MESSAGE FROM RON RIVEST VIA JIM BIDZOS VIA STEVE KENT VIA STEVE CROCKER: Thanks to Robert Silverman for keeping many people honest. As an additional effort to that end, I attach an analysis of the recent factoring effort, done by Ron Rivest. The early reports of RSA's demise have been greatly exaggerated... Note: Be sure and read the end of Rivest's note. Jim Bidzos, RSA Data Security To: Whom It May Interest (Feel free to distribute further...) From: Ronald L. Rivest Date: June 21, 1990 Re: Recent Factoring Achievement (Preliminary draft; may contain typos or other inaccuracies. Please send corrections to firstname.lastname@example.org) This note is in response to the numerous inquiries I've received regarding the recent factoring of a 155-digit number by A. Lenstra, M. Manasse, and others. (See the New York Times article of 6/20/1990 by G. Kolata.) This note attempts specifically to correct some of the misimpressions that may arise from a reading of such popular press articles. Using an ingenious new algorithm, Lenstra, Manasse, and others have factored the 155-digit number known as "F9", the ninth Fermat number: F9 = 2^(2^9) + 1 = 2^(512) + 1 . In binary, this number has the form 100000....000000001 where there are 511 zeros altogether. (F9 is a 513-bit number.) This is a fascinating development, and the researchers involved are to be congratulated for this accomplishment. The algorithm used is known as the "number field sieve", or "NFS" (not to be confused with a network protocol of the same acronym!). The NFS algorithm is described in the Proceedings of the 1990 ACM STOC Conference. The NFS algorithm is based on an idea due to Pollard, as developed further by Arjen Lenstra, Hendrik W. Lenstra, and Mark S. Manasse. The NFS algorithm is specifically designed to factor numbers that, like F9, have a very simple structure: they are of the form a^b + c where c is relatively small. (For F9, we have a=2, b=512, and c=1.) Some simple extensions of this algorithm are also possible, to handle numbers whose binary representation has many zeros, and related kinds of numbers (ternary, etc.) Numbers that have such a special structure are extremely rare and are unlikely to be encountered by chance. That is, the NFS algorithm does not apply to the kind of "ordinary" numbers that arise in practical cryptography, such as using RSA. They only apply to numbers with "sparse" representations having few nonzero components. (Let us call such numbers "rarefied".) When working on a rarefied number, the NFS algorithm has an estimated running time of the form (for an input number n): exp(1.56 (ln n)^1/3 (ln ln n)^2/3) (1) For n = F9, this evaluates to 4.1 x 10^15 operations, which, at 3.15 x 10^13 operations/year for a 1 MIP/sec machine (i.e. a MIP-year), gives a workload estimate of 130 MIP-years, only off by a factor of two from the actual work of 275 MIP-years. (That is, formula (1) may be roughly too low by a factor of two.) It is instructive to see the effect of doubling the size of the number being dealt with. A 1024-bit (332-digit) rarefied number requires an estimated 1.54 x 10^21 operations = 4.9 x 10^7 MIP-years, a dramatic increase in difficulty. The NFS algorithm algorithm is not a "polynomial-time" algorithm; the difficulty of factoring still grows **exponentially** with a polynomial function of the length of the input. What has this to do with RSA and cryptography? I think there are three basic points: -- This development indicates that the status of factoring is still subject to further developments, and it is wise to be conservative in one's choice of key-length. -- The NFS algorithm may yet be generalized to handle "ordinary" numbers, and the potential impact of this should be considered. -- Factoring is still a very hard problem, despite everyone's best efforts to master it. Regarding the further extensions of NFS to handle ordinary numbers, this is judged to be a reasonable possibility by those working on NFS, so it is helpful to consider what impact this may have. It is conjectured (see the ACM STOC paper referenced above) that a successful extension of the NFS algorithm to ordinary numbers would have a running time of the form: exp(2.08 (ln n)^1/3 (ln ln n)^2/3) (2) This is similar to equation (1) except that the constant 1.56 is replaced by the constant 2.08. Note that a practical version of such an extension does NOT yet currently exist (to the best of my knowledge), but even granting its plausibility we arrive at an estimate of the tie required to factor a 512-bit number of 6.5 x 10^20 operations = 2 x 10^7 MIP-years which (in my opinion) is a substantial degree of security. It is interesting to note that this work factor is actually GREATER than that required by the ``standard'' factoring algorithms (e.g., the quadratic sieve), which have a running time of exp((ln n)^1/2 (ln ln n)^1/2); for a 512-bit number, this gives a work-factor estimate of only 6.7 x 10^19 operations. Indeed, the NFS algorithm (when extended) will be asymptotically superior than the quadratic sieve algorithm, but will be slower for numbers with less than about 200 digits. That is, assuming that (2) is indeed the correct running-time estimate for any extension of NFS, then NFS will not affect the security of any numbers of less than about 215 digits. So any "standards" that have been considered using 512-bit RSA moduli are not likely to be affected by any NFS extensions. (At most, one could imagine that the RSA key-generation process might be extended to check that the resulting modulus n is not a rarefied number.) In the truly worst-case scenario, we would have that an extension of NFS would be found that allows ordinary numbers to be factored with a work-factor that is governed by equation (1); in this case one would need to adjust the sizes of moduli used by RSA upwards by a factor of less than two to more than offset the new algorithm. A factor of two in size affects the running time of public-key encryption (or signature verification) by a factor of four and the running time of private-key encryption (or signature generation) by a factor of eight. Noting that the speed of workstations has increased by a factor of over 100 in the last decade (indeed, such factors have been the technological advance that made the successful implementation of NFS possible!), such performance penalties, if necessary, seem to be easily absorbed by expected technological advances in the speeds of the underlying RSA implementation technologies. That is, the NFS-like factoring algorithms do not, even in this worst-case scenario, prevent successful implementations of the RSA cryptosystem. As a cryptographer, I am actually very happy with all the effort that is being spent trying to determine the exact level of difficulty of factoring. Achievements such as the recent development of NFS help to pin down the best-possible rate of growth of the difficulty of factoring, so that users of cryptographic schemes can pick key sizes with an increased degree of confidence that unforeseen developments are unlikely to occur. The best way to ensure confidence in a cryptographic system is to have it attacked vigorously and continuously (but unsuccessfully) by well-qualified attackers. If, despite their best efforts, the difficulty of cracking the system remains intrinsically exponential, then one can have a reasonably high degree of confidence that the system is actually secure. This is the process we have been seeing at work in the recent work on factoring. The results of the attacks can be used to guide the selection of the necessary key size for a desired level of security (with an appropriate margin of safety built in, of course). (As a closing note, here's a prediction: I expect that the 128-digit ``challenge RSA cipher'' published in the August 1977 issue of Scientific American to be cracked (probably by the quadratic sieve algorithm or a variant, not NFS) during the next 1-3 years. This accomplishment will require substantially more computer time than the 275 MIP-years required to factor F9.)
I'm concerned by the increased use of laptop computers with cellular phones, to connect to remote host machines. Given the ready availability of scanner radios capable of receiving cellular phone frequencies, this practice seems to amount to broadcasting your passwords, proprietary information, etc., to anyone who cares to listen. (Yes, I know that federal law prohibits listening to such calls, but everyone who has a scanner or knows someone with one knows that it's done all the time.) Is there any record of anyone breaking systems this way yet? Has any company adopted a policy discouraging its employees from using laptops with cellular phones, or, for that matter, from broadcasting other sensitive information over cellular phones? -- Jan Wolitzky, AT&T Bell Labs, Murray Hill, NJ; 201 582-2998 att!mhuxd!wolit or email@example.com (Affiliation given for identification purposes only)
At the risk of turning this into a medical forum I wanted to add to the excellent summary on carpal tunnel syndrome that there is a related problem called ulnar nerve syndrome. As with CTS, those of us who spend lots of time using keyboards are prone to ulnar nerve syndrome, too. The main difference in the symptoms is that with CTS the numbness and tingling is in the thumb and first two fingers, with UNS it is in the other two fingers and along the outside of the forarm. (When you "hit your funny bone", what you've done is bash your ulnar nerve. The symptoms are similar to that feeling, but they don't go away.) My wife's neurologist (she's been going through a bout of it) said there are many causes, but a common one is some activity that constantly rubs the elbow. In her case it is almost certainly typing while resting her elbows on the arms of a chair. One danger is that carpal tunnel syndrome is so hot, such a fad right now, that a physician might automatically connect "numbness" with "programmer" and say "carpal tunnel". My wife's physician did this, and only changed his mind after we had looked in some medical books and found that CTS symptoms did not match hers. (Physicians deny that they reason this way. Experience tells otherwise.) Treatment is basically like CTS: ice, avoid bending the elbow as much as possible (not easy to do), don't do things that rub the elbow a lot, anti-inflammation drugs. There is surgery, but unlike CTS, ulnar nerve surgery is not a good option. Apparently it has about a 50% chance of making things worse, according to my wife's neurologist. (That might have been an estimate that took the nature of her condition into account, i.e., there are cases where it has a high probability of success, but hers isn't one of them.) -- mike
>One point the author does not mention is that the "force-depression curve" of >your keyboard may also play a role. It is better to have a linear >relationship between force and depression... >... People have suggested that this sort of dynamic may >aggravate or even induce CTS... What was the incidence of CTS twenty years ago, when electric typewriters routinely had non-linear force-depression curves? Or before that, when manual typewriters required far more finger pressure than any modern keyboard? Yet again, we have here a case of a "computer risk" that isn't really new, and data from olden days could be very useful in deciding what *really* causes it. Data would be particularly useful because it's easy to construct an argument that points in precisely the opposite direction! Once you've pushed a non-linear keyboard key "over the hump", you can relax pressure. But with a linear keyboard, you have to push all the way down, since you get no "that's enough" feedback until the key hits bottom. Some of this may be a risk not of nonlinear keyboards, but of lack of proper training. Pre-computer typing courses taught you to *strike* the keys rather than *pushing* them, so your muscles were already relaxing when the key bottomed out. One side effect of the proliferation of keyboards is that far more people are using them without formal training, or with training from "touch typing" programs that teach you which keys to hit but don't teach posture or hand position. Henry Spencer at U of Toronto Zoology
> Tonight's news was really distressing -- that there is a fundamental > mirror flaw that cannot be repaired until 1993 or so. It's not *quite* that bad. The optics do seem to have a significant case of spherical aberration -- so perfect that it's almost certainly the result of an incorrectly-figured shape, rather than a manufacturing deviation from the desired shape -- and this will hurt the cameras badly, but the spectrographs and the photometer (3 out of 5 of the instruments) shouldn't be badly affected. Unfortunately, the affected instruments are the ones that would produce all the nice crowd-pleasing pictures, so the public-relations disaster is much worse than the scientific one. The Wide Field / Planetary Camera was slated for replacement in 1993-4 anyway, and it may be possible to build the replacement in time for a 1991 flight (about the earliest NASA could launch a repair mission anyway). The big question is, how did this happen? Henry Spencer at U of Toronto Zoology utzoo!henry
The Hubble space telescope, which cost $1.5e9 to build (and a lot more to store and launch), isn't working properly. There appears to be a problem with one of the two mirrors being flawed. Weren't they tested? Yes, they were -- but they were tested individually, not in the final assembly. It seems that the test jig would have cost too much. Lessee -- the components all worked individually, but not as a system. Where have I heard that one before? --Steve Bellovin
Just to get the record straight, I originally wrote and posted the entire story to rec.aviation, a newsgroup I felt would appreciate the humor without undermining the serious aviation-related discussions also posted there. I considered submitting the story to RISKS orginally, but decided against it because of the harm it might have done to RISKS serious nature. Robert Dorsett, I believe, passed the submission on to RISKS (through PGN?) much to my delight. Personally, considering the flood of mail I've received, I think we might well start a serious discussion in RISKS about the risks of computer types who cannot recognize a blunt-faced satire when it hits them. Gregory Travis
Robert L. Smith <firstname.lastname@example.org> wrote: > I'm afraid Robert Dorsett is mistaken when he states "And nobody's >crashed a 757/767 yet. . ." On July 23, 1983, a 767 operating as Air >Canada Flight 143 crashed at Gimli, Manitoba. Actually, it's difficult to classify that as a crash. Part of the dif- ficulty is terminology: "crash" is not defined by either ICAO or the FAA. The two relevant terms are "accidents" and "incidents." US NTSB 830-3 defines "aircraft accident" as "an occurrence associated with the operation of an aircraft which takes place between the time any person boards the aircraft with the intention of flight and all such persons have disembarked, and in which any person suffers death or serious injury, or in which the aircraft receives substantial damage." "Incidents" are anything other than crashes, which could affect the future safety of operations. The Air Canada landing at Gimli can certainly qualify as an accident. How- ever "crash," as used by the industry in general, has more permanent connot- ations. The airplane was under control at impact, did not suffer substantial damage, and nobody was seriously injured. I therefore hesitate to classify it a crash. On the other hand, there *was* a rather nasty incident in Florida a few years ago, a result of a "hard landing." The airplane was damaged sufficiently enough that it was a marginal-recovery issue, but it remained under pilot control, and nobody was seriously injured. Both the Gimli 767 and that 767 are back in the air. Perhaps I should rephrase my original comment: there have been two A320 hull losses. There have been no 757 or 767 hull losses. > That crash may well have been a proper "risk" subject. Indeed; it's been popular in the past. Here's the review of _Freefall_ I posted to rec.aviation last year. It may be of interest to RISKS readers. ====== _Freefall_ (by William and Marilyn Hoffer, St. Martin's Press: New York, 1989) deals with the near-crash of Air Canada Flight 143, a Boeing 767-200 which ran out of fuel over near Winnipeg, Manitoba in 1983. The book is partially investigative reporting, partially schlock: while it provides a detailed accounting of the events leading up to the eventual landing, it also wastes an enormous amount of space on what the passengers think, feel, etc--and in that respect rather closely resembles the style Arthur Haily used in _Airport_. In other words, it's light reading, and tries to be something for everyone. Fortunately, though, the authors kindly segregate the chapters into what's happening in the cockpit, and what's happening elsewhere. If one sticks to the "Cockpit" (clearly labelled) chapters, it's tolerable (but since the book itself is only 263 pages of double-spaced large print and large margins, and less than half of it deals with the technical issues, the $17.95 price tag isn't exactly worth it). To save people the some time and effort, here's what it boils down to: 1. A fuel-metering device failed on a previous flight. A maintenance worker, through trial and error, got the system to work again. A subsequent worker mistakenly flagged the system as inoperative. 2. To verify how much fuel was on board, "dripsticks," a series of fuel- measurement devices mounted under the wings, were used. A flotation device ("donut") is mounted on the top of the stick; when the stick is released, it drops. Where it stops (a length unit, represented by how much of the stick is pointing out) can be used to determine how much fuel is in that compartment. The procedure then is to use a chart to convert the unit read into liters, then to kilos, then, finally, to use a conversion factor to get pounds. All such measurements are then added to determine the total load on board. In Flight 143's case, an incorrect conversion factor was used for the last step. Confusion among ground workers and the flight crew as to the correct conversion factor (the 767 was the only metric aircraft in Air Canada service) induced them to launch with half tanks. 3. The captain launched with his fuel metering system inoperative, in violation of the minimum equipment list (but there may have been justification for doing so, as several references contradicted each other on this point). To overcome the loss of the totalizer, they set the flight management system with what they thought was a full load, which then provided them with a fuel burn total for the rest of the flight. GIGO. 4. They ran out of fuel near Winnipeg, Manitoba, which resulted in a loss of nearly all of their electrical power, including the EFIS and EICAS systems (electronic flight instrumentation and engine control and monitoring in- strumentation). Radios and backup instrumentation was supplied by battery power (and, initially, APU power); some hydraulic power was obtained from a ram air turbine (optional equipment, I believe, on the 767--Air Canada bought them in anticipation of extended-range twin oceanic operations). The turbine serviced basic flight controls, but did not provide power for other surfaces, such as flaps. They were eventually forced to deadstick the plane to an abandoned military base near Gimli. They landed on top of a social event being held on a disused runway, fortunately not killing anyone. The nose gear collapsed, there was a small fire (from the nose brakes) and there was some damage to the airplane, but the plane was eventually put back in service. 5. The flight crew was initially fired, then rehired. The first officer, according to the book, is starting his captain's training.
Robert L. Smith in his posting re: the A320's attack of nerves refers to the crash of an Air Canada Boeing 767 at Gimli Manitoba. He seems to be using the word 'crash' in a sense with which I was previously unfamiliar. The plane ran out of fuel because of a combination of factors. The fuel guages (computerized?) were malfunctioning, and Air Canada was in the early stages of conversion to metric. Reports at the time suggested that the refuelling was measured in pounds and the measurement was used un- or incorrectly converted in a formula that expected kilos. When the engines flamed out, the rather small WW 2 training field at Gimli was the closest airport, 100 miles away. The pilot was highly praised for making a successful dead stick landing, and the plane became known as the Gimli glider. I believe it was flown out a few days later, without the passengers and with a very carefully calculated fuel load. The subsequent investigation was somewhat acrimonious, with the airline pointing out that the pilot had ultimate responsibility for the fuel on board, opposed by a strong public sympathy for the pilot and a feeling that the system was designed to fail. I suppose the risks are pretty much the same whether a computer was involved or not - large organizations occasionally screw up when they try to run complicated systems. This case also shows the value of properly trained people retaining ultimate control of the system.
On June 19-21, 1990, IBM held some kind of a development conference for GDR universities, in the research center of the ministry for science and technology in (east) Berlin-Koepenick. Similar to an annual conference for West German universities (`IBM university forum'), invited speakers from West and East German universities as well as from IBM informed about their actual work. A broad diversity of areas was covered, from CD-ROM based 'Thesaurus Linguae Graecae' to CAD, simulation of complex molecules and synthetic speech. The conference was accompanied by an exhibition where many additional applications and software products of scientific interest were shown by East and West German scientiests as well as IBM people, on IBM owned PS-2s. Many demonstration diskettes were freely available. Among the exhibitors, the Virus Test Center demonstrated how to detect and eradicate viruses. In many discussions, we were surprised to learn that many scientists regarded viruses as some kind of a joke as they had suffered mainly from viruses of the funny kind, e.g. playing Yankee Doodle in the Bulga- rian version "TP 44" or "legalizing marijuana"; only a few seemed to have experiences in really damaging viruses such as Israeli or Dark Avenger. Yet at the end of the exposition, our essential task was to eradicate some damaging viruses such as Dark Avenger (the Bulgarian "Eddie" which broadly migrates through Eastern Europe) from most of IBM's PS-2 as neither protection nor careful work had been practized nor prescribed. With surprise we learned that there existed a secret research unit in GDR to which every virus or other threat had to be reported; this secret group would then produce an antivirus and send it to concerned institutions. In its latest version (which we hope to receive afterwards), 11 viruses could be detected and eradicated. Lesson learned: there should be a special antivirus service for exhibitions, not only for large ones (in FRG's CeBIT and Systems exhibitions, about 15-20% of the workstations and PCs were found to be infected *at exhibition's end*). Klaus Brunnstein University of Hamburg
Recently, the LISTSERV which I get my risks digests from sent me a copy of one of those unsolicited mailings that I always get via the US MAIL. You know the type, "You can make 50,000 bucks overnight with my ...". Yeah, yeah right. Those are one of the reasons (besides bills) why I dont go to my mail box any more, and deal with electronic mail. Now I get to deal with this kind of crapola via the BITNET? Regardless of whether or not the message was accidently sent to a LISTSERV and redistributed to everyone that the LISTSERV knows about, I never should have been able to receive the message. LISTSERV software should be modified to include some sort of authentication. Granted, an individual with enough know-how can bypass a simple "authentic address" in the "Received From/By:" headers, but it would prevent those "whoops, I accidently did its...". Maybe the author did, maybe he didn't. Either way, I really dont enjoy reading that stuff, and am sure that others don't as well. Also, the proposed scheme is a pyramid, and is quite illegal here in Il. Anyone for prosecution? Greeny [Abuses of the BITNET and USENET RISKS addresses have happened before, although they could easily be blocked were it not for the intentional sandbox orientation. Fortunately for the rest of us, we did not see the abuse. The author did apologize to me that it was not intentional, however... PGN]
Please report problems with the web pages to the maintainer