Talk:List of cryptographers
|
|
| Contents |
Ross Anderson or RJ Anderson?
Ww made several tidying up copyedits, but also changed the link from Ross Anderson to "R J Anderson". I can't see why this was done, as the page "R J Anderson" doesn't exist, while the page "Ross Anderson" already existed (and is definitely the same person). Also, in my experience Professor Anderson is generally referred to as "Ross" rather than "R J", and certainly usually signs his books, papers etc. as either "Ross Anderson" or "Ross J. Anderson". So, I changed the link back. Securiger 23:34, 22 Dec 2003 (UTC)
Securiger,
I was trying to fully distinguish Ross J from other possible Rosses. I didn't realize there was an article for plain Ross. Should have checked. Perhaps there may be a need for a disambiguity reference?
ww
lost book on Roman military crypto by JC?
in re Julius Caesar writing a book on crypto. My understanding about this lost book was that it discussed military crypto -- with emphasis on Julius' use of same -- but that we didn't know the author. On what basis do you claim Julius as the author?
ww
Probably on the basis that I don't know my arse from my elbow ;-) -- Cimon Avaro on a pogostick 16:51, Feb 24, 2004 (UTC)
- On the assumption from your comment then, that your attribution was 'conventional' and not ascribable (lousy wording, but like Pascal I'm too rushed to do better), I've revised the Caesar entry here. Is this adequate in your view?
- ww
- According to Kahn, the book is attributed to a grammarian named Probus, probably Valerius Probus. --Imran 18:02, 24 Feb 2004 (UTC)
- Imran, thanks for the ref. Will change entry soon. Have to go. ww
Well done, folks. -- Cimon Avaro on a pogostick 07:54, Feb 25, 2004 (UTC)
how about Lord Lyon Playfair?
Now that we got Julius fixed, for which I offer kudos once more; there are still quite a few significant omissions here. In light of the dogs breakfast I made of Ceasar, I think I will refrain from editing tha article itself directly, but kibitz from the sidelines, sort of thing.
- Lord Playfair (probably not the correct link, but should probably be made a redirect or disambiguation page; too lazy to preview) I won't add him, cause I have absolutely no understanding of his relative significance beyond the fact that he invented the Playfair Cipher (again, that should be a redirect, if it is not the correct link)
- The Playfair cipher was invented by Sir Charles Wheatstone. Lord Lyon Playfair only recommended it to the British government - allegedly on a table napkin, at dinner. Wheatstone should be on the list, and I will add him. However IMHO Playfair doesn't belong there; his minor contribution is sufficiently acknowledged in the Playfair cipher article, and excessively acknowledged in its name!. Securiger 07:56, 28 Feb 2004 (UTC)
include 'mere' cypher users? (& Kasiski comments)
- Someone removed Edgar Allan Poe from the list. He is not a significant serious theorist, but a very notable amateur. More notable for his too high opinion of himself as a cryptographer than anything else, but still notable. And he did write Goldbug. I would err on the side of listing him, but explaining that his significance to cryptogaphy is more cultural than scientific.
- The matter of notable users of cryptography; I would like comments on whether these belong to the list:
- Mary, Queen of Scots, Or is it sufficient to list Babington?
- Queen Christina of Sweden used ciphered text to send indiscreet love-letters to Cardinal Azzolino; and had a general keen intrest in alchemical and heraldic codes and injokes.
- Neal Stephenson?
- Abraham Abulafia?
-- Cimon Avaro on a pogostick 12:14, Feb 27, 2004 (UTC)
- Well, before cryptology became too complicated for interested amateurs, it seems most prominent people used it to some degree; a few references even describe it as one of the arts that every Renaissance gentleman should master, in order to be effective in service to his lord. So, a list of notable users could be very long indeed. I guess it comes down to the question of what this page is for. So far, I've used it mainly as a device for clicking "related changes" so my watchlist doesn't need to include every cryptographer! I didn't really think of it as a research tool on "usage of cryptography throughout history", although I guess it could be. Perhaps we need to really rethink the purpose and design of the page. Securiger 07:56, 28 Feb 2004 (UTC)
- Perhaps one solution would be to be ambitious and bold and start more general articles titled Timeline of steganography, Timeline of cryptology or something? (please de-link those if you think they are not the most apposite title for such an article) -- J-V Heiskanen 04:57, 8 Mar 2004 (UTC)
- Poe probably belongs here as he was not merely a user but a cryptanalyst, though an amateur and egotistical one. His claim was foolish at the time he made it, and as his technique appears to have been frequency analysis, his claim would have been out of line (for anyone with some sense of things) since Alberti. Nevertheless, his story is famous, his claim is infamous, and he actually did public cryptanalysis. Mere users, I agree (from the Renaissance or any other time) need not be included.
- On another subject. Securiger, I noted that Kasiski had published decades later from a memory that his publication was in the 80s or some such (I didn't check, though) and the timing of the Crimean War. Babbage's work was just before the War, if memory serves in this instance, and so the weasel word choice of decades made some sense. On the other hand, as you may observe, a person from Porlock may have interferred lest I would have done better (or more accurately). I do note, however, that "a few years" has that same weasel word aroma about it. Did you look his pub date up? If we collectively keep each other on the straight and narrow, perhaps Sanger/Wales assertion will be borne out in full! ww 17:11, 22 Mar 2004 (UTC)
- I looked up Kasiski in Kahn; his book was published in 1863. I have not yet found any referenced date for Babbage, although several web sites state 1854, without references. I am a little concerned at the claim that Babbage's and Kasiski's methods were the same. I have so far been unable to find a detailed description of Babbage's method, but several sources claim that it was also useful against the autokey, which Kasiski's method is not. I am therefore inclined to think it was probably something a lot more sophisticated like the index of coincidence or somesuch, and Kasiski's method was actually unique to him - it certainly has more the feel of someone who approached the problem by hacking on lots of ciphertexts rather than approaching it from the mathematical end. But I could be completely wrong, perhaps Babbage's method is identical to Kasiski's and he got into autokeys by also discovering a quite separate attack for them like probable plaintext a.k.a. the Friedman attack. Until someone can point me to a good description I will just feel a little disturbed, as I usually do anyway 8^) Securiger 18:31, 22 Mar 2004 (UTC)
- Oh, and yes, definitely put Poe in. Always a good idea to be reminded of hubris in this field. 8^) Securiger 18:33, 22 Mar 2004 (UTC)
- My information on Babbage comes from Singh's Code Book. It was new to me when I came across it, and I've basically trusted him to have gotten his bibliographic ducks properly in a row. It was my impression from him that Babbage and Kasiski had found just about the same thing, a general attack method on polyalphabetic cyphers. I would not find it hard to believe (but have absolutely no information of any kind supporting) that 1) Babbage found more than that, or a more general method (though what this might be is more than a little obscure as that's a very general result all by itself), or 2) that his additional work managed to get lost. After all, it seems that much of the reason he couldn't get the Analytical Engine any farther along than he did was that his machinist (or at least one of them) was essentially holding him up for ransom via the equivalent of a mechanic's lien on parts. Oh, would that Babbage had been as good with his hands as Turing, and not as it appears he was, another Poincare or Oppenheimer! Anyone who let London organ grinders get so thoroughly under his skin was a little too tightly wound. ww 19:40, 22 Mar 2004 (UTC)
"No home page"
Following comment snippet imported from Talk:Topics in cryptography:
>>>
- "We have an instance in your recent edit at the cryptographer list. The intro noted that home pages would be given when known (and we can discuss why -- or why not -- that was included), and in several of the medieval entries, I noted explicitly that there is no home page. Same sort of harmless humor, of which you disapprove I gather. Do you like limericks, or maybe puns, or knock knock jokes, or ... instead? You certainly hastened to correct an unintended pun recently. I take the position, with Douglas Adams, that it's "mostly harmless". ww 18:39, 23 Mar 2004 (UTC)"
<<<
- Well, I missed the joke, to be honest. My mirthless condition non-withstanding, the explicit statement of "no home page" is surely redundant, and the list is easier to browse when brief. — Matt 19:39, 23 Mar 2004 (UTC)
Suggested entry for Poe
- Edgar Allan Poe, an unremarkably talented amateur cryptanalyst, whose public claim of being able to decipher any monoalphabetic substitution cipher nevertheless brought him significant public awareness in his time. Despite the actual triviality of the claim, and the palpable fallacy of his general assertion that: "human ingenuity cannot concoct a cipher which human ingenuity cannot resolve"; his short story The Gold Bug remains a seminal literary depiction of cryptanalysis.
-- J-V Heiskanen 22:00, 25 Mar 2004 (UTC)
- Looks good to me. Petty nitpicks — I would quibble with the "palpable fallacy of..assertion..human ingenuity..cannot resolve" sentence. Of course, the general assertion is indeed a fallacy because we have provably secure ciphers such as the one time pad. But the ciphers used in practice have no such proof, and it remains an open question whether "human ingenuity" can resolve them; I dont think this would be a "palpable fallacy" to a non-technical reader. Also, we could probably chop out "unremarkably talented", as the absence of any of his achievements in cryptography probably says it well enough... — Matt 22:29, 25 Mar 2004 (UTC)
- J-V and Matt, I liked the entry for Poe; the sharp phrasing is pleasing. But it's a bit NPOV. Matt has caught the problematic points I think. As for Matt's comments, they are anachronistic as one-time pads hadn't been invented in Poe's time -- much less was there any proof of their unbreakability. And, while we probably have (other) encryption algorithms which are 'proof against human ingenuity' now, no one did then. I would agree that 'unremarkably talented' is a bit strong, if entirely accurate and quite nicely phrased, and that 'palpable fallacy' is not entirely obvious and a bit strong, if also accurate. I would disagree that readers will note the absence of achievement by Poe in this field -- people characterisitically don't notice absence of <whatever> as an observable. Some positive statement of his status (ie, nil) is indicated, whihc implies that some reason for including him should be provided. His 'fame' as a cryptanalyst is based on his public boasting in the public press (Baltimore newspapers, if memory serves) and his high frequency of success against reader proposed cyphertexts. Which certainly says more about general reader knowledge of the time than about Poe's abilities. ww 15:32, 26 Mar 2004 (UTC)
unbreakable cyphers known to be possible when?
- According to the Friedman paper (see below) provably unbreakable cryptosystem were known by cryptographers of Poe's time, but none of them were considered practicable. --Imran 11:18, 28 Mar 2004 (UTC)
- Imran, I don't think this will belong here, but I'm not sure where to put it, so... I had understood that Shannon (in secret work during WWII first published 48/49) produced the only adequate basis from which to prove any result on 'breakability', and then did so in re the one time pad. If so, Friedman must have been referring to someting else, such as "believed unbreakable in practice by us who are in the (state of the art) know". Which would not be the same thing; notwithstanding Friedman's (quite deserved) reputation. Or he may have been referring to some progress in the theory (the knowledge of which is still secret) achieved by him or his folks at SIS; thus anticipating Shannon who would, on this reading, not have been informed of such a result while doing official and classified War research or, if so informed, decided to take credit for it. Given the status of SIS and its funding at the time, I find this quite difficult. Still less plausible do I find that this position could have been reached prior to the Civil War here, as I have not seen the slightest hint of such knowledge (as opposed to "belief without sufficient foundation") prior to Shannon's work. For instance, Bazieres (who had sterling crypto credentials) thought he'd come up with something unbreakable ca 1900 only to be disabused rather promptly by a rival.
- Unfortunately I have no access to your URL listed below via JSTOR. I will try to find it another way, read the paper, and get back.
- Comments? ww 16:20, 29 Mar 2004 (UTC)
- Here's the relevant section "So far as the professional cryptographer is concerned, there has never been any question about the theoretical possibility of constructing at least one or two cipher systems, which are mathematically demonstratable as being absoltuely indecipherable. It is far from being the case that the invention of such ciphers had to await modern advances in cryptographic science; their devising was possible from the very earliest days of secret writing. The difficulty has been to make such systems practicable for regular usage by persons having a need for the highest degree of cryptographic security." - I think the reasoning must have been along the line of a OTP, that basically if all possible messages can be generated from a ciphertext then the system would be secure. --Imran 21:00, 29 Mar 2004 (UTC)
- Imran, I'm having trouble counting the ':' and so will not indent further to the right. We're piling up in any case. Haven't found a copy of the article yet, though there was, apparently, a reprint(?) in a Army Signal Corps publication about the same time.
- The extract you provide is interesting, but I fear inconclusive. From this it's not possible to tell whether F is making a pleasant point about how cyphers could be 'real good' in practice and could have been so for a long time, or is making a provable claim about unbreakability. Perhaps F and co. had anticipated Shannon's result (which you note later, in effect), but this gets us little farther on as nearly as I can tell.
- Descending to the level of speculation from textual analysis, I would tentatively guess that he wasn't speaking from solid theoretical grounds. He says "...at least one or two..." which is sufficiently weasel wordy to be a hint toward the 'in practice' side of things. But..., the "...mathematically demonstrable as being absolutely indecipherable." bit is tantalizing, isn't it?
- Certainly F is correct that at any time since Alberti (ie, ca 1460) it had been possible to produce an effectively unbreakable cypher (ie, "...the highest degree of cryptographic security...") -- by doing a polyalphabetic one right, and further doing the key scheduling sensibly. It was failures in key scheduling which made Purple so porous (after the 'reverse engineering' breakthrough by SIS), and it was certainly poor key scheduling which contributed to Enigma's breakability. Naval Enigma was not much different for much of the war than the Werhmacht/Luftwaffe versions (more rotor choice being the biggest difference), but that the key management was much better on the whole and so Hut 6 had a much harder slog than the others. There are several statements from BP folk who were in a position to competently comment (and, I seem to remember, from some of the SIS/OP-20-G folk as well), that 'had they managed their keys better, we never would have been able to break xyz at all'.
- For example, recall the Swedish count <insert name here, which has fallen out of my head> from about 1750 (more or less, anyway) who seems to have produced a mechanical cypher which could have been better than any of the rotor machines (save perhaps SIGABA and similar) of the 20th century. As near as I can make out, with good key scheduling it would have been extremely good. Too bad it didn't have any particular influence I'm aware of.
- But the really fascinating issue is whether F was commenting from actual knowledge of what has come to be known as Shannon's results. As one interested in the history of this stuff, I'd love to know for sure! Wouldn't you?
- ww 16:20, 30 Mar 2004 (UTC)
Imran's refs for Poe as cryptographer/cryptanalyst
Searched JSTOR and it turns out Poe's crypto ability has been well analyzed, I'll go through the papers later but if anyone else wants to route through them,
"The King of `Secret Readers'": Edgar Poe, Cryptography, and the Origins of the Detective Story
Shawn Rosenheim
ELH, Vol. 56, No. 2. (Summer, 1989), pp. 375-400.
Stable URL: http://links.jstor.org/sici?sici=0013-8304%28198922%2956%3A2%3C375%3A%22KO%60RE%3E2.0.CO%3B2-C
The Code for Gold: Edgar Allan Poe and Cryptography
Terence Whalen
Representations, No. 46. (Spring, 1994), pp. 35-57.
Stable URL: http://links.jstor.org/sici?sici=0734-6018%28199421%290%3A46%3C35%3ATCFGEA%3E2.0.CO%3B2-G
What Poe Knew about Cryptography
W. K. Wimsatt, Jr.
PMLA, Vol. 58, No. 3. (Sep., 1943), pp. 754-779.
Stable URL: http://links.jstor.org/sici?sici=0030-8129%28194309%2958%3A3%3C754%3AWPKAC%3E2.0.CO%3B2-7
Edgar Allan Poe, Cryptographer
William F. Friedman
American Literature, Vol. 8, No. 3. (Nov., 1936), pp. 266-280.
Stable URL: http://links.jstor.org/sici?sici=0002-9831%28193611%298%3A3%3C266%3AEAPC%3E2.0.CO%3B2-2
"The Language of the Cipher": Interpretation in "The Gold-Bug"
Michael Williams
American Literature, Vol. 53, No. 4. (Jan., 1982), pp. 646-660.
Stable URL: http://links.jstor.org/sici?sici=0002-9831%28198201%2953%3A4%3C646%3A%22LOTCI%3E2.0.CO%3B2-I
Poe's Solution of the "Frailey Land Office Cipher" (in Notes, Documents, and Critical Comment)
W. T. Bandy
PMLA, Vol. 68, No. 5. (Dec., 1953), pp. 1240-1241.
Stable URL: http://links.jstor.org/sici?sici=0030-8129%28195312%2968%3A5%3C1240%3APSOT%22L%3E2.0.CO%3B2-6
Back to Poe
Curiously, possibly the most pertinent book is lacking in the above list.
Shawn Rosenheims article: "King of the 'Secret Readers'" is in fact subsumed into and expanded to an overview of most of the sources mentioned above, in the book: The Cryptographic Imagination: Secret Writing From Edgar Poe to the Internet.
I hasten to note, that it is pretty thin gruel; perhaps only three substantial claims to specific significance of Poe regarding cryptology are proffered:
- That many cryptographers have been inspired by Poes fiction, most notably Friedman. This can be substantiated to the degree that Friedman did actually write a book about Poe, but he had definite intrests in Baconian scholarship, Jules Vernes' crytpography-laden fiction &c., &c., &c.
- That the U.S. Military used/(uses?) Poes fiction as training material for cryptographers.
- That a (still classified) Polish treatise about breaking the Enigma machine has substantial, extended and crucially significant references to Poe.
I don't really think there is much there to help in constructing a concise bullet-point entry for Poe.
-- J-V Heiskanen 15:50, 31 Mar 2004 (UTC)
- Perhaps we should make an article about Poe as a cryptographer and just link to it from this page ?
- I'm curious to know if there's actually any solid evidence that Friedman was inspired by Poe, as he doesn't mention it in his paper. His interest in Baconian scholarship is almost certainly due to his days working for Fayban (initially as a geneticist). --Imran 16:03, 31 Mar 2004 (UTC)
- J-V, I'm very curious about the Polish reference. I have read most of the literature on Enigma (at least that which is in English) and this is the first I've heard of such a thing. It seems implausible on fundamental grounds as Rejewski's triumph was only trivially related to the only cryptanalytic technique Poe knew (ie, simple frequency analysis) and relied on maths that I feel morally confident Poe was innocent of. Group theory and the sorts of statistics Rejewski used were either unknown in Poe's time or were mathematical rocket science, being wrestled with by such as Galois, Abel, and so on. Poe didn't play in that league.
- As for the US military using Poe's fiction as training, I think this has no significance beyond good writing in the illustration of very elementary cryptanalysis methods. The Dancing Men (Sherlock Holmes) would serve equally well.
- And on Friedman having been inspired by Poe, I am with Imran. Never heard of it, and want to know more! ww 16:57, 31 Mar 2004 (UTC)
On the list of modern cryptographers
I looked at the list of modern cryptographers and thought to myself that it is a pretty small list at present. What are the criterias for adding someone to that list? Just off the top of my head I can think of many people that should be added to that list. For example (just off the top of my head): Gilles Brassard, CAN, co-invented quantum cryptography and did allot of other work. David Chaum, US, did allot of work on anynimity systems, blind signatures Claude Crépeau, CAN, Zero-knwoledge proofs, multi-party computations, oblivious transfer, quantum information theory Stefan Brands, CAN, digital credentials Adam Back, UK, hashcash, implemented Eternity server, RSA in 3 lines of perl on t-shirt for exportation Ian Goldberg, US, broke many cryptosystems with David Wagner. Mihir Bellare, US, Random oracles, provable security Philip Rogaway, US, Random oracles, provable security
Maybe a better approach would be to take one of the existing list of cryptographers that can be found on the web, and add all of those names for starters. Then, each name can be anotated little by little.
- Yup, reasonable suggestion. Been on my list for quite a while. Anyone willing to undertake it? If not, i'm sure I'll get around to it... . ww 14:05, 10 May 2004 (UTC)
- I'd recommend a modicum of filtering if importing a web-based "list of cryptographers", as the ones I've seen tend to be quite exhaustive. I would suggest that we include modern cryptographers only if we have an existing article on them already, or if there should be an article on them, however that's decided. (See also: Wikipedia:Criteria for inclusion of biographies). — Matt 14:20, 10 May 2004 (UTC)
Need two cryptographers?
I fished the following people from the list of mathematical topics:
Étienne Bazeries -- military cryptanalyst
Whitfield Diffie -- cryptographer
William F. Friedman -- cryptologist
Martin Hellman -- cryptologist
Bruce Schneier -- cryptographer
The first and the fourth do not show up in the list of cryptographers. Would anybody be interested in adding them? I will remove them from the list of mathematical topics (I will remove the mathematicians too, but the latter will go to the list of mathematicians). Cheers, Oleg Alexandrov 01:08, 29 Apr 2005 (UTC)