Wikipedia talk:WikiProject Cryptography/Archive 3

"MD series"? Looking for suggestions

Hi guys,

I'm not (yet) familiar with the guidelines of this project. Someone asked on the MD4 talk page whether MD4 was really the fourth in a series. As you can see from my reply there I'm looking for a way to add that historical information without duplicating it in all MDx articles. Perhaps you have already an established way to do this. Otherwise I was thinking to a little "The MD series" infobox, with (e.g.) the year of publication/creation of each algorithm and its inventor. Opinions? —Gennaro Prota^•Talk 23:26, 18 November 2006 (UTC)

I quickly put up a simple infobox to cope with the problem (see MD4). I know it's far from professional at the moment, but being unfamiliar with this project guidelines, I wanted to hear some comment first. If it is basically ok I'll make refinements. BTW, any reason why the MD4 article has no {{CryptographyProject}}/{{CryptographyReader}} tags? —Gennaro Prota^•Talk 15:04, 21 November 2006 (UTC)

Perhaps the infobox could be turned into a generic infobox for hash functions, and the series could be linked together through the 'derived from' and 'derived to' fields. Infoboxes should generally outline the factual characteristics of the subject, not provide context for the article. The latter is typically achieved through navboxes (navigation boxes) — we can use those for this purpose. In addition, a new "MD series" article could be created to serve as the main article for this infobox. Thoughts? -- intgr 17:35, 24 November 2006 (UTC)

I agree, although I'd query whether we actually need a new MD series article? Could we not cover them all in Ronald Rivest? Only MD2, MD4 and MD5 are algorithms of note, and of those (I believe) only MD4 and MD5 are related. — Matt Crypto 17:43, 24 November 2006 (UTC)

If we're not going to have articles on MD1 and MD3 then is it really worth a navbox? Although the existence of such hash functions should be documented somewhere. Or what do you think should be the criteria for the navbox, would it be "MD series hash functions", "hash functions by Ronald Rivest" or something else? Also, wouldn't it be duplicating {{crypto hash}}? This problem is harder than I thought. :) -- intgr 19:16, 24 November 2006 (UTC)

Stablepedia

Beginning cross-post.

See Wikipedia talk:Version 1.0 Editorial Team#Stablepedia. If you wish to comment, please comment there. ★TWO YEARS OF MESSEDROCKER★ 03:44, 26 November 2006 (UTC)

End cross-post. Please do not comment more in this section.

Extremely poor entry on Secure Hash Standard: let's improve or remove it

Hi guys,

I just came upon Secure Hash Standard and I don't think I need to comment on how bad it is. Of course I'll try to help in as far as I can but I'm looking for others to work together. I'd say the first thing to do is to prepare a todo-list; or just decide to make the page a redirect to SHA hash functions. What isn't acceptable is the no-contents entry we have now. —Gennaro Prota^•Talk 19:14, 23 December 2006 (UTC)

You may want to click on the "Edit this page" tab and fix the issues in question. Samboy 20:55, 13 February 2007 (UTC)

Wikipedia Day Awards

Hello, all. It was initially my hope to try to have this done as part of Esperanza's proposal for an appreciation week to end on Wikipedia Day, January 15. However, several people have once again proposed the entirety of Esperanza for deletion, so that might not work. It was the intention of the Appreciation Week proposal to set aside a given time when the various individuals who have made significant, valuable contributions to the encyclopedia would be recognized and honored. I believe that, with some effort, this could still be done. My proposal is to, with luck, try to organize the various WikiProjects and other entities of wikipedia to take part in a larger celebrartion of its contributors to take place in January, probably beginning January 15, 2007. I have created yet another new subpage for myself (a weakness of mine, I'm afraid) at User talk:Badbilltucker/Appreciation Week where I would greatly appreciate any indications from the members of this project as to whether and how they might be willing and/or able to assist in recognizing the contributions of our editors. Thank you for your attention. Badbilltucker 17:46, 30 December 2006 (UTC)

Interesting crypto primitives

I found the following set of crypto primitives yesterday Fast Flex. It includes a dtream cipher, a block cipher, and a hash primitive. Samboy 21:01, 13 February 2007 (UTC)

Proposed rename of Category:WikiProject Cryptography templates

I have proposed that Category:WikiProject Cryptography templates be renamed to Category:Cryptography templates; please see Wikipedia:Categories_for_discussion/Log/2007_March_12#Category:WikiProject_Cryptography_templates for my reasons and to comment. Thanks. Mike Peel 23:22, 4 March 2007 (UTC)

Hi Mike. I have nothing against the name change since the category holds both article templates and wikiproject templates. The reason we put all our crypto templates into one single category is that we don't have that many templates, and template categories are just there to help us editors to keep track of our templates. Thus it is easier to have them in one place.

But what I do disagree about is that you seem to prefer to organize categories exclusively in a tree fashion. The purpose of categories is NOT to label things, but to help us editors to FIND things. (And for the article name space also to help the general public to find things.) Thus having several paths to a category helps us find things and is a good thing. For instance we usually organise cryptography something like this: (Computing->Computer security->Cryptography) (Maths->Applied maths->Cryptography) (Communication->Computer communication->Protocols->Cryptography) and so on. And since the "Category:Cryptography templates" contains both article and project templates it also fits under "Category:WikiProject templates". So I re-added some categories to "Category:Cryptography templates".

--David Göthberg 14:20, 5 April 2007 (UTC)

Revitalizing the project

This project has felt dead for a while. Coverage of Cryptography on Wikipedia isn't that great and could use a lot of improving in a lot of areas. As a project we have few featured articles and only two good articles (see the project page). I'd like to start up an article improvement drive for Crypto articles. The idea would be for us to pick an article and improve it to at least good article status, preferably featured status, before moving on to do the same for another article. A shortlist of candidates I started was taken from the list of articles on Wikipedia:WikiProject Cryptography/Featured. Here it is: Alan Turing, anonymous remailer, digital signature, frequency analysis, grille (cryptography), one time pad, pretty good privacy, purple code, secret sharing, substitution cipher. In particular, I'd like to see us work hard on one of digital signature, one-time pad, or substitution cipher, as these are all very central topics in cryptography that need to be covered well. Thoughts? And, is anyone actually active around here? Mangojuice^talk 19:59, 23 March 2007 (UTC)

I suppose you got your question answered. All we need is a new Matt Crypto. Is that too much to ask for? :(

Anyway, I'm still somewhat active, and a few other people have also appeared on the radar of my watchlist, but I'm personally not very good with longer articles. I guess people only edit what they know enough about and are interested in. -- intgr 05:46, 5 April 2007 (UTC)

I can assist, time and knowledge permitting of course. Mmernex 14:57, 20 April 2007 (UTC)

I think many of us are still around and new editors join every now and then. But just as Intgr said: I guess most of us prefer to edit what we know about and are interested in. And you know, we all need some wiki vacation every now and then. (That is, to get away from Wikipedia sometimes to see the real world, or work on our crypto software projects or similar.) --David Göthberg 22:43, 19 May 2007 (UTC)

Wikipedia:Articles for deletion/StartCom -- notable or not?

The following articles have been nominated for deletion as a group:

• StartCom
• StartCom Certification Authority
• StartCom Linux

While there's a conflict of interest concern that these were added by User:Startcom, the main issue as I see it is one of notability. I encourage Linux-knowledgeable editors to weigh in on this topic at the AfD, especially if you know of reliable sources that would establish notability per the Notability Guideline for Companies (and their products). If there are no reliable sources, that would be helpful to know, too. I don't know enough about the topic to understand which if any of the hits I'm turning up in a Google search meet the Reliable Sources Guideline -- for instance, does a listing on DistroWatch[1] count? Likewise, does their connection[2] to the XMPP Federation[3] confer notability? Thanks, --A. B. ^(talk) 13:21, 5 April 2007 (UTC)

Requesting a peer-review of the VEST website for NPoV, etc.

I have made an extensive revision of the VEST cipher page on behalf of Synaptic Laboratories Limited, in response to false and/or misleading statements that have progressively entered that page over the last 6 months, to update the latest attack status and make other textual and editorial improvements. Please refer to the VEST talk page for detailed information on the recent changes to the page.

We have strived to maintain a NPoV in our revision and have made use of extensive verifiable references. However, we are aware that we may not have fully achieved this objective and are proactively encouraging an open peer review and discussion of the page. I am unfamiliar with the specifics of the peer-review process in Wikipedia and have opted to post here before attempting to follow the steps on Wikipedia:Peer_review.

I encourage discussions on the VEST talk page preceding significant editing.

Thanks.,

Benjamin Gittins 14:39, 20 April 2007 (UTC) :: Chief Technology Officer :: Synaptic Laboratories Limited.

Is verbatim copying of U.S. government crypto sources plagiarism?

I came across a very recent Wikipedia article (inserted within the last few weeks) titled "Duquesne Spy Ring" which is a near verbatim copy of a 14-page article with photos that's currently on the FBI's website.  I posed a question about the article on the Wikipedia:Village Pump, policy section (#35. When does verbatim copying from a government website become plagiarism?").  Both the FBI's article and the identical Wikipedia article are linked in the text of that policy question.

I have disputed the author's use of the {USGovernment} template because (a) his article is a verbatim copy of a lengthy U.S. government article, and therefore an inappropriate citation, and (b) that U.S. government source he cited in the template is clearly not the source he used to create his Wikipedia article.  However, I haven't found any clear policy statement in Wikipedia that prohibits or even discourages extensive verbatim copying of sources in the public domain.

My question here relates directly to eleven good, A-class cryptography articles from U.S. Government public domain sources than I could easily Wifify and insert verbatim in Wikipedia.  I obtained all eleven publications at the National Security Agency's National Cryptologic Museum.  All eleven publications are U.S. Government publications.  All include the author's name, but none have copyright notices.  And because they are U.S. government publications I presume them all to be in the public domain — and can be freely copied and used.  If there's no real policy concerns by the Wiki administration about wholesale copying of U.S. government sources, how do members of this project feel about inserting (Wikified) verbatim U.S.government cryptologic articles in Wikipedia with only the current (and IMHO inadequate) {USGovernment} template used as a source cite at the end of each article?  Do you believe it's plagiarism to do that?  Or at the very least unethical?  Is there some other template or flag that can be used to alert readers that such articles are direct copies from another source?

Don't get me wrong.  I have no intention of copying verbatim the government's articles I have in my possession, and inserting them in Wikipedia.  What I'm looking for is suggestions and guidance on what to me seems to be plagiarism — the verbatim copying of U.S. government publications — and its prevention if it's considered to be an undesireable practice in Wikipedia.

Any help?

K. Kellogg-Smith 02:46, 12 May 2007 (UTC)

My personal opinion here: do it!! It's free content, it's probably a vast improvement on the articles in question, and once submitted, the articles can likely be improved by editing. Watch out for WP:OR, though - are all those articles adequately sourced? As long as we attribute the information to the source and state that it's public domain, we're covered, and the authors of the government stuff get proper credit because we'll attribute them (which we don't even have to do). Personally, if I was one of those authors, I would be proud that my work was reaching as many people as possible. However, I would be cautious in one regard: double check that the documents in question are really produced by the US Federal Government, not merely documents they have permission to display or whatever. But yeah, if they really are public domain, I don't think there's an ethical concern.. but if you do, you could always write to the author and ask permission (not that you'd need to, if it's public domain). I'm not sure about the tagging templates. I suppose I would just begin with a "* This article incorporates text from ''Foo bar'', a US Federal Government publication from the [[National Cryptologic Museum]], which is in the [[public domain]]." at the end of the article. Mangojuice^talk 03:46, 12 May 2007 (UTC)

Well, it is a long standing tradition here at Wikipedia to initiate an article by using a verbatim but wikified copy of some public domain article. It is not considered plagiarism, it is simply reusing good stuff. I seen any amount of articles that state something like "This article was based on an article in XXXX encyclopedia from 1911 which is now in the public domain". But as Mangojuice pointed out, check carefully that the articles you want to reuse really are free to use. Just because they are published on a US government site doesn't necessarily mean they were produced by the US government. --David Göthberg 22:36, 19 May 2007 (UTC)

When does verbatim copying from a government website become plagiarism? Since plagiarism is defined as "Plagiarism is the practice of claiming, or implying, original authorship of (or incorporating material from) someone else's written or creative work, in whole or in part, into one's own without adequate acknowledgement.", verbatim copying becomes plagiarism only when there is inadequate acknowledgement. So I agree with Mango -- if the question is "Should encylopedic public-domain content be added to Wikipedia?", the answer is Yes -- please do add encylopedic public-domain content to Wikipedia, acknowledging it appropriately. Or are you really asking "What is the appropriate way to acknowledge that the entire article was copied from some other source?"? --75.48.165.135 14:33, 1 October 2007 (UTC)

Wikipedia:Articles for deletion/List of passwords used in fiction

If your interested, please see the discussion there.--agr 16:34, 14 June 2007 (UTC)

Ouch, now it is deleted. Would have been interesting to take a look at it. Ah well, it was probably junk. --David Göthberg 16:44, 27 July 2007 (UTC)

Template:Espionage

Hi everyone, today I stumbled on the new Template:Espionage. It seems nice but is currently not categorised and not handled by any WikiProject. After a quick look around I think it should either be handled by WikiProject Military history (talk) or WikiProject Cryptography (talk) or perhaps even jointly handled? So what do you guys think?

I left this message on the talk pages of both WikiProjects. (I am a participant in the WikiProject Cryptography but not the WikiProject Military history.)

--David Göthberg 15:09, 27 July 2007 (UTC)

Page delete? Applications in cryptography

I have suggested that the article Applications in cryptography be deleted. Discuss it at: Wikipedia:Articles for deletion/Applications in cryptography --David Göthberg 16:44, 27 July 2007 (UTC)

New cryptography article search

Today I discovered that Mangojuice recently asked the robot AlexNewArtBot to search for and list new articles that contain cryptography related words. The robot logs what it finds at this page: Wikipedia:WikiProject Cryptography/New articles. So now we can easily find and take care of new articles that might be crypto related. Thanks Mangojuice!

Mangojuice: Is this ready to be fully deployed? That is, can I add information about this on our WikiProject page under "Editor oriented lists" etc?

And how often does the robot run? It seems to be more or less every day, right?

And if we want more search words etc can we ask you for help? Since editing the search words for that bot looked really tricky.

Anything else we need to know about the bot?

--David Göthberg 22:36, 27 July 2007 (UTC)

It's the same bot that does lots of other of these lists. Go to User:AlexNewArtBot to learn about it. I think the filter is working okay: there were some false positives having to do with Rap music (another meaning of "cipher"). The parameters can be tweaked further, too. Yeah, I think it's ready for "prime time." Mangojuice^talk 01:28, 28 July 2007 (UTC)

Extraordinary claims

Moved here by David Göthberg from the talk page of Cryptography:

Please check-out the new Wikipedia page dedicated to a person named Ed Trice. Specifically, the "encryption inventions" section. He claims to have created an "Inexhaustible One-Time Pad Cipher Protocol" (IOTPCP).

You will encounter such unusual quotes as-

In order to never disclose the exact nature of this encryption technology, a patent for it was never filed.

The encryption engine works like a digital version of an Enigma machine but of functionally infinite size ...

... with an inexhaustible key size, the IOTPCP encryption method is virtually immune from brute force attack.

This is original research of the most unverifiable kind with the most grandiose claims. There are no references, no program has been written to implement its "infinite" cryptographic algorithm and no paper has been published in any journal. --AceVentura 05:31, 29 July 2007

Yeah, seems to be a terrible article. Thanks AceVentura for "sounding the alarm" on this. --David Göthberg 14:47, 29 July 2007 (UTC)

Cryptography article naming

This discussion was moved here from the talkpages of David Levy and David Göthberg:

Hello David Levy. I noticed that you started renaming/moving a lot of cryptography articles. Those articles are handled by the Wikipedia:WikiProject Cryptography. The naming of those articles have been discussed and agreed upon by the cryptography editors here on Wikipedia. If you disagree with how we name cryptography articles please discuss that on the talk page of WikiProject Cryptography instead of just mass renaming articles.

For instance I disagree that you renamed Tiger (hash) to Tiger (cryptography). "Hash" is more specific than "cryptography" and it is shorter. And you know, we actually do have categories for hashes. Why do you want to "categorise" the name on the more general level of "cryptography"? Then I could claim you should instead categorise it even more generally as say "algorithms" or "applied maths" or why not "computing"?

Besides, if you look at Category:Cryptographic hash functions you see that we categorise hashes under several areas: "Cryptographic algorithms", "Error detection and correction" and "Hash functions". Since even cryptographic hashes like Tiger is used in more areas than just cryptography.

--David Göthberg 19:01, 3 August 2007 (UTC)

Hello, David! I should explain the context in which I renamed these articles.

Someone left a message on my talk page noting that the WHIRLPOOL article's title did not comply with Wikipedia:Manual of Style (trademarks) and needed to be changed. I noticed that PANAMA had the same problem. As I wasn't familiar with our disambiguation conventions for cryptography articles, I checked to see how others were named and found MD2 (cryptography) (another article about a cryptographic hash function) and numerous other articles with the "(cryptography)" disambiguation. Tiger (hash), conversely, was the only article in Category:Cryptographic hash functions with the "(hash)" disambiguation (and the only one that turned up in my search).

I certainly am not an expert in this area and only sought to follow the established naming convention. If I was mistaken, I apologize. I will, however, note that "hash" is not a particularly good identifier (because the word "hash" has multiple meanings). Please keep in mind that the purpose of titular disambiguation is to indicate the article's general nature (usually via the use of a broad class), not to categorize it. For example, we wouldn't use the title Paul Jones (Canadian actor) unless there were a notable actor of another nationality named Paul Jones. We would simply use the title Paul Jones (actor). —David Levy 22:45, 3 August 2007 (UTC)

Oops, seems we have been sloppy. I would prefer that the MD2 article be named MD2 (hash) instead of MD2 (cryptography). Although I should admit that the difference is really minor.

In the case of Tiger (hash) I definitely would prefer to use (hash) instead of (cryptography) since in computer security and cryptography "Tiger" can also mean two other things. (Tiger trees and Tiger teams.) For me using Tiger (hash) instead of Tiger (cryptography) is like using Kiss (band) instead of Kiss (music), it just feels more clear.

By the way, most names of cryptography related things are not trademarks, they are just names. Therefore I think Wikipedia:Manual of Style (trademarks) does not apply. Although personally I don't know if I prefer WHIRLPOOL or Whirlpool (hash).

So hereby I bring up the naming question again at this talkpage. But I think (and hope) we will come to the same conclusions as the last time. Although it seems last time the discussion ended before we had come to a definitive agreement on how to use the naming. My guess is that most of us felt it doesn't matter that much since we have redirects and have disambig links on top of pages etc.

Here are the old discussions in the archives:

• Archive 1, Crypto_article_naming_conventions
• Archive 2, Disambiguation_naming

(Don't edit the archives, instead discuss here.)

Oh, I should probably mention that Mangojuice suggested using more descriptive naming instead of ( ), so his suggestion was Tiger hash function. I think a nice shorter version could be Tiger hash too.

--David Göthberg 01:42, 4 August 2007 (UTC)

I have no strong opinions on the "hash" vs. "cryptography" issue. (As I said, I was following what appeared to be the established convention.) I can only state that to me, "hash" is not a clear identifier. It's a term familiar to people knowledgeable on the subject of computer programming, but not to others. (Three other definitions of the word "hash" come to mind.) The meaning of the word "cryptography," conversely, is quite clear (despite the fact that I know very little about it). If "cryptography" lacks the necessary specificity, there must be a term that would be clearer to outsiders than "hash" is.

Please see User talk:Matt Crypto#WHIRLPOOL for an explanation of why these are trademarks to which Wikipedia:Manual of Style (trademarks) applies. —David Levy 02:32, 4 August 2007 (UTC)

Yes, I have seen that discussion. And I disagree with you, those names are in most cases not trademarks. (Well, some are patented and then I think they might be trademarks.) But most of them are just names on public domain algorithms designed by academic researchers and any one can implement them, modify them and do whatever they want with them. They are no more trademarks than words like "tractor" or "lorry". However, I think the question if they are trademarks or not doesn't matter. Instead as I see it the question is rather if Wikipedia should enforce "English spelling rules" in all cases, not only for trademarks. Such as no words/names in all upper-case instead of spelling a word/name as it is usually spelled in the literature (like RIPEMD.) Personally I prefer to spell things as we who work with it are used to spell it.

By the way, "Whirlpool" indeed is a trademark, but not for a cryptographic hash function, instead for the Whirlpool Corporation which makes home appliances.

Oh, and regarding the term "hash". Yes, it probably is unclear to outsiders, but we have no better name for it. (Well, the full name is "hash function" and cryptographers sometimes call it a "message digest".) But as I see it the disambig naming such as Tiger (hash) doesn't have to be clear to outsiders, it is enough that outsiders who are for instance looking for the animal see that Tiger (hash) probably is not the animal. And the insiders that actually are looking for the hash knows what it means. We have even worse cases, like there are two different ABC ciphers, ABC (block cipher) and ABC (stream cipher). I just fixed the disambig naming on those.

--David Göthberg 12:38, 4 August 2007 (UTC)

1. You evidently share Matt Crypto's misconception that something must be commercial in nature for its name to be trademarked. That simply isn't true. It makes absolutely no difference whether an algorithm is patented and sold for a profit or released into the public domain and freely distributed. You note that anyone "can implement them, modify them and do whatever they want with them," but that's beside the point. What someone cannot do (without causing problems) is refer to an unrelated cryptographic hash function as "WHIRLPOOL" or "Panama." These are not generic terms like "tractor" and "lorry," but distinctive indicators that uniquely identify the sources of the cryptographic hash functions and distinguish them from those of other entities. In other words, they're trademarks.

You also need to understand that trademarks don't necessarily apply to unrelated categories. Yes, there is a Whirlpool appliance brand, and this trademark prevents others from using the name "Whirlpool" in conjunction with such appliances. It does not however, bar the use of the word in other contexts that are unlikely to cause consumer confusion. This is because "Whirlpool" is a dictionary word (not an invention of that company). "Coca-Cola," conversely, is distinctive enough that no one other than the Coca-Cola Company (or someone operating under the Coca-Cola Company's authority) would be permitted to market shoes or light fixtures under that brand.

2. I agree with you that this is irrelevant to the real issue at hand. Matt attempted to cite a technicality, but he didn't attempt to provide an actual reason why the principle behind the MoS entry in question shouldn't apply to this situation (other than the fact that he disagrees with it).

You note that you "prefer to spell things as [those] who work with it are used to spell it," but you need to keep in mind that this is a general-interest encyclopedia (not a technical journal). Therefore, we use standard English instead of industry jargon. (This applies to my interests as well.)

3. As I said, I have no strong opinions on the "hash" vs. "cryptography" issue. I'm just explaining why I renamed the pages and offering some outside perspective. While significantly longer, the term "cryptographic hash function" would be considerably more informative to readers than "hash" is. (Again, please keep in mind that we're writing this encyclopedia for the general public, not merely for computer enthusiasts.) —David Levy 20:59, 4 August 2007 (UTC)

1) It seems, from our article on trademarks, that, firstly, "The essential function of a trademark is to exclusively identify the commercial source or origin of products or services". Secondly, that one establishes a trademark either by using it in the marketplace, or by registering it with a trademark office. I don't believe that the authors have done either for the algorithms in question. The proposition that merely giving something a name in academic research makes that name into a trademark is a claim that still requires substantiation.

2) But moving beyond trademarks, you say that I "didn't attempt to provide an actual reason why the principle behind the MoS entry in question shouldn't apply to this situation". Well, I did imply that a more important principle should take precedence, namely, that "Wikipedia should use the names and styles that everyone else uses in the literature of that topic", in preference to general purpose rules on capitalisation of English. I don't see how any reader is served by Wikipedia using one capitalisation if everyone else uses another. Would you, for example, insist that "ISO" be written "Iso" everywhere on Wikipedia, even though in actual usage people normally use all-capitals?

3) "(cryptographic hash function)" isn't a bad disambiguation suffix, but perhaps "(hash)" is a bit better just because it's shorter. I don't think it's particularly important for comprehensibility: a non-technical reader can readily be informed as to what the general topic is at the article or the relevant disambiguation page. — Matt Crypto 22:15, 4 August 2007 (UTC)

1. The word "commercial" has multiple definitions. When I noted that a trademark needn't describe something commercial, I was referring to the context that you (and subsequently David) introduced to the discussion ("prepared, done, or acting with sole or chief emphasis on salability, profit, or success"). Our article merely refers to something "of, pertaining to, or characteristic of commerce."

No, our article does not indicate "that one establishes a trademark either by using it in the marketplace or by registering it with a trademark office." That's how someone establishes proprietary rights in relation to a trademark, a concept that's entirely irrelevant to the matter at hand. Whether the people behind the WHIRLPOOL hash function could successfully sue to defend their trademark rights has absolutely no bearing on the manner in which the name is used. I will, however, note that "use in the marketplace" does not mean "use for monetary gain."

Indeed, merely giving something a name in academic research doesn't make that name into a trademark. Distributing that thing to consumers and using a name as a distinctive indicator that uniquely identifies the thing's source and distinguishes it from those of other entities does.

2. Again, you're explaining why you disagree with the MoS entry and believe that it should be changed. You aren't citing a substantial difference between this situation and others to which we apply the rule in question. (You're merely attempting to invoke a technicality with no practical significance.)

"ISO" is an initialism. Styling an acronym or initialism in all-uppercase lettering is a normal convention in the English language. (This is why the MoS advises us to use the spelling "MCI" instead of "Mci.") Styling a word in this manner is not a normal convention in the English language. —David Levy 04:04, 5 August 2007 (UTC)

ISO is not an initialism; neither are a number of other pseudo-acronyms. — Matt Crypto 08:03, 5 August 2007 (UTC)

The ISO that you're referring to is commonly treated as an initialism (regardless of the name's actual origins), and "ISO" is an initialism for numerous other entities. WHIRLPOOL is simply a nonstandard styling of "whirlpool." —David Levy 08:12, 5 August 2007 (UTC)

If you were me, you'd tell me that "You aren't citing a substantial difference between this situation and others to which we apply the rule in question." The reason we use ISO and not Iso is simple: everyone uses ISO. We'd be crazy to use a novel capitalisation that nobody else uses just to satisfy some general-purpose advice about how to style trademarks. The main reason the MoS trademark rule exists is to avoid Wikipedia sounding like a COMPANY™ advertisement, where every mention of a PRODUCT™ is unnaturally over-emphasised. Perfectly reasonable that we avoid that. In the crypto world, however, some algorithms are just normally rendered capitalised. Let's not rigidly apply a rule outside of its intended context. — Matt Crypto 08:35, 5 August 2007 (UTC)

This is its intended context. This is a general-interest encyclopedia, so the rules of the "crypto world" don't override those used by everyone else. What's appropriate for Cryptopedia isn't necessarily appropriate for Wikipedia (just as there are linguistic conventions that are widely used within industries of interest to me that wouldn't be appropriate here.)

Again, ISO (the one that you're referring to) is commonly treated as an initialism in everyday use. It serves as an indirect abbreviation of the organization's name, and it's used alongside actual initialisms of related organizations. It was selected, in fact, as an interlingual alternative to the English initialism "IOS" (International Organization for Standardization) and the French initialism "OIN" (Organisation internationale de normalisation) as a means of promoting the underlying principle of international standardization.

WHIRLPOOL, conversely, is simply "whirlpool" rendered in all-uppercase letters for decorative/stylistic purposes (just as "DIRECTV" and "REALTOR" are). —David Levy 09:02, 5 August 2007 (UTC)

I would suggest that your argument that ISO is "commonly treated as an initialism in everyday use" is an application of the principle that Wikipedia should favour the common terms in everyday use. Further, there a bunch of other pseudo-acronyms which demonstrate how we apply this principle: e.g. UNISON. The question boils down to whether Wikipedia should A) adopt the styling that people actually use, or B) use a different styling. For me, the first seems the most helpful approach, even for a general audience. — Matt Crypto 09:49, 5 August 2007 (UTC)

1. I disagree with your argument (which I view as an oversimplification). An important consideration is the reason behind a term's styling. If it's because it's an acronym/initialism (or commonly treated as such), it complies with the standard rules of the English language. If a term is arbitrarily written in all-uppercase letters purely for decorative purposes, it does not comply with the standard rules of the English language.

2. Nonetheless, your argument is reasonable. Again, however, it's an argument in favor of modifying the MoS. —David Levy 10:05, 5 August 2007 (UTC)

1: Regarding trademarks: Yes, I know that trademarks don't necessarily apply to unrelated categories. My family and friends own several registered trademarks so I have some understanding of the issue. (Yes, they paid the government for the right of exclusive use within their area of business.) But I should mention that I live in Sweden and thus just know how trademark rules work in Sweden. And at least over here just naming something you invent/design does not give it automatic protection, unless you register it at the government patent office and pay the fees. You only get automatic protection if you use the name in business frequently and publicly enough. (So if you only used the name when talking to some customers but have not marketed it publicly you are not protected.) So as far as I see most of the cipher names at least here in Sweden are not trademarks since they are not used in marketing and business by the cipher designers. Right here on Wikipedia we have a splendid example of that cipher names often are not protected, there are two totally unrelated ciphers with the same name: ABC (stream cipher) and ABC (block cipher).

2: And yes if cipher names are trademarks or not doesn't matter that much. (Actually, I feel that if they are trademarks that would actually be a stronger incentive to not spell them differently. After all "trademark" means it has some kind of protection.) Anyway, the real question is if we at Wikipedia feel we can enforce other spelling than what people in the field call something. And my feeling is that in most cases we should remember we are an encyclopaedia, which to me at least means that we document the reality, we don't do original research and don't strive to change facts and word usage. (Although I admit I have sometimes broken that rule and chosen a much less used name as the article title because I and other people I have talked with think that that was the better clearer name.)

3: And yes Tiger (cryptographic hash function) would be much clearer than just Tiger (hash). However I feel that becomes too long. See, I frequently paste Wikipedia URLs into chat rooms, emails or even spell them out over telephone and thus know the value of short article names. And if anyone seeing Tiger (hash) wonders what that means they are just one click away from finding out! And you know "cryptographic hash function" already is a short form since we got tired of writing the full correct name "cryptographically secure hash function". Hehe, let's see:

http://en.wikipedia.org/wiki/Tiger_%28cryptographically_secure_hash_function%29

Now, that is just hideous! :))

Ah well, as I always say to my Internet and crypto students: Naming is messy business. Different people use different names for the same thing and the same name for different things, and names tend to change over time. So it is more important to know how something works than what it is called.

--David Göthberg 01:25, 5 August 2007 (UTC)

1. Indeed, a prospective trademark holder's rights and responsibilities vary from one jurisdiction to the next. That, however, isn't relevant to the matter at hand. The MoS entry in question has nothing to do with legal issues. It refers to the tendency of some trademark holders to apply nonstandard styling to their trademarks (and our decision to instead apply normal English-language conventions).

2. That's a reasonable argument in favor of changing the MoS. It does not, however, justify a special exception in this case.

3. No matter what title we assign to the article, Tiger (hash) (http://en.wikipedia.org/wiki/Tiger_%28hash%29) will continue to function as a redirect. —David Levy 04:04, 5 August 2007 (UTC)

2: Regarding uppercase spelling like WHIRLPOOL: So, the question seems to be: Are Wikipedia supposed to be an encyclopaedia (which to me at least means that we document the reality, we don't do original research and don't strive to change facts and word usage) or is it supposed to be a schoolbook to teach people "proper" English? I vote for it to be an encyclopaedia.

Then we have the issue that applying Wikipedia rules to hard like that will make experts in a field like us feel unwelcome here. I don't think scaring away the experts in a field is a good strategy for Wikipedia. (Even though I probably am an "übergeek" in some of the fields I write about I do strive to write in such a way that even laymen should understand the articles. Or at least the first half of the article.)

3: Well, sure Tiger (hash) will still be a redirect even if we choose another longer name for the article. However, I think that when cutting and pasting URLs to articles most users don't first check "what links here" to find a shorter and nicer URL to use.

--David Göthberg 14:21, 5 August 2007 (UTC)

2. It isn't "[changing] facts" or "do[ing] original research" to drop a purely decorative stylistic device (especially when we note this in the article). "WHIRLPOOL" is merely a fancy rendering of "Whirlpool." If it were customary to always write the name as "WHIЯLPOOL" (with alternating green and purple letters and a Я simulating a backwards "R"), we wouldn't do that either.

I seriously doubt that being asked to follow normal English conventions will scare away many editors. I haven't been scared away by Wikipedia's failure to adopt the linguistic conventions commonly applied in the industries of interest to me.

But again, despite my obvious disagreement, it's perfectly reasonable to present the above arguments as reasons to change the MoS.

3. I would argue that titular clarity is more important than short URLs. Someone seeking the latter can use one of the numerous services that create convenient redirect links. Something like "[http:// snurl . com /tigerhash http:// snurl . com /tigerhash]" is "nicer" than any URL that we can offer directly. —David Levy 15:39, 5 August 2007 (UTC)

I respectfully disagree. We are supposed to be a tertiary source. The reliable sources on which we base our content, english technical journals and book publishers, have competent editors who make decisions on how the rules of english apply in these contexts. We should not be second guessing them. Yes we are addressing a general audience, but lay people who seek out these article are looking for clear, accurate explanations of technical topics. We do not do them a service by introducing typographical conventions that differ from those accepted in the field. As for what to use to disambiguate MD2 and the like, I would suggest "hash function." That is the name of our overall article on the topic. I would avoid the adjectives "cryptographic" or "secure" as the imply an evaluation of the quality of the hash function. --agr 19:46, 5 August 2007 (UTC)

1. If there were some technical reason why the name should be styled as "WHIRLPOOL," I would agree. It appears, however, that this is simply a branding decision (just like "DIRECTV" and "REALTOR") made by its creators. It evidently has no scientific basis or significance; it's nothing more than flashy typography (with official sources noting that the name is a reference to the Whirlpool Galaxy).

2. Once again, you're presenting a reasonable argument for changing the MoS (but not for ignoring it in this specific instance).

3. "Hash function" certainly would be clearer than "hash" is, so I (from an outsider's perspective) endorse such a compromise. —David Levy 22:00, 5 August 2007 (UTC)

David, why don't you bring up this discussion on the talk pages of iPod, iTunes and iMac and see what happens? Those names are definitely trademarks so shouldn't those articles be named Ipod, Itunes and Imac? And go ahead, be bold, change the spelling of those words in the article texts too while your at it. That's what you did to the WHIRLPOOL article didn't you? Oh wait, the Wikipedia:Manual of Style (trademarks) has an exception for them. So, could you explain to us why they are exempt? --David Göthberg 03:11, 6 August 2007 (UTC)

Nowadays, the use of CamelCase and similar forms of case is widely regarded as a normal convention in the English language. The use of all caps for most dictionary words is not. The former isn't disruptive to most people, while the latter is perceived by many as the typographical equivalent of shouting. —David Levy 03:50, 6 August 2007 (UTC)

Technical terms of art are not the same as trademarks. We should follow the conventions used in published sources. MOS is a guideline, V and NOR are policy. That said, I followed the links from the WHIRLPOOL article and I'm not convinced that all caps is the preferred usage. The hash functions home page http://paginas.terra.com.br/informatica/paulobarreto/WhirlpoolPage.html uses big caps/small caps font (there is a better term for this but I can't recall it). The standards pages that the home page links to do not use all caps. So, while I disagree with David's reasons, Whirlpool may be right.--agr 04:35, 6 August 2007 (UTC)

1. The only reason that anyone has cited why "WHIRLPOOL" isn't a trademark is that the hash function in question is freely distributed at no charge instead of patented and sold for a profit. That is irrelevant. (Otherwise, "Wikipedia" wouldn't be a trademark either.) The product in question meets all of the criteria. (But as David Göthberg noted, this isn't even the key issue.)

2. It's entirely verifiable (and not original research) that "WHIRLPOOL" is merely a decorative typographical variant of "Whirlpool"; its creators publicly acknowledge this. If it were customary to always write the name as "WHIЯLPOOL" (with alternating green and purple letters and a Я simulating a backwards "R"), would you expect us to do this as well? Where do we draw the line?

3. I'll also note that the first page from the External links section in the Panama (cryptography) (formerly PANAMA) article refers to the product as "Panama." —David Levy 05:41, 6 August 2007 (UTC)

"A trademark is a word, phrase, symbol or design, or a combination of words, phrases, symbols or designs, that identifies and distinguishes the source of the goods [or services] of one party from those of others." -- from http://www.uspto.gov/web/offices/tac/doc/basic/trade_defin.htm . Trademark is not a synonym for name. The reason typographic questions come up with regard to trademarks is that the mark must be distinctive in some way. There is no suggestion that the authors of this algorithm are assserting trademark rights (unlike Wikipedia). However on their home page, the authors are using a small caps font. The W is clearly in uppercase in the font they use. (Just like the way Wikipedia is displayed in the upper left of each page, by the way). The corrrect way. I believe, to translate that to the font we use is Whirlpool, not WHIRLPOOL. That seems to be what others do as well. So the trademark question is moot. --agr 11:15, 6 August 2007 (UTC)

Okay, but where are you seeing the argument that "trademark" is a synonym for "name"? I've claimed nothing of the sort, and the definition that you quoted is very similar to the one that I've repeatedly cited. I also noted above that this has nothing to do with the assertion of legal protection. I get the impression that you've read only some of the discussion. —David Levy 11:55, 6 August 2007 (UTC)

Whirlpool is the name of a mathematical function or algorithm, just like cosine or Ackerman function or bubble sort. New algorithms are usually given names, either by their creator, or by others who use them. I am not a lawyer, but my understanding is that such names do not automatically acquire trademark status. One has to do certain things with the name, such as registering it or using it in such a way as to acquire trademark rights under common law. Indeed, the practice in the pharmaceutical industry suggests that algorithm names may not even be eligible for trademark status unless there is also a generic name for the algorithm, in which case we would probably use the generic name, as we do with, say, Ridalin. --agr 19:18, 6 August 2007 (UTC)

Again, the matter at hand has nothing to to with the ability to invoke legal protection. In this context, that's irrelevant. Regardless of whether the "WHIRLPOOL"/"Whirlpool" name is legally protected, it's used as a trademark (a distinctive indicator that uniquely identifies its source and distinguishes it from those of other entities). For our purposes, there is no relevant distinction between the creators' decision to style "WHIRLPOOL" in all caps (assuming that this were so, which it appears not to be) and the decision to style REALTOR in all caps. It makes no difference who's making money and who could defend the name in court. What matters is how this unorthodox typographical convention is to be addressed (as prescribed by our MoS). The MoS can be changed, but it should not be overridden on the basis of an irrelevant distinction. —David Levy 23:12, 6 August 2007 (UTC)

The MoS you keep citing is Wikipedia:Manual of Style (trademarks). Trademarks have everything to do with "the ability to invoke legal protection." You are attempting to extend that guideline to cover all technical jargon. Guidelines are supposed to represent a consensus among Wikipedia editors. You are entitled to your views, of course, but there is no reason we should believe a guideline on trademarks represents a consensus on anything but legal trademarks that are clearly marked as such. --agr 13:13, 7 August 2007 (UTC)

You're missing the point. All that matters is how the name is used. If it's used as a distinctive indicator that uniquely identifies the product's or service's source and distinguishes it from those of other entities, it's being used as a trademark. It needn't be a registered trademark or a trademark that will hold up in court. None of that is contextually relevant, and you haven't cited one reason why it should be.

Why does it matter that a product is freely distributed instead of patented and sold? What bearing does this have on how we apply English rules to its name? —David Levy 01:53, 8 August 2007 (UTC)

How American to think of an algorithm as a product. Over here in Europe algorithms are not patentable and have no legal protection as products what so ever (as far as I know).

By the way, I have worked for many years as a researcher (mainly in computer communications but some in security) and invented many different kinds of algorithms and discovered many phenomena and that means I have had to come up with names for all of them. So you think I hold about a hundred trademarks or so? Oh, and sometimes other researches have invented/discovered the same algorithms. Then usually after some time we tend to converge/agree upon a common name for that algorithm. But wait, doesn't that create trademark problems? I mean, you say it is supposed to "identify the product's or service's source and distinguish it from those of other entities". So how can we then use the same name for the same algorithm if several of us invented it independently?

Anyway, it is clear that it is NOT consensus here that names of (cryptographic) algorithms are trademarks. In fact, it seems to be a clear majority here thinking they are not. So David Levy, why don't you sit back and ponder for a while why all of us thinks different from you, instead of you wasting your time here trying to convince all of us we are wrong.

--David Göthberg 04:06, 8 August 2007 (UTC)

1. I don't appreciate having my viewpoint dismissed/degraded as "American," nor do I appreciate being told that I'm "wasting [my] time" by expressing my honest opinions in good faith.

2. As I have written over and over again, this matter has ABSOLUTELY NOTHING to do with the ability to invoke legal protection. I'll note, however, that the inability to patent a product doesn't preclude the possibility of obtaining a registered trademark under which to market it. See, for example, Alavert, a trademarked brand of loratadine introduced in the U.S. by Wyeth after the drug's patent expired. So your prior statement that "some are patented and then [you] think they might be trademarks" is based upon an incorrect premise (though it certainly is possible that only the patented algorithms bear registered trademarks). This is an irrelevant aside, however, as the matter at hand has absolutely nothing to do with the ability to invoke legal protection.

3. Algorithms invented independently by multiple individuals still can be marketed in a manner that uniquely identifies their sources and distinguishes them from other algorithms.

4. I sincerely hope that you aren't attempting to cite this "clear majority" (consisting of three WikiProject members) as community consensus. Also note that my involvement in this discussion was triggered by a complaint from another individual, so you're actually citing a 3–2 majority. This sampling, of course, is not a valid cross section and is far too small to be meaningful.

I also hope that you aren't suggesting that such a debate is to be settled via majority voting. I'm still waiting for someone to cite a reason why the distinction in question is relevant. Again, why does it matter that a product (an item offered to satisfy a want or need) is freely distributed instead of patented and sold? What bearing does this have on how we apply English rules to its name? You previously wrote the following:

"However, I think the question if they are trademarks or not doesn't matter. Instead as I see it the question is rather if Wikipedia should enforce 'English spelling rules' in all cases, not only for trademarks."

Therefore, I don't understand why you've adopted an argument that amounts to the attempted exploitation of a technicality. —David Levy 08:29, 8 August 2007 (UTC)

1: Ehm, I was intending to be a humorous.

2: Anyway, for me it is weird to think of an idea (an algorithm) as a product. But if someone patents an algorithm it makes it clear to me that they view it as a product and thus perhaps makes the name a trademark. (Note, I think algorithm patents are silly and bad.)

3: Ehm, I meant invented by several independent researchers. Thus there is no unique source and thus no "ownership".

4: What I meant is that you keep pointing to the Wikipedia:Manual of Style. But such guidelines are supposed to be built on consensus. But our discussion here clearly shows that there is no consensus on this matter.

Anyway, we have tried to explain to you why we think that algorithm names are not trademarks. Your failure to understand our explanations does not mean we are wrong. (But it of course also does not necessarily mean we are right.)

So understand that we disagree with you, both about if algorithm names are trademarks or not and how we should spell algorithm names. I don't think you can convince us otherwise and it seems we can not convince you otherwise. So accept that there is no consensus on this matter.

--David Göthberg 17:11, 8 August 2007 (UTC)

1. Sorry, I mistook that remark for the type of sincere anti-American sentiment that I frequently encounter.

2. As I noted, I'm using the word "product" to mean "an item offered to satisfy a want or need." I'm not commenting on the appropriateness of patenting an algorithm, as this isn't relevant to my argument.

3. Again, this isn't about ownership. As I noted, someone can legally trademark a name for a product that someone else invented and owned until the patent expired. But that's irrelevant too, because this isn't about trademark protection.

I'll try an analogy. Suppose that a young child sets up a lemonade stand and labels it "MAЯY'S LEMONADE." This name would carry no legal protection, but that wouldn't change the fact that Mary was using it in the same way that the supermarket down the street uses its name—as a trademark. Of course, Mary's lemonade stand probably wouldn't be notable enough for Wikipedia to have an article, but suppose that it were notable for some reason. Would we title the article MAЯY'S LEMONADE? No, it would be called Mary's Lemonade. The fact that this isn't a legally protected trademark wouldn't stop us from honoring a MoS entry that clearly applies in spirit.

Now please embrace your original instinct and set aside the fact that the MoS entry even mentions trademarks. What relevant (from the perspective of writing an encyclopedia article) distinction exists between a soft drink that a company markets as "ULTIMATE" and free algorithm designated "ULTIMATE" by its creator? It's perfectly reasonable to assert that the MoS should be changed to indicate that the all-uppercase spelling should be retained for both, but why should we apply the existing rule to one and not the other?

4. This discussion has clearly established a lack of consensus among three WikiProject members, the editor who left a message on my talk page, and me. The failure of five people (four if we exclude the one who hasn't directly participated in the discussion) to reach consensus on a WikiProject's talk page has little bearing on the consensus (or lack thereof) present within the project as a whole and the resultant guideline.

Also note that I seek to build consensus on this talk page. I'm not demanding that you blindly accept my word as gospel truth. I'm simply conveying my arguments (just as you are). I understand your explanations and respectfully disagree with them. —David Levy 19:43, 8 August 2007 (UTC)

Cryptography article disambig naming, examples/voting

Since we had a lot of discussion in the previous section about several matters lets look at just the disambig naming here. First of all, disambig naming only seems to be necessary when there is a risk of the name being mixed up with some other (usually non-cryptography related) name. So most cryptography articles should not need any disambig naming. --David Göthberg 11:58, 6 August 2007 (UTC)

Here are some examples that we can vote/discuss on:

General

When we do not have any more specific disambiguation naming for a cryptography related article:

1. Block size (cryptography)

• This form was agreed upon long ago by us crypto editors. --David Göthberg 11:58, 6 August 2007 (UTC)
• I agree, but (cryptography) should be reserved for general topics like this.--agr 19:41, 6 August 2007 (UTC)

• Well, I think it can be used as fallback to. Articles can be moved to more specific disambig naming if/when we figure out what to name them. --David Göthberg 05:01, 8 August 2007 (UTC)

Hash functions

1. Tiger hash
2. Tiger hash function
3. Tiger (hash)
4. Tiger (hash function)
5. Tiger (cryptography)

• I prefer Tiger (hash) or Tiger (hash function). --David Göthberg 11:58, 6 August 2007 (UTC)
• Tiger (hash) or Tiger (hash function) would be my preference too. — Matt Crypto 18:09, 6 August 2007 (UTC)
• I prefer Tiger (hash function). Where possible, the the reader should be able to type in the disambiguation term, in this case hash function, and be taken to an article that explains what the term refers to.--agr 19:41, 6 August 2007 (UTC)
• From an outsider's perspective, (hash function) is vastly preferable to (hash). The former at least informs someone that the article doesn't pertain to the other meanings of the word "hash." —David Levy 23:12, 6 August 2007 (UTC)
• Ok, you guys are right, Tiger (hash function) seems best. --David Göthberg 05:01, 8 August 2007 (UTC)
• Per invitation from David Gothberg, my two 1/100 Euros. From a reader's perspective the point of a name qualifier is to be both informative and as fully classifying as [possible. On that ground, I think Tiger (crypto) is better than tiger (hash xxx) as there is a kind of hash function which has no crypto application but which is important in computer science. Let us avoid confusing hte reader, at least those readers who notice. the convention, whatever form it takes, was adopted to evade, on the reader's behalf, herds of disamig pages. Almost any of these would do that, but only one of those listed would avoid further confusion. ww 09:02, 8 August 2007 (UTC)
  • I disagree. Non-cryptographic hash functions should also be disambiguated as (hash function). Cryptographic strength is an attribute that is subject to change, based on new analysis. The word "crypto" is not a synonym for cryptography, except in military slang, and should not be used for disambiguation. I see not reason not to use reasonably specific terms for disambiguation. Lay readers will find out what the term means from the article's hopefully clear introduction, not the words in parentheses. "Algorithm" is far too general. Who knows how many algorithms named tiger are out there.--agr 01:35, 9 August 2007 (UTC)
    • OOPS! Sorry, agr. In every case I meant crypto to be equivalent to cryptography, save for the extra, and treacherous, finger motions. Sorry ot have misled. I think I did something similar a few times in these comments. ww 01:55, 9 August 2007 (UTC)

Ciphers

1. Blowfish cipher
2. Blowfish block cipher
3. Blowfish (cipher)
4. Blowfish (block cipher)
5. Blowfish (cryptography)

• I prefer Blowfish (cipher). --David Göthberg 11:58, 6 August 2007 (UTC)
• Blowfish (cipher) or Blowfish (block cipher). — Matt Crypto 18:09, 6 August 2007 (UTC)
• I also prefer Blowfish (cipher). To a lay reader, the important point is that it is a cipher.--agr 19:41, 6 August 2007 (UTC)
• As above, I would favor Blowfish (cryptography) in the interest of a Reader who might notice it. That it's also a kind of ghoti is merely confusing when what's meant is the crypto sense. Using blowfish (block cipher) is certainly more precise, but also less informative to our assumed Gentle Reader, likewise assumed to be innocent of crypto mavenry. ww 09:19, 8 August 2007 (UTC)
• Blowfish (cryptography) is a clear link to cryptography related titles. Blowfish (cipher) is less clear - 'is there a ciphers project?' Dan Beale-Cocks 17:11, 8 August 2007 (UTC)

Stream ciphers vs. block ciphers

We have at-least one case where there are a stream cipher and block cipher with the same name. Then we need a little extra disambiguation:

1. ABC block cipher
2. ABC stream cipher

or

1. ABC (block cipher)
2. ABC (stream cipher)

• I prefer ABC (block cipher) / ABC (stream cipher). --David Göthberg 11:58, 6 August 2007 (UTC)
• Yup, agree with David. — Matt Crypto 18:09, 6 August 2007 (UTC)
• Agree, but this is an exceptional case. --agr 19:41, 6 August 2007 (UTC)
• Agree with all. Further qualification required as the ido... esteemed inventors of these algortihms didn't have sense enough to check for conlicts, even in the crypto world. ww 09:21, 8 August 2007 (UTC)

Pseudorandom number generators (PRNGs)

(Note that some PRNGs are also stream ciphers and then they probably should be disambiguated as ciphers.)

1. Fortuna pseudorandom number generator
2. Fortuna (pseudorandom number generator)
3. Fortuna PRNG
4. Fortuna (PRNG)
5. Fortuna (CSPRNG)
6. Fortuna (cryptography)

• Very tough choice. I think I prefer Fortuna (cryptography) since Fortuna (PRNG) is to "cryptic" for the general audience. --David Göthberg 11:58, 6 August 2007 (UTC)
• Either Fortuna (cryptography) or Fortuna (pseudorandom number generator) for me. — Matt Crypto 18:09, 6 August 2007 (UTC)
• Fortuna include an entropy accumulator, so I don't agree that it is a pseudorandom number generator. I would prefer Fortuna (random number generator). --agr 19:41, 6 August 2007 (UTC)

• Oh, that is an interesting suggestion, I like it. I see Fortuna as "pseudorandom", but simply "random" is shorter and probably less cryptic for the general audience. So perhaps Fortuna (random number generator) or Fortuna (cryptography). --David Göthberg 05:01, 8 August 2007 (UTC)

• Would Fortuna (number generator) be suitable? That would be shorter and avoid the "random" vs. "pseudorandom" issue. —David Levy 08:40, 8 August 2007 (UTC)

• Our Gentle Reader cannot be assumed to be much informed by random number generator or pseudo random number generator. And, it's not just any random number generator, it's intended to be a CSPRNG, something else our Gentle Reader is not likely to be very clear about. Even lots of us manage to get caught in that gator swamp. So I'd prefer Fortuna (crypto) as filling the disambig purpose and stressing the crypto connection if noticed. ww 09:25, 8 August 2007 (UTC)

Cipher machines

(Note, some cipher machines are rotor machines, some are not.)

1. Mercury machine
2. Mercury cipher machine
3. Mercury rotor machine
4. Mercury (machine)
5. Mercury (cipher machine)
6. Mercury (rotor machine)

• I prefer Mercury (cipher machine). We have used "rotor machine" for some articles about rotor machines but that might be to "cryptic" for the general audience. --David Göthberg 12:56, 6 August 2007 (UTC)
• Either Mercury (cipher machine) or Mercury (rotor machine) — Matt Crypto 18:09, 6 August 2007 (UTC)
• I prefer Mercury (cipher machine) too, but that point out we need a cipher machine article, per my earlier comments.--agr 19:41, 6 August 2007 (UTC)

• Yeah, a cipher machine article has been suggested many times over the years. But pretty much all text we could put in such an article is already in other articles like History of cryptography and so on. But it could probable be a short overview article with links to further reading in those other articles. --David Göthberg 05:01, 8 August 2007 (UTC)

• Surely not one of our Gentle Readers are ciphers, but if one or more would be this might be confusing. I'd prefer Mercury (crypto). It does the job (disambig) and flags the special connection, leaving to the article to discuss just what a cipher machine might be and why this is an example of one. ww 09:28, 8 August 2007 (UTC)

Protocols etc

1. Firefly key exchange protocol
2. Firefly (key exchange protocol)
3. Firefly (protocol)
4. Firefly (cryptography)

• I prefer Firefly (cryptography) or Firefly key exchange protocol. --David Göthberg 12:17, 6 August 2007 (UTC)
• Me, I'd prefer Firefly (cryptography) or Firefly (key exchange protocol). I'm generally in favour of using parentheses to show which part of the article title is the name of the topic, and which part is just there for disambiguation. You can also do the "pipe trick" when writing internal links. — Matt Crypto 18:09, 6 August 2007 (UTC)
• Per my proposed rubric, I would suggest Firefly (key-agreement protocol) or we should change the name of the Key-agreement protocol article.--agr 19:41, 6 August 2007 (UTC)

• Ah, right you are. Now I read up on Firefly, it seems to be a key-agreement protocol. So perhaps Firefly (key-agreement protocol). But if we are in doubt what it is then I think we could fallback to Firefly (cryptography). --David Göthberg 05:01, 8 August 2007 (UTC)

• Is there another notable protocol called "Firefly" from which this needs to be disambiguated? If not, why do we want the added specificity? To me, this doesn't jibe with the Blowfish (cipher) opinions expressed above. —David Levy 08:45, 8 August 2007 (UTC)

• Not that I know of. But I just find "protocol" to be to unspecific since it can mean so many different things, see for-instance Protocol (diplomacy). --David Göthberg 16:23, 8 August 2007 (UTC)

• Devoted readers of my comments (there might be one!) will realize that my favorite here is nr 4, Firefly (crypto). for the same reasons as before. We are not after full explanation of the topic, or even partial explanation in this qualifier business, but rather 1) evasion of disambig pages, and 2) avoiding misleading the non crypto types, as that's the hardest case). ww 09:40, 8 August 2007 (UTC)

• Hehe. I am one of your devoted readers. You often bring up new interesting views on matters. :)) --David Göthberg 16:23, 8 August 2007 (UTC)

(algorithm)

Would it be feasible to use the disambiguation (algorithm) for any of the above? It suddenly occurred to me that this is a fairly specific word recognizable to far more people than something along the lines of "hash function." —David Levy 08:58, 8 August 2007 (UTC)

Not if the point is to distinguish between gerernal algorithms of which there are multitudes and cryptographic algorithms of which there only seem to be multitudes. Too general for anyone who actually notices the disambig qualifier. ww 09:43, 8 August 2007 (UTC)

Well, in a case where there is another algorithm with the same name, this obviously wouldn't work. I was thinking of a case in which the disambiguation is needed because non-algorithm-related things with the same name exist.

I agree with you that it makes sense to use (cryptography) unless further disambiguation is required (as in the case of the two "ABC" ciphers). This is far simpler and less confusing than devising a complicated system of unnecessarily specific descriptions for all of the individual cryptographic topics. —David Levy 09:54, 8 August 2007 (UTC)

Ah yeah, you guys have a very good point here. It might be more important to tell what an article is not than telling exactly what it is. Since that will be easier to understand for the general audience and if anyone want to know exactly what it is they just have to click on the link. And as David Levy points out, it is far simpler for us editors. (Thus avoiding instruction creep.) So for those articles that are about cryptographic algorithms then Tiger (algorithm) perhaps is the most general and well known word we could apply? (Good suggestion David Levy.) And for other non-algorithm cryptography related articles we could use Mercury (cryptography). It all depends on if we see it top down or bottom up and to which audience we want to cater the most. Although I think most people have at least a faint understanding what the words "cipher" and "cryptography" means so perhaps (algorithm) is unnecessarily unspecific? Thus perhaps making (cryptography) a decent compromise even for cryptographic algorithms? Thus perhaps we only need one single disambig word for the whole cryptography area. Of course, another argument is that for instance old cryptographic stream ciphers over time move on to just become insecure pseudorandom number generators, thus making (algorithm) a more timeless choice. And some bad algorithms are questionable to call "cryptographic" already from starters. Argh! It is hard to choose the right resolution for this. (Resolution as in "zoom level" that is.) --David Göthberg 16:23, 8 August 2007 (UTC)

From my perspective, it makes sense to append (cryptography) unless further disambiguation is required. The idea is not to identify the article's precise nature, but to differentiate it from one or more articles that otherwise would have the same title. In doing so, it's a good idea to avoid selecting a term that's likely to confuse or mislead readers, but said term needn't inform the reader of anything beyond the topic's general nature. Unless we need to differentiate two or more cryptography-related articles (as in the case of the "ABC" ciphers), I like the idea of sticking with (cryptography) whenever possible.

What I lack, of course, is expert knowledge of when this term is technically accurate. It seems to me, however, that all of the topics discussed fall under the general heading of "cryptography" (even if they are outdated or aren't particularly good from a cryptographic standpoint).

I suggested (algorithm) because it seems like a recognizable term that satisfies the desire for specificity expressed on this page, but I disagree that such specificity is desirable and believe that the use of (cryptography) in most cases is a better solution. —David Levy 19:43, 8 August 2007 (UTC)

Conclusion?

So what's our conclusion? Should it be Tiger (hash function) and Blowfish (cipher) or should it be Tiger (cryptography) and Blowfish (cryptography)? Or perhaps it is too hard to decide and we'll leave it for now? Personally I am not sure which disambig naming I prefer. --David Göthberg 13:19, 21 August 2007 (UTC)

Encryption scheme vs cryptosystem vs (encryption) algorithm vs ...

This seems as good a time as any to bring this up. I've noticed some inconsistencies with what to call these things. E.g., ElGamal encryption (the article goes on to begin "The ElGamal algorithm is..."), Cramer-Shoup cryptosystem (which begins "The Cramer-Shoup system is..."). There are many articles other that should be made consistent, both in their title and terminology used within the article. I have been hesitant to make any changes because I wanted some consensus on this. I guess other options would be to drop the specifier and just have ElGamal and Cramer-Shoup, or to put the specifiers in parens like ElGamal (cryptosystem), etc..

My vote is for either "encryption scheme" or "cryptosystem". "System", "scheme" are much too generic. I've always disliked "encryption", "encryption algorithm", and "algorithm" because an encryption scheme is three algorithms: encryption, decryption, and key generation. // Blokhead 16:07, 8 August 2007 (UTC)

Hi Blokhead. You bring up a BIG question. So I'll try to answer part of it. First of all, the article name (URL) doesn't have to be exactly the same as the word we use inside the article. So the article normally can be named say ElGamal but if it collides with some name for something else we can use for instance ElGamal (cryptography), but in this case there are both ElGamal encryption and ElGamal signatures so the article probably has to be named ElGamal encryption just as it is now. Then in the first paragraph of the article we can have something like what the article has now: "In cryptography, the ElGamal encryption algorithm is ..." (Or perhaps "ElGamal encryption system" since that seems to be what you prefer.)

Usually in the first paragraph we mention several or even all of the names and bold each of them and fix redirect pages to the article using those names. See for instance the hash tree article.

Then in the rest of the article we often choose to use a shorter name (if the name is long), in this case for instance "ElGamal encryption". I Usually try to be fairly consistent in the article text. But it is not uncommon to be slightly inconsistent and use several of the common names and that can be seen as a way to get the reader used to the different names. When I teach cryptography verbally to engineers I usually am "inconsistent" on purpose so they learn and get used to all the names for the same thing.

Next thing is the choice of words. Personally I dislike the word "scheme", since for me it sounds like "evil scheme". (But that might be because I am not a native English speaker and mostly heard the word in B-movies... :)) Then there's the word "encryption": I think we can use the word "encryption algorithm", "encryption system" or simply "encryption" for all three things (encryption, decryption and key generation). Of course, in most cases it probably is better to use the word "cipher" when meaning all three. But since there are two ElGamal ciphers (encryption/decryption and signing) again I think in this case it is right to use the words "encryption", "encryption algorithm" or "encryption system".

Of course, we should probably also do the "Google test" to see which are the most commonly used words for it. Although I believe that we as "experts" have the right to choose the best word for it in the article, as long as we mention the other words for it in the first paragraph.

--David Göthberg 18:16, 9 August 2007 (UTC)

ROT13

ROT13 has been nominated for a featured article review. Articles are typically reviewed for two weeks. Please leave your comments and help us to return the article to featured quality. If concerns are not addressed during the review period, articles are moved onto the Featured Article Removal Candidates list for a further period, where editors may declare "Keep" or "Remove" the article from featured status. The instructions for the review process are here. Reviewers' concerns are here. hbdragon88 21:57, 5 August 2007 (UTC)

The crypto navboxes

I am currently doing some testing with a slightly new look and new code for the crypto navboxes. The primary visual change is to use dots "·" instead of pipes "|" between the links since that seems to be the new fad for navboxes. I also tried to use the new standard {{navbox}} although that didn't really make our code simpler, but it slightly changed the appearance of our navboxes and added the [hide] feature. (Although personally I find the [hide] feature unnecessary in our case.) So take a look at the old crypto stream box and the new test example. Tell what you think about it on this talk page.

Oh, by the way, I of course will keep the "box in a box" thing we use now. If you don't know what that means take a look at the examples at Wikipedia:WikiProject_Cryptography#Navigational templates.

--David Göthberg 12:00, 21 August 2007 (UTC)

Proposed deletions

• Definite Avalanche (via WP:PROD on 6 October 2007) Deleted
• Encryption Anywhere Hard Disk (via WP:PROD on 4 October 2007) Deleted

--User:Ceyockey (talk to me) 01:47, 3 October 2007 (UTC)

• Message Weaving Algorithm (via WP:PROD on 3 October 2007) Kept; subsequently deleted via AFD
• Imran Shaik (via WP:PROD on 6 October 2007) Deleted

Mangojuice^talk 02:39, 3 October 2007 (UTC)

updated --User:Ceyockey (talk to me) 19:37, 21 October 2007 (UTC)

Concern: Kleptography and Cryptovirology

I'd like to draw attention to Kleptography and Cryptovirology, articles about terms that have not (as far as I know) entered into mainstream usage outside the publications of Young and Yung. There's a hint of self-promotion about them, too. — Matt Crypto 21:24, 20 October 2007 (UTC)

Sort of related articles?

I've been doing a good deal on intelligence disciplines, with much more to do, such as an article that pulls together the disciplines of intelligence and how they work with the political process. SIGINT, which has several daughter articles, was in the cryptography category, and I see no reason to take it out although I only make external references to things such as cryptanalysis.

Nevertheless, I'm sure there are people in this Project that would have expertise, and, hopefully interest. Feel free to look through this set, and, if you are interested, the MASINT set of articles. Howard C. Berkowitz 21:53, 20 October 2007 (UTC)

Proposed deletion: ViPNet Safe Disk and ViPNet Technology

ViPNet Safe Disk (via WP:PROD) ViPNet Technology (via WP:PROD)

--User:Ceyockey (talk to me) 19:41, 21 October 2007 (UTC)

There's a spam filter notice on this talk page...

Spam filter notice From Wikipedia, the free encyclopedia Jump to: navigation, search This is a talk page. Please respect the talk page guidelines, and remember to sign your posts by typing four tildes (132.205.99.122 (talk) 23:12, 29 November 2007 (UTC)). The spam filter blocked your page save because it detected a blacklisted hyperlink. You may have added it yourself, the link may have been added by another editor before it was blacklisted, or you may be infected by spyware that adds links to wiki pages. You will need to remove all instances of the blacklisted URL before you can save. Blacklists are maintained both locally and globally. Before proceeding, please review both lists to determine which one (or both) are affecting you. You can request help removing the link, request that the link be removed from the blacklist, or report a possible error on the local or global spam blacklist talk page. If you'd like to request that a specific link be allowed without removing similar links from the blacklist, you can request whitelisting on the local spam whitelist talk page.

The following link has triggered our spam protection filter: http:// snurl . com Either that exact link, or a portion of it (typically the root domain name) is currently blacklisted.

Return to Wikipedia talk:WikiProject Cryptography.

removed

I've disabled it, but it prevented new sections from being posted here. That's pretty bad, when you're adding a new section or whatever. 132.205.99.122 (talk) 23:12, 29 November 2007 (UTC)

WP:AFD on Error-correcting codes with feedback

Error-correcting codes with feedback has been nominated for deletion, see Wikipedia:Articles for deletion/Error-correcting codes with feedback 132.205.99.122 (talk) 23:13, 29 November 2007 (UTC)

Infoboxes

I know infoboxes full of links are popular, but are we sure they are valuable in this instance?

Why does each page on, for example, stream ciphers benefit from having a big list of stream ciphers on it? I wonder how often these are used - on the rare occasions someone does want to navigate from one stream cipher to another unrelated one, the stream cipher category will do that just fine.

And it only leaves us open to big, pointless fights about which stream ciphers should be in the list.

What do folk think? — ciphergoth 23:30, 2 December 2007 (UTC)

Which "fights" are you talking about? I haven't witnessed any. -- intgr [talk] 03:18, 3 December 2007 (UTC)

Maybe I'm wrong about the fight, in which case the other points still stand. I'm anticipating one, though, since I just edited some lists to match my perception of which ones are considered interesting in the community, and dropped one cipher which was at one point mentioned in practically every article in Wikipedia thanks to a vigorous publicity campaign by its designer. Though I have no vested interest in which ciphers go in the list, I can't reference any Official List of stream ciphers the community finds interesting to back up my choices, so if challenged, we'll have to either put every cipher in the appropriate category in the list or drop the list altogether. Since I don't think the list is that useful anyway, I think we should do the latter, and that we shouldn't wait for a fight before we do it. — ciphergoth 09:04, 3 December 2007 (UTC)

Have pics, need help

(Cross-posting to wikiproject cryptography, wikiproject computer science, and Wikiproject military history)

Over my Christmas vacation, I happened to get out to the National Cryptologic Museum at the NSA. I photographed basically everything in the musuem, (including all the nameplates so that I could later identify the photographs). There's a *LOT* of neat stuff there.

I've uploaded them to commons, and now I need help insterting them into articles and whatnot. You can find the gallery at User:Raul654/favpics/National Cryptologic Museum. Please insert captions into the gallery (red links for any nonexistent articles) and insert the pictures into relevant articles. (talk) 04:39, 30 December 2007 (UTC)

Hi Raul, thanks for uploading these. I'd love to visit the NSA museum some day, so I'm quite envious. I think we've illustrated some articles using NSA museum photos already, but I'm sure there's places we can use these new ones. — Matt Crypto 15:03, 30 December 2007 (UTC)

I've seen a few of the ones we already have - they're mostly pretty low res shots (typically 400x600 or less). The ones I just uploaded are much higher resolution and typically I have two or three of the same exhibit (just to make sure that one of them came out well) Raul654 (talk) 16:09, 3 January 2008 (UTC)