December 2003 - cypherpunks-legacy

Re: DES-Busting Screen Savers?
by Steve Reid 17 Dec '03

17 Dec '03

On the subject of choosing keys randomly, rather than dividing up the keyspace... This seems like a very good idea to me. One potential problem is actually choosing the random keys. Have to be able to get entropy at a fast rate, and/or use a good PRNG. Both of these things would take time away from the actuall cracking (more time than just incrementing the key). And if a weak PRNG is used in order to save time, it's possible that it might favour certain keys and thus interfere with the attempt. Any ideas for a fast and good PRNG? ===================================================================== | Steve Reid - SysAdmin & Pres, EDM Web (http://www.edmweb.com/) | | Email: steve(a)edmweb.com Home Page: http://www.edmweb.com/steve/ | | PGP (2048/9F317269) Fingerprint: 11C89D1CD67287E68C09EC52443F8830 | | -- Disclaimer: JMHO, YMMV, TANSTAAFL, IANAL. -- | ===================================================================:)

1 1

Exporting cryptosystems in pieces: Re: Question [NOISE, mostly]
by Bill Stewart 17 Dec '03

17 Dec '03

Cc: vice.president(a)whitehouse.gov, The cypherpunks list has been discussing ITAR again. :-) >> If it is illegal (by our governments standards...) to export programs >>like PGP, etc., and you can send the whole source code in a message >>because that is also considered illegal, then could you send the code, >>broken up into many pieces, and send THAT in Email, would that be >>illegal? Exporting components of military hardware, including cryptosystems, is also specifically banned by ITAR. How big a piece is enough to get you prosecuted is a question for the nastiest of the N prosecutors out there, and whether you can be convicted is a question for the best of the 12 jurors you'll have.... Vince Cate's arms exporter page lets you export a highly-useful fully working cryptosystem in three lines with one mouse click (developed by Adam Back and an international cast of dozens.) Adam Back's export-three-lines-of-PGP-at-once is a more blatant test of this; go see his web pages. Here's my two bits worth - the following bits are components of PGP, Netscape 3.0b5, and also of RSAREF. 0 1 # Thanks; Bill # Bill Stewart +1-415-442-2215 stewarts(a)ix.netcom.com # http://www.idiom.com/~wcs # Confuse Authority!

2 1

Re: Digital Watermarks for copy protection in recent Billbo
by Alex F 17 Dec '03

17 Dec '03

> > People buying CDs at a garage sale & getting arrested for > >piracy. Wonderful. > > Arrests like this are uncommon. Even buying "cheap bikes" and other "cheap" > (= probably stolen and fenced) merchandise almost never subjects the > purchaser to criminal sanctions. Yes, but concievably if (whoever would be incharge, FBI?) *could*, under law do this, even if they are wrong. It is a lot harder to prove that they intentionally harrassed *you* than it is for them to say that they were following leads and show evidence. Yes, this may never happen, but the mere fact that it *could* is uncomforting. Kinda like the CDA an Clinton saying "we will not enforce this...." Fine, but there is still a law on the books that allows them to if they decide to change their mind. Even if you can't be found guilty, you could still be ruined by legal fees, job loss, etc. > ("Alice's Restaurant" not to the contrary; the confession to Officer Obie > cinched his fate.) Don't wanna end up on that "Group-W" bench! :) > > It would be trivial to add a digital ID signal at, > >say 30,000 or 15 or something like that. This could then be decoded, > > Doubtful. The existing CD standard tops out at a Nyquist limit of about > 20KHz, with the actual sampling at 44 KHz--but there is simply "nothing" at > above 20-22KHz. Putting a signal in at "30 KHz" is simply not possible, > given the Nyquist Theorem and the CD sampling rate. Cds are often sampled at 48 these days. Mine was, and we had to reduce it to 44.1 for mass producing (much to our surprise, since many CD manufacturers love getting stuff at 48 over 44.1) Not familiar with the Nyquist limit w/ regards to sampling rate vs frequency :( Alex F =-=-=-=-=-=-=-=-=-=-=-=-=- Alex F alexf(a)iss.net Marketing Specialist Internet Security Systems =-=-=-=-=-=-=-=-=-=-=-=-=-

1 0

Wiesenthal Center (Canada) and the Ostrich Syndrome (fwd)
by Rich Graves 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- For Skippy. - ---------- Forwarded message ---------- Date: 23 Jul 1996 08:26:15 -0700 From: Ken McVay OBC <kmcvay(a)nizkor.almanac.bc.ca> Newsgroups: alt.revisionism, can.general, alt.censorship, comp.org.eff.talk, misc.legal, ont.general, bc.general, talk.politics.misc Subject: Wiesenthal Center (Canada) and the Ostrich Syndrome The July 19 article in the Vancouver Sun, "B.C. Internet provider is the largest Canadian site for racist material" is alarming. (See URL http://ftp.nizkor.org/ftp.cgi?orgs/american/wiesenthal.center/press for a copy of the article.) It is not the material found on the websites mentioned, nor the fact that the host server is Canadian - nor indeed that the server is physically located in BC - which give cause for alarm. It is alarming because - either by deliberate design or by abject ignorance - - Sol Littman (and the Simon Wiesenthal Center [SWC] whom he represents in Canada) is fostering and promoting the spread of the "Ostrich Syndrome." Their actions represent a counter-productive denial of reality - akin to the Ostrich burying its head in the sand. Littman is quoted as saying: "We found the longer you leave these groups unexposed, the longer they fester and the more they infect others and the only way to deal with them honestly and forcefully is to expose them to the light of truth." Yet the article concludes, "Littman said he wants to see if Klatt will remove the groups from Fairview voluntarily before the centre takes any other action." The only truth that seems to emerge from such a veiled threat is that Littman has no understanding of the Internet. One is at a loss to determine how removal of the "groups" from one Internet Provider's server would in any way "expose them to the light of truth." If Littman had any knowledge of the Internet, he would know that the Nizkor Project [http://www.nizkor.org/] is an award winning website that is accessed daily by hundreds. In addition to being a source for those who seek information about the facts of the Holocaust, it is used as an electronic resource for those who wish to "deal with [these groups] honestly and forcefully" thereby exposing Lemire and many others - of whom Littman may not even be aware. Nizkor is arguably the "host" to more hate literature than any other website. Since we also include a link to the Zundelsite amongst others, will Littman next be targetting Nizkor and demanding that we remove such links? Or that we hide from public view the mountains of archived material which meticulously documents and uses the "arguments" put forward by these groups as instruments of their refutation and exposure? If Littman and the SWC prefer to bury their heads in the sand while engaging in this futile exercise in darkness, let them do so. But the martyrdom they hand on a platter to Lemire and others whom they find offensive stands in marked contrast to their failure to use the Internet on their own website. Conspicuously absent on the SWC website are links to the increasingly growing number of useful resources for those who truly do wish to participate in the battle to expose racist and anti-semitic groups on the Internet. The Ostrich syndrome is far more dangerous to society than any white supremacist group on the Internet. Such groups will fester in darkness, but wither in light. Kenneth McVay, OBC The Nizkor Project - -- Nizkor Canada | http://www.nizkor.org - -----------------------| Prince Myshkin's Troll Bait Sold Here |-------------------------------------- http://www.nizkor.org/hweb/people/g/giwer-matt/ -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQBVAwUBMfVcLJNcNyVVy0jxAQENJQIAwGhMdP82l3S0Ac3uSaYaUcj1TowW6J9F K+l+wRmWwpQBuxq+INEPghNcY54thnc6iQIqGfR5KVoUsFNLAk9lcg== =12JK -----END PGP SIGNATURE-----

1 0

Re: Netscape
by shamrock＠netcom.com 17 Dec '03

17 Dec '03

At 1:23 7/23/96, Jeff Weinstein wrote: > If you know that the recipient can read a message encrypted with >3DES, IDEA, or RC2-128, then you can send the message using one of >these strong algorithms. Given that you need someones public key >to send them a message, there are several obvious ways to transmit >information about what algorithms they accept along with it. Granted. What about the signature bug? Will Netscape encrypt the outside signature? -- Lucky Green <mailto:shamrock@netcom.com> PGP encrypted mail preferred. Defeat the Demopublican Unity Party. Vote no on Clinton/Dole in November. Vote Harry Browne for President.

2 1

Re: Bare fibers
by jim bell 17 Dec '03

17 Dec '03

At 03:20 AM 7/23/96 -0400, Rabid Wombat wrote: > >> >> Doesn't that make it vulnerable (detectable) to Tempest attacks? >No. >Transmitting light via fiber doesn't emit EM. >Anyway, the original post, as I recall, was about keeping sensitive data >on a second hard drive, connected via (very thin, therefore harder to >notice) fiber. Tempest monitoring was not a factor. It occurs to me that a bare fiber could actually be (randomly) hung across treetops, roofs, power lines, and various other structures, over a many-block distance in suburban areas. Such a fiber wouldn't be protected very well, but it would probably last a few months. It would also be exceedingly hard to find its terminations, and tracing it would be a real pain. (It probably wouldn't be visible against a bright sky more than a meter or two away.) Jim Bell jimbell(a)pacifier.com

4 3

RE: Brute Force DES
by Matt Thomlinson 17 Dec '03

17 Dec '03

well, first of all, I wasn't computing 2^40, I was computing something like 2^(56*0.667) * 8bytes/each or somesuch, I don't recall exactly as I don't have the paper in front of me. In any case, it was the tradeoff between time and space -- 2^40 in time was something like 2^38th in space. But don't trust my numbers, get the paper and rattle them off yourself. It just seems like if we're going to browse all of the way through the DES keyspace, we _ought_ to take notes along the way -- that means building a table. I don't care how big the table is; if it is only 2^30 entries (about 1G entries, each 8 bytes = 8Gig) we reduce our next DES crack by a factor of 8. If we keep 2^31 entries (16Gig) we can cut it down to 2^50, or a factor of 16. If we have 4 - 9 Gig drives (or perhaps three drives and some wiggling, described below) we can save about 2^32 entries and the search becomes a measly 2^48. :) To whittle this down to a 40-bit workload, we'd have to save 2^36 entries* 2^8 bytes/entry = 2^39 Bytes = 512 Gig. Yes, admittedly large. What's the cheapest form of storage, magtape? How much can you store on magtape? The entries can be sorted so that lookup doesn't take long even when you have to mount tapes. Wiggling: We may be able to save less than 2^8 bytes/entry because we know the quality that made the point interesting enough to save (say we only keep points where the top 32 bits were zeros -- no need to save these zero prefixes) but I suppose we'd only be able to cut this storage factor down by a factor of two at most. If you don't have Hellman's paper handy, the apropos formulas are: Total time=T, memory=M, search space=N Time/space tradeoff: M=mt , T=t^2, m*t^2=N In our case, N=2^56; M, T variable. mattt >---------- >From: Ray Arachelian[SMTP:sunder@dorsai.dorsai.org] >Sent: Tuesday, July 23, 1996 3:16 PM >To: Matt Thomlinson >Cc: 'trei(a)process.com'; 'cypherpunks(a)toad.com' >Subject: RE: Brute Force DES > >On Tue, 23 Jul 1996, Matt Thomlinson wrote: > >> why not put together (a LOT of) disk space and we can build a table >> (read: "a cryptanalytic time-memory tradeoff") for cracking DES? Using >> the table, we could brute-search the DES keyspace in less time than it >> would take to do an exhaustive search of a 38 bit keyspace, according to >> the paper. 4 gigs is what, a couple of hundred nowadays? >> >> Making DES equivalent to a 40-bit crack would take approx. 500Gig, but >> publishing the table would push DES out usefulness. Certainly we could >> scale back (make DES equivalent to a 45-bit crack?) if we don't have >> enough disk... > >IMHO it's more expensive to go this route than to build a machine with >dedicate DES cracking chips. 2^40 = 1,099,511,627,776 or about 1 terabyte >worth of space, not 500G. 2^39 would be 500Gb. A 4Gb drive these days is >$800, hardly a couple of hundred dollars. :) > >Even so, that's a ton of hard drives. Further you need machines to hook >these drives up to. Infact, you need a farm of machines. Why? You can >only put 7 on a chain, and maybe if you're lucky four chains in a machine >using four controllers. > >A better idea might be to make small cheap computers, say based on 8086's >or 68000's that replace the drive's controller card, or if that drive >controller card is intelligent enough to be a CPU or contain one, burn >EEPROMs. Have the EEPROMs be able to generate DES (or any other >cypher's) keyspace given a range, and then have them able to search the >whole drive for a match. > >Even so, if you build these drive boxes, all you've accomplished is to >create a nice huge big searchable array. You still will need some sort >of logic to figgure out when it finds the right key, and you still can't >do 3DES or recusively encrypted files, nor know when you've found the key >for data you can't recognize - or rather have these drives recognize. > >However: Reading a 4Gb drive end to end takes less than 2 hours. I know >this because I have a RAID array of them, and it takes 2 hours to >rebuild, so since rebuilding an array requires reading from two drives >and writing to one drive, reading a whole 4Gb drive at full speed would >be something like maybe 1 to 1.5 hours(???) > >You might be better off with 9.0Gb drives if you can afford them because >you then have less controller logic cards to build. > >The drives alone will cost $204,800. $800*(2^40/(4*1024*1024*1024)). >You could get a nice big discount if you buy that many, but this will >also mean however much it will cost you to build the cpu cards for >multiplied by 256 drives, plust the R&D cost, plus the network connection >between all the CPU boards. > >At that point you also run into the MTBF of the drives which means that >your drives will fail quite often. > >If you want to go dirt cheap on the CPU's while using this huge space >method, you could just buy something like 37 Mac IIsi's, hook each up to 7 >of the drives (you'll have to partition the drives as they won't support >volumes that huge.) and network the machines using localtalk. You won't >need a faster connection because all you need for networking is keyspace >distribution and success reporting. But then IIsi's are sloooow machines >and your searches will suffer a hit from the lack of the machine's speed, >plus all the overhead of having an operating system and using the SCSI >chain to talk to the drives. > >IIsi's go for $350-$500 nicely loaded... $14800 for 37 at $400 a pop, >add the drives to that, plus the cost of writing the program and hooking >all of this crap together and that'll be $219,600. Ya got that kinda >dough to spare? > >========================================================================== > + ^ + | Ray Arachelian |FL| KAOS KERAUNOS KYBERNETOS |==/|\== > \|/ |sunder(a)dorsai.org|UL|__Nothing_is_true,_all_is_permitted!_|=/\|/\= ><--+-->| --------------- |CG|What part of 'Congress shall make no |=\/|\/= > /|\ | Just Say "No" to|KA|law abridging the freedom of speech' |==\|/== > + v + | Janet Reno & GAK|AK| do you not understand? |======= >===================http://www.dorsai.org/~sunder/========================= > Key Escrow Laws are the mating calls of those who'd abuse your privacy! >

1 0

Re: DES-Busting Screen Savers?
by Bill Stewart 17 Dec '03

17 Dec '03

Tim and others have discussed the effectiveness of random search vs. centralized servers, problems of cheating, scaling, etc. My take is that, if you can ignore scaling, the best approach is probably to have a central server that doles out keyspace and wraps around when it reaches 100%, and doesn't worry too much about collecting results - even if there are cheaters, machine failures, etc., and people don't finish their keyspace, it'll be more likely to cover the whole space than randoms. (Make it a web page, and use cut&paste to transfer to the search programs so they don't need to be network-equipped.) To support scaling, make it easy for people to run subset servers; grab a chunk of keyspace from the main server and dole it out to people who ask you for it. If you want to get fancy, hack a DNS server to allow people to register their machines as NNN.descrack.org, 0<=NNN<1000, so that people can find subsets without having to ask the main server. >>> a brute force "screen saver" for >-- By the time the total amount of computons expended has equalled the >amount that would have been expended in a "no duplications" allocated >search, the Poisson probability distribution says that 1/e = 36.8% of the >keyspace will not have been searched; the rest of the probabilty lies in >keyspace searched once, twice, three times, etc. # Thanks; Bill # Bill Stewart +1-415-442-2215 stewarts(a)ix.netcom.com # http://www.idiom.com/~wcs # Confuse Authority!

2 1

re:SHI_fty
by Alan Olsen 17 Dec '03

17 Dec '03

At 08:13 PM 7/22/96 -0800, Vinnie Moscaritolo wrote: >Uh I think apple had that for a while, it's called cyberdog >http://cyberdog.apple.com screw this ole fud. Unfortunatly, this is only available for Macs at this time. Hopefully we will see Windows and/or Unix versions sometime soon. (Of course if you expect Microsoft PR to resemble *ANYTHING* vaguely like reality...) --- Alan Olsen -- alano(a)teleport.com -- Contract Web Design & Instruction `finger -l alano(a)teleport.com` for PGP 2.6.2 key http://www.teleport.com/~alano/ "We had to destroy the Internet in order to save it." - Sen. Exon "Microsoft -- Nothing but NT promises."

1 0

RE: Distributed DES crack
by pjn＠nworks.com 17 Dec '03

17 Dec '03

> I've a few machines around that could be dedicated almost full time to the > task. What are the bandwidth requirements? Specifically, could the > keycracker be run over a 28.8 (with a 486 running linux)? If so, how many > 486's could I get over a single 28.8 (i.e. 28.8 -> multiple 486's daisy > chained with ppp over direct serial connection)? > It's not a factor of the bandwidth, you search offline and send in > your results to a central server. > But first, a little reality check is in order. According to libdes, > the 200Mhz Pentium Pro on my desk will do 1,827,997 ECB bytes/sec, or > 228,499 ECB blocks. A DES crack would have to try, on average, 2^55 > blocks. That would take my machine 43,798,875 hours, or 1,824,953 > days. > OK, so let's be reasonable and say that a week would be a good time to > come up with a DES key. We would need 260,707 200Mhz Pentium Pro's to > achieve this. > Looking at that, 30 days seems not such an unreasonable target. We > would need 60,831 200Mhz Pentium Pro's to achieve this. > It seems obvious to me that DES is still *way* out of reach of > anything other than special purpose hardware. In> - Andy (hoping he got his sums right) (I think you did :) One small thing... You are assuming that we will not get the right equasion/code/whatever untill the very end. There is a good chance that it will only take half that time...and a slim chance that we will get it with the first try... P.J. pjn(a)nworks.com (BTW, I have a 486 DX/4 100MhZ That I will put to the effort...) ... As easy as 3.14159265358979323846264338327950288419716 ___ Blue Wave/QWK v2.20 [NR]

1 0