User talk:David Eppstein/2017b

DYK for Mark Barr
Mifter (talk) 12:02, 1 April 2017 (UTC)

Marcia Barbosa
Hi, Professor David Eppstein. I would like to have your help. It is on an article within your area of interest here on Wikipedia (women scientists). I would like to improve the article about Marcia Barbosa. The reason for this is because I am one of her students in a course at UFRGS, and her classes are very good, and she is a very nice person. But, given all of this, I may have a problem of conflict of interest and not be an editor as neutral as Wikipedia requires (unconsciously). Would you like to help me? (please? xD). By the way, I admire very much your work here on Wikipedia. Cheers, all the best, MathKeduor7 (talk) 14:00, 1 April 2017 (UTC)
 * Thanks, but her article already looks in reasonable shape and her specialty is not close enough to subjects I know much about for me to have any special expertise in improving her article. To make any significant improvements in the article, I think you'd need published sources that are independent of Barbosa and that describe her or her work in-depth; do those sources exist? And are they in English? (English is not a requirement on Wikipedia for sources, but I don't personally read Portuguese and machine translations are often not good enough.) —David Eppstein (talk) 03:23, 19 April 2017 (UTC)
 * Thank you! The best source I could find: http://revistapesquisa.fapesp.br/en/2013/10/17/the-bizarre-side-of-water/ . MathKeduor7 (talk) 07:10, 19 April 2017 (UTC)

Re: Geometric series
About Special:Diff/773542597, I am fairly confident that that is not what the formula does. I am not a mathematician, but it is my understanding that "this is a formula for calculating x of y" should usually be followed by an equation where on the left side is a mathematical restatement of the thing to be calculated, "y(x)", and on the right side is a mathematical expression for computing the thing, involving the variable x. Here, you claim that it sums the first n terms, but while the left hand side contains a summation symbol, neither the left nor right sides contain the letter n (not even "sigma", "equals", or "divided by"). How, then, can it be a formula to calculate based on n? ⁓ Hello  71  13:05, 3 April 2017 (UTC)
 * I suspect that this formula actually computes the sum of the ath to bth terms of the series, including a but excluding b and is arrived at by a trivial subtraction of the parts of the series by the formula derived earlier in the article. I do not think that this is notable (in the informal sense of notability) and will thus re-remove it. Please feel free to re-insert it if you edit it so that it actually makes some sense. ⁓ Hello  71  13:08, 3 April 2017 (UTC)
 * (brought here by the link in your edit summary) Yes, there is a presentation problem here; but the right approach is not to remove the formula wholesale -- here it is simply more general than what is announced. I have reverted in the meantime. — Gamall Wednesday Ida (t · c) 16:49, 3 April 2017 (UTC)
 * User:Hello71. A few remarks: 1° you were right that it does not do what it said on the tin 2° I think you are right that it is not interesting, and thus 2'° I think it should be removed as trivial 3° This discussion should have been on the article's talk page rather than here. 4° This discussion should have been had before removal, because a brute deletion of content lit up all the red flags.  — Gamall Wednesday Ida (t · c) 17:08, 3 April 2017 (UTC)
 * After looking at this more carefully, I agree — a separate formula for the sum of a consecutive subsequence within a geometric series (not the partial sum of a prefix) is not interesting, because it's just a partial sum with a different start and can be handled by the other formula. So I'm not going to contest the re-removal. —David Eppstein (talk) 18:15, 3 April 2017 (UTC)
 * User:Hello71 I have put the formula here. As David Eppstein said it gives the sum of the numbers from $$r^a$$ to $$r^b$$, extremes included. There is also a proof of that. It's not a remarkable math result of course, but what's wrong in keeping it in the article? Raffamaiden (talk) 08:51, 24 May 2017 (UTC)

NP-completeness of Sudoku and other Puzzles
David Eppstein, I noticed you removed Sudoku from the list of List of NP-complete problems (games and puzzles). Isn't it correct that a Sudoku is a type of latin square, and a latin square is NP-complete (see note 10 at Latin square). Also, the following puzzles are also listed as NP-complete: Fillomino, nonograms, kakuro, and Nurikabe (puzzle) and they are of bounded size. Finally, the articles about Sudoku and Mathematics of Sudoku say it is NP-complete with an academic reference. Based on these articles and the related references shouldn't Sudoku be in the list?—LithiumFlash (talk) 14:19, 8 April 2017 (UTC)
 * Maybe, but there is a technicality here: Sudoku puzzles are required to have unique solutions, and that puts them in different complexity class than NP. Instead, the type of problem one wants to solve is that there is a promise of having at most one solution and you want to know whether there is zero or one. I don't know whether Sudoku has actually been proved hard for this class.  —David Eppstein (talk) 15:12, 8 April 2017 (UTC)
 * David Eppstein, I don't see anything in NP-completeness that requires such problems to have multiple solutions. In fact it says "...any given solution to an NP-complete problem can be verified quickly," and that would be easier to do when there is a unique solution.
 * These problems also have unique solutions and are in the list: Longest path problem, Vehicle routing problem, and Shortest total path length spanning tree. Also, please let me know if you dispute the reference in the Sudoku articles that it is NP-complete.
 * How about we add Sudoku (and Latin squares) back to the list of NP-complete games (unless you feel NP-completeness needs to be corrected to define such problems as requiring multiple answers). Thank you for your consideration and advise.—LithiumFlash (talk) 17:04, 8 April 2017 (UTC)
 * How about we don't add them. Latin square completion can be added. "Latin squares" is not a decision problem, and the sudoku-like version of Latin square completion is NP-complete but it is not accurate to call it sudoku. The NP-completeness reduction from SAT to these problems can and will produce invalid puzzles with multiple solutions, so it is not a reduction to sudoku. The longest path problem has no restriction that the longest path be unique (there can be multiple equally-long longest paths), so I don't know what you're talking about when you say that those problems have unique solutions. —David Eppstein (talk) 17:59, 8 April 2017 (UTC)
 * Can you clarify "The sudoku-like version of Latin square completion is NP-complete but it is not accurate to call it sudoku"?—LithiumFlash (talk) 18:17, 8 April 2017 (UTC)
 * When you use the reduction to translate instances of something else into things that sort of look like Sudoku problems, there is no guarantee that the results will really be Sudoku problems, because they may not have unique solutions. It may be the case that the reduction produces things with two or more solutions instead, and I don't think it's fair to call those Sudoku problems. It's also necessary to state that it is only NP-complete for boards of arbitrarily large size, whereas Sudoku proper is usually only 9x9. —David Eppstein (talk) 20:09, 8 April 2017 (UTC)
 * The grid size does not affect the classification as to whether it is NP-complete or not. The definition of NP-completeness makes the distinction. You can't add "Generalized Sudoku" to the list without adding "Sudoku", unless you say "Generalized Sudoku but exluding Sudoku with properties (yet to be defined)". I have not seen Sudoku partitioned into such sub-classes in any academic paper. We need to stick with published academic material, and not research or theories conceived on Talk pages.—LithiumFlash (talk) 20:28, 8 April 2017 (UTC)
 * Wrong. The grid size does affect the classification, because only problems with an unbounded size can be NP-complete. If "Sudoku" as it is generally understood means 9x9 boards, then it is certainly not NP-complete (unless P=NP), because it can be solved in constant time. 3^81, the number of ways of filling in the grid, is a constant. This is a very very standard part of the application of NP-completeness to games and puzzles, clearly described in the earliest papers on the subject (on hardness of generalized chess and generalized checkers), so the fact that I have to spend so much time on my talk page explaining it to you is also irrelevant. But if you really want a published reference, you can see my paper "Solving Single-digit Sudoku Subproblems" (published in FUN 2012) which states "when generalized to n × n grids, solving Sudoku is NP-hard". Note (1) the generalization to larger grids, and (2) that I said NP-hard, not NP-complete, in part because of this uniqueness issue. —David Eppstein (talk) 20:37, 8 April 2017 (UTC)
 * Your comments here about Sudoku aren't substantiated with the academic material on the subject. But I'm OK with the way you left the list, mainly because a Sudoku is within the class of generalized Sudoku. I also appreciate you leaving the reference attached, for the case other mathematical study might be explored on the topic. Thank you for your time.—LithiumFlash (talk) 20:46, 8 April 2017 (UTC)
 * Um, I'm beginning to wonder whether you really understand what the various complexity classes mean.  E Eng  21:17, 8 April 2017 (UTC)
 * I'm not working on any article related to that. My editing is done so that Wikipedia correctly identifies if the problem of solving Sudoku is NP-complete based on reliable academic resources.—LithiumFlash (talk) 21:24, 8 April 2017 (UTC)

I don't see why you are claiming that my FUN paper is not part of "the academic material on the subject". But anyway, thank you for your comment "I also appreciate you leaving the reference attached" as it led me to check that reference more carefully. What it actually shows is that, given a solved $n &times; n$ Sudoku puzzle, it is NP-complete to test whether there is a second solution. As you might guess if you thought about it, this doesn't actually say much about the complexity of finding a solution of an unsolved puzzle. —David Eppstein (talk) 22:21, 8 April 2017 (UTC)
 * Your paper is good and it supports the claim that Sudoku is NP-complete. In particular in (6) it reads "The time bound for our Nishio algorithm is exponential" which is in superpolynomial time. Your revised statement doesn't make sense: "Testing whether a solved n × n Sudoku puzzle has a second solution". A solved Sudoku has no more room for manipulation. Proper Sudokus have only a single solution. So I'm not sure what your statement means. Will you help clear it up in some way? (Or just say "Sudoku"). I did add two more references. Thanks.—LithiumFlash (talk) 05:33, 9 April 2017 (UTC)
 * If you don't understand the difference between "this problem has an exponential time algorithm" and "this problem is NP-complete" you should not be editing articles about NP-completeness. The problem that your reference proves hard is one where the input is something that looks like a sudoku puzzle (but may be lsarger than 9x9 and have multiple solutions) together with one of its solutions. The output is a Boolean value: True, if there is another different solution, and False if the given solution is the only one. —David Eppstein (talk) 07:06, 9 April 2017 (UTC)
 * What I said.  E Eng  07:30, 9 April 2017 (UTC)
 * Can you cite the page number and paragraph where Sudoku is found to not be NP-complete? It can't be for a lack of terminology. In the abstract you say "We also show that single-digit deduction in Sudoku is NP-hard." (which does not exclude it from the class of NP-complete).
 * EEng: Is there any reason you nor anyone else have not used David Eppstein's paper as a reference to show a different result on this topic? If anyone feels someone's work has changed the decision then of course an editor can update Wikipedia. (Perhaps not Eppstein if using Eppstein's work due to non-neutral POV). I don't think the list is the best place to first show the result. Mathematics of Sudoku may be better. In any case, the math community would be interested in knowing how Sudoku is classified, not "Testing whether a solved Sudoku has a second solution". The statement currently in the list makes no sense.—LithiumFlash (talk) 12:49, 9 April 2017 (UTC)
 * Speaking for myself, the answer is: Because I didn't know the paper existed, I didn't know the article existed, I find sudoku boring, and for now I'm limited to a cell phone with a tiny screen. There's nothing wrong with DE citing his own papers on this.  E Eng  16:39, 9 April 2017 (UTC)
 * My paper doesn't really show what LithiumFast wants to show, that the problem of computing the whole solution to the original (9x9, unique solution) version of Sudoku is NP-complete (because it's not true, for multiple reasons). It does contain an NP-completeness proof (for a different problem related to Sudoku) but its main contribution is an implementation technique for a specific Sudoku deduction rule. So I have no intention of adding it as a reference. As long as LithiumFast goes into this with the attitude of finding a reference to show what he wants to show, rather than understanding the references well enough to accurately describe what they actually say, his edits are not likely to see much success. —David Eppstein (talk) 17:19, 9 April 2017 (UTC)
 * Your paper has good information and might be a good source of material for Sudoku solving algorithms. But it can't be used as a basis to say Sudoku is not NP-complete. As usual thanks for you attention and contributions.—LithiumFlash (talk) 17:35, 9 April 2017 (UTC)
 * Of course it can't; nobody publishes papers of the form "X is not NP-complete" (among other reasons, because you can't prove such a thing without also proving P≠NP). And I didn't say that's what it contained, so I don't understand why you are trying to tell me what's in my own paper. But likewise, the fact that my paper doesn't say this cannot be used as a reason to re-add the bogus claims you have been repeatedly adding to the list. Since it doesn't seem to be sinking in, let me be explicit: You should not be editing articles about NP-completeness. Your comments here demonstrate that your understanding of this subject is too poor for your edits to be accurate. —David Eppstein (talk) 18:23, 9 April 2017 (UTC)
 * The entry still requires a "needs clarification" tag. Readers may want to know if Sudoku is in the list of problems that is NP complete. People don't care if a finished grid has another solution. Proper Sudokus never do.—LithiumFlash (talk) 19:17, 9 April 2017 (UTC)
 * True. That's why the proof you referenced is unsatisfactory for a claim that Sudoku is NP-complete. And that's why I edited the entry to clarify what they prove rather than claiming what you would like to be true, with a bad reference. There is no clarification needed; the entry is correct as I stated it, unless you have messed it up yet again. Do you have another reference that contains a better proof? (Not merely one that surveys a bunch of results but doesn't prove any.) —David Eppstein (talk) 19:27, 9 April 2017 (UTC)
 * Here, plus material to show there has been peer review and consensus on the topic:
 * http://www-imai.is.s.u-tokyo.ac.jp/~yato/data2/MasterThesis.pdf
 * http://theinf1.informatik.uni-jena.de/publications/sudoku-weller08.pdf
 * https://www.cs.wmich.edu/elise/courses/cs431/icga2008.pdf
 * http://geevi.github.io/2014/puzzles.html
 * —LithiumFlash (talk) 19:55, 9 April 2017 (UTC)

D.E., I'll leave the rest to you. I gotta finish Lewis.  E Eng  20:08, 9 April 2017 (UTC)
 * Your first source says "In the 1970’s, Stephen Cook, Leonid Levin and Richard Karp found that sudoku is not just a one off case, for which polynomial time algorithms cannot be found", which is preposterous (literally).
 * Your last source is an MA thesis. MA theses are never reliable sources -- even PhD theses must be used with extreme caution.
 * Too busy myself today to do much more than revert Lithium's bad edits. But thanks for the backup and for finding the problems with these sources. —David Eppstein (talk) 21:01, 9 April 2017 (UTC)
 * Is there anything published that disputes the findings in those papers (not speculation or conjecture on a Wikipedia Talk page)?—LithiumFlash (talk) 21:12, 9 April 2017 (UTC)
 * That's not the way reliable sourcing works. We need a paper that actually proves the claim, not just papers that vaguely wave their hands around it and nothing that refutes them. And again, YOU ARE NOT QUALIFIED TO EDIT THIS SUBJECT. It is very clear from your earlier comments. Please stop trying until you learn this subject better. Wikipedia does not require credentials, but it does require competence which you appear not to have. Continued incompetent editing may lead to a block. —David Eppstein (talk) 21:21, 9 April 2017 (UTC)
 * I'm not changing the list anymore and my last three edits were either adding a tag or adding a reference. Also please see Conflicting sources and Fringe theories. But I'm not going to worry about it anymore. I'll leave the list as it is now with the statement by you that doesn't make sense. It's fine with me.—LithiumFlash (talk) 22:07, 9 April 2017 (UTC)

About references in Kathryn Hess
I saw your changes using rma and ran. What a great way to make an academic's "Selected publications" available as references without duplicating them. I've been fixing references since 2012, but there is always something new! StarryGrandma (talk) 22:50, 10 April 2017 (UTC)
 * Glad you liked it!, see, maybe your weird format will catch on after all. —David Eppstein (talk) 02:35, 11 April 2017 (UTC)
 * And here I thought, like most great artists, I'd never be truly appreciated until after I'd died. Listen, input is needed at Talk:Harry R. Lewis.  E Eng  04:03, 11 April 2017 (UTC)

list of people by Erdős number‎

 * Why did you revert my edit? You did not provide any comment or explanation to your revert.  Lawrence Krauss and James Hartle are cited as being Erdős number three at " ."  Nicole Sharp (talk) 15:57, 11 April 2017 (UTC)
 * Krauss and Hartle can stay but we have long since decided not to list people with number 4 here. Also your edit summary "no response on user talk page" is ridiculous — I am not glued to my watchlist every minute of the day. Give people time to come back from whatever else they're doing before deciding they're nonresponsive. A day might be reasonable, or a week in some cases, if you want a response. A half hour is way too short to expect a response. —David Eppstein (talk) 17:01, 11 April 2017 (UTC)
 * I figured it might have been accidental (reverting all of the edits instead of just removing the new section), so I went ahead and unreverted. If it looked more controversial I normally would have waited longer, yes.  I didn't mean any offense; sorry.  Nicole Sharp (talk) 17:59, 11 April 2017 (UTC)

Horning in
Re
 * The problem of assigning a sign to each variable in such a way that no two variables end up as positive in the same clause as each other, making...

-- why not
 * The problem of assigning a sign to each variable in such a way that no two variables end up as positive in the same clause, making...

-- ? Is there some subtle function to as each other?  E Eng  04:21, 13 April 2017 (UTC)


 * No, the simpler version is still ok. —David Eppstein (talk) 04:27, 13 April 2017 (UTC)

Suggest semiprot
...for User:ElKevbo. This has been going on a while.  E Eng  09:53, 15 April 2017 (UTC)
 * You mean his talk page, right? I'd hesitate to protect a user talk page without the owner's request, but if ElKevbo would like to try that it might help. —David Eppstein (talk) 15:37, 15 April 2017 (UTC)
 * ?  E Eng  15:49, 15 April 2017 (UTC)
 * I appreciate the offer but I'm reluctant to prevent all unregistered or logged in users from participating in my Talk page just because of one jerk. I'm fine with escalating blocks to prevent harassment by one individual unless he or she begins using multiple IP addresses to continue.  Thanks for asking and offering!  ElKevbo (talk) 23:29, 15 April 2017 (UTC)
 * Problem is, he's hopping IPs. If you look at my talk page you'll see my approach is to wear such attacks proudly.  E Eng  23:57, 15 April 2017 (UTC)

Gosh, it seems I'm your only visitor lately
If he hasn't by the time you're skulking about in the morning, could you take care of this ?  E Eng  08:51, 18 April 2017 (UTC)

https://en.wikipedia.org/wiki/Squared_triangular_number
If it's not important, why you keep mentioning it? Actually, this is more than just not important, this is ridiculous, saying that mathematical proof could possibly need a source to back up its credibility, regardless of where it is presented, in Encyclopedia, or in the Evening News. Your article maybe mentions at least nine proofs, but presents two, at best. What's their significance? Among those mentioned ones is proof by induction too, for which Stein observes that it is "uninformative". Yeah, well, this is why this one stands out. I wanted to present just how uninformative it is, to clarify that point. Something like this: https://math.stackexchange.com/questions/61541/are-proofs-by-induction-inferior-to-other-proofs
 * You are adding an uninformative proof, to show just how uninformative it is?? See WP:POINT. —David Eppstein (talk) 22:00, 18 April 2017 (UTC)

DYK for Harry R. Lewis
Mifter (talk) 00:02, 19 April 2017 (UTC)

Fibonacci
This is about the portrait of Leonardo Fibonacci that you insist on removing. There are hundreds of images of ancient figures that are romanticized and probably inaccurate, this doesn't mean the image cannot be placed in the article with the right caption. This is what I did with the edit of Fibonacci. An example is Christopher Columbus, no known contemporary images of him exist, only posthumous depictions, which are used today to represent him nonetheless, even in wikipedia. In the edit to Fibonacci, the image was not placed as the portrait, it was placed throughout the article just like the statue. Your edit in my opinion is groundless. Walnut77 (talk) 20:38, 1 May 2017 (UTC)
 * The image is a generic and non-notable one, drawn many centuries later, by someone who had no idea what Fibonacci looked like. We have no sources about this specific image to use to provide any commentary on the image, and it is an egregious example of what WP:PORTRAIT specifically discusses, generic 19th-century illustrations that serve more to obfuscate (by imposing 19th-century ideas of what this sort of person must have looked like) than to inform. —David Eppstein (talk) 20:43, 1 May 2017 (UTC)
 * The source of the image is 'I benefattori dell'umanita by L. Ducci. It is a book from Florence, Italy from the 19th century. The statue is also a 19th century depicition, which you do not question. The caption clearly stated this. The images of Columbus are also very different, yet they are still used throughout the article. Walnut77 (talk) 20:46, 1 May 2017 (UTC)
 * In addition, you are not supposed to judge the artistic value of an image. What may seem to you as a bunch of scribbles is considered modern art by others. Walnut77 (talk) 20:53, 1 May 2017 (UTC)
 * See WP:WAX. I haven't tried to clear out all bad portraits on Wikipedia (there are many), only the ones on articles that interest me, and I have no strong opinion on whether the statue should stay or go (it at least implies that someone in that town cares enough to erect a statue to him, more than we can say for the portrait). As for "the source of the image": that's not what I'm asking for. I'm asking for reliably published sources about the image, by other people. Without those we have nothing we can say about it and no reason to include it in the article. You know, something like "Ducci's notorious anti-intellectualism explains the weak chin and stupefied expression on his imagined depiction of Fibonacci" only, you know, something actually backed by scholarship rather than just made up. As for your "you are not supposed to judge" attitude: bullshit. I can judge all I want. And I can also demand that pictures included in articles live up to some sort of standards; in this case of bringing actual information to the article, in others (especially living people) of not insulting them by being gratuitously ugly. —David Eppstein (talk) 20:55, 1 May 2017 (UTC)
 * This article is not about the art, it's about the person. It is still a representation of the person which has been used almost universally. Both images used in the article are certainly inaccurate (the statue and the drawing), yet both are artistic depictions of the historical figure. I think if one illustration stays, so should the other, like it is done in nearly every other similar article. Otherwise, there should be no images of him. Walnut77 (talk) 21:18, 1 May 2017 (UTC)
 * So your position is that, when it comes to visual depictions of people, you are incapable of distinguishing significant and notable artworks from drivel, and you think everyone else is too? —David Eppstein (talk) 21:22, 1 May 2017 (UTC)

Bret Stephens
Welcome to Wikipedia. Although everyone is welcome to contribute constructively to the encyclopedia, we would like you to assume good faith while interacting with other editors, which you did not do on Talk:Bret Stephens. Take a look at the welcome page to learn more about contributing to this encyclopedia. Thank you.

Your recent editing history at Bret Stephens shows that you are currently engaged in an edit war. To resolve the content dispute, please do not revert or change the edits of others when you are reverted. Instead of reverting, please use the talk page to work toward making a version that represents consensus among editors. The best practice at this stage is to discuss, not edit-war. See BRD for how this is done. If discussions reach an impasse, you can then post a request for help at a relevant noticeboard or seek dispute resolution. In some cases, you may wish to request temporary page protection.

Being involved in an edit war can result in your being blocked from editing&mdash;especially if you violate the three-revert rule, which states that an editor must not perform more than three reverts on a single page within a 24-hour period. Undoing another editor's work—whether in whole or in part, whether involving the same or different material each time—counts as a revert. Also keep in mind that while violating the three-revert rule often leads to a block, you can still be blocked for edit warring&mdash;even if you don't violate the three-revert rule&mdash;should your behavior indicate that you intend to continue reverting repeatedly. —Danwroy (talk) 01:44, 3 May 2017 (UTC)
 * Wow. The "I'm rubber you're glue" strategy. What will they think of next? —David Eppstein (talk) 01:48, 3 May 2017 (UTC)

Christopher Morley Collection
I added it because it is a part of a permanent collection in a university library that is based around a notable author. Are the 128 other libraries the same way or just available for checkout? SL93 (talk) 21:43, 6 May 2017 (UTC)
 * Sure, but inclusion in the collection means merely that the library looked for everything they could find that was published about Morley and found this. It adds little useful information about it, beyond "it's about Morley", which we already knew. —David Eppstein (talk) 22:40, 6 May 2017 (UTC)
 * I understand where you're coming from. Also, thanks for the copy edit and new content. I know no one "owns" articles, but I like seeing an article get good help. SL93 (talk) 01:45, 7 May 2017 (UTC)

DYK for Donald G. Saari
Vanamonde (talk) 05:08, 9 May 2017 (UTC)

You better watch it...
Prickly is awful close to prick. I suggest that in future you use Arbcom-approved terms such as cunt and asshole.  E Eng  18:15, 9 May 2017 (UTC)
 * Heh, good point. I should have called them "cunty interactions"; I'm sure that would have gone over much better. —David Eppstein (talk) 18:21, 9 May 2017 (UTC)

A barnstar for you!

 * Thanks! —David Eppstein (talk) 00:07, 10 May 2017 (UTC)

DYK for Anne Penfold Street
Mifter (talk) 04:53, 10 May 2017 (UTC)

Thank You!

 * You're welcome! —David Eppstein (talk) 00:10, 12 May 2017 (UTC)

Hey
What are you doing? — Preceding unsigned comment added by Gwoegi (talk • contribs) 06:47, 12 May 2017 (UTC)
 * Uh, removing somewhat negative unsourced claims from your biography, according to our policy on that? It looked like vandalism. Was that actually you? —David Eppstein (talk) 06:50, 12 May 2017 (UTC)
 * Yes, it is me. I want to be open with my problems. — Preceding unsigned comment added by Gwoegi (talk • contribs) 06:52, 12 May 2017 (UTC)
 * Well, if this is actually you (it's hard to tell with a brand new account), you should find out about Wikipedia's policies on writing about yourself (short version: don't) and biographies of living people (every claim needs to come from a published source). —David Eppstein (talk) 06:56, 12 May 2017 (UTC)
 * Why are you thinking it is not me myself? :((( — Preceding unsigned comment added by Gwoegi (talk • contribs) 07:26, 12 May 2017 (UTC)
 * Well, your choice of user name is convincing enough, but anyone could make an account with that name. Anyway, as I forgot to say it before, welcome to Wikipedia! Outside your biography, there is plenty of content here where your expertise would be welcome. —David Eppstein (talk) 14:32, 12 May 2017 (UTC)

Another one bites the dust

 * I've ruined your clean block log, David. :( clpo13(talk) 07:55, 12 May 2017 (UTC)
 * It happens to the very best of us. You've been blooded.  E Eng  12:51, 12 May 2017 (UTC)
 * I slept through all the excitement — what was this about? —David Eppstein (talk) 14:33, 12 May 2017 (UTC)
 * I believe he meant to block Gwoegi but got you instead.  E Eng  14:50, 12 May 2017 (UTC)
 * That's exactly it. Gwoegi started blanking your talk page and wouldn't stop. I went to block him, but I was on the wrong user page when I made the block. clpo13(talk) 16:09, 12 May 2017 (UTC)
 * The talk page blanking goes a long way towards convincing me that this is an impersonator, not the real gwoegi. But this is something that can be verified through off-wiki channels. —David Eppstein (talk) 16:22, 12 May 2017 (UTC)

Erdős number and James Parry
All right, what kind of documentation would you accept for the "Kibo number" in the "Variations" section of Erd%C5%91s_number? Given that it is/was almost entirely a usenet phenomenon, and that FAQ was regularly posted to relevant newsgroups over a long period of time? There's also http://wiki.c2.com/?KiboNumber, and of course the WP article on Parry himself which was being linked to in that edit. There are numerous references to people's "Kibo number" in old usenet posts. BunsenH (talk) 00:11, 13 May 2017 (UTC)
 * Mention in a published book would do. There have been various books of hacker lore published; one of them might include this. Google books gives a couple of relevant looking hits but I think to use them you'd need to find a copy elsewhere as they are only shown by Google in snippet view, which is not generally good enough to produce an accurate citation. —David Eppstein (talk) 00:22, 13 May 2017 (UTC)
 * The "Hacker's Dictionary" was the first place I looked, but that doesn't mention the "Kibo number" in its entry on Parry. There are a couple of books written by Carla G. Suratt, with snippets indexed in Google Books.  Both of those cite the above FAQ as their original source for that info.  It seems to me that using them as a source would, if anything, be a step backward -- obfuscating the real source rather than establishing its credibility. BunsenH (talk) 01:11, 13 May 2017 (UTC)

Hollow Moon
Hi David. You added a dubious tag to Hollow Moon a while back. I'm passing through briefly, and thought I'd try again to get the article up to GA, and your tag is standing in the way of that.

You were questioning whether it's accurate to call it a conspiracy theory. The first thing to note is that Hollow Moon (as a "reality", rather than in fiction) is now exclusively promoted by people like David Icke ("Icke has made his name since the 1990s as a professional conspiracy theorist"). Your reason for the tag was:This term is used only for theories that secret groups of people have done things — who is supposed to have hollowed the moon? I suspect the answer is the intergalactic confederation who are monitoring our thoughts from the spaceship that you call the moon, silly, but I must admit I don't care to spend too much time wading through his pronouncements to establish that :) .  It's generally described as a conspiracy theory in the press, although I accept that just because sources say something, doesn't make it actually true.

Is there a preferred wording that you'd have, or is the fact that the secret group that are meant to have done it are aliens rather than an earth government/clique acceptable? Cheers, Bromley86 (talk) 21:42, 14 May 2017 (UTC)
 * You could call it a fringe theory or pseudoscience. I think that would be better than a conspiracy. —David Eppstein (talk) 04:14, 15 May 2017 (UTC)


 * Cheers, will do. Bromley86 (talk) 04:35, 15 May 2017 (UTC)

The English Language
Just FYI: "avowed uncertainty" is a paradox, not an oxymoron. Certain uncertainty would be an oxymoron. Avowed =/= certain. — Preceding unsigned comment added by 68.175.127.198 (talk) 14:07, 22 May 2017 (UTC)
 * It's also a stupid choice of language to use on a highly politicized and heavily argued point of a Wikipedia article. —David Eppstein (talk) 14:32, 22 May 2017 (UTC)
 * What's your position on unknown knowns and known unknowns and unknown unknowns?  E Eng  15:19, 22 May 2017 (UTC)
 * Say what you will about the tenets of neoconservatism, at least they understood nuance. —David Eppstein (talk) 15:48, 22 May 2017 (UTC)

Sorry about my recent edits
Sorry about my recent edits on Ryan Ross I was not thinking straight, you were right, I will stop with those edits. Thanks. Bowling is life (talk) 00:24, 22 May 2017 (UTC)
 * I think it can be sourced — the Panic! article also says "emo pop" but with three footnotes, two of which support that claim and are from the time when Ross was with that band — so mostly what I was objecting to was the addition of unsourced claims to a WP:BLP. —David Eppstein (talk) 00:39, 23 May 2017 (UTC)

I'll start working on finding a source for all the genres. Bowling is life (talk) 00:55, 22 May 2017 (UTC)

I found sources for most of the genres for Ryan Ross, the only genre I could not find a source for is alternative rock. Bowling is life (talk) 1:45, 22 May 2017 (UTC)
 * Ok, thanks. I'll stop reverting. —David Eppstein (talk) 03:03, 23 May 2017 (UTC)

Golden ratio
Concerning this revert: actually, these artist used the golden ratio to construct their compositions, as opposed to seemingly random constructions, of the type designed by Picasso and Braque. That is precisely the reason they chose the name Section d'Or. It was this ratio that gave to the works a 'classical' or 'traditional' foundation. Thus Cubism was, for them, a continuation of a long tradition, rather than an abrupt break from the past. They used the number more so than Salvador Dalí or Le Corbusier, both of which are amply mentioned in the article. Perhaps that was not explained well in my post. Cheers. Coldcreation (talk) 10:31, 25 May 2017 (UTC)
 * So it should be included in List of works designed with the golden ratio, then? —David Eppstein (talk) 15:36, 25 May 2017 (UTC)
 * Already there. Coldcreation (talk) 17:33, 25 May 2017 (UTC)

Knowledge graph embedding
Dear David, thanks for your constructive revert! I can understand graph embedding might have different meanings, i.e. spatial vs semantic (hence, the off-topic and undue). I also can understand the idiosyncratic, as this is actually one of my first efforts here in Wikipedia. However, I cannot understand the dubious, since there is now a consistent literature on the topic (also known as Knowledge representation learning) from conferences such as AAAI, IJCAI, ACL, and ICML. So, given your excellent expertise in this field, may I know where and how I should insert this novel topic in Wikipedia? - Mommi84 (talk) 16:47, 30 May 2017 (UTC)
 * Graph embedding has a very specific mathematical meaning, that your addition seemed quite far from. But the "dubious" was in reference to claims in your edit such as the one that directed graphs "are usually referred to as knowledge graphs". My experience, on the contrary, is that directed graphs are never referred to as knowledge graphs. —David Eppstein (talk) 17:29, 30 May 2017 (UTC)
 * I accept that. Still, you have not answered my question, although it seems quite obvious I will have to reformulate the concept of knowledge graph embedding in vector spaces in an entry of its own. Surprisingly, its definition has a very specific mathematical meaning too (see Knowledge Graph Embedding by Translating on Hyperplanes). I would be glad if you helped me with that, maybe editing or suggesting instead of deleting. Anyhow, if the community will ask to merge the page knowledge graph embedding with graph embedding, I am afraid we will have to discuss again. - Mommi84 (talk) 23:06, 30 May 2017 (UTC)

Invitation to San Diego wiknic and bonfire
You are invited to join the discussion at Wikipedia:Meetup/San Diego/July 2017 Wiknic. RightCowLeftCoast (talk) 01:12, 31 May 2017 (UTC)

Eeng
Helping people making personal attacks? And I should be happy that you didn't template me? "Can you imagine the bullying she suffered during her formative years? " is continuing the theme of mocking her name, which you are supporting here. Please undo your revert of my BLP violation removal instead of encouraging EEng to violate our policies with his childish behaviour. If you can't recognise BLP violations and attacks, just leave this alone in the future. Fram (talk) 20:40, 31 May 2017 (UTC)
 * Removing non-attack comments that explained why the earlier comment you complained about was also not an attack? As an administrator, you should know better. —David Eppstein (talk) 21:46, 31 May 2017 (UTC)
 * If you can't recognise an attack on a BLP when it is spitting you in the face, then don't bother undoing removals of said attacks. I'll go and reremove the problematic part now, if you disagree don't bother coming to my talk page but either brush up on your knowledge of what is acceptable or take it to a noticeboard. Fram (talk) 06:37, 1 June 2017 (UTC)

Hello, I'm Fram. I wanted to let you know that some of your recent contributions have been reverted or removed because they could seem to be defamatory or libellous. Take a look at our welcome page to learn more about contributing to this encyclopedia. Thank you.
 * Was there a point to this, or do you just like to think of yourself as rebellious by flouting WP:DTR? —David Eppstein (talk) 18:09, 1 June 2017 (UTC)
 * Oh no, I flouted an essay, how daring. The point was that you seem to need the basics explained again, so that's where we start. Explaining the difference between policies, guidelines and essays seems to be necessary as well then. Fram (talk) 19:32, 1 June 2017 (UTC)
 * [rolls eyes] I guess I should be happy that Wikipedia attracts people who like to explain things. —David Eppstein (talk) 19:37, 1 June 2017 (UTC)

Red dwarf
Hello, hope you are well... Your posts and math topics are very interesting! I have been studying a certain type of probability for a while now relating to random subsets of random sets.

This question might not make sense, but do you know what I mean by "red dwarf" I'm searching for a few answers and my journey into the magical world of Math has lead me to asking you this question. — Preceding unsigned comment added by 86.27.104.135 (talk • contribs)
 * The only mathematical meaning I can find of "red dwarf" is in the work of A. Blokh and co-authors on Julia sets. Is that what you mean? —David Eppstein (talk) 20:02, 1 June 2017 (UTC)

Unfortunately no it was not, I was specifically looking for someone who may have focused a lot of time on cycles of random subsets. — Preceding unsigned comment added by 86.27.104.135 (talk) 20:07, 1 June 2017 (UTC)

Although it could be? I just had a look at the references you linked too...

Many thanks for your reply BTW... Whilst I am no-where near your level of mathematical knowledge I just wanted to say that I admire your work and am finding it very interesting. — Preceding unsigned comment added by 86.27.104.135 (talk) 20:17, 1 June 2017 (UTC)

Stirling's approximation
Hi,

I am writing regarding Striling's approximation, because you reverted my recent changes.

I can understand that you don't like random people removing references. But the truth is that I myself added this reference earlier, because this was the best I found, but this book had its focus on something else. Later I realized that the book probably also got the result from Robbins', although it does not specify it. I thought it is better to go back to the original reference.

And it is true that I changed the bound to a less precise one, but I think it is also more intuitive, and since the better bound is also on the page below I think it is the most beneficial way of presenting it to the audience of the page. — Preceding unsigned comment added by 2001:4898:80E8:F:0:0:0:1C9 (talk) 17:07, 6 June 2017 (UTC)
 * It would be helpful if you made some explanation for these things in your edit summaries. Because as it is, a not-logged-in editor removing sourced material without any explanation at the time of the edit, looks a lot like vandalism. —David Eppstein (talk) 17:15, 6 June 2017 (UTC)

Ok, fair enough, I just did so. — Preceding unsigned comment added by 2001:4898:80e8:f::1c9 (talk • contribs)

Subuey's bad habits
"due to Subuey's bad habit of going back to stale comments and re-editing them." You have defended active members removing months old information. My "bad habits" are none of your concern. :) — Preceding unsigned comment added by Subuey (talk • contribs) 18:29, 7 June 2017 (UTC)
 * The context. —David Eppstein (talk) 18:38, 7 June 2017 (UTC)
 * If something needs to be said, it is better said later than never. and since you have a bad habit of not replying to arguments, it makes perfect sense to edit. "If something should be removed, it is better later than not at all.  How long something has been up is not an argument for it staying." — Preceding unsigned comment added by Subuey (talk • contribs)
 * Do not change talk page content added by other people, as you did in the edit in which you made that comment. Continuing to do so may cause you to be blocked. —David Eppstein (talk) 20:58, 7 June 2017 (UTC)
 * I am going to seriously contest your ability to do anything, as you have proven to be derogatory, the header above is one small example. Subuey (talk) 00:19, 8 June 2017 (UTC)
 * Huh? The header is a direct quote from the headerless comment you added to start this thread. —David Eppstein (talk) 00:24, 8 June 2017 (UTC)
 * Which was actually a direct quote from you... Subuey (talk) 00:28, 8 June 2017 (UTC)
 * Describing a behavior I've seen you do repeatedly ("habit") which I think you should not be doing ("bad") because, as the context makes clear, you are misleading people about the timing of your messages. Telling people they are doing something wrong when, in fact, they are doing something wrong is not something I think I should apologize for or stop doing. —David Eppstein (talk) 00:33, 8 June 2017 (UTC)
 * Oh I think you should apologize AND stop doing it. I'm not amused, and I'm serious. Change the header and change your attitude on Jacob's page. As far as the above "context", it is nonsense and you are seriously wasting my effort and time.  Subuey (talk) 00:41, 8 June 2017 (UTC)
 * My impression is that any unanswered comment you make will be the subject of continued rewording from you, indefinitely. So to head that off, I'll give you one last answer to your demands: no. —David Eppstein (talk) 00:54, 8 June 2017 (UTC)

10 years of adminship, today
Wishing David Eppstein a very happy adminship anniversary on behalf of the Wikipedia Birthday Committee! Mz7 (talk) 21:29, 7 June 2017 (UTC)
 * Thanks! I didn't know anyone kept track of such things; I wouldn't have known myself without this reminder. —David Eppstein (talk) 21:54, 7 June 2017 (UTC)

NPR
Since, you frequently participate in deletion disc. etc. concerning academic personalities, can you kindly review Richard G. Salomon (academic).Thanks! Winged Blades Godric 06:48, 9 June 2017 (UTC)
 * The selected publication list is longer than I'd prefer, but he looks clearly notable and the article looks neutral and well-formatted. So I marked it as reviewed, but I don't think it needs any special attention. —David Eppstein (talk) 14:40, 9 June 2017 (UTC)

DYK for Babette Hughes
Mifter 00:02, 11 June 2017 (UTC)

My plan is working puhrrrrfectly
I see you've been thoroughly infected by the ran/rma virus. But someday, you know, you and I will be the defendants in some Wikipurity show-trial about it.  E Eng  20:02, 11 June 2017 (UTC)

Andrásfai chromatic number
How would you suggest that the Chromatic number be writteen in the sidebar? It's 3 except for $$n=1$$, which is covered in the mathworld source. Original $$n$$ came from the Hungarian wiki which I used as a basis for this article, so that was a good catch, thank you. Porphyro (talk) 15:35, 14 June 2017 (UTC)
 * MathWorld doesn't actually say the chromatic number is 3; it has a formula for the number of 3-colorings, but no statement that this is the chromatic number. In general, content like that in an infobox should be supported with sentences that say the same thing in the actual body of the article, with inline footnotes indicating where the information came from. In this case, I think $$\min(2n,3)$$ would be a correct formula, but we need a reference, not just "I think". —David Eppstein (talk) 15:39, 14 June 2017 (UTC)
 * Ok, point taken. It's pretty trivial to prove they're not 2-colourable but I guess that still does count as original research. I might look for a concrete source later if I have time. Porphyro (talk) 15:43, 14 June 2017 (UTC)

Allahverdiyev
Hello David Eppstein. You have just accused me of making a "copy-and-paste revert" when in fact all I did was undo your earlier redirection and your unexplained preference of the obviously erroneous spelling of this person's name. I did not copy and paste a single character. Hence my question: why is undoing an edit is considered "a copy and paste move" but this is not? Parishan (talk) 04:07, 16 June 2017 (UTC)
 * The original copy-and-paste move was the one done on May 17. The article was created by, but then copied its content to the other spelling and changed the original article into a redirect. That's the wrong thing to do, because it loses the authorship information and makes it look like Samiq is taking credit for Araz's work. So I undid those changes. What you did today was to go back to the wrong versions, the ones that made it look like Samiq wrote the article when actually Araz wrote it. The correct thing to do is to move the article rather than copy-and-paste, but by now with all the page history on both sides that has to be done through WP:RM. That's why I reverted you. The erroneous spelling can be fixed by a page move, but not by what you are trying to do. —David Eppstein (talk) 04:12, 16 June 2017 (UTC)

Jacob Barnett
--Neil N  talk to me 23:11, 16 June 2017 (UTC)

Your expertise is requested...
...casual dress. Uhm...I don't think my ping worked, so I brought the ping to you. Atsme 📞📧 00:11, 20 June 2017 (UTC)

Broken iterative implementation of DFS
Regarding your edit comment in Depth-first search: > I think including a *correct* non-recursive version is important. If you have a source for a correct+simple O(|V|)-space version, feel free to use it.

The version I deleted was not correct, since it doesn't respect the space complexity in the infobox at the top of the article. I agree it would be nice to have a correct iterative version, but that requires a reference which I don't have at the moment.

Regardless of there being a correct version, the version which is there now is not correct, so it should be deleted.

If you're not convinced that the current iterative code is wrong, ponder for a moment: would a O(2^N) version of, say quicksort be acceptable in an article about quicksort, presented as if it's an actual quicksort algorithm? Rbarreira (talk) 18:31, 24 June 2017 (UTC)
 * Perhaps you misunderstand my point, so let me try to make it more clearly. By "correct" I mean that the algorithm produces the desired output. Using as little space and time as possible are also worthwhile goals, but I would describe achieving them as being "efficient" rather than "correct". Many people have described DFS and BFS as being the same, except that DFS uses a stack and BFS uses a queue. Doing this naively produces an algorithm that is incorrect, in that it produces the wrong ordering of the vertices. I think it's important to describe how to substitute a stack for a queue in a way that produces the correct vertex ordering. The space usage is larger, but that's not a correctness issue (the algorithm still produces the correct output), it's an efficiency issue. And linear time and linear space is still adequate for most DFS-based tasks, especially as you need O(|E|) space just to represent your graph explicitly. —David Eppstein (talk) 18:38, 24 June 2017 (UTC)
 * By that standard, one could present a bubble-sort algorithm as a "correct" quicksort, since it produces the desired output. You're a Wikipedia administrator and I'm not, so I'm not going to start a revert war with you, but I'd like to ask you to rethink what "correct" means here. In my opinion, the time and space usage of an algorithm are both properties of the algorithm (and the article agrees as it presents them in the infobox). Whether the algorithm is "adequate" or not is subjective and beside the point so I won't even debate that. Rbarreira (talk) 18:44, 24 June 2017 (UTC)
 * Bubble sort is a correct sorting algorithm; it sorts its input correctly, always. It is a different algorithm from quicksort, but that does not make it incorrect. "Correct" is a technical term here, and you are not using it correctly. Perhaps part of the confusion here is that there are two different things called "depth-first search": traversing the vertices in the correct order, and the specific algorithm for doing that with O(|E|) time and O(|V|) space. The stack-based non-recursive algorithm is a different algorithm (there would be no point in giving pseudocode for the same algorithm twice in a row) but one that produces the same traversal, just like lexicographic breadth-first search is an even-more-different algorithm that produces a valid breadth-first traversal. —David Eppstein (talk) 18:50, 24 June 2017 (UTC)
 * Incidentally, I have a non-recursive O(|V|)-space implementation of DFS at http://www.ics.uci.edu/~eppstein/PADS/DFS.py but I don't think that counts as a reliable source. —David Eppstein (talk) 18:42, 24 June 2017 (UTC)

Need your expertise, please?
Structured_kNN is in the NPR que, and from what I can tell, it looks like OR as it is pretty much unsourced, so I don't quite know what to do with it. It is clearly far too technical for the average reader. . Atsme 📞📧 17:33, 26 June 2017 (UTC)

Precious two years!
--Gerda Arendt (talk) 21:49, 26 June 2017 (UTC)

Regarding Irvine Density Numbers
They're not off, they're using the Census estimate populations for 2016 because the area numbers are for 2016. While this isn't an inherent problem for Irvine (as it hasn't annexed or lost any territory since 2010), it's a real problem for, say, Jurupa Valley, CA which didn't even exist in 2010, so the choices are either inconsistency across US place pages or using the 2016 numbers everywhere. It also won't match my planned updates of the "location within county and state" svg maps as those maps will be based on 2016 Census geography too. DemocraticLuntz (talk) 12:04, 28 June 2017 (UTC)

Graph matching
Hi, while voting in Articles for deletion/NeuronDotNet I was surprized to find that there is no article "Graph matching". So I started a minimal stub. Unfortunately to write a more decent article is way above my head. Can you please lend a hand here? Either add something by yourself or invite other wikipedians you may know to be experts on the subject?

While we are at this, due to this another meaning of the term "matching" in graph theory, I would suggest to consider renaming "matching (graph theory)" to a more specific disambiguator, such as matching (edge set). Similarly Category:Matching begs for renaming even more. What do you think? Staszek Lem (talk) 17:13, 28 June 2017 (UTC)

Decision stream
Hi David,

According to request of professor draft for Decision stream was created and employees of several laboratories who successfully use this technologies in their investigations added appropriate citations in Wikipedia. It looks quite unusually that guys here demonstrate aggressive actions against scientific community, this bright idea, and provides strange comments even into the draft. Are you professional in this field?

What's the reason for this activity?

Cold you please give reference to the Wikipedia rules that supports such behavior? — Preceding unsigned comment added by 46.39.231.205 (talk • contribs)


 * WP:COI, WP:SOAPBOX, WP:GNG, and WP:REFSPAM all look relevant, for a start. As for who I am, you can easily look that up, but in general you are not allowed to ask this of Wikipedia editors; see WP:OUTING and WP:CRED. —David Eppstein (talk) 18:35, 28 June 2017 (UTC)

WP:REFSPAM - Citation spamming is a subtle form of spam and should not be confused with legitimate good-faith additions intended to verify article content and help build the encyclopedia. — Preceding unsigned comment added by 46.39.231.205 (talk • contribs)
 * Yes, I agree. Your spamming of this reference to a dozen or more unrelated Wikipedia articles was definitely unsubtle. And as your message above makes clear, your intent was to promote your professor's work rather than to build the encyclopedia. By the way, when you use talk pages here, please sign your messages by adding four tildes ( ~ ) at the end of your message. —David Eppstein (talk) 19:02, 28 June 2017 (UTC)

Several employees of different labs carefully added accurate text fragments into appropriate places in related articles. Is it possible to suspend this strange war started by Joel B. Lewis?

46.39.231.205 (talk) 19:18, 28 June 2017 (UTC)
 * Ok, add WP:MEATPUPPET to your reading list. —David Eppstein (talk) 20:07, 28 June 2017 (UTC)

AN/I
As you participated in Administrators' noticeboard/IncidentArchive957, you may be interested in Administrators' noticeboard/Incidents. — Godsy (TALK CONT ) 03:30, 29 June 2017 (UTC)