Re: Wiretap Act Does Not Cover Message 'in Storage' For Short Period(was Re: BNA's Internet Law News (ILN) - 2/27/03)
Will Rodger wrote: John says: Wireless is a horse of a different color. IANAL but the last time I looked, there was no federal law against intercepting most wireless signals, but you were (generally) not allowed to disclose the contents to anyone else. No longer, if it ever was. It's a crime, as evidenced by the wireless scandal a few years back when some Democrat partisan intercepted communications of Republican leadership in Florida, then talked. The simple act of interception was illegal. Next time, before disagreeing with someone: a) Please read what he actually wrote, and b) Don't quote snippets out of context. Three sentences later, at the end of the paragraph that began as quoted above, I explicitly pointed out that cellphone transmissions are a more-protected special case. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: Wiretap Act Does Not Cover Message 'in Storage' For Short Period
Tim Dierks wrote: In order to avoid overreaction to a nth-hand story, I've attempted to locate some primary sources. Konop v. Hawaiian Airlines: http://laws.lp.findlaw.com/getcase/9th/case/9955106pexact=1 [US v Councilman:] http://pacer.mad.uscourts.gov/dc/opinions/ponsor/pdf/councilman2.pdf Well done. Thanks. I'd be interested in any opinions on how this affects the government's need to get specific wiretap warrants; I don't know if the law which makes illicit civilian wiretapping illegal is the same code which governs the government's ability (or lack thereof) to intercept communications. 0) IANAL. But as to the question of same code, the answer is clearly no. 1) As to government-authorized intercepts, see http://www.eff.org/Privacy/Surveillance/Terrorism_militias/20011031_eff_usa_patriot_analysis.html which gives a plain-language discussion of at least eight different standards under which some sort of authorization could be obtained. Also note that neither Konop nor Councilman involved government intercepts, so you can't learn anything about authorized intercepts by studying them. Also note that post-9/11 laws have superseded everything you might previously have known on the subject. 2) As to intercepts by civilians, it's wrong, and it may be punishable under many different theories and standards, including invasion of privacy, copyright infringement, computer trespass, computer vandalism, simple theft of things of value, and who-knows-what else. 3) As to unauthorized intercepts by government agents, in theory it is exactly the same as item (2), but in practice your chance of seeing anybody punished for it is comparable to your chance of seeing a State Trooper ticketed for speeding, tailgating, weaving, and failing to signal turns enroute to the donut shop. They're doing God's work, you know; why should mere laws and bills of rights apply to them? About the best you can realistically hope for is the exclusionary rule (illegally siezed evidence can't be used against you) but I wouldn't necessarily count on that. 4) Crypto-related sidelight: I wonder what would have happened if Konop had encrypted his sensitive data. (eBook format or the like. :-) Then could he have used the draconian provisions of the DMCA against his opponent (Hawaiian Airlines)? - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Columbia crypto box
As reported by AP: | Among the most important [debris] they were seeking was | a device that allows for the encryption of communication | between the shuttle and NASA controllers. A NASA spokesman | in Houston, John Ira Petty, said Friday that NASA feared | the technology could be used to send bogus signals to the | shuttle. Apparently some folks skipped class the day Kerchhoffs' Principle was covered. One wonders what other shuttle systems were designed with comparable disregard of basic principles. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: Patents as a security mechanism
Matt Blaze wrote: Patents were originally intended, and are usually used (for better or for worse), as a mechanism for protecting inventors and their licensees from competition. That's an oversimplification. Patents were originally intended as a bargain between the inventors and the society at large. Under the terms of this bargain, the inventors make public (which is the root meaning of patent) the details of the invention, rather than thereby advancing general knowledge and permitting follow-on inventions. In exchange the inventor was granted limited protection from competition. In the absence of a patent system, inventors will try to keep everything a trade secret, which is another way of fending off competition for a while. From society's point of view, patents are generally better than trade secrets. From the inventors' point of view, patents are generally better than trade secrets. So we have a mutually-beneficial bargain. Patents were originally intended to be a win/win proposition. Of course it is axiomatic that whatever you're doing, you can always do it wrong. We can debate whether the current system fulfills the original intention, but let's not go there right now. But I've noticed a couple of areas where patents are also used as a security mechanism, aiming to prevent the unauthorized production of products that might threaten some aspect of a system's security. OK. ... mechanical locks ... Many users actually prefer these patented products because even though it means they might have to pay monopoly prices for their keys, it makes it less likely that a thief will be able to get a duplicate at the corner hardware store. An interesting observation. I'm a bit skeptical about whether this really is effective So am I. (and at least one legal case, Best v. Ilco, casts some doubt on the validity of many of the key blank patents) It's amusing that Best had a utility patent and a design patent, both of which were held invalid (on different grounds). It is the design patent which I think speaks most clearly to the point Matt is making. http://www.law.emory.edu/fedcircuit/aug96/95-1528.html == One example close to home is the DVD patents, which, in addition to providing income for the DVD patent holders, also allows them to prevent the production of players that don't meet certain requirements. This effectively reduces the availability of multi-region players; the patents protect the security of the region coding system. The following sounds like a nit, but I think it is more than that: I think it is the _CSS licenses_ rather than the DVD patents that play the role of protecting the region coding system and reducing the availability of multi-region players. This gets back to the bargain discussed above, because the CSS license is based, as far as I can tell, on trade secrets. No particular patents are mentioned in the CSS license forms I've seen; instead there is much mention of Highly Confidential Information. Perhaps a more important point is the economic angle. Let's re-examing the statement: Many users actually prefer these patented products We need sharper terminology. We need to unbundle the products; that is, we have a _lock_ product and a _key_ product. It is unsafe to assume that whoever buys the lock product is the same person who buys the key product. Whoever pays for the locks has a vested interest in high-security locks that open to as few keys as possible. Whoever pays for the keys, on the contrary, has a vested interest in keys that are extra-powerful and/or cheap and extra-widely available. Suppose some party Alice controls a restriction, such as a patent or trade secret. Alice will try to sell the restriction to the lock-buyer, Larry, who benefits directly from the security. Larry won't buy it unless he is convinced that Alice is willing and able enforce the restriction against key-makers and key-buyers such as Kathy. Are there other examples where patents are used as a security mechanism? Not that I know of. So we have a grand total of less than one valid examples. -- CSS depends on secrecy, which is by definition the opposite of patentcy. -- Best v. ILCO held that patenting key-blanks is an abuse of the design-patent law. I think this is as it should be. That's not the proper purpose of patent law. Of course if you ask about non-patent laws, there are many examples: -- in some jurisdictions it is illegal in general to carry lock picks. -- in some jurisdictions it is illegal in general to copy a key marked do not duplicate. -- copyright law is sort of a do not duplicate stamp protecting original creative works against certain types of duplication. -- DMCA makes it a federal criminal offence to circumvent triple-rot-13. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL
DeCSS, crypto, law, and economics
Regarding the acquittal of Jon Johansen, I quoted CNN as saying: The studios argued unauthorised copying was copyright theft and undermined a market for DVDs and videos worth $20 billion a year in North America alone. Some elements of the industry did indeed claim that, but such claims are grossly irrelevant, and to bring them up is foolish or dishonest. This case was never about unauthorized _copying_ of DVDs. You can make a bit-for-bit perfect copy of a DVD without decrypting it. Indeed it's easier to copy if you don't decrypt it! The main thing the industry really had at stake in this case is the zone locking aka region code system. The studios like to release videos in different parts of the world at different times, and to charge different royalty fees in different places. This is called market segmentation. The idea of an open-source player was abhorrent to them, because it makes it easy to buy a DVD in one region and play it in other regions. This is an example of arbitrage. For normal products, market segmentation is neither forbidden by law nor protected by law. Mushrooms that cost $4.00 per ounce at the supermarket can be purchased for $4.00 per pound at the Asian grocery down the street. The stores are free to charge whatever they like, and I am free to shop wherever I like. The law is silent on the issue. People who engage in market segmentation are always looking for ways to prevent arbitrage. For instance, airlines make sure tickets are non-transferable, to prevent some ticket agent from stocking up on tickets at excursion prices and reselling them to business travelers. Movie studios never had a really good market segmentation system, because -- I can legally own region-1 or region-4 DVDs or some of both, no matter whether I live in the US or Australia. -- I can legally own a region-1 or region-4 DVD player, or both, no matter whether I live in the US or Australia. To be clear: the industry was never able to erect a legal barrier to arbitrage of disks _or_ arbitrage of players. The closest they could come was to make it slightly hard to get a _multi-region_ player. The manufacturers of player hardware had to do the studios' bidding because of the the controversial (to say the least) anti-circumvention provisions of the 1998 DMCA law. If we somewhat charitably assume the studios knew what they were doing, their whole market segmentation scheme was predicated on the lack of multi-region players _and_ on the assumption that players would remain sufficiently expensive that users couldn't just buy a stack of players, one per region. Less charitably the scheme was predicated on the foolish assumption that nobody would ever discover the possibility of inter-region arbitrage of player hardware. I repeat, the practical issue in this case was never about cheating the studios out of their per-disk royalties on DVDs. At this point you might be wondering about per-player royalties. First, let's dispose of an irrelevant side-issue. The rights to patents on raw DVD hardware are held by a consortium of hardware companies, not movie studios. These people presumably collected their cut when Mr. Johansen purchased his raw DVD drive hardware. So this case was never about patent infringement. The studios arguably hold intellectual property rights in the CSS decoding keys, and they can collect per-player royalties from hw mfgrs who incorporate such keys in their products. AFAIK Mr. Johansen never copied any such key (or even had one he could have copied), so this case was never about illegal copying even on a per-player basis. The truly amazing thing about this case is that the crime would not have occured if the studios had used decently-strong crypto. It's ironic that in an age when for cryptographers enjoy a historically-unprecedented lopsided advantage over cryptanalysts, the industry adopted a system that could be cracked by amateurs. This probably wasn't simply due to stupidity in the industry; it is more plausibly attributed to stupidity in the US export regulations which induced the industry to use 40-bit keys. So what we have here are remarkably intrusive laws: under US regulations the crypto must be easy to break, while under US law it is illegal to break it. The latter is dressed up as a copyright law even if no illegal copying is involved. This strikes me as analogous to requiring everyone to use pin/tumbler locks with only a single pin, so that all locks can be picked using a popsicle stick, and then arresting people for burglary whenever they are caught carrying a popsicle stick. US law is not the same as Norwegian law. You should not imagine that this case sets a precedent for US courts. Additional remarks: We should try to avoid overwrought arguments about the morality of market segmentation and/or arbitrage. Producers and retailers will always try to benefit themselves by segmenting the market; consumers and arbitrageurs will always try to
Re: did you really expunge that key?
1) This topic must be taken seriously. A standard technique for attacking a system is to request a bunch of memory or disk space, leave it uninitialized, and see what you've got. 2) As regards the volatile keyword, I agree with Perry. The two punchlines are: if, for example, gcc did not honor [the volatile keyword], the machine I am typing at right now would not work because the device drivers would not work. If they haven't implemented volatile right, why should they implement the pragma correctly? 3) However, a discussion of compilers and keywords does not complete the analysis. A compiler is only part of a larger system. At the very least, we must pay attention to: -- compiler -- operating system -- hardware architecture -- hardware physics At the OS and hardware-architecture levels, note that a device driver accesses a volatile device register only after beseeching the OS to map the register to a certain address in the driver's logical address space. In contrast, for some address that points to ordinary storage, the OS and the hardware could (and probably do) make multiple copies: Swap space, main memory, L2 cache, L1 cache, et cetera. When you write to some address, you have no reason to assume that it will write through all the layers. Swap space is the extreme case: if you were swapped out previously, there will be images of your process on the swap device. If you clear the copy in main memory somehow, it is unlikely to have any effect on the images on the swap device. Even if you get swapped out again later (and there's no guarantee of that), you may well get swapped out to a different location on the swap device, so that the previous images remain. The analogy to device drivers is invalid unless you have arranged to obtain a chunk of memory that is uncacheable and unswappable. To say the same thing in other words: a compiler can only do so much. It can generate instructions to be executed by the hardware. Whether that instruction affects the real physical world in the way you desire is another question entirely. 4) In the effort to prevent the just-mentioned attack, a moderately-good operating system will expunge memory right before giving it to a new owner. It would be more secure (but vastly less efficient) to expunge it right after the previous owner is finished with it. To see this in more detail, consider swap space again: a piece of used swap space need not be expunged, unless you are fastidious about security, because the operating system knows that it will write there before it reads there. Clearing it immediately would be a waste of resources. Leaving it uncleared is potentially a security hole, because of the risk that some agent unknown to the operating system will (sooner or later) open the swap-space as a file and read everything. 5) We turn now to the hardware-physics layer. Suppose you really do manage to overwrite a disk file with zeros. That does not really guarantee that the data will be unrecoverable. As Richard Nixon found out the hard way, the recording head never follows exactly the same path, so there could be little patches of magnetism just to the left and/or just to the right of the track. An adversary with specialized equipment and specialized skills may be able to recover your data. 6) To reduce the just-mentioned threat, a good strategy is to overwrite the file with random numbers, not zeros. Then the adversary has a much harder time figuring out what is old data and what is new gibberish. (To do a really good job requires writing your valuable data always in the middle, and overwriting gibberish twice, once offset left and once offset right.) This is one of the reasons why you might need an industrial- strength stretched random symbol generator: http://www.monmouth.com/~jsd/turbid/paper/turbid.htm#sec-srandom Note that the random-number trick can be used for main memory (not just disks) to ensure that the compiler + OS + hardware system doesn't optimize away a block of zeros. This actually happened to me once: I was doing some timing studies, and I wanted to force something out of cache by making it too big, so I allocated a large chunk of memory and set it to zero. But no matter how big I made it, it fit in cache. The system was using the memory map to give me unlimited copies of one small page of zeros (with the copy-on-write bit set). 7) Terminology: I use the word expunge to denote doing whatever is necessary to utterly destroy all copies of something. Clearing a memory location is sometimes far from sufficient. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: Optical analog computing?
R. A. Hettinga wrote: ... the first computer to crack enigma was optical the first synthetic-aperture-radar processor was optical but all these early successes were classified -- 100 to 200 projects, and I probably know of less than half. -- Do these claims compute?! is this really a secret history, or does this mean holography, of am I just completely out of the loop?1 Gimme a break. This is remarkable for its lack of newsworthiness. 1) Bletchley Park used optical sensors, which were (and still are) the best way to read paper tape at high speed. You can read about it in the standard accounts, e.g. http://www.picotech.com/applications/colossus.html 2) For decades before that, codebreakers were using optical computing in the form of superposed masks to find patterns. You can read about it in Kahn. 3) People have been doing opto-electronic computing for decades. There's a lot more to it than just holography. I get 14,000 hits from http://www.google.com/search?q=optical-computing Optical info is a complex-valued wave (spatial frequency, amplitude and phase) It isn't right to make it sound like three numbers (frequency, amplitude, and phase); actually there are innumerable frequencies, each of which has its own amplitude and phase. lenses, refractions, and interference are the computational operators. (add, copy, multiply, fft, correlation, convolution) of 1D and 2D arrays and, of course, massively parallel by default. and, of course, allows free-space interconnects. Some things that are hard with wires are easy with light-waves. But most things that are easy with wires are hard with light-waves. Here's a commercialized effort from israel: a space integrating vector-matric multiplier [ A ] B = [ C ] laser- 512-gate modulator - spread over 2D 256 Teraflop equivalent for one multiply per nanosecond. People were doing smaller versions of that in the 1980s. Unclassified example: acousto-optic spectrometer, 500 Gflops equivalent (for 12 watts!) doing continuous FFTs. Launched in 1998 on a 2-year mission. Submillimeter wave observatory. Not FFTs. FTs. Fourier Transforms. All you need for taking a D=2 Fourier Transform is a lens. It's undergrad physics-lab stuff. I get 6,000 hits from: http://www.google.com/search?q=fourier-optics Of course, the rest of the talk is about the promise of moving from optoelectronic to all-optical processors (on all-optical nets with optical encryption, so on). All optical??? No optoelectronics anywhere??? That's medicinal-grade pure snake oil, USP. Photons are well known for not interacting with each other. It's hard to do computing without interactions. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: Quantum computers inch closer?
AARG!Anonymous wrote: The problem is that you can't forcibly collapse the state vector into your wished-for eigenstate, the one where the plaintext recognizer returns a 1. Instead, it will collapse into a random state, Sorry, that's a severe mis-characterization. David Honig wrote: I thought the whole point of quantum-computer design is to build systems where you *do* impose your arbitrary constraints on the system. David Wagner wrote: Look again at those quantum texts. That's good advice. Quantum doesn't work like the original poster seemed to wish it would; state vectors collapse into a random state, Random is not the right word. not into that one magic needle-in-a-haystack state you wish it could find. C'mon folks, let's cut down on extreme statements like the-whole-point-is-this or the-whole-point-is-that and using words like magic to describe finding the right answer. 1) Computer design has many points that must be taken into consideration. Quantum computer design is in some ways more powerful but in other ways more constrained than classical computer design. 2) One of the points is that yes, the computer should compute what you want it to compute. OTOH it takes more than wishing to bring such a computer into existence. 3) A sufficiently well designed quantum computer can, in principle, find some needles in some haystacks, precisely because the structure of the machine, acting according to the laws of quantum mechanics, does in fact collapse the wave-function into a representation of the wished-for answer. (PS most of what has been written about collapse of wave-functions is baloney, but we need not pursue that tangent just now.) = A general remark about parallel computing: For every parallel algorithm (running on P processors) there exists a corresponding uniprocessor algorithm: just set P=1 and turn the crank. The converse does not hold. The existence of a uni- processor algorithm may or may not be a guide to the creation of a parallel algorithm. As Brooks famously said, creating a baby requires nine months, no matter how many mothers are assigned to the task. The same applies even more strongly to quantum computing: It would be nice if you could take a classical circuit, automatically convert it to the corresponding quantum circuit, with the property that when presented with a superposition of questions it would produce the corresponding superposition of answers. But that cannot be. For starters, there will be some phase relationships between the various components of the superposition of answers, and the classical circuit provides no guidance as to what the phase relationships should be. So let's not guess about what quantum algorithms exist. It is possible to construct such algorithms, but it requires highly specialized skills. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: get a grip on what TCPA is for
bear wrote: ... I have one box with all the protection I want: it's never connected to the net at all. I have another box with all the protection that I consider practical for email and web use. Both run only and exactly the software I have put on them, That is trusted computing sir, and TCPA/Palladium is a huge step *backward* from it. Brother Bear belabors one obvious point while missing a more-important obvious point. What some people want is not what other people want. The TCPA/Pd designers don't much care whether the person who has custody of the machine trusts it. They've been shipping untrustworthy software for years. The thing they care about, probably the only thing they deeply care about, is whether _they_ can trust the machine while it is in _somebody else's_ custody. To a first approximation, TCPA/Pd is for !!their!! direct benefit, not for yours. But to a second approximation, they are not entirely wrong when they say consumers will benefit, because there are indirect benefits of having some sort of system whereby authors, performers, and inventors get paid for their work. Things that are simply not available now would become available if there were a way people could get paid for creating them. You can wish for some Land of Cockaigne where you get paid but nobody has to do any paying, but that's a long way from reality. == Most of us know how to secure a machine that is disconnected from the net. We can probably even combine some limited networked functionality with some degree of security -- !!provided!! we retain physical custody of the machine. But how to trust a machine when you don't have physical custody? Even the most-skilled members of this list would find that a challenge (depending, as I have emphasized before, on what your threat model is). I guarantee you will not understand TCPA/Pd unless you walk a while in the proponents' moccasins. If you can't stand the smell of those moccasins, OK, but prepare yourself for perpetual ignorance and irrelevance. For example: Imagine you are the owner of a valuable copyright and you want to protect it. You want consumers to be able to use your work in some ways, but you want to prevent rampant infringement. What will you do??? It's not an easy problem. If your powers of imagination are not up to the task in the previous paragraph, here's an alternative: Suppose you want to spend a few weeks visiting Outer Zambonia, but you want to communicate securely with your colleagues back home during this time. Alas, the Zambonian Ministry of Friendship has been looking forward to this as an opportunity to trojanize your laptop. You simply don't have the resources to guard your laptop 24 hours a day. You can't travel with a GSA-approved safe in your carry-on. You can't take your laptop with you when you go swimming. The idea of hardware with !!some!! degree of tamper-resistance might eventually start to appeal to you. Of course, our task of understanding what TCPA/Pd is trying to do is made more difficult when proponents lie about what they are trying to do. === The most interesting technical point AFAICT is figuring out how to _vet_ a piece of tamper-resistant hardware. Presumably you want it to detect the early stages of tampering and react by expunging all its private keys. Alas essentially identical behavior could be used to cover the tracks of built-in trojan beasties. Here are some partially-baked thoughts: 1) You have to allow it to expunge things. That's the only way it can really protect your secrets. 2) So allow that. It should be possible to verify that the box is in a tabula-rasa state -- if the trojan is gone, it's gone, and if it's not gone, it should be detectable if you probe hard enough. We require the hardware to allow certain types of probing. 3) After you're satisfied that the hardware is not infested, load the software, and the keys, from a trusted source. Replace the tamper-evident seals and latches. This isn't a complete design, but you can where it's going: It should be possible to design hardware with some degree of tamper-resistance !!without!! creating a monopoly as to who decides who trusts whom. Alas it is also possible to design the hardware so that it becomes a monopoly-enhancer of Orwellian proportions. We need to be vigilant to prevent this. This will require nuanced, non-extremist thinking. Those who exhibit the knee-jerk response that all tamper-resistant hardware is bad will be ignored. Such hardware, like most things, can be used for good or ill, depending on details. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: Translucent Databases
David Wagner wrote: It seems to me that a much more privacy-friendly solution would be to simply refrain from asking for sensitive personal information like SSN and date of birth -- name and a random unique identifier printed on the application form ought to suffice. (If SSN is later needed for financial aid purposes, it could be requested after the student decides to matriculate.) Am I missing anything? I think the problem is a lot harder than that. Let me clarify by telling a story: Once upon a time, Hansel designed an online-forms system that collected credit-card info, encrypted it using PGP, and mailed it to Goldylocks (the secretary) with a backup copy going to Tweedledee. Despite the fact that Hansel had installed PGP on her computer and indoctrinated her on how to use it, Goldylocks was unable to decrypt the info. So at her request, Tweedledee decrypted it -- a whole conference's worth of registrations -- and sent it to her in the clear. In a clear violation of Murphy's law, no harm came of this, but otherwise it was a worst-case use of cryptology: just secure enough to be a nuisance to the authorized users, but in the long run providing no real protection for the card- holders. The sad fact is that most people on this planet cannot get PGP to work in a way that suits them. The future of security depends at least as much on user-interface research as it does on mathematical cryptology research. Oh, BTW, a preprinted number on the admissions form doesn't really do the trick. Forms are printed on printing presses, in batches of several thousand, all alike. After they are mailed out, the guidance counselor at Podunk South High School will make copies as needed. A web-based approach won't work unless you are making computer-savviness an entrance requirement. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
David Wagner [EMAIL PROTECTED] writes: I don't know of any good cryptographic hash function that comes with a proof that all outputs are possible. What about the scheme Pad - Encipher - Contract described at http://www.monmouth.com/~jsd/turbid/paper/turbid.htm#sec-uniform-hash if we apply the Luby-Rackoff construction (i.e., 3 rounds of a Feistel cipher), with ideal hash functions in each round, does this have the desired properties? It might. Paul Crowley wrote: This seems to define a block cipher with no key, which is collision free but not one-way. Am I misunderstanding what you're proposing? David Wagner wrote: You understood it perfectly. Good point. I didn't notice that problem. Harrumph. There is only the most minor of problems here, namely that DAW mentioned a symmetric cipher. The problem goes away if you use asymmetric crypto. You want a cipher with no _deciphering_ key, as described in my paper. (op. cit.) - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
1) There were some very interesting questions such as -- whether one can construct a hash function that generates all possible codes. -- ditto, generating them as uniformly as possible. -- Whether off-the-shelf hash functions such as SHA-1 have such properties. The answers are respectively yes, yes, and very probably. I wrote up a discussion of this, with examples, at http://www.monmouth.com/~jsd/turbid/paper/turbid.htm#sec-uniform-hash 2) David W. suggested (off-list) that I clarify the relationship of entropy-based information-theoretic arguments to computational- feasibility arguments. I took some steps in this direction; see http://www.monmouth.com/~jsd/turbid/paper/turbid.htm#sec-objectives - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
Barney Wolff asked: Do we even know that the popular hash functions can actually generate all 2^N values of their outputs? David Wagner replied: It seems very unlikely that they can generate all 2^N outputs (under current knowledge). I was temporarily astonished, but he clarified as follows: Q2: If we cycle through all messages (possibly very long or very short), are all 2^N output values possible? ... I'd guess that the answer to Q2 is probably Yes, or close to it. 1) Consider the following function H0: Divide the input into chunks N bits long. Calculate the XOR of all the chunks, and use that as the output. This meets the definition of hash function, although it would not be a one-way hash function. And it would most certainly be capable of generating all 2^N possible outputs. 2) I can't prove that a standard hash function such as SHA1 generates all possible codes, but I consider it likely. It would be quite shocking if a strong hash function such as SHA1 generated fewer codes than a weak function such as H0. 3) For a one-way hash function should not expect a _constructive_ proof that it generates all possible codes; such a construction would violate the one-way property. 4) Here is a rough plausibility argument. Consider two hash functions H_1 and H_2 that are independent. We define H'(x) := H_1(x) XOR H_2(x) (1) which implies H_1(x) = H'(x) XOR H_2(x) (2) Now let's look at the truth table for equation (2), where the row-index is the H' code, the column-index is the H_2 code, and each entry represents the H_1 code required to uphold the equation: 00 01 10 11 00 00 01 10 11 01 01 00 11 10 10 10 11 00 01 11 11 10 01 00 Now let's suppose H_2 is missing one code (say the 10 code) and H' is missing one code (say the 11 code). Then H_1 must be missing at least three codes! Otherwise there would be a way of combining a non-missing H_1 code with a non-missing H_2 code to create the missing H' code. 00 01 10m 11 00 00 01 10 11 01 01 00 11 10 10 10 11 00 01 11m 11m 10m 01 00m We can extend this argument by combining lots and lots of independent hash functions H_i. The combination has far fewer missing codes than any of the ingredients. So either you conclude that a) there is a conspiracy that prevents us from constructing independent hash functions to use as ingredients, or b) we can produce hash functions with very few missing codes. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
I wrote: a) if the hash function happens to have a property I call no wasted entropy then the whitening stage is superfluous (and you may decide to classify the hash as non-simple); David Honig responded: Not wasting entropy does not mean that a function's output is white ie uniformly distributed. E.g., a function which doubles every bit: 0-00, 1-11 preserves total entropy but does not produce uniformly distributed output. (It also reduces entropy/symbol.) That's a red-herring tangent. I'm not talking about any old function that doesn't waste entropy; I'm talking about a !!hash!! function that doesn't waste entropy. The hash function has a hard constraint on the word-size of its output. If it starts doubling bits, or otherwise putting out redundancies, then it is wasting entropy. See the discussion of the BADHASH-2 function in the paper. http://www.monmouth.com/~jsd/turbid/ And remember: in addition to having a non-entropy-wasting hash function, we are also required to saturate its input. Then we can conclude that the output is white to a very high degree, as quantitatively discussed in the paper. The simple-hash's --lets call it a digest-- function is to increase the entropy/symbol by *reducing* the number of symbols while preserving *total* entropy. Total entropy is preserved in the non-saturated regime. This is documented in upper rows of the table: http://www.monmouth.com/~jsd/turbid/paper/turbid.htm#tab-saturation In the saturated regime, some entropy is necessarily lost. This is documented in the lower rows of the table. This is only a small percentage, but it is mandatory. I don't consider this to be wasted entropy; I consider it entropy well spent. That is, these are necessary hash collisions, as opposed to unnecessary ones. A function like xor(bit N, bit N+1) which halves the number of bits can do this. While its output is whiter than its input, its output will not be perfectly white unless its input was. See the discussion of BADHASH-1 in the paper. Two scenarios where SHA is contraindicated came up: 1. hardware 2. interrupt handlers Sometimes a hardware RNG will feed raw data into a crypto-WEAK analogue of SHA, ie a LFSR which does the bit-reduction and mixing functions, but doesn't mix as well as SHA. (LFSR are hardware-friendly) Separating the bit-reduction from the mixing can be useful for analyzing what's really going on. Well, I tend to agree that systems that separate the bit-reduction from the mixing are easier to analyze, in the sense that it is easier to find flaws in them. But that's because they're so flawed! - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
Amir Herzberg wrote: So I ask: is there a definition of this `no wasted entropy` property, which hash functions can be assumed to have (and tested for), and which ensures the desired extraction of randomness? That's the right question. The answer I give in the paper is A cryptologic hash function advertises that it is computationally infeasible for an adversary to unmix the hash-codes. A chosen-plaintext (chosen-input) attack will not discover inputs that produce hash collisions with any great probability. In contrast: What we are asking is not really very special. We merely ask that the hash-codes in the second column be well mixed. We ask that the data acquisition system will not accidentally produce an input pattern that unmixes the hash-codes. We believe that anything that makes a good pretense of being a cryptologic hash function is good enough for our purposes, with a wide margin of safety. If it resists attack when the adversary can choose the inputs, it presumably resists attack when the adversary can't choose the inputs. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
David Honig helped focus the discussion by advocating the block diagram: Source -- Digitizer -- Simple hash -- Whitener (e.g., DES) Let me slightly generalize this to: ! Source -- Digitizer -- hash -- Whitener (e.g., DES) i.e. we defer the question of whether the hash is simple or not. I continue to claim that a) if the hash function happens to have a property I call no wasted entropy then the whitening stage is superfluous (and you may decide to classify the hash as non-simple); otherwise b) if the hash function does not have that property, this is a defective Random Symbol Generator and b1) the whitener will _at best_ conceal, not remove the defects, and b2) this is not the best way to conceal defects. Very definitely not. To illustrate my point, I will accept David's example of a simple-hash function; he wrote: Parity is the ultimate hash. Well, then, suppose that the raw data coming off my digitizer consists of an endless sequences of even-parity words. The words have lots of variability, lots of entropy, but the parity is always even. Then the output of the simple-hash is an endless sequence of zeros. I encrypt this with DES. Maybe triple-DES. It's not going to help. The generator is defective and doesn't even have satisfactory error-concealment. I like my design a lot better: + Source -- Digitizer -- good hash where I have chosen SHA-1 as my hash function. Finally, since SHA-1 is remarkably computationally efficient, I don't understand the motivation to look for simpler hash functions, especially if they are believed to require whitening or other post-processing. = Thanks again for the questions. This is a good discussion. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: understanding entropy (was: building a true RNG)
At 10:59 PM 7/22/02 -0700, [EMAIL PROTECTED] wrote: Entropy is not quite a physical quantity -- rather it is on the slippery edge between being a physical thing and a philosophical thing. If you are not careful, you will slip into a deep epistemic bog and find yourself needing to ask how do we know what is knowable, and what is the whichness of why? To avoid such deep waters, know where your entropy is coming from. Right. Then David Honig wrote: We agree on your substantive points re RNGs, I think, I join in the agreement. but you're interestingly wrong here. I don't think jamesd's point was wrong. One could quibble about some of the wording, especially if it were taken out of context, but the passage as a whole makes an important, valid point. Entropy is a physical quantity, it even figures into chemistry. Yes, it is a physical quantity. Yes, it enters into chemistry. But it also contains an element of subjectivity. For a careful discussion of what entropy is, including the element of subjectivity, see http://www.monmouth.com/~jsd/physics/thermo-laws.htm#sec-second-law The physics-of-computation people (Bennett? Landaur? etc) Charles H. Bennett and Rolf Landauer. have written about thermodynamics information. Not to mention Leo Szilard, Ed Fredkin, Wojciech Zurek, and others. Modulo Chaitin-type mindgames about measuring it :-) Chaitin's work is profound and well-regarded. Referring to it as mindgames is, well, nothing but name-calling and won't advance the scientific discussion. If anybody has a thoughtful objection to Chaitin's work I would be extremely interested to hear it. Anyway we're cryptographers, not philosophers, so we should be safe.. A lack of understanding of what entropy is has gotten more than one cryptographer into trouble. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG
Eugen Leitl wrote: ... framegrabber with a 640x480 24 bit/pixel camera. It doesn't compress, is rather noisy, and since self-adjusting I get the maximum entropy at maximum darkness. OK. Evidently it's dominated by thermal noise, not to be confused with the Poisson noise recently featured in another thread. Not a problem. Is there any point in compressing the video before running it through a cryptohash? There might be a minor point, namely computational efficiency. A well-chosen compressor might eliminate low-entropy bytes rather quickly. Make sure it's a lossless compressor, perhaps GIF or PNG ... as opposed to a perceptual coder (e.g. JPEG) that would persumably throw away some of the entropy. Calling SHA-1 on low-entropy bytes doesn't waste entropy, but wastes CPU cycles. How does e.g. SHA-1 fare with very sparse bitvectors? 1) In any good hash function, any input bit should have about as much effect on the output as any other input bit. SHA-1 has been analyzed by experts (of which I am not one :-) and I would imagine they checked this. 2) There are 5 one-bit shifts in the fivefold expansion, and lots of 5-bit shifts in the main loop, so it shouldn't matter that the sparse input bits are clustered in the bottom of the 32-bit words. 3) I performed an amateur kick-the-tires test, namely cobbling up some sparse input vectors, calling SHA-1, and applying standard statistical tests including Diehard and Maurer's universal statistical test. No nobody's surprise, the tests didn't detect anything. Arnold Reinhold wrote: ... with a portable TV set and a video digitizer should be a good source of high bandwidth noise. In both cases you are just using the receivers as high gain amplifiers of the thermal noise at the antenna terminals. Thermal noise is good. Antennas are bad -- just an invitation to be attacked that way. Get rid of the antenna. Keep the high gain preamp. Better yet, do as Eugen has done: Use a framegrabber !!without!! the portable TV set. No RF section at all. Plenty of entropy, lower cost, greater simplicity, and less vulnerability to attack. For that matter, an audio card (without microphone) produces more than enough entropy for most applications. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: It's Time to Abandon Insecure Languages
[EMAIL PROTECTED] wrote: Most security bugs reported these days are issues with application semantics (auth bypass, SQL injection, cross-site scripting, information disclosure, mobile code execution, ...), not buffer overflows. Really? What's the evidence for that? What definition of most are we using? One out of 20 doesn't count as most in my book. When I look at the reports for 2002 year-to-date, at http://www.cert.org/advisories/ there are 20 advisories. Depending on how you count multi-bug reports, it appears that 19 out of 20 involve buffer overflows and related issues -- things that could easily be prevented by using a language that has a built-in string type and automatic object management. Exotic languages are not required; C++ would make a huge impact. And of course in any language a modicum of skill and care is required; it's hard to make a language foolproof because fools are so ingenious. My evidence: http://www.cert.org/advisories/ 20- multiple, including writing out-of-bounds 19 buffer overflow 18 multiple, including buffer overflow 17 stack overflow 16 multiple, including stack overflow 15= DoS: internal consistency check 14 buffer overflow 13 buffer overflow 12- format string 11 heap overflow 10- format string 9 multiple, including buffer overflow 8 multiple, including buffer overflow 7- double free 6 multiple, including buffer overflow 5 multiple, including heap overflow 4 buffer overflow 3 multiple, including buffer overflow 2 buffer overflow 1 buffer overflow - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: It's Time to Abandon Insecure Languages
[EMAIL PROTECTED] wrote: This is more indicative of CERT's focus than the relative frequency of security issues. The fact that a large fraction of e-commerce merchants let you set the price for the goods you buy is in practice a larger threat than the widely publicized buffer overflows. Semantic security bugs in individual web sites do not rate highly enough on Cert's seismograph, but are in practice far more common. Interesting.. Earlier he wrote Most security bugs reported these days are issues with application semantics We are talking about _reported_ bugs. If CERT is not the right place to look for reports, please tell us where we _can_ find appropriate reports. I was trained as a scientist. I like to look at data. Listening to other people's summaries and conclusions is nice, too, but sometimes it pays off to take a look at the real data. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: building a true RNG (was: Quantum Computing ...)
David Honig wrote: The thread here has split into QM True Randomness and what do you need to build a true RNG... Yup. Specifically: The executive summary of the principles of operation of my generator is: -- use SHA-1, which is believed to be resistant to collisions, even under chosen-input attack. -- use it under conditions where the adversary cannot choose the input. -- the rest is just physics and statistics. Sure. There are many examples of this kind of generator, using physical sources from video'd lava lamps to radioactive decay (incl. semiconductor junctions, resistors, microphone, detuned FM radio cards). For the humor-impaired, let me point out that the lava lamp is a joke. What it conspicuously lacks is a proof of correctness -- that is, a nonzero lower bound on the entropy rate of the raw data. The lava could turn out to have a not-very-complicated periodic pattern. Secondarily, the pattern changes so slowly that there must be rather strict upper bounds on the entropy rate, small out of all proportion to the cost of the contraption. A detuned FM card is a bad idea, because it is just begging the opponent to sit next door with an FM transmitter. A microphone causes users to worry about privacy, and in any case doesn't add much beyond what you'd get with the same input circuitry open-circuited, i.e. everything except the microphone itself. Radioactive decay has a poor price/performance ratio, and isn't nearly as random as neophytes might think, when the data-acquisition hardware is taken into account. 2) Vetting a generator by trying to detect patterns in the output is like kicking the tires on a used car ... go ahead and do it if you want, but it is far from sufficient for establishing any reasonable standard of correctness. You can't vet a RNG by looking at its output, We agree. which is likely whitened anyway, Depending on what whitening means; see below. but you can gain confidence by looking at its design Yes! and measuring the entropy in the raw-physical-source derived bitstream. That's the point where I would like some more detail. If measuring means applying statistical tests, then I've never seen such measurements done in a way that is really convincing. Constructive examples would be welcome. Just saying Joe Schmoe applied all the tests he could think of and couldn't compress it more than XY% isn't going to convince me. I recommend _calculating_ the entropy from physics principles, rather than trying to measure the entropy using statistical tests. The calculation is based on a handful of macroscopic physical parameters, such as temperature, gain, and bandwidth. If the raw source has 1 bits/symbol (and it will), Commonly but not necessarily 1 bit/symbol. Depending on what you mean by symbol, a 24-bit audio card provides a low-cost counterexample. it'd be nice if a later stage distilled this to near 1 bit/symbol, before whitening. We need to be more specific about what the symbol alphabet is. If the symbols are ASCII characters, 1 bit per symbol is not nearly good enough. More importantly, I don't know what whitening means in this case. The output of a good distiller has virtually 100% entropy density, i.e. 8 bits per byte. I say virtually because perfection is impossible, but 159.98 bits in a 160 bit word ought to be good enough for most applications :-). I see no point in whitening the output of such a distiller. If whitening means encrypting the output of the distiller, I consider that just a more-complicated hash function ... just another few rounds. Of course, no one outside the box will know, since you're whitening, but it yields resistance to (albeit difficult) attacks (e.g., your hash turns out to be attackable). I assume that means know [that I'm using a distiller] Well, in principle nobody outside the box knows _anything_ about the mechanism (unless they read the documentation). One random symbol stream looks a lot like another :-). Attackers can always check to see whether the generator is broken or not. But if it's not broken, all they can do (from outside) is measure the output-rate. I also fail to see harm in measuring/monitoring entropy as the RNG operates. Certainly there's no harm. It's like kicking the tires on the used car. It gives some people a warm fuzzy, but it's far from sufficient for establishing any reasonable level of confidence. I recommend monitoring the aforementioned macroscopic (non-statistical) physical parameters, both to detect gross hardware failure and to detect attempted jamming. But that's very different from the traditional (and hereby deprecated) procedure of measuring the entropy using statistical tests. For lots more detail, see http://www.monmouth.com/~jsd/turbid/ - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL
Re: building a true RNG (was: Quantum Computing ...)
David Honig wrote yet another nice note: So work in a Faraday cage... Tee, hee. Have you ever worked in a Faraday cage? Very expensive. Very inconvenient. Depending on what whitening means; see below. You can imagine simple-hashing (irreversible compression) as distinct from whitening which is related to cryptographic strength, avalanche, mixing, etc. I'm not trying to be dense, but I'm totally not understanding the distinction here. The following block diagram is excellent for focussing the discussion, (thanks): Source -- Digitizer -- Simple hash -- Whitener (e.g., DES) OK, we have DES as an example of a whitener. -- Can somebody give me an example of a simple hash that performs irreversible compression of the required kind? -- Isn't the anti-collision property required of even the simplest hash? Isn't that tantamount to a very strong mixing property? If there's strong mixing in the simple hash function, why do we need more mixing in the later whitening step? -- What is meant by cryptologic strength? Strength against what kind of attack? If this means in particular the one-way property, why do I need it? I can understand why a !!pseudo!! random symbol generator needs the one-way property, to protect its internal state, but since my generator has no secret state to protect, why do I need any cryptologic properties other than mixing? -- BTW note that the term avalanche is probably not helpful to this discussion, because it is usually defined (see e.g. Handbook of Applied Cryptography) in terms of single-bit flips. The anti-collision property of the hash function demands resistance to multi-bit flips. In this view, SHA combines the compression (aka digestion) function with the crypto-strength whitening If you want to think of it that way, then we come to the same final state. I assume digestion means the same as distillation? You collect some representative raw data, and run a number of entropy measurements on that sample. You find 1bit/baud. In particular you have found an upper bound. To paraphrase Dykstra: testing can find an upper bound on the entropy density, but it can never find a lower bound. You run the data through an algorithm which produces fewer bits. You measure the entropy of the result. When successive (or 'stronger') runs measure at 1b/bd, you have distilled entropy from that sample. No, you have just reached the limits of your chosen number of entropy measurements. You cannot convince a skeptic that a new test, discovered tomorrow, will not greatly lower your new (1b/bd) upper bound. To use this in crypto, you'd put it through a whitener --feed blocks to DES-- for belts-and-suspenders assurance. And because you don't want someone looking through your simple-hashing logic back to your source. I say again that I don't need the one-way property. At each iteration, the input to the hash function is unrelated to all past and future inputs. We agree that the belt-and-suspenders approach is a standard way of achieving high reliability. It works in crypto and in many unrelated fields http://www.monmouth.com/~jsd/how/htm/misc.html#sec-layers-safety If you want to XOR the output of the HRNG with some nice PRNG, that will give excellent error-concealment in the case of a gross failure of one or the other. But (minor opoint) I recommend we continue to call a belt a belt, and call a suspender a suspender. Call a HRNG a HRNG, and call a PRNG a PRNG. Adding a strong one-way function to my HRNG seems like a total waste of CPU cycles. Once you put it through DES, anything (eg the integers) appears random. That's why you measure before whitening, if possible. We agree that measuring after whitening is pointless, given the current state of the art, namely that encryption algorithms are incomparably stronger than automatic measurement (pattern-finding) algorithms. I recommend _calculating_ the entropy from physics principles, rather than trying to measure the entropy using statistical tests. The calculation is based on a handful of macroscopic physical parameters, such as temperature, gain, and bandwidth. You measure because your model may have overlooked something. Measure what? Measure how? If I run diehard on my raw data it will tell me that the data has entropy density far less than 8 bits per byte -- but duh, I already knew that. Running standard compression algorithms (Lempel-Ziv) or whatever will give me an upper bound that is much, much, higher than my calculated lower bound -- so even if I've overlooked something or made an error in the calculation the measurement is not sensitive enough to catch it. The output of a good distiller has virtually 100% entropy density, i.e. 8 bits per byte. I say virtually because perfection is impossible, but 159.98 bits in a 160 bit word ought to be good enough for most applications :-). I agree with the first statement (by definition), I think in crypto you have to be
vulnerability in Outlook PGP plugin
http://www.eeye.com/html/Research/Advisories/AD20020710.html This vulnerability can be exploited by the Outlook user simply selecting a malicious email, the opening of an attachment is not required. ... [NAI] have released a patch for the latest versions of the PGP Outlook plug-in to protect systems from this flaw. Users can download the patch from: http://www.nai.com/naicommon/download/upgrade/patches/patch-pgphotfix.asp = By TED BRIDIS, Associated Press Writer WASHINGTON (AP) - The world's most popular software for scrambling sensitive e-mails suffers from a programming flaw that could allow hackers to attack a user's computer and, in some circumstances, unscramble messages. The software, called Pretty Good Privacy, or PGP, is the de facto standard for encrypting e-mails and is widely used by corporate and government offices, including some FBI ( news - web sites) agents and U.S. intelligence agencies. The scrambling technology is so powerful that until 1999 the federal government sought to restrict its sale out of fears that criminals, terrorists and foreign nations might use it. The new vulnerability, discovered weeks ago by researchers at eEye Digital Security Inc., does not exploit any weakness in the complex encrypting formulas used to scramble messages into gibberish. Instead, hackers are able to attack a programming flaw in an important piece of companion software, called a plug-in, that helps users of Microsoft Corp.'s Outlook e-mail program encrypt messages with a few mouse clicks. Outlook itself has emerged as the world's standard for e-mail software, with tens of millions of users inside many of the world's largest corporations and government offices. Smaller numbers use the Outlook plug-in to scramble their most sensitive messages so that only the recipient can read them. It's not the number of people using PGP but the fact that they're using it because they're trying to safeguard their data, said Marc Maiffret, the eEye executive and researcher who discovered the problem. Whatever the percentage is, it's very important data. Maiffret said there was no evidence anyone had successfully attacked users of the encryption software with this technique. He said the programming flaw was not totally obvious, even to trained researchers examining the software blueprints. Network Associates Inc. of Santa Clara, Calif., which until February distributed both commercial and free versions of PGP, made available on its Web site a free download to fix the software. The company announced earlier it was suspending new sales of the software, which hasn't been profitable, but moved within weeks to repair the problem in existing versions. The company's shares fell 50 cents to $17.70 in Tuesday trading on the New York Stock Exchange ( news - web sites). Free versions of PGP are widely available on the World Wide Web. The flaw allows a hacker to send a specially coded e-mail - which would appear as a blank message followed by an error warning - and effectively seize control of the victim's computer. The hacker could then install spy software to record keystrokes, steal financial records or copy a person's secret unlocking keys to unscramble their sensitive e-mails. Other protective technology, such as corporate firewalls, could make this more difficult. You can do whatever you want - execute code, read e-mails, install a backdoor, steal their keys. You could intercept all that stuff, Maiffret said. Experts said the convenience of the plug-ins for popular e-mail programs broadened the risk from this latest threat, since encryption software is famously cumbersome to use without them. Even the creator of PGP, Philip Zimmermann, relies on such a plug-in, although Zimmermann uses one that works with Eudora e-mail software and does not suffer the same vulnerability as Outlook's. A plug-in for Microsoft's Outlook Express - a scaled-down version of Outlook - is not affected by the flaw. Maiffret said his company immediately deactivated the vulnerable software on all its computers, which can be done with nine mouse-clicks using Outlook, until it could apply the repairs from Network Associates. The decision improved security but makes it kind of a pain to send encrypted e-mails, he said. Zimmermann, in an interview, said PGP software is used quite extensively by U.S. agencies, based on sales when he formerly worked at Network Associates. He also said use of the vulnerable companion plug-in was widespread. Zimmermann declined to specify which U.S. agencies might be at risk, but other experts have described trading scrambled e-mails using PGP and Outlook with employees at the FBI, the Energy Department and even the super-secret National Security Agency. In theory, only nonclassified U.S. information would be at risk from this flaw. Agencies impose strict rules against transmitting any classified
Re: privacy digital rights management
I wrote: Perhaps we are using wildly divergent notions of privacy Donald Eastlake 3rd wrote: You are confusing privacy with secrecy That's not a helpful remark. My first contribution to this thread called attention to the possibility of wildly divergent notions of privacy. Also please note that according to the US Office of Technology Assessment, such terms do not posess a single clear definition, and theorists argue variously ... the same, completely distinct, or in some cases overlapping. Please let's avoid adversarial wrangling over terminology. If there is an important conceptual distinction, please explain the concepts using unambiguous multi-word descriptions so that we may have a collegial discussion. The spectrum from 2 people knowing something to 2 billion knowing something is pretty smooth and continuous. That is quite true, but quite irrelevant to the point I was making. Pick an intermediate number, say 100 people. Distributing knowledge to a group of 100 people who share a vested interest in not divulging it outside the group is starkly different from distributing it to 100 people who have nothing to lose and something to gain by divulging it. Rights Management isn't even directly connected to knowledge. Suppose I know by heart the lyrics and music to _The Producers_ --- that doesn't mean I'm free to rent a hall and put on a performance. Both DRM and privacy have to do with controlling material after you have released it to someone who might wish to pass it on further against your wishes. There is little *tehcnical* difference between your doctors records being passed on to assorted insurance companies, your boss, and/or tabloid newspapers and the latest Disney movies being passed on from a country where it has been released to people/theaters in a country where it has not been released. That's partly true (although overstated). In any case it supports my point that fixating on the *technical* issues misses some crucial aspects of the problem. The only case where all holders of information always have a common interest is where the number of holder is one. Colorful language is no substitute for a logical argument. Exaggerated remarks (... ALWAYS have ...) tend to drive the discussion away from reasonable paths. In the real world, there is a great deal of information held by N people where (N1) and (Ninfinity). - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: Commercial quantum crypto product - news article
Kossmann, Bill asked: Anybody familiar with this product? A Swiss company has announced the commercial availability of what it says are the first IT products which exploit quantum effects rather than conventional physics to achieve their goals. (05/31/2002) http://itworld.ca/rpb.cfm?v=20021510001 Actually a couple of products, I will comment only on one of them, the quantum random number generator. It reminds me of using a sport-utility vehicle to drive to the neighbor's house, ten feet away. There are easier ways to get to there. There is no reason to believe that quantum noise has any practical advantage over thermal noise. This point is not discussed in Quantique's principles-of-operation paper http://www.idquantique.com/files/paper-qrng.pdf and indeed they say there goal is to avoid thermal noise. You can harvest industrial-strength randomness from the thermodynamics of electrical circuits, costing next to nothing. A draft writeup can be found at: http://www.monmouth.com/~jsd/turbid/paper/turbid.htm - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Microsoft to shift strategy toward security and privacy
WASHINGTON -- Microsoft Chairman Bill Gates announced to employees Wednesday a major strategy shift across all its products, including its flagship Windows software, to emphasize security and privacy over new capabilities. http://www0.mercurycenter.com/breaking/docs/039127.htm - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]
Re: CFP: PKI research workshop
[EMAIL PROTECTED] wrote: ... People running around in business selling products and services and then disclaiming any liability with regard to their performance _for_their_intended_task_ is, IMHO, wrong. IMHO this presents an unsophisticated notion of right versus wrong. By way of analogy: Suppose you go skiing in Utah. A rut left by a previous skier causes you to fall and break your leg, or worse. Now everybody involved has been using the ski area _in_the_intended_manner_ yet something bad happened. So who is liable? The ski area could have groomed that trail, but they didn't. They could have enforced a speed limit, but they didn't. They could at least have bought insurance to cover you, but they didn't. They simply disclaimed all liability for your injury. Not only is this disclaimer a matter of contract (a condition of sale of the lift ticket) it is codified in Utah state law. Other states are similar. If you don't like it, don't ski. Returning to PKI in particular and software defects in particular: Let's not make this a Right-versus-Wrong issue. There are intricate and subtle issues here. Most of these issues are negotiable. In particular, you can presumably get somebody to insure your whole operation, for a price. In the grand scheme of things, it doesn't matter very much whether you (the PKI buyer/user) obtain the insurance directly, or whether the other party (the PKI maker/vendor) obtains the insurance and passes the cost on to you. The insurer doesn't much care; the risk is about the same either way. The fact is that today most people choose to self-insure for PKI defects. If you don't like it, you have many options: -- Call up some PKI vendor(s) and negotiate for better warranty terms. Let us know what this does to the price. -- Call up http://www.napslo.org/ or some such and get your own insurance. Let us know the price. -- Write your own PKI. Then defray costs, if desired, by becoming a vendor. -- Et cetera. In general, there is a vast gray area between Right and Wrong. Most things in my life can be described as not perfect, but way better than nothing. - The Cryptography Mailing List Unsubscribe by sending unsubscribe cryptography to [EMAIL PROTECTED]