Tor Project Petition in the link you posted. Not to you. Same with when I said I thought you were the fucking expert I was refering to the article you posted and addressing the Tor Project, not you. That wasn't clear my sincere apologies.
Many of these guys worked at the NSA as interns and still work with the NSA.
>Everything you're saying is made up.
No. Just the fractional share of tor node relays and exit nodes that I proposed, I don't know what those proportions actually are. I would expect that soveriegn nations with healthy economies would have the lions share of their tor infrustructure run by their own people, it would be surprising to discover that 80% of Tor nodes in Romania were actually run by the CIA, that would be shocking, but not impossible. I'm making an assumption for rhetorical / computational value. The assumption is that the largest share of tor nodes in a foriegn country are administered by people in that country. Might not be true, its an assumption, it seems reasonable.
My statements are backed up by logic and probability. You're the person claiming that by being in NATO that that country's telecom infrastructure is completely owned by the US military. It could be true. I don't know, but I don't think that is true.
>I'm sure they don't share anything with the U.S. Because you believe it, it must be true.
They do share intelligence, just not everything. And certainly this intelligence sharing is predominantly about Russian and Chinese economic influence, and millitary positioning (Troop movements, missle defense, estimates of nuclear stock piles ).
My formating is messed up because I'm posting from links browser.
I assumed that you are US military intelligence based upon your links to these articles (which normal people don't read), and your technical proficiency with computers and your quickness to dismiss the technical merit of what I'm saying with ad hominem attacks rather than discussing the technical details themselves. And then boogie manning with "NSA is in everything so don't even try it".
What I'm saying makes sense, people who read this thread will come to the same conclusion, based upon the technical points that I've raised, not based on your joker image and the belittling "You're a Fool" commentary.
I've raised techical objections to the standard way Tor Project wants you to perform routing, and gave technical reasons why my proposal is better. I could care less if you are Tor Project member, or US military Intelligence.
>>3522 >it's what you're implying.
It's not.
>NATO is an Anti-Soviet mutual defense organization
The Soviet Union does not exist anymore. And, yet, NATO still does. Really gets the ol' noggin' joggin'.
>Point number 2 is that what I say is backed up by combinatorics, and basic probability and statistics.
What you said was incoherent and full of made-up numbers.
>>3518 >My guess is
>>3519 >suppose
>>3522 >I don't know
>I would expect
>I'm making an assumption
>Might not be true
>its an assumption
>I don't know
You're just making shit up. You have no idea what you're talking about, and even if you did, I have no confidence that you could present it in a coherent manner. In addition, your entire worldview reflects an unwarranted confidence in your understanding of the way the world works, the way international intelligence sharing works, and the way threats to the Tor network work. You are desperately foolish.
>>3522 >Many of these guys worked at the NSA as interns and still work with the NSA.
When you say "many of these guys" and don't specify who they are, you can say anything you like. Many of these guys work at a tyre shop. Many of these guys work at Seaworld. What guys? "These guys", of course. If you're saying that the core contributors to Tor
>worked at the NSA as interns and still work with the NSA
feel free to post your evidence, but it would only be of potential benefit to other posters, as I've tired of your incoherence, your ignorance, and your arrogance, and will be ignoring your further ramblings.
>My statements are backed up by logic and probability.
lol
>You're the person claiming that by being in NATO that that country's telecom infrastructure is completely owned by the US military.
I claimed no such thing. Please feel free to engage in whatever additional straw-man argumentation soothes your autistic rage.
>They do share intelligence, just not everything. And certainly this intelligence sharing is predominantly about Russian and Chinese economic influence, and millitary positioning (Troop movements, missle defense, estimates of nuclear stock piles ).
Interesting! Nanochan's first leak! See, the particular details of what intelligence is shared between NATO members, and especially particular NATO members, would be TS/SCI information. Clearly you have such a security clearance and are sharing your knowledge with us here. I thought I was the Tor Project/CIA/military guy, but it turns out it was you all along.
>My formating is messed up because I'm posting from links browser.
I am utterly unsurprised that you've found yet another way to make your browsing through Tor look different from almost everyone else's. Bravo.
You're welcome to have the last word, or last words, in case your response comes out as a series of feverish, semi-literate posts again. You are ineducable, and anyone who comes away from this exchange with confidence that you know what you're talking about is in similarly sorry shape, so I'm done tolerating your make-believe numbers, your risible understanding of the world, and your incompetent prose.
>I am utterly unsurprised that you've found yet another way to make your browsing through Tor look different from almost everyone else's. Bravo.
You go ahead and keep using that piece of junk known as Tor Browser. And I'll do things "the wrong way". And you can make apologetics for them when they do bullshit like leak my OS in the user-agent as Linux to make my traffic stand out.
Relays are around 6000 so the correct computation should have been
6000 Choose 3 = 6000 C 3 = 3.5982002 E10
Pretty good, now lets try 6000 C 6
=6.463815 E 19
Both are large numbers. If the relays are chosen at random both sample spaces are reasonably large
35 trillion vs 6.5* 10^19. Locking out 9 eyes will lower that substantially. Anyone who passed high school math knows what I'm talking about when I refer to combinatorics (Perms and Combs) and probability and sample spaces. These aren't made up numbers but rather an estimate based on data. My original estimate was wrong because I used 1000 relays instead of 6000 as shown in the graph. It was a reasonable estimate.
>. You are ineducable, and anyone who comes away from this exchange with confidence that you know what you're talking about is in similarly sorry shape, so I'm done tolerating your make-believe numbers, your risible understanding of the world, and your incompetent prose.
"Make believe numbers"... I just grabed the data from their (Tor Project) metrics page and re-calculated the estimate. It seems reasonable. Anyone who reads this exchange will realize that you aren't comfortable with high school level math, probably because you flunked out and joined the military... Which is great, there's nothing wrong with that, we all value your service to your country, just not the deception and lies.
>I've tired of your incoherence, your ignorance, and your arrogance, and will be ignoring your further ramblings.
Good, piss off, you've just been name calling and have added nothing of technical merit to this discussion. But I do appreciate it, as contrarian opinions and dialogue are necessary for progress in discovery.
I'll read over the rest of this in the morning, and comment if I think of somthing.
I don't really have a problem with you in general, but the name calling isn't adding value here. I've been talking about techincal stuff, and you're just name calling and avoiding explaining why you think my technical approach is wrong. I think you're wrong and I've stated why in technical terms. Your respnse has been non technical, didn't address any of my points and resorted to name calling. You're wasting my time. I do enjoy reading your posts however.
So why don't you address the point that TOR does work only under certain conditions, that is nodes not cooperating?
Uncovering of TOR user happened because their entry-exit nodes were in the same hand:
Harvard Student Receives F For Tor Failure While Sending 'Anonymous' Bomb Threat What Kim didn’t realize is that Tor, which masks online activity, doesn't hide the fact that you are using the software. In analyzing the headers of the emails sent through the Guerrilla Mail account, authorities were able to determine that the anonymous sender was connected to the anonymity network. Using that conclusion, they then attempted to discern which students had been using Tor on the Harvard wireless network around the time of the threats. By going through network logs and looking for users who connected to the publicly-known IP addresses that are part of the Tor network, the university was able to cross-reference users that were using both Tor and its wireless internet around the time the bomb threats were received. https://www.forbes.com/sites/runasandvik/2013/12/18/harvard-student-receives-f-for-tor-failure-while-sending-anonymous-bomb-threat/ http://archivecaslytosk.onion/r16vS
Did you look at what routes an fresh out of the box TOR-Browser chooses? While there very many nodes in a lot of different countries TOR does choose always the same high bandwidth nodes in the "Echelon" countries.
>You are a fool.
Who is a fool, or in this case more a trickster, that would recommend to have entry and exit at the same "Echelon" countries.
So how high is the probability of France, Netherlands, USA and Germany (NSA sits right on the main DE-CIX exchange) collaborating and that of Russia and the USA?
Russia and USA do collaborate, but close enough in exchanging sensitive packet and timing information? For sure less likely, than with her vassals.
>It is trivial for a country's intelligence services to rent servers in a non-cooperating country through a front company and spin up Tor relays and exits.
You mean the FSB let the CIA/NSA run nodes in Russia and monitoring them? As likely as the NSA does the GRU do that in the USA.
>>3515 > So even if Germany is able to figure out that I'm the guy who exited from Turkey, they aren't bugging me about it.
Be careful, Germany is a colony of the USA and the NSA has full access to their comms.
https://en.wikipedia.org/wiki/Dagger_Complex
>>3516 >I just don't want any nodes in 9 eyes.
In an ideal world that would be cool, but under the constrains we live, there are only a very limited number of actors and we don't know how they are collaborate, that isn't feasible.
I'd say using an NSA server to connect to an FSB one isn't the worst if the third hop goes to say Indonesia, for example.
>3520
> Its not likely that Germany will have the legal authority to Compel Romanian telecoms to allow them to install a black box tap.
Both are in the EU.
EU was founded by the glow niggers of the CIA:
>>3521 >Also, Romania is a NATO member.
In addition, but that makes the principle not wrong (the whole TOR system is based on).
You're attempt of sarcasm doesn't refute that.
>>3509 >>3507 >6 hope bounce
>Will be more difficult to get both the entry and exit to be correlated or owned by NSA.
True, but did you check that your 6 hop self compiled TOR packets look identical to that of a out of the box TOR client?
Since that means 6 times the encryption, I suspect the individual packets are bigger, just for the additional routing information.
But if your packets are different the other TOR packets, this would make identifying you trivial.
Generally I think it would be a good idea to have some variation in length of the Onion chain, just like the original Onion mail remixer had.
I think it would a good idea to send individual requests with different, random routing and chain length. More difficult routing and latency, sure, but in addition with changes in the nodes that at the moment work in a FIFO manner, would make it much more difficult to trace back the request.
>>3515 I would make the nodes to increase latency by a small random time. For that nodes would collect a number of incoming packets and resend them in random order instead of order of ingress.
The Onion remailer did that. They waited for a number of mails to receive before resending them further. An external observer would not be able to distinguish so easy where from to where to.
That works best if there is a minimum traffic. To improve that I would make the nodes interchange packets at random, to create a "noise floor". The "noise" could be reduced with traffic, but always kick in if there is a "hole" in the transport stream. It would "pad" and "salt" the transport streams to make traffic analysis difficult. So if one request has send, all packets passed, the node would add packets for some time to be discarded at the next hops.
At the moment, if there is not much traffic, one can interfere who is sending to whom.
>>3512 >Change the number 3 to the number 6 and enforce it for everyone. Done. Now everyone is doing 6 hops and your argument is now irrelevant.
Makes to much sense and increases latency, which is good for anonymity (if some random, jitter is added)
>>3513 [D]
>The more mixing, the more variable the time delays between servers, the more variety in distance between hops that changes every circuit, the more dificult it becomes to do successful cross correlation,
Yup, but it shouldn't be by chance, but by design.
"The core principle of Tor, "onion routing", was developed in the mid-1990s by United States Naval Research Laboratory employees, mathematician Paul Syverson, and computer scientists Michael G. Reed and David Goldschlag, with the purpose of protecting U.S. intelligence communications online. Onion routing was further developed by DARPA in 1997.[22][23][24][25][26][27] The alpha version of Tor, developed by Syverson and computer scientists Roger Dingledine and Nick Mathewson[20] and then called The Onion Routing project, or Tor project, launched on 20 September 2002.[1][28] The first public release occurred a year later.[29] On 13 August 2004, Syverson, Dingledine, and Mathewson presented "Tor: The Second-Generation Onion Router" at the 13th USENIX Security Symposium.[30] In 2004, the Naval Research Laboratory released the code for Tor under a free license, and the Electronic Frontier Foundation (EFF) began funding Dingledine and Mathewson to continue its development.[20] In December 2006, Dingledine, Mathewson, and five others founded The Tor Project, a Massachusetts-based 501(c)(3) research-education nonprofit organization responsible for maintaining Tor.[31] The EFF acted as The Tor Project's fiscal sponsor in its early years, and early financial supporters of The Tor Project included the U.S. International Broadcasting Bureau, Internews, Human Rights Watch, the University of Cambridge, Google, and Netherlands-based Stichting NLnet."
"Tor was developed by Dingledine—with Nick Mathewson and Paul Syverson[4][better source needed]—under a contract from the United States Naval Research Laboratory.[1] As of 2006, the software they developed was being distributed using proceeds from the Electronic Frontier Foundation, by the Tor Project.[1] As described at the end of 2015"
Me:
I'm pretty sure that I heard Nick Matthewson say ( Out of his mouth at a talk that I watched on youtube probably "state of the onion" ) that he interned at the NSA as a summer student. I might be mis-remembering that, but I'll look for something in print. You don't get on this type of project with a contract from the US Naval Research Laboratory, without having a security clearance.
I didn't claim their code was bad, or that it doesn't work as stated, I just have made some design choices that I think improve things, and you claim my "Improvements" are down grades. I dispute what you're saying and I've stated why. Your response has been name calling.
>>3526 I have no particular opinion on 3 vs more hops, because I use Tor to hide in the crowd of other Tor users. But I had to reply to this.
>Uncovering of TOR user happened because their entry-exit nodes were in the same hand:
Wrong. The article you quote, even the very excerpt you pasted here, refutes what you've said:
>What Kim didn’t realize is that Tor, which masks online activity, doesn't hide the fact that you are using the software.
He fucked up by being the _only_ Tor user on the Harvard network at the time (probably didn't even use bridges). This is why you try to sell Tor browsing to as many people as possible around you - and not use private university networks to send in bomb threats like a braindead retard. He could have used any free Wi-Fi or Internet cafe away from the place he's most likely to be identified in. This is OpSec failure, not related to software or who's controlling the nodes.
The NSA or any other agency don't need to run Tor nodes, since they monitor the entire Internet, the infrastructure of Tor. So the premise of using six hops instead of three to circumvent LEA-run nodes, or whatever it is you're trying to do (I've forgotten after reading all those walls of text) is flawed.
>>3536 [D] >>3537 The NSA or any other agency don't need to run Tor nodes, since they monitor the entire Internet,
They don't, just partial.
So the problem is a little bit more complex, auspicious.
>>3537 "The entire Internet" is, of course, an exaggeration, though they might be able to monitor the entire Tor network, depending on its size, which is why it is crucial that everybody use Tor for daily communications, and not just some people sometimes to post on obscure imageboards. In spite of this, Tor does still work, since it is a mixnet, and all the data it shares is encrypted. More than three hops is redundant and futile, and only serves to single you out and cause further stain on the network. Also, using a browser other than the Tor Browser doesn't benefit anyone. One of the methods Tor achieves anonymity is by having all Tor users look the same; the less Tor users look the same, the more identifiable everyone becomes. If you don't want to use it properly, don't use it at all. On how the Tor Project addresses the issue of infrastructure surveillance, refer to their documents, because I don't really know, I'm just a user.
X will be your input into the entry node, Y will a particular exit node, perform this sum over the time series where X(t) is packet rate or some other metric, (signal burst durration, packet length),and pairwise compute the cross correlation between your signal (which was recorded at the listening post) and each exit node that is under surviellance or is being recorded re-entering the USA system.
You can also do a 2D or 2 variable cross correlation on the two most identifiable variables. The cross correlation with the largest peak indicates your delay time, and the largest pairwise cross correlation indicates the probable correct signal pair.
This could be completely automated or specifically targeted.
The more dissimilar the input and output signals are, the weaker the correlation, so random jitter in delay times, variable transmission rate changes between input and output, more layers of encryption changing the size of the packets and making them dissimilar, will help to make these correlations coefficients be smaller. If the delay times between input and output are always changing this will also mess up the correlation coefficient.
With 6 hops it should look the same as 3 hops after the first 3 hops, so if there are any observable differences between packet length with a 6 hop and 3 hop signal, this discrepancy should disappear once you have 3 hops left. So at worst this is as good as 3 hops, and no better. At best it could be alot better due to the positive effects of mixing (increasing your node sample space), and variable delay times (from variable travel time distances between nodes, and circuits changing periodically switching ).
>>3525 >35 trillion
6000 C 3 = 3.5982002 E10
10^ 6 is Million, 10^9 is Billion. 10^12 is Trillion. My mistake, correction:
So that's 35.9 Billion possbile three node combinations/path selections.
6000 C 6 =6.463815 E 19
That's 64.6 *10^12 * 10^6
or 64.6 Million Trillion.
The first sample space seems tractable, the bigger sample space will require more computing power to model all possbile paths.
That's another possible type of attack, or analysis method. I would model all three node paths for travel time, or just model / send ping packets through each of the 64 Billion paths and make a lookup table of travel times. Then if you have all entry and exit nodes monitored (which they don't) then you could use this lookup table and compare it to what you see in real life. This will rule out certain paths.
Joe connects at node A at time t0, and simultaneously within a 5 second window traffic emerges at 1000 distinct nodes and was observed. Calculate the time difference between these connections, and compare with the lookup table values that contain these entry and exit nodes. This will weed out several of the possbile paths. So say the emperical travel time / delay time to these 1000 nodes from node A ranges from 100ms to 400ms, then you can cross off all paths with node A and the exit node (in the lookup tables) that are greater than 400ms from your ping tests, and less than 100ms and come to a smaller subset of 3 node paths.
Form there examine (from your listening posts) if any of those potential 3 node paths matches connections between node A and the remaining possible exit nodes. Namely look at all of Node A's connections in that time window and compare this to the remaining paths in the lookup table. That will give an even smaller subset of paths to examine.
At that point you could calculate probabilities of the path selections being correct and rank them by the most probable.
Seems like a lot of work but with only 34 Billion possible paths this seems tractable for a supercomputing cluster.
So say Joe is an important person and has been placed under surviellance. All of his connections to node A are timestamped and logged. Given that you know Joe connected to Node A, you really only have to look at 5999 C 2 possible paths = 1,799,1001, or about 2 million paths. Very tractable.
So from all exit nodes that are monitored from the trafic bursts in that 5 second window, calculate delta t from Joe's connection to node A at t0 to the first burst of data coming from the 1000 exit nodes. From this calculation 100ms<dt_Joe<400ms
Go to the lookup table for Node A connections to those 1000 exit nodes and cross out any paths that have dt A->C < 100ms and dt A->C > 400ms
This should bring your 2 million possioble paths down to tens of thousnds, then rank them by distance of dt_joe to dt_model.
A -> B_j -> C_k dt_model dt_observed
From the top 100 closest rankings ( smallest absolute difference | dt_model - dt_model | , examine the connections from Node A to middle Node B_j and see if any of those node A connections match with the top 100 paths. If so you now have the complete path. If not look at the top 200. This would weed the paths from 10^4 down to maybe 100 or so. Then from these 100 possible connections observe what they do and correlate this to things you already know about Joe and his habits. This should weed the connections down to 2 or 3.
If all three of thes connections happened inside of the surviellance grid ( pass through listening posts (logged routers) along their path) then all of this analysis could be performed, probably automated and within a day or two.
Error
> 1,799,1001, or about 2 million paths.
17,991,001 or 18 million paths.
But now since you only observed 1000 exit nodes making traffic and you know node A was used by Joe, you only have to search 5999*1000 = 5,999,0000 possible paths from C observed to B possible middle nodes. So you only need to do 6 million delta t travel time comparisons.
>From the top 100 closest rankings ( smallest absolute difference | dt_model - dt_model |
| dt_model - dt_observed | where dt_model is from the lookup table of values containing node A and C, and dt_observed are the actual calculated dt values from Joe's first connection to the first output burst at those 1000 exit nodes.
It would be nice if this page displayed point counts by country circle. But estimating by the size of the circles about 1/8th of world relays are in USA, and 1/5th of world exits are in USA.
US relays 6676/8 = 834.5 ~ 835
US Exits 908/5 = 181.6~182
So if all 3 nodes in your path stay in the United States, you have 835 C2 * 182 possible paths.
this is 63,371,490 or 64 million
If you are under surviellance as mentioned then only 835*182 = 151,970 possible paths need to be investigated. This is now an extremely tractable problem.
182 * # number of new connections in 5 second window.
measure all of those dts and compare to the model values. remove all model paths that are bigger or smaller than your max and min. Rank paths by smallest travel time distance. Examine all connections from Node A and match with your remaining model paths.
This could go from 150,000 -> 10000 -> 100.
Then repeat this process for the next hour, until you're left with 1 or 2 paths by building statistics.
>>3576 >Germany, US, Netherlands, France are the largest pools of relays.
Imagine, them all in one box. Not just one box, but a single program simulating a network.
France is a very crypto hostile regime. They have some very high bandwidth nodes. Suddenly.
Secret documents reveal: German foreign spy agency BND attacks the anonymity network Tor and advises not to use it
The German spy agency BND developed a system to monitor the Tor network and warned federal agencies that its anonymity is „ineffective“. This is what emerges from a series of secret documents that we are publishing. The spies handed a prototype of this technology over to the NSA, in expectation of a favor in return.
A few weeks prior to the conference, the BND hackers from Unit 26E „developed the idea of how the Tor network could be monitored relatively easily“, according to internal BND documents.
Both NSA and GCHQ expressed „a high interest“ and offered support. The three spy agencies decided on further meetings and the creation of a project group, while the BND planned to set up its own Tor exit node server, as well as a „test capture“ and „evaluation with the NSA“.
How exactly the spy agencies want to crack Tor remains vague.
The BND hackers based their attack on „a paper by an American university“, which they handed over to the NSA. During the video conference in Bad Aibling, the BND responded to questions and presented a timetable with further steps. The Germans planned to set up their own Tor network in a lab within „six to eight weeks“ in order to better understand the system and to verify the research paper.
The NSA agreed to contact the university to learn more about the research paper. The BND started its work, set up the test network and developed a „proof of concept“ for the attack, a prototype.
Precisely how the BND plans to „chop“ Tor is unfortunately redacted in the document we obtained. But as before, the spy agency refers to public research.
The Americans gladly accepted his offer – the NSA and GCHQ took over the project.
One and a half years later, the BND warned German federal agencies not to use Tor. The hacker unit „IT operations“ entitled its report: „The anonymity service Tor does not guarantee anonymity on the internet“. The six-page paper was sent to the chancellery, ministries, secret services, the military and police agencies on 2 September 2010.
############### BEGINNING OF GUIDE ##########################
In this thread I document and revise how to set up a tor hidden service email server, you may substitute the servers that you are most comfortable with.
STEP 0) Collect the relevent files from The Endware Hidden Service
Set up a tor mail server using postfix or OpenSMTPd, with dovecot for imap or pop.
You may contact me anonymously at endwall@zvdcyrpole74oo24gqkx2wh6rmrthrhexzik5dm6xf7ewtiekxmvqwqd.onion
Use endmail.sh to send mail to this account.
http://nguipxnkrp3qrzrlduhsatpcpwehnblzmlkc5ifiumxq4z5jlh4lwvid.onion/endwall_pgp.asc
Encrypt with pgp and send messages and files by email on a tor hidden mail service on port 25.
Everyone should do this. Then just share your hidden service address and handle.
http://nguipxnkrp3qrzrlduhsatpcpwehnblzmlkc5ifiumxq4z5jlh4lwvid.onion/endware/endmail.sh
http://nguipxnkrp3qrzrlduhsatpcpwehnblzmlkc5ifiumxq4z5jlh4lwvid.onion/endware/endfix.cf
http://nguipxnkrp3qrzrlduhsatpcpwehnblzmlkc5ifiumxq4z5jlh4lwvid.onion/content/dovecot/ <--- download all of these config files
Try it out!
STEP 1) Install Postfix
$ su
# torsocks pacman -S postfix dovecot
# ...
# cd /etc/postfix/
# cp endfix.cf /etc/postfix/main.cf
STEP 2) Install Dovecot
# cd /etc/dovecot/
# mkdir -p conf.d
# cp dovecot.conf /etc/dovecot/dovecot.conf
# cp *.conf /etc/dovecot/conf.d/
# cp *.ext /etc/dovecot/conf.d/