- cypherpunks-legacy - lists.cpunks.org

Re: [tahoe-dev] keeping private grids privatey
by sickness 06 Jul '18

06 Jul '18

> > Setting up a private i2p network might be getting OT. But I'd be > interested as well in finding out how to setup a private i2p network > if you ever figure out or find documentation on how to boot strap one > for your community, please share the documentation on how its done ;) > > Thanks, > Jimmy. > sure, it was discussed multiple times on i2p forums: http://forum.i2p2.de/viewtopic.php?p=13603#13603 http://forum.i2p2.de/viewtopic.php?t=2547 http://forum.i2p2.de/viewtopic.php?t=2381 (the same could be done with tor and freenet, iirc but you don't have tahoe over those, and even if freenet is able to store content, it's fixed to k=1 h=3 s=3 speaking in tahoe-lafs language, and you don't really have a way to repair/check the shares if not by redownloading the entire content over time, it's more like a big cache... but differences between tahoe and freenet were already discussed by zooko in another post) _______________________________________________ tahoe-dev mailing list tahoe-dev(a)tahoe-lafs.org http://tahoe-lafs.org/cgi-bin/mailman/listinfo/tahoe-dev ----- End forwarded message ----- -- Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE

1 0

Re: [p2p-hackers] Pirate Pay
by Raphael Manfredi 06 Jul '18

06 Jul '18

Quoting "Vijay K. Gurbani" <vkg(a)bell-labs.com> from ml.p2p.hackers: :This looks like Pirate Pay is injecting multiple sybils into the :DHT with node-IDs close to the info-hash of the file, thus making :the sybils responsible for the file. This assures that all queries :are sent to the sybils, allowing the sybils to sent out false :information to the other peers. See: "Efficient DHT attack mitigation through peers's ID distribution" by Thibault Cholez et al., published in June 2010. This is a protection that can be added to every DHT nodes and which will significantly increase the cost of widespread Sybil attacks. This DHT protection algorithm has been implemented in gtk-gnutella, with a few enhancements of mine (transmitted back to the author of the above article). Here's my summary of the article, as C comments within gtk-gnutella: /* * The idea is that Sybil attacks will necessarily change the statistical * distribution of the KUIDs surrounding the target. By comparing the * actual KUID prefix distribution with the theoretical one, we can detect * that something is unusual. * * The divergence between the theoretical distribution of prefixes and * the actual one is measured by computing the Kullback-Leibler divergence * (K-L divergence for short). * * The measured distribution of prefixes sharing "i" leading bits with the * target is M(i). For instance, if 6 nodes from the 20 closest share * 13 bits with the target, M(13) = 6/ 20 = 0.3. * * The prefix length b_min at which we start looking is the theoretical * k-ball furthest threshold, which dht_get_kball_furthest() gives us. * It is computed as: * * b_min = E[log2(estimated_DHT_size / KDA_K)] * * with E[x] being the integer part of x. * * The maximum prefix length we consider, b_max, is computed as: * * b_max = b_min + KDA_C * * where KDA_C is the "closeness factor", an arbitrary amount of extra * bits we allow to have in common with the key before looking suspicious. * * Starting at b_min, the theoretical distribution of prefixes, T(i) is * computed as: * * T(i) = 1 / 2^(i - b_min + 1) * * So if b_min is 13, T(13) = 1/ 2, T(14) = 1/2^2, T(15) = 1/2^3, etc... * Up to T(b_max) = 1 / 2^(KDA_C + 1). * * The K-L divergence of T from M is given by: * * Dkl(M|T) = SUM(M(i) * log2(M(i) / T(i)) * i = b_min to b_max * * Intuitively, the larger M(i)/T(i), the larger the divergence added * by the i-bit prefix. Given that an attack will focus on getting close * to the KUID target, T(i) will get smaller and smaller as "i" increases * and a large M(i) will indicate a potential attack. * * Since Dkl is a summation, we can determine which prefix length * contributes the most towards the divergence and therefore remove these * nodes from the path as a counter-measure. * * The beauty of that protection is that the more efficient the Sybil * attack is designed to be, the more we'll spot it and defeat it! */ Raphael _______________________________________________ p2p-hackers mailing list p2p-hackers(a)lists.zooko.com http://lists.zooko.com/mailman/listinfo/p2p-hackers ----- End forwarded message ----- -- Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE

1 0

Re: [liberationtech] Disruption at the Intersection of Technology and Human Rights - Forbes
by Collin Sullivan 06 Jul '18

06 Jul '18

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi Julian and Eugen, I'm Collin, I work on Martus training and outreach at the Human Rights Program at Benetech. I wanted to clarify a few concerns coming from the Forbes piece. You're right to question cloud-based security, but Martus is not a traditional cloud computing application--I'd hesitate to call it a cloud application at all, really. Martus is a java desktop application that encrypts information (stored as "bulletins," semi-structured rich documents) locally on the user's machine. The encryption keys (RSA public-private key pairs used to encrypt per-bulletin AES keys) stay on the user's computer, and in her 5-choose-2 secret-share backups. The "cloud" part comes from our server network. The Martus application replicates the ciphertext (through SSL) to a publicly-available server; the servers replicate to each other so that each has a full copy of all bulletins. Note that only ciphertext is ever stored on the servers, unless a user chooses to publish <https://martus.ceu.hu/servlet/SimpleSearch> some of her data through Martus--then the public data (and only the public data) is stored in plaintext on the servers. It's up to the user to decide whether to make any data public, and it's very easy to set Martus preferences such that data will always be private and stored in ciphertext. And there are never any keys stored out of the user's control. Martus users can share information securely through the network. Each bulletin's author can authorize other users to read a given bulletin by including the other user's key when the bulletin is saved. Again, the keypairs are created and stored locally, and there is no web portal for access to a Martus user's private bulletins. Also mentioned in the Forbes piece, Martus comes with some wipe features for the attacker-at-the-door use case. One is an account and data wipe, the other is account/data wipe plus uninstall. These are designed to be quick-erase features ("panic button" functionality, as it's been called), with time constraints precluding overwriting the data several times, and we're careful to explain this to users during trainings and support. Of course, by using her backed-up key, the user can retrieve her data whenever she needs to -- whether she's wiped the data, or had her computer lost or stolen. We first released the software in 2003, and it's in use by human rights monitors all over the world. It runs in ten languages, including Russian, Arabic, Thai, and a number of other non-latin-character languages. There are over 250,000 bulletins saved in the server network. Of course the software is available under the GPL, always has been, and always will be. Hope this helps. More info is available at martus.org <https://www.martus.org/>, and I'm happy to answer any questions. Cheers, Collin - -- Collin Sullivan Human Rights Program Associate Benetech Human Rights Program Email: collin.s(a)benetech.org Skype: collin.w.sullivan GPG: 0x78657D4D https://www.benetech.org - Technology Serving Humanity https://www.martus.org - Martus Human Rights Bulletin System https://www.hrdag.org - Human Rights Data Analysis Group Julian Oliver: > ..on Tue, Nov 13, 2012 at 10:50:04AM +0100, Eugen Leitl wrote: >> On Mon, Nov 12, 2012 at 09:41:05PM -0800, Yosem Companys wrote: >>> http://www.forbes.com/sites/skollworldforum/2012/11/12/disruption-at-the-in… >> >>> Look at Benetechbs development of Martus, a human rights database, >>> based in the cloud with highly secure encryption and eraser technology >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > If you can cartwheel through the lasers while wearing the appropriate 3d printed > face it's a snap. Just don't look down: Cloud computing is high-altitude stuff.. > -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJQo99YAAoJEN3b1xdSAv8h5FcP/1kBSuXGZJcyvSD6qvIUfKKp w9ckp2Hs6exfY1zghhw0oL1+fpRzfhP3uoj06vUtv6jOSYkyH8s7lZATylMy0La7 vnJOlM21/r0pdc/kh4QZKgZBmgchSdWMiO+RnGEXSWtfpzidJ3uT4Zx9ip/p6MLa PnNqywKCyACd8CKYnx5ESVbWXW/764PDnwEn+biCYDdEoMUsNXsW9so40ritwEdt S7Pm4rl6rAklzHRdeiMpSBhx165cszlZjrPmEkRQRbQmbelhsXkPR4kJ+gEgljlr TDQ2iO46MgUViwHH47KkvsGV+sj24BbhrilPrgptxkIoANWiHD8J9Omt5+/Txffu FwzNKD07wk/10BwFgjYnvGviSJyH+c+no1TvIHBLRsLgV2/LeC4nFwRe7uSYT48L vPD6b3qCT3O8sWkTpOqMHFiRrqtEE7Lfjn1N6ymmqrUxIOOuoQY0K63Do6zHvemW PT8NQFlsoqJIAC26dGJAi2xIcch1kvpds2ORijnf/4M5vvPjYa2vNDv+9L6np9AO 5l87XACiDO5A6AWujdpz55ZfLa6lMcuJycTXgxzu/tsKif41A30pBQnDi6k3TY3M J+zhJOQ9WE5DAl8CKDFIPmInWg3dGJcakwj7T9ZxjRFeaPmOSjWyz+BBOIdrpGAu GDG5zb1MzBAhaeLLR7ZH =z3vc -----END PGP SIGNATURE----- -- Unsubscribe, change to digest, or change password at: https://mailman.stanford.edu/mailman/listinfo/liberationtech ----- End forwarded message ----- -- Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0

None
by James A.. Donald 06 Jul '18

06 Jul '18

> Origami world. To be honest, I was thinking about lots dots - no easily inferred shapes, but pretty obvious for a human. If the state-of-the-art software using neural nets and what not is capable of reading even *letters* made up of many dots (circles) in different colors, shapes, and textures (and with many "false" dots in colors that blend in the background for the human eye), then the programmers are much more advanced than anything I read about. > Computer generates a random 3D object out of large polygons with fairly > sharp angles of contact, subject to various limits on the way in which the > object is generated. Displays 2D image of 3D object. > > Human infers 3D object from 2D image, infers unseen portions of the image > from rules by which the 3D image is generated -- for example that the > object must make sense mechanically -- that it should be stable resting on > a plane. I don't know how well people would manage to do that. One requirement was to be easy for a human to answer the puzzle. Mark

1 0