July 2018 - cypherpunks-legacy

[Press Release List] Zero-Knowledge Systems Unveils Free,
by Craig Silverman 06 Jul '18

06 Jul '18

Easy-To-Use Freedom 2.0 Inte rnet Privacy Suite: The Most Comprehensive Internet Privacy And Security Tool Available Date: Tue, 12 Dec 2000 11:32:47 -0500 Sender: owner-pressrelease(a)zeroknowledge.com Reply-To: pressrelease(a)zeroknowledge.com ====================================================================== The Zero-Knowledge Systems Press Release http://www.zeroknowledge.com ====================================================================== FOR IMMEDIATE RELEASE ZERO-KNOWLEDGE SYSTEMS UNVEILS FREE, EASY-TO-USE FREEDOM 2.0 INTERNET PRIVACY SUITE: THE MOST COMPREHENSIVE INTERNET PRIVACY AND SECURITY TOOL AVAILABLE --Freedom 2.0 gives consumers free access to the most easy- to-use privacy and security features, including a personal firewall, form filler, cookie manager, ad manager and keyword alert-- Montreal - December 12, 2000 - Zero-Knowledge(R) Systems Inc., the leading developer of privacy solutions for consumers and business, today unveiled the next generation version of its award-winning Freedom(R) Internet Privacy Suite. Freedom 2.0 now offers five standard privacy and security features as a free download in addition to enhanced paid premium services of untraceable private email and anonymous private browsing and chat. "The Freedom 2.0 Internet Privacy Suite affirms Zero- Knowledge Systems' commitment to creating strong, transparent and easy-to-use solutions that protect individual privacy and security on the Internet," said Austin Hill, president of Zero-Knowledge Systems. "From its new free standard features to the unmatched privacy protection of its enhanced premium services, Freedom addresses every need of Internet users who want to protect themselves and control their personal information, security and privacy." Freedom 2.0 builds upon the first version's unmatched premium services by offering five essential privacy and security features as a free download and substantial enhancements to its paid premium services of untraceable encrypted email and anonymous private browsing and chat. Users who upgrade to the premium services gain access to the global Freedom Network of servers and create untraceable pseudonyms or "nyms." The nyms allow customers to select different digital identities for their online activities. All nym traffic is encrypted and routed through the Freedom Network, ensuring complete privacy for browsing, chat, email and Telnet. Freedom's already intuitive ease-of-use and striking appearance have also been improved with this version's completely redesigned user interface. Freedom 2.0 Internet Privacy Suite Free Standard Features * Personal Firewall - Protect against malicious hackers Freedom's personal firewall controls both incoming and outgoing Internet connections and alerts users of any unauthorized attempts to connect to their computer by malicious hackers or outside entities. This feature is critical for consumers with an "always-on" DSL or cable modem connection. * Form Filler - Speed up and secure online registrations and transactions Freedom automatically fills out forms and remembers passwords, making online registrations and purchases a snap. Users can either use information from their true identity, create a pseudonymous identity or have the Form Filler generate random information. Unlike other products that require users to store their personal information on a company server, Freedom's Form Filler encrypts and stores sensitive information on the user's computer, where only they have access to it. * Cookie Manager - Prevent websites from tracking activities Freedom lets Internet users take advantage of the cookies they want while blocking or deleting the ones they don't. Unlike other stand-alone or browser based solutions that lump cookies together, Freedom's Cookie Manager sorts cookies into individual "Cookie Jars." This allows users to manage each online identity separately, and control how much information websites record about browsing habits. * Ad Manager - Control ads and speed up browsing Freedom's Ad Manager speeds up the Web browsing experience by getting desired information to the user faster, while removing the clutter of unwanted ads. It also prevents activity-tracking cookies and "Web bugs" from being dropped onto their computer. * Keyword Alert - Prevent personal information from leaving the computer Freedom's Keyword Alert instantly scans all outgoing communications and warns users before sending sensitive or identifying information. For example, parents can set Freedom to scan for the real name or telephone number of their children and Freedom will show an alert before releasing that information. Freedom 2.0 Internet Privacy Suite Premium Services * Untraceable Encrypted Email - Secure and privatize email Freedom lets Internet users send and receive encrypted email from existing email applications to ensure that no one, including their ISP, can open or read it. The true identity of Freedom users is protected since no one can associate the user with their nym's address. Freedom also blocks unsolicited bulk email (spam) from reaching their inbox. * Anonymous Browsing & Chat - Go online undetected Freedom lets users enjoy untraceable access to the Internet by way of private routes through the Freedom Network. These private access routes prevent anyone from tracing a user's online activities or from determining their physical location, guaranteeing a completely private Web browsing, Telnet and chat room experience. Zero-Knowledge's commitment to releasing the source code to the Freedom system has set the standard for consumer privacy products by allowing anyone to verify how the system performs. This assures Freedom provides the promised protection. As the recognized privacy leader, Zero- Knowledge is fulfilling its commitment to deploy consumer products that are open and transparent. Among other accolades, Freedom 1.0 was named 2000's "Most Promising Internet Newcomer" by PC World, was called "the Rolls-Royce of privacy software" by Yahoo! Internet Life and received a 5/5 editor's rating from PC Magazine. ZDNet also gave Zero-Knowledge's Customer Support group a 10/10 rating for technical support. Availability and Pricing Freedom for Windows 95/98, 2000, Me and Linux is available as a free download at http://www.freedom.net. Premium services can be purchased for US $49.95 per year. A Macintosh version of Freedom is expected in 2001. About Zero-Knowledge Systems, Inc. Zero-Knowledge Systems (http://www.zeroknowledge.com) is the leading provider of privacy enabling technologies and services for both consumers and enterprise. Zero-Knowledge Managed Privacy Services (http://www.zeroknowledge.com/business) provides expert consultation and privacy-enhancing solutions that enable businesses to comply with privacy legislation, maximize customer relationships and build consumer trust without violating privacy. Zero-Knowledge Freedom Internet Privacy Suite provides consumers free access to the most easy-to- use privacy and security features -- including a personal firewall, form filler, cookie manager, ad manager and keyword alert -- in addition to its leading paid premium services offering untraceable private email and anonymous browsing and chat. Freedom 2.0 for Windows 95/98, 2000, Me and Linux is available as a free download at http://www.freedom.net and Freedom source code is available at http://opensource.zeroknowledge.com. Journalists can visit the Zero-Knowledge pressroom at http://www.zeroknowledge.com/media. (Freedom, Zero-Knowledge and Managed Privacy Services are registered trademarks of Zero-Knowledge Systems, Inc. All other names may be trademarks of their respective owners.) For more information Dov Smith Zero-Knowledge Systems 514.350.7553 dov(a)zeroknowledge.com Kristy Cory Red Whistle Communications 503.552.3749 kcory(a)redwhistle.com To unsubscribe from this list, please visit: http://www.zeroknowledge.com/media --- end forwarded text -- ----------------- R. A. Hettinga <mailto: rah(a)ibuc.com> The Internet Bearer Underwriting Corporation <http://www.ibuc.com/> 44 Farquhar Street, Boston, MA 02131 USA "... however it may deserve respect for its usefulness and antiquity, [predicting the end of the world] has not been found agreeable to experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'

1 0

None
by None 06 Jul '18

06 Jul '18

1 0

A hard rain
by jw＠bway.net 06 Jul '18

06 Jul '18

Every morning these days I take the subway from Brooklyn Heights to the World Trade Center, where I catch the PATH train to Newark. I arrived at WTC at 9 this morning, exited the subway and walked through the long underground passageway to WTC One. Usually there is a violinist there, and he often is playing the Godfather theme around the time I get there. I didn't notice him this morning. I could tell something was going on by the way people were milling around, and by the unusual crush of people trying to exit to the street up a staircase not usually so crowded. A uniformed subway employee was standing there, telling us we could not proceed through the door to the World Trade Center, and for a public official directing human traffic, she was unusually hysterical: "There are terrorists....go back that way," gesturing towards the subway. And most of the crowd started off obediently back towards the trains. At that moment the people going up the stairs began running down the stairs back into the passageway. I decided to go upstairs, against the flow, partly out of idiot curiosity, partly because if something was happening I felt safer in the open than in a tin can in a narrow tunnel. On the street, I could see that WTC One was burning from its upper stories; curlicues of paper were floating down, glittering in the light, an eerily beautiful and incongruous effect. I saw people running past, some of them crying, as I stood immobilized, watching the flames. Then there was an explosion, and fragments of glass rained down on my head. I saw a huge hole in the roof of a building two blocks from World Trade Center. I thought there were incoming projectiles of some kind, missiles or mortars, and began running away from the burning buildings, up Church Street to Chambers, wondering what would blow up next. At some point, as emergency vehicles sped past with screaming sirens towards the disaster, I concluded that the best thing I could do was make a run for the Brooklyn Bridge and walk home to Brooklyn Heights, so I headed that way down Chambers. My brain was clicking over very quickly: the municipal buildings might be another target, and I had to pass close to them to get to the bridge; or terrorists might blow up a van in the middle of the bridge itself and try to bring it down with the hundreds of people crossing it on foot, escaping lower Manhattan. There's another click that happens: fuck it, I could blow up anywhere, I'm just going to do the best I can. I hesitated a moment at the entry to the bridge, watching the two towers burning, thinking about the people who were dying and suffering there. A van, driven by two Arab-looking men, proceeded slowly onto the bridge; I let it get far ahead of me then started to walk, looking back over my shoulder repeatedly at the burning towers. Someone behind me shouted, "There are people jumping!" and I turned to see a black dot, almost certainly a human dot, fall from the top of a tower. I was on the bridge about a half-hour. The lane into Manhattan was closed to everything but emergency vehicles but the roadway to Brooklyn was packed with cars, some of them stopping midway on the bridge while the occupants got out to look at the flames. I heard a bang and flinched around to see that a driver opening his door had hit a motorcycle, throwing its riders to the ground; motorcyclist and driver started screaming into each other's faces in typical New York style--"You shouldn't have been white-lining!"--an expression I had never heard before. A few minutes after I got home I was watching live coverage on CNN as World Trade Center 2 collapsed, then building 1 a few minutes later. This afternoon a report that no attempt is being made to search the wreckage of the building looking for survivors until the fires burn out....Two hundred firefighters are missing, and 78 police. No-one will even venture to say how many people total are entombed in the wreckage tonight. I went to give blood at an over-whelmed Red Cross center where they sent us away and told us to come back in a few hours. But when I came back they had run out of supplies and closed, as had the other center in the neighborhood. I am watching the endless recycling of the same news on CNN. Where is our government? With the exception of the White House counsel, who gave a brief press conference, the only talking heads on CNN are ex-government types; they found Henry Kissinger, but the President, the Secretary of State, the Secretary of Defense are all in hiding, perhaps in some simalacrum of the War Room from Dr. Strangelove: "Gentlemen, you can't fight here, Its the War Room." Right now an expert on CNN is saying, "The plane was the bomb." The terms of reference changed today. Planes are bombs, the World Trade Center can fall, and the President is hiding. The police and firefighters, who are supposed to rescue us, are standing by and letting the flames burn out because it is too dangerous to go in. I thought of Bob Dylan: the glass I picked out of my hair an hour later was a hard rain. I've stepped in the middle of seven sad forests, I've been out in front of a dozen dead oceans, I've been ten thousand miles in the mouth of a graveyard, And it's a hard, and it's a hard, it's a hard, and it's a hard, And it's a hard rain's a-gonna fall. I'm left with a few mixed, jangling thoughts. As I wrote in last month's Spectacle, http://www.spectacle.org/0801/shield.html, why bother building the National Missile Defense, a classic example of fighting the last war, when the next war--the current war--will be fought using our own planes as bombs? I think that monsters exist among us. Considering how easy it turned out to be to bring down the federal building in Oklahoma City with a truckful of fertilizer, the only reason more people haven't done something like it is because we have some sort of internal governor that makes most of us not want to. And planes haven't previously been used as bombs this way because people who could fly them were restrained, possibly by this governor, and possibly by a stronger one, the desire not to die. But the governors are weaker than they were, while the technology gets stronger. CNN kept reporting that the Trade Center buildings were built to resist a hit from a 707, but when 737's and 767's were invented, there was no way to upgrade them. But we are not unresponsible for the monsters; they are responsible for themselves, certainly, as moral actors, but we did our part in their creation. We funded and trained the Afghanistan resistance against the Russians, who later turned on us; but more than that, when we squeeze people and crush them, make them live in humiliation and without the air and water, autonomy and self respect necessary to sustain them as humans, we make monsters. Any chaotic outsider on the margins can kill himself in order to take out others, but when the first Israeli Arab fifty-three year old politician does it (as happened last week) or the first pilot does it (as happened today) we must take notice. What happens next I don't really know, but I know what I fear: that we will embark, like the Israelis and Palestinians, on a mindless spasm of killing and counter-killing; that civil liberties and freedom of speech will be curtailed; that the president, with his powerful backers, his deer-in-the-headlights look and his primitive philosophy, will not see the line before crossing it, or won't care. Getting out of Manhattan this morning, out from the shadow of the burning towers, was like a walk in the park, compared to what I think is going to come. I tried to read this morning's newspaper this afternoon, but the pre-bombing news seemed irrelevant, like news from twenty years ago. [Note from Matthew Gaylor: Jonathan Wallace is a long time subscriber to Freematt's Alerts, a NYC Software executive and publisher of The Ethical Spectacle http://www.spectacle.org. ] ************************************************************************** Subscribe to Freematt's Alerts: Pro-Individual Rights Issues Send a blank message to: freematt(a)coil.com with the words subscribe FA on the subject line. List is private and moderated (7-30 messages per week) Matthew Gaylor, (614) 313-5722 ICQ: 106212065 Archived at http://groups.yahoo.com/group/fa/ **************************************************************************

1 0

Say goodbye to the ISS
by Steve Schear 06 Jul '18

06 Jul '18

1998/05/13) Sun Feb 2 20:40:39 2003) Received: from psmtp.com (exprod5mx6.postini.com [64.75.1.146]) by sphinx.got.net (8.12.2/8.12.2/Debian -5) with SMTP id h134WVIk017383 for <tcmay(a)got.net>; Sun, 2 Feb 2003 20:32:31 -0800 Received: from source ([209.157.136.81]) by exprod5mx6.postini.com ([64.75.1.245]) with SMTP; Sun, 02 Feb 2003 23:32:32 EST Received: (from majordom@localhost) by gw.lne.com (8.12.5/8.12.5) id h134PUZq020838 for cypherpunks-goingout345; Sun, 2 Feb 2003 20:25:30 -0800 X-From_: owner-cypherpunks(a)lne.com Sun Feb 2 20:32:33 2003 Return-Path: <owner-cypherpunks(a)lne.com> X-Authentication-Warning: slack.lne.com: majordom set sender to owner-cypherpunks(a)lne.com using -f Message-Id: <5.1.0.14.2.20030202200948.046d5fb8(a)mail.attbi.com> X-Sender: schear(a)mail.attbi.com X-Mailer: QUALCOMM Windows Eudora Version 5.1 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: owner-cypherpunks(a)lne.com Precedence: bulk

1 0

[mnet-devel] DOS in DHTs
by Some Guy 06 Jul '18

06 Jul '18

Ok as promissed, here's a bit of analysis on DOS in DHTs and trade offs with performance. Have a nice read. I'm going to discuss Denial of Service Attacks (DOS) from within P2P networks, not the underlieing network (TCP/IP), independently of what routing system is used. The resistance R of a network to a DOS censure attack is the number of nodes/resources an adversary is forced to attack in order to make it hard to retrieve a particular piece of data. ---- Really stupid routing (RSR) Let's talk about the most resistant network style there is first, the old Gnutella network. Since data can be anywhere, a adversary is pretty much forced to attack all the nodes in the system. R = O(N) ---- Global Hashes Let's talk about some of the pretty academic DHTs, which provide nice peformance by mapping each key onto a specific target node via global hash. Many of these systems have the resistance of only one or a fixed number of nodes. R = O(1) ---- Intermediate networks In between these two networks we have systems like Entropy which break thier hashspace down into s (s=16) specialized cells via a global hash, which then use simple RSR. http://entropy.stop1984.com/en/entropy.html I'm going to do some analysis on this type of network, which should generally be valid for DOS in all DHTs. s = specialization of the network r = redundancy with which a piece of data is stored d = search depth or number of specialized nodes asked p = probablity that the data is found N = total number of nodes R = Resistance to a DNS attack Since an adversary would not know which nodes in one of Entropy's cells had the data, he would be forced to attack all of them. R = O(N/s) This should be clear: p = 1-(1-r*s/N)^d At constant p we can evaluate design trade-offs at large N. s*r*d = O(N) or R = O(r*d) regardless of N. There is a three way geometric trade-off between redundancy of storage (insert HTL*frequency, or popularity), query depth (request HTL), and resistance to DOS. This is what I'm calling the holy trinity of DOS in DHTs. I believe it holds true for all DHTs using global hashes. You could for example design a network where: R = N^0.5 and r = d Thus making r and d O(N^0.25), which you might be able to live with. Here are some sample parameters and p, to give you an idea of the trade-offs in a million node net: n r d s p 1000000 100 100 100 0.633967659 1000000 100 10 1000 0.65132156 1000000 10 100 1000 0.633967659 1000000 100 100 100 0.633967659 1000000 1000 10 100 0.65132156 1000000 1000 100 10 0.633967659 1000000 10 1000 100 0.632304575 1000000 100 1000 10 0.632304575 1000000 1 1000 1000 0.632304575 1000000 10 10 10000 0.65132156 1000000 1000 1 1000 1 1000000 100 20 1000 0.878423345 1000000 20 20 1000 0.332392028 ---- Non-Global Hashing I've got another neat trick to improve things a bit. A trusted group of nodes could use a private hashing function to redistribute data between them. As long as the adversary can not infiltrate the group, he is forced to flood the whole group. A request or insert need only travel one hop through the group. In networks where there is believed to be a certain fraction of adversarial nodes f, you can calculate the optimal clustering c to throw random nodes together in this way. This local hashing cann't be used to get around the trinity, but just win a considerable boost to performance. In each specialized cell you could have clusters of nodes, instead of single ones. I've got other ideas that involve key sharing to try to make clusters that can scale higher, but I'm not sure it's feasible/nessesary. ---- One more point Replication (r) doesn't have to mean coping the data. I could premix route to r nodes and only tell them I have the data. I still will have done O(r) work though. Think of it this way. <work done to DNS> = O( <work done by insertion> * <work done by request> ) Likewise "depth" (d) for searches doesn't have to be HTL; you can search several nodes in parallel, but the system will still have used O(d) resources. __________________________________________________________________ Gesendet von Yahoo! Mail - http://mail.yahoo.de Logos und Klingeltvne f|rs Handy bei http://sms.yahoo.de ------------------------------------------------------- This SF.net email sponsored by: Enterprise Linux Forum Conference & Expo The Event For Linux Datacenter Solutions & Strategies in The Enterprise Linux in the Boardroom; in the Front Office; & in the Server Room http://www.enterpriselinuxforum.com _______________________________________________ mnet-devel mailing list mnet-devel(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mnet-devel ----- End forwarded message ----- -- Eugen* Leitl <a href="http://leitl.org">leitl</a> ______________________________________________________________ ICBM: 48.07078, 11.61144 http://www.leitl.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE [demime 0.97c removed an attachment of type application/pgp-signature]

1 0

Re: [Mac_crypto] MacOS X (Panther) FileVault
by David Shayer 06 Jul '18

06 Jul '18

>From: "Ralf-P. Weinmann" <weinmann(a)cdc.informatik.tu-darmstadt.de> >Are there any whitepapers available on the design of FileVault? Except for >impressive words from marketing droids (AES-128, industry-standard cipher, ><yawn>) I have seen absolutely zilch on the implementation yet: i.e. is >encryption done on a per-file basis or is rather blockwise underneath the >filesystem layer (ala loop-aes under Linux)? AES-128, fair enough; but what >mode is used for encrypting the files/blocks? ECB? CBC? CTR? CCM? I was told that FileVault replaces your home directory with an encrypted disk image, much like PGP Disk, so its probably blockwise underneath the file system layer. Files in your home directory are copied into the disk image, and some file system links redirect calls to the home directory to the disk image, and keep the user from seeing it as another mounted disk. File Vault will automatically expand or contract the disk image at certain points. It creates a new image, copies everything over, and deletes the old image. I don't know what mode of AES-128 it uses. -- David "If tyranny and oppression come to this land, it will be in the guise of fighting a foreign enemy." - James Madison _______________________________________________ mac_crypto mailing list mac_crypto(a)vmeng.com http://www.vmeng.com/mailman/listinfo/mac_crypto --- end forwarded text -- ----------------- R. A. Hettinga <mailto: rah(a)ibuc.com> The Internet Bearer Underwriting Corporation <http://www.ibuc.com/> 44 Farquhar Street, Boston, MA 02131 USA "... however it may deserve respect for its usefulness and antiquity, [predicting the end of the world] has not been found agreeable to experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'

1 0

[Politech] Feingold, Kennedy ask AT&T and others if they're in bed with NSA [priv]
by Declan McCullagh 06 Jul '18

06 Jul '18

The letter the two senators sent to AT&T, Sprint Nextel, and Verizon: http://www.politechbot.com/docs/feingold.letter.telecom.020906.pdf A survey of telecommunications and Internet firms about NSA participation: http://news.com.com/2100-1028_3-6035305.html -Declan _______________________________________________ Politech mailing list Archived at http://www.politechbot.com/ Moderated by Declan McCullagh (http://www.mccullagh.org/) ----- End forwarded message ----- -- Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE [demime 1.01d removed an attachment of type application/pgp-signature which had a name of signature.asc]

1 0

Re: Copyright infringement notice
by Anne P. Mitchell, Esq. 06 Jul '18

06 Jul '18

> On Thu, Aug 16, 2012 at 6:16 AM, groupstudytac groupstudytac > <groupstudytac(a)gmail.com> wrote: >> I get copyright notices from companies like Irdeto , saying that one of my >> customers IP is downloading unauthorized material using bittorent. I also >> have processes in place to handle such notices . >> >> Can anyone share how he handles such notices in his ISP environment , i am >> ready to adapt some valid steps to improve the existing process. >> >> Or should i just ignore such messages ? > > If you're in the U.S., the process for handling these notices is > prescribed by law, specifically the Digital Millennium Copyright Act > (search: DMCA takedown notice). It details what the infringement > notice must include in order to be actionable and what steps the ISP > must take on receipt of an actionable notice. It also prescribes > procedures for the alleged infringer to object and for the ISP to > restore the material following an objection. > > Follow the procedures described in the law to retain your immunity as > an ISP. Consult a local lawyer if you don't find them sufficiently > obvious. The thing that muddies this is that, as I understand it, the notice was not for takedown (i.e. there is not an allegation that they are *hosting* infringing material) - it is a notice that one of their users *downloaded* copyrighted material (IP, do I have that right?) This is part of the RIAA's "graduated response" program, to which several major ISPs, including AT&T, Verizon, and Comcast, have agreed. Basically, the accuser contacts the ISP, and the ISP sends a warning (a "copyright alert") to their user (without giving up the user to the accuser). If the same user is accused subsequently, they get another, sterner warning. In total there is a series of six warnings, with "mitigation measures" accompanying the fifth and sixth warning. If I were counseling an ISP - whether one that was part of the agreement, or not - I would say that the first order is to *put your policy around copyright alerts in writing* - asap - and make it as specific as possible - and then *ALWAYS FOLLOW IT EVERY SINGLE TIME*. It almost (I say almost) doesn't matter what the policy is so long as it's reasonable, but it matters that it be followed to the letter every time, no exceptions. And, if you are an ISP that isn't part of the agreement with the RIAA, it's still not a bad idea to structure your policy to follow the six "copyright alert" structure, because there is some precedent there, and then you come off looking like you are trying to do the right thing, which will make you a less easy target. These two articles give a pretty good explanation of the deal: https://www.eff.org/deeplinks/2012/03/graduated-response-deal-steamrollers-… http://arstechnica.com/tech-policy/2011/07/major-isps-agree-to-six-strikes-… Anne Anne P. Mitchell, Esq CEO/President Institute for Social Internet Public Policy http://www.ISIPP.com Member, Cal. Bar Cyberspace Law Committee ISIPP Email Accreditation: http://www.SuretyMail.com ----- 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

[tahoe-dev] verification of subset of file == proof of retrievability
by Zooko Wilcox-O'Hearn 06 Jul '18

06 Jul '18

Folks: Over on the Bitcoin discussion forums (warning: wretched hive of scum and villainy), someone was asserting that they wanted a "proof of retrievability" protocol and saying that, while they hadn't looked, they were pretty sure Tahoe-LAFS didn't do it right: https://bitcointalk.org/index.php?topic=2236.msg847771#msg847771 I was mildly annoyed by this, because actually we have some extremely strong features along those lines. However, when I wrote a reply explaining exactly what we do have, I was forced to admit that it isn't fully there yet. We already have verification of complete files (although see #568), but we don't have verification of a randomly-chosen subset of a file, which would be a "Proof of Retrievability". See below for the message I posted to the Bitcoin forum. See also an old rant of mine complaining that academic cryptographers have failed to study the papers and documentation of Tahoe-LAFS closely enough to realize that there is most of a proof-of-retrievability in there: https://lafsgateway.zooko.com/uri/URI:DIR2-RO:d73ap7mtjvv7y6qsmmwqwai4ii:tq… https://tahoe-lafs.org/trac/tahoe-lafs/ticket/568# make immutable check/verify/repair and mutable check/verify work given only a verify cap ------- message I posted to the Bitcoin forum > Correct, to verify every bit of every share you need to download every bit of every share. It's expensive - hopefully future versions of Tahoe-LAFS will implement a probabilistic "proof of retrievability" protocol like the one you suggest. Downloading only a subset of a file is already implemented, but there isn't a command implemented that says "pick a random segment of this file, download it from that server, and let me know if it passed integrity checks". You can approximate it with the current Tahoe-LAFS client, like this (the following lines that begin with "$" is me typing in stuff as though I were using a bash prompt): 1. Pick a random spot in the file. Let's say the file size is 10 MB: $ FILESIZE=29153345 $ python -c "import random;print random.randrange(0, $FILESIZE)" 2451799 2. Fetch the segment that contains that point. Segments are (unless you've tweaked the configuration in a way that nobody does) 128 KiB in size, so this will download the 128 KiB of the file that contain byte number 2451799 and check the integrity of all 128 KiB: $ curl --range 2451799-2451799 http://localhost/uri/URI:CHK:jwq3f6lkcioyxeuxlt3exlulqe:sccvpp27agfz32lqjgh… | hexdump -C % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1 0 1 0 0 0 0 --:--:-- 0:00:01 --:--:-- 0 00000000 a9 |.| 00000001 As you can see it took only a second to download and emitted only one byte to stdout, but it downloaded and verified the integrity of the 128 KiB segment that contained that byte. If you are using multiple servers, using Tahoe-LAFS's awesome erasure coding feature to spread out the data among multiple servers, then this will download the data from the 3 fastest servers (unless you've changed the default setting from "3" to some other number). There is no good way to force it to download the data from specific servers in order to test them -- it always picks the fastest servers. You can see which server(s) it used by looking at the "Recent Uploads and Downloads" page on the web user interface, which will also tell you a bunch of performance statistics about this download. In short, this feature is *almost* there. We just need someone to write some code to do this automatically in the client (which is written in Python) instead of as a series of bash commands. Also this code should download one (randomly chosen) block from every server it can find instead of from just the three fastest servers, and it should print out a useful summary of what it tried and which servers had good shares. Oh, there is a different function which does print out a useful summary of results -- the "verify" feature. But, that downloads and tests every block instead of just one randomly chosen block. Another way to implement this is to add an option to that to indicate how many blocks it should try: $ time tahoe check --verify URI:CHK:jwq3f6lkcioyxeuxlt3exlulqe:sccvpp27agfz32lqjghq2djaxetcuo7luko5dhrpdgs7bfidbasa:1:1:29153345 Summary: Healthy storage index: 7qhuoagk4z4ugsjkjgjcre6sx4 good-shares: 1 (encoding is 1-of-1) wrong-shares: 0 real 1m2.705s user 0m0.570s sys 0m0.060s _______________________________________________ tahoe-dev mailing list tahoe-dev(a)tahoe-lafs.org https://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

[tahoe-dev] verification of subset of file == proof of retrievability
by Zooko Wilcox-O'Hearn 06 Jul '18

06 Jul '18

Folks: Over on the Bitcoin discussion forums (warning: wretched hive of scum and villainy), someone was asserting that they wanted a "proof of retrievability" protocol and saying that, while they hadn't looked, they were pretty sure Tahoe-LAFS didn't do it right: https://bitcointalk.org/index.php?topic=2236.msg847771#msg847771 I was mildly annoyed by this, because actually we have some extremely strong features along those lines. However, when I wrote a reply explaining exactly what we do have, I was forced to admit that it isn't fully there yet. We already have verification of complete files (although see #568), but we don't have verification of a randomly-chosen subset of a file, which would be a "Proof of Retrievability". See below for the message I posted to the Bitcoin forum. See also an old rant of mine complaining that academic cryptographers have failed to study the papers and documentation of Tahoe-LAFS closely enough to realize that there is most of a proof-of-retrievability in there: https://lafsgateway.zooko.com/uri/URI:DIR2-RO:d73ap7mtjvv7y6qsmmwqwai4ii:tq… https://tahoe-lafs.org/trac/tahoe-lafs/ticket/568# make immutable check/verify/repair and mutable check/verify work given only a verify cap ------- message I posted to the Bitcoin forum > Correct, to verify every bit of every share you need to download every bit of every share. It's expensive - hopefully future versions of Tahoe-LAFS will implement a probabilistic "proof of retrievability" protocol like the one you suggest. Downloading only a subset of a file is already implemented, but there isn't a command implemented that says "pick a random segment of this file, download it from that server, and let me know if it passed integrity checks". You can approximate it with the current Tahoe-LAFS client, like this (the following lines that begin with "$" is me typing in stuff as though I were using a bash prompt): 1. Pick a random spot in the file. Let's say the file size is 10 MB: $ FILESIZE=29153345 $ python -c "import random;print random.randrange(0, $FILESIZE)" 2451799 2. Fetch the segment that contains that point. Segments are (unless you've tweaked the configuration in a way that nobody does) 128 KiB in size, so this will download the 128 KiB of the file that contain byte number 2451799 and check the integrity of all 128 KiB: $ curl --range 2451799-2451799 http://localhost/uri/URI:CHK:jwq3f6lkcioyxeuxlt3exlulqe:sccvpp27agfz32lqjgh… | hexdump -C % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1 0 1 0 0 0 0 --:--:-- 0:00:01 --:--:-- 0 00000000 a9 |.| 00000001 As you can see it took only a second to download and emitted only one byte to stdout, but it downloaded and verified the integrity of the 128 KiB segment that contained that byte. If you are using multiple servers, using Tahoe-LAFS's awesome erasure coding feature to spread out the data among multiple servers, then this will download the data from the 3 fastest servers (unless you've changed the default setting from "3" to some other number). There is no good way to force it to download the data from specific servers in order to test them -- it always picks the fastest servers. You can see which server(s) it used by looking at the "Recent Uploads and Downloads" page on the web user interface, which will also tell you a bunch of performance statistics about this download. In short, this feature is *almost* there. We just need someone to write some code to do this automatically in the client (which is written in Python) instead of as a series of bash commands. Also this code should download one (randomly chosen) block from every server it can find instead of from just the three fastest servers, and it should print out a useful summary of what it tried and which servers had good shares. Oh, there is a different function which does print out a useful summary of results -- the "verify" feature. But, that downloads and tests every block instead of just one randomly chosen block. Another way to implement this is to add an option to that to indicate how many blocks it should try: $ time tahoe check --verify URI:CHK:jwq3f6lkcioyxeuxlt3exlulqe:sccvpp27agfz32lqjghq2djaxetcuo7luko5dhrpdgs7bfidbasa:1:1:29153345 Summary: Healthy storage index: 7qhuoagk4z4ugsjkjgjcre6sx4 good-shares: 1 (encoding is 1-of-1) wrong-shares: 0 real 1m2.705s user 0m0.570s sys 0m0.060s _______________________________________________ tahoe-dev mailing list tahoe-dev(a)tahoe-lafs.org https://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