Wikipedia talk:WikiProject Mathematics/Archive/2014/Aug

Draft:Deformation tensor
Could someone take a look at this Draft to see if should be approved. I'm just not sure there is enough there for its own article or if it should be approved or something else...Naraht (talk) 02:07, 2 August 2014 (UTC)

René François Walter de Sluse
I believe those two are the same person, but my English is not good enough to start merging discussion: - Vlsergey (talk) 10:56, 29 July 2014 (UTC)
 * René-François de Sluse
 * René François Walter de Sluse


 * Both articles are about René François Walter de Sluze. Merging should be uncontroversial.  Deltahedron (talk) 06:51, 2 August 2014 (UTC)

McShane
Should we have an article titled McShane integral? Michael Hardy (talk) 17:43, 3 August 2014 (UTC)
 * (It appears that the McShane integral and the McShane–Whitney extension theorem are both named after the same person, but McShane's identity may be named after someone else. Michael Hardy (talk) 17:46, 3 August 2014 (UTC))
 * The first appears to be Edward James McShane, the second Gregory McShane . Deltahedron (talk) 18:34, 3 August 2014 (UTC)

Richard Gosselin
I've proposed deletion for the article about mathematician Richard Gosselin after failing to find sources that could be used to verify the information in this (very short) article and demonstrate some notability for its subject. Perhaps someone else here can do better in this regard. —David Eppstein (talk) 22:54, 3 August 2014 (UTC)

Articles for deletion/Ankur Tiwari
People may be interested in this 19-year-old Indian who has invented a way of dividing by zero, and developed a supercomputer which maintains "the complete 100% accuracy in all the mathematical operations." JohnCD (talk) 15:25, 4 August 2014 (UTC)


 * I was a bit interested, and have read his article about dividing by zero. Three ideas that I understood are:
 * let us waive continuity condition on the function 1/x, and then we can define it at 0 as we like, with no contradiction;
 * let us treat division as division with remainder, then we may treat 1/0 as 0 with remainder 1;
 * let us define 1/0 as the limit of $$\frac12(\frac1{-x}+\frac1x)$$ as x tends to 0; then we get 1/0 = 0.
 * Wow indeed... :-)
 * Important copyright notice: I did not quote any line of his article (which is explicitly prohibited by him, "not even to personal blogs and sites"). Rather, I presented my impressions in my own words (and formulas). Thus, hopefully, I avoid legal actions promised by the author.
 * Boris Tsirelson (talk) 20:22, 4 August 2014 (UTC)
 * Alice has prior art . Deltahedron (talk) 20:33, 4 August 2014 (UTC)
 * We could also define 1/0 as $$i$$. After all it is an imaginary amount. I'm sure that would be just as useful and consistent. ;-) By the way APL defined 0/0 to be 1 whereas the J programming language has it as zero - you can see their wonderful list of reasons at . Dmcq (talk) 21:16, 4 August 2014 (UTC)
 * Even when programming in C++ you don't necessarily get that blue-screen these days that you so well deserve (for the sake of quality (and even safety) for the consumer of your software) when screwing badly. It is a while ago since I programmed now, but are floating-point exceptions and divide by zero exceptions even issued these days? My impression has been (in fairly recent years) that the underlying language and/or OS try to smooth over programming mistakes to give an impression of "stability". And oh, your "J-link" is marvelous reading. YohanN7 (talk) 21:54, 4 August 2014 (UTC)
 * See IEEE 754, particularly the "Exception handling" and "Recommendations" sections. No such user (talk) 08:22, 5 August 2014 (UTC)
 * Boris gave a pretty good summary. Also, according to that paper, the formula 1/0=0 leads to important applications in intellectual property.  (I'm not kidding; this claim is actually in there!)   Sławomir Biały  (talk) 22:40, 4 August 2014 (UTC)

Tensor Flight Dynamics
Notable? The author of the article appears to have written most its references. Might be legit, have no idea, but the suspicion of self-promotion is automatically there when this happens. The article was also linked (by the author) from Tensor, but I just reverted that. YohanN7 (talk) 17:54, 4 August 2014 (UTC)


 * While we're cleaning out non-notable self-promotion, Laplacian of the indicator is also worth a look. (And this article is already a few years old, so there's less of a worry about possible biting of the newcomers.)   Sławomir Biały  (talk) 22:20, 4 August 2014 (UTC)
 * Nominated for deletion here. -- Kinu  t/c 20:17, 7 August 2014 (UTC)

Avoid png rendering at all costs
It was pointed out at the talkpage of Lie group–Lie algebra correspondence that the png renderings of math formulae do not look good. My question is: do they really? I usually edit Wikipedia on my phone and on my screen the png rendering doesn't look bad. I'm wondering about what other people see. -- Taku (talk) 15:15, 20 July 2014 (UTC)


 * As a general rule, I prefer HTML for inline mathematics because this has the right size and alignment with the surrounding text, and latex for math appearing on its own line. Sometimes I have noticed also that the PNG images are of an insufficient resolution, rendering the resulting images indecipherable blurry.  That having been said, the promise of functioning MathJax on the indefinite horizon will hopefully make such discussions fruitless.  As for the article under discussion, most of the inline formulas there should be converted to HTML, the most glaring example being $$\operatorname{Lie}(G)$$.  Also, I agree that for some reason inline latex seems to look better on an iphone. Perhaps a more responsive design would help to make things look better for all users.  Then again I am too cynical to think this would do any good.  Sławomir Biały  (talk) 18:51, 20 July 2014 (UTC)


 * It just occurred to me that the computer should be able to convert latex to html, no? It requires the humans to do conversion seems [fill the blank]. -- Taku (talk) 16:52, 24 July 2014 (UTC)


 * The software used to be more aggressive about converting the tag to HTML when possible, but now Template:math is preferred for explicit HTMLing with <math for explicit PNGing. I think it the template works better and gives more control. For example, the  tag never HTML'd subscripts, mathrm, spacing, or greek letters. But with $$, I can say $2x × πy = cos(φ_{B})$. SamuelRiv (talk) 01:10, 10 August 2014 (UTC)

Algorithm examples
See Talk:Algorithm examples for a discussion on the naming and use of this article -- 65.94.169.222 (talk) 05:18, 9 August 2014 (UTC)
 * Thus article is also discussed for deletion. See Articles for deletion/Algorithm examples‎‎. D.Lazard (talk) 19:16, 10 August 2014 (UTC)

Packing fraction
I have changed Packing fraction and Packing Fraction, which previously redirected to Atomic packing factor to redirect to Packing density, which I recently created. It is possible that a disambiguation page or other solution is needed, such as merging Packing density with Atomic packing fraction under the title Packing fraction. Judging from the incoming links to Packing fraction, they should all indeed point to Packing density as none discuss atoms. Eigenbra (talk) 05:05, 11 August 2014 (UTC)

Mathoid
Members here interested in mathematics rendering and editing may like to look at Mathoid, a MediaWiki application which takes various forms of math input and converts it to MathML + SVG output. Deltahedron (talk) 08:27, 26 July 2014 (UTC)
 * See also this posting. Deltahedron (talk) 18:09, 11 August 2014 (UTC)

Kathleen Ollerenshaw‎
A user is persistently adding a claim to this article that she has died. Can anyone verify that? I can't find anything online. Deltahedron (talk) 20:25, 11 August 2014 (UTC)
 * BBC Radio Manchester appears to have reported it on Twitter, but I'm not finding any sort of news item that definitively qualifies as a WP:RS at this time. -- Kinu  t/c 23:00, 11 August 2014 (UTC)
 * Reported now, sadly: . Deltahedron (talk) 06:31, 12 August 2014 (UTC)

Tech help required to improve categories
Please see Village pump (policy) and User:Paradoctor/CatVisor if you are willing and able to assist this innovative WP project move along it would be greatly appreciated. Thank you, IZAK (talk) 23:56, 12 August 2014 (UTC)

Wikidata inter language links
An editor has quoted in talk:quartic function that quartic function and quartic equation are two different wikidata items. This is normal, as these are different items, that may correspond, in some languages to different articles. However, as is a redirect to Quartic function, it seems natural to link both wikidata items to quartic function (possibly through a redirect). It seems impossible. This has the consequence, that most languages, which have only the equivalent of "quartic equation" have not the inter language link to the English article on the subject. Any idea to solve this problem that is certainly not specific to quartic equation/function? D.Lazard (talk) 11:01, 13 August 2014 (UTC)
 * Is it not possible to have the Wikidata item for quartic equation on the redirect page? Deltahedron (talk) 17:15, 13 August 2014 (UTC)

Moving inactive participants to the relevant list
I am writing a bot to move inactive users out of participants list (and add them to the relevant "inactive participants" list). Your list of participants is huge and my bot detects lots of inactive participants. With your consent, I will move them to the inactive list. The participants page says that inactive users "have either left the project or have not edited in the last three months." Should I keep the same criterion? My bot can also handle constraints like "not edited a page in the scope of the WikiProject in the last X months" or "not edited an internal page of the project (portal, project talk, …) in the last X months". If I proceed, then I will most likely merge the sections of the table, but I think they can be easily restored by hand afterwards. Cheers. Pintoch (talk) 09:26, 15 August 2014 (UTC)

Curvature &rarr; Curvature (mathematics)
This edit, although I have mixed feelings about it overall, got me thinking that perhaps the main link curvature should be the disambigation page and curvature should be moved to curvature (mathematics) over the redirect. As there's probably a lot of links coming in there, it seemed to me to be a good idea to raise the matter here in order to make sure that there is a solid consensus before inflicting the community with the inevitable hellstorm of disambiguation cleanup. Sławomir Biały (talk) 18:20, 12 August 2014 (UTC)


 * Hmm, the mathematical type of curvature seems the dominant type of curvature to me. Just googling for non-WP pages leads mostly to mathematical curvature topics, with a few pages on spinal curvature and image field curvature in optics. That edit seems to be an attempt to make the company a little more prominent, as it is already mentioned in the dab page. --Mark viking (talk) 20:54, 12 August 2014 (UTC)


 * I think that the output of a Google search is probably not the most accurate indicator of what the typical reader is expecting when they type "curvature" into the "Search" bar. But I will not insist too much on this point.   Sławomir Biały  (talk) 20:56, 12 August 2014 (UTC)
 * (edit conflict) Yes, I think WP:PRIMARYTOPIC applies, and that the mathematical meaning is clearly primary. A hatnote pointing to the disambiguation page is appropriate, but I don't see the need to mention the company separately in the hat. —David Eppstein (talk) 20:58, 12 August 2014 (UTC)
 * I removed the reference to the company and cleaned up the hatnote prior to seeing your post; glad to see we agree on that. -- Kinu  t/c 21:04, 12 August 2014 (UTC)
 * I also agree that the hatnote should not mention the company and that the current article is the primary topic. --JBL (talk) 21:23, 12 August 2014 (UTC)
 * I think that the company is hardly the primary topic: indeed it's far from clear that it is at all notable. Its article appears to have been written by a single-purpose account with a name uncannily reminiscent of that of someone in the company's marketing department.  Deltahedron (talk) 21:35, 12 August 2014 (UTC)
 * "the company is hardly the primary topic" You appear to be disputing a position taken by no one in this discussion. --JBL (talk) 21:38, 12 August 2014 (UTC)
 * I was under the impression that I was reinforcing the point by suggesting a reason why the company was not the primary topic. Deltahedron (talk) 21:41, 12 August 2014 (UTC)
 * Regardless of what was intended, I believe that the lack of notability of the company is something that should also be addressed in addition to the subject under discussion.  Sławomir Biały  (talk) 21:44, 12 August 2014 (UTC)

I too agree with the solution expressed in unanimity on this page, and I have advised the editor who made the initial/controversial edit to read WP:COI because the rest of his edits don't look very kosher either. JMP EAX (talk) 17:50, 15 August 2014 (UTC)

Autoparallel curve
There's vague def in parallel curve which makes a bit of sense but I can't find it in any reliable sources. That may happen because there's a deluge of sources about autoparallel curves in the sense defined in parallel transport (in reliable sources like, or , Wikipedia fails to define autoparallel curves in there right now). Can anyone find references for the former meaning/def? The (unreferenced) material was added by an IP in 2004, so there's little chance of getting sources from him now. JMP EAX (talk) 16:58, 15 August 2014 (UTC)


 * Regarding the request for a formal definition, I would guess: "a curve is autoparallel if it is a nontrivial parallel of itself." I am not qualified to have an opinion about whether this is consistent with the word's use in some narrow corner of the literature, however.   Sławomir Biały  (talk) 19:12, 15 August 2014 (UTC)


 * Rather than guess, I would suggest removing the material which is unreferenced and unclear, particularly since it is in the introduction, which should simpply summarise what is already written, and referenced, in the main text of the article. Deltahedron (talk) 19:35, 15 August 2014 (UTC)


 * You have no argument from me on this point. I interpreted the original post in part as a genuine request for a formal definition (notice how the definition in our article is decorated with clarification neededs), rather than a request specifically for sources on the matter.   Sławomir Biały  (talk) 19:42, 15 August 2014 (UTC)


 * Now gone. --Salix alba (talk): 20:27, 15 August 2014 (UTC)
 * Thanks. I've also removed a small section later in the article about the same topic because it made no sense without the definition of "autoparallel". JMP EAX (talk) 21:20, 15 August 2014 (UTC)

And a related issue: Can anyone familiar with tubular neighborhood opine if the 3rd figure there is correctly depicting the concept? JMP EAX (talk) 19:06, 15 August 2014 (UTC)


 * Yes, there is nothing requiring the tubular neighborhood to be given by some particular exponential map. So I would say it is correct.  However, I think a more useful illustration though would be of a tubular neighborhood of a curve on a surface, or at least a tubular neighborhood in which the red lines are replaced by more general non-intersecting curves.   Sławomir Biały  (talk) 19:20, 15 August 2014 (UTC)
 * Ok, thanks. JMP EAX (talk) 21:20, 15 August 2014 (UTC)

Bézier spline (vs. B-spline and polybezier)
This one is a mess because resulting from the morass of inconsistent terminology between authors. In general the latter two terms are consistently defined by those using them, but "Bézier spline" isn't. I've left some comments on the talk page there. More participation would not hurt, both in terms of comments and actually fixing the darn page... JMP EAX (talk) 21:20, 15 August 2014 (UTC)

An observation about quantifier complexity
An observation about Lebesgue's number lemma. It says:
 * If the metric space (X, d) is compact and an open cover of X is given, then
 * there exists a number δ > 0 such that every subset of X having diameter less than δ is contained in some member of the cover;

that is,
 * ∃δ>0 ∀x∈X ∃α ∀y∈X ( |y-x|<δ ⇒ y∈Uα ).

Wow: a ∃∀∃∀ formula! Does anyone remember a higher (or even equal) quantifier complexity?

Strangely, I did not find any discussion of such cases in "usual" mathematics (rather than math logic). I mean, the number of alternations of like quantifiers. Boris Tsirelson (talk) 21:23, 16 August 2014 (UTC)


 * The pumping lemma for regular languages is of the form "for all regular A, there exists a p such that, for all w, there exist x, y, z such that, for all i, ...". (I'm omitting all of the content, and leaving only the quantifiers.) See also Pumping lemma for context-free languages. You might classify this stuff as math logic though? I'd be surprised if we couldn't find really mainstream examples. Mgnbar (talk) 21:59, 16 August 2014 (UTC)


 * Indeed. By the way, Lebesgue's number lemma is (as far as I understand) a ∀∃∀∃∀ formula, since the first quantifier is "for every metric space"; and the compactness probably is ∀∃∀∃ (for every sequence there exists an accumulation point, whose every neighborhood contains...). But that may depend on the choice of one of equivalent definitions of compactness. Boris Tsirelson (talk) 12:03, 17 August 2014 (UTC)
 * "To grasp the meaning of ∀x∃y∀z ψ(x,y,z) is not so easy. In natural speech it is never used. When it is needed it is somehow circumvented. Consider, for example, the sentence:
 * In every town there is the tallest building.
 * To imagine the meaning of four alternations of quantifiers seems as difficult as to imagine four-dimensional space."
 * To imagine the meaning of four alternations of quantifiers seems as difficult as to imagine four-dimensional space."


 * (Pavel Pudlák, "Logical foundations of mathematics and computation complexity", page 75.) Boris Tsirelson (talk) 12:18, 17 August 2014 (UTC)


 * "White has a forced checkmate in n moves" has a string of 2n+1 alternating quantifiers. According to http://timkr.home.xs4all.nl/chess2/diary.htm there is a position with KQP v KRBN that is a forced mate in 549 moves, giving well over 1000 alternations of quantifiers. r.e.b. (talk) 14:24, 17 August 2014 (UTC)


 * Yes, I know of this "cheap" way to get high quantifier complexity via games. But that is another story. Boris Tsirelson (talk) 18:49, 17 August 2014 (UTC)

This might be a stupid question
It is mathematically accurate to refer (in general) to a "set of formal languages"? Or does one usually need a class (set theory) for that? The CS literature varies widely in what it calls such a bunch: set/class/family are some terms I found. JMP EAX (talk) 16:01, 17 August 2014 (UTC)


 * A formal language is a set of strings over a fixed finite alphabet &Sigma;. Therefore each language over &Sigma; is a set. Further, each set of languages over &Sigma; is a set. For example, the set of regular languages over &Sigma; = {0, 1} is a set.
 * Textbooks and practitioners often omit mention of &Sigma; in stating theorems. They make statements such as, "The union of two regular languages is a regular language." Implicit in this statement is that the two regular languages are defined over the same alphabet. Or they may be defined over different alphabets, but then the union is defined over the union of the alphabets. Practitioners neglect such details because they are boring.
 * If you want to talk about the class of "all regular languages over all alphabets", then you have to let &Sigma; vary over all finite sets. So I believe that the question boils down to: Is the class of all finite sets a set? This is outside my expertise, but I'm guessing not.
 * Surely not. Boris Tsirelson (talk) 18:53, 17 August 2014 (UTC)
 * Consider the class S of all singleton sets, and the subclass R = { x ∈ S : ∩x ∉ ∩x }. Is {R} ∈ R?  Deltahedron (talk) 19:23, 17 August 2014 (UTC)
 * This is all in relation to some article editing? Mgnbar (talk) 16:53, 17 August 2014 (UTC)

E.g. one can read in a book the following def(s):

1. A set L of languages is mildly context-sensitive iff a) L contains all context-free languages. b) L can describe cross-serial dependencies: There is an n ≥ 2 such that {wk |w ∈ T∗} ∈ L for all k ≤ n. c) The languages in L are polynomially parsable, i.e., L ⊂ PTIME. d) The languages in L have the constant growth property. 2. A formalism F is mildly context-sensitive iff the set {L|L = L(G) for some G ∈ F} is mildly context-sensitive.

(As an aside: "Formalim" is not formally defined in that book as far as I can tell, but apparently the author can formally write membership in it with "∈" regardless. Also, the "constant growth property" is defined in a def after that one [not before]. So it all kinda rings alarm bells of rather sloppy writing to me.) JMP EAX (talk) 17:15, 17 August 2014 (UTC)


 * I don't know that material, but perhaps there is some fixed &Sigma; that's going unmentioned, as I described above. For example, the definition mentions P (complexity), and I think that a fixed &Sigma; is implicit in that definition. Or maybe not, and nobody cares?
 * If these questions are not in relation to planning or editing Wikipedia math articles, but rather for your curiosity or work, then please raise them at Reference desk/Mathematics instead of here. Mgnbar (talk) 17:57, 17 August 2014 (UTC)


 * We allow some latitude here in discussions here, especially for discussions on topics like this, the language used to describe mathematical objects. After all, we need to understand how the language we're using works before we can write the articles.  Deltahedron (talk) 18:07, 17 August 2014 (UTC)

Up to isomorphism, every finite alphabet is a subset of the natural numbers, every string is a sequence of natural numbers, every language is a set of strings, and the collection of all languages forms a set. So there's a perfectly valid way of talking about "the set of all languages". My suspicion is that the people who work in this area tend not to care enough about foundations to care whether they're doing things up to isomorphism or over all finite alphabets, nor whether using all alphabets causes these things to become a proper class rather than a set. —David Eppstein (talk) 18:46, 17 August 2014 (UTC)
 * Indeed from the class article Outside set theory, the word "class" is sometimes used synonymously with "set". I guess that might be the case here.--Salix alba (talk): 19:19, 17 August 2014 (UTC)
 * David Eppstein's answer is basically what I thought I understood of this issue, i.e. for a fixed alphabet (or up to equivalence classes [isomorphisms] thereof) one can speak of the set of languages (e.g. ), but otherwise [over all possible alphabets] it's improper to speak of a set... Some authors probably have the best punt when they use neither class nor set but "family" . JMP EAX (talk) 20:14, 17 August 2014 (UTC)
 * Searching Google Books for a similar expression such as "the set of Turing machines" finds plenty of examples where it's used in the general sense, presumably glossing over or taking as self-understood the fixing [or taking equivalence classes of] the alphabet, states etc. JMP EAX (talk) 20:42, 17 August 2014 (UTC)


 * By the way, I assumed that the original poster's question was not "up to isomorphism". Then clearly the set of all finite sets is just the set of natural numbers. Mgnbar (talk) 21:10, 17 August 2014 (UTC)

Heh, the article in question just got nuked from space (entirely rewritten). JMP EAX (talk) 21:48, 17 August 2014 (UTC)

René François Walter de Sluse
Would someone who understands the maths please merge René-François de Sluse and René François Walter de Sluse? Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 17:12, 19 August 2014 (UTC)

Boxcar function and Crenel function
Boxcar function and Crenel function have been tagged for a merger since March 2014. There is no discussion on the talk pages. -- Alan Liefting (talk - contribs) 21:48, 19 August 2014 (UTC)

Kleene algebra
Can someone suggest good parentheticals to add to add to the disambiguation(s) for Kleene algebra? I'm thinking that the first one (in the order given on that page) should perhaps be called while the 2nd one should could perhaps be called Any other ideas? I think the first one probably only needs to redirect to De Morgan algebra presently, because the rest of the article doesn't cover it.
 * Kleene algebra (De Morgan algebra) or Kleene algebra (with involution)?
 * Kleene algebra (with closure) or Kleene algebra (regular expression generalization)

Any other suggestions? JMP EAX (talk) 17:35, 20 August 2014 (UTC)

Fields Medal Table
I prepared a new table to replace with current one in Fields Medal page.I want to hear your invaluable comments about it.here it is:

Discussion
There seems to be a editing dispute at Fields Medal in which multiple parties have reverted each other over the table issue. That article is now even fully edit-protected. It might be helpful if someone could summarize what the difference between these tables is, and why it matters so much. Just posting one of the two versions of the tables here does not convey much information about the difference. Also, I think we need to understand why the matter was so controversial as to demand canvassing this over multiple user talk pages. Sławomir Biały (talk) 12:51, 19 August 2014 (UTC)
 * @ Sławomir Biały  Thank you for your attention. Canvassing is the result of my fault.Sorry for that.I don't want to be disruptive,and of course,there is not an urgency or anything else. The major differences between this table and previous versions are:1)in the old table,there was no information about fields medalists current institution(s),2)The current table which is now fully edit-protected, contains some wrong and weak information(e.g. It states that many of medalists were born in USSR, which has been dissolved,instead of stating for example that they were born in today Russia),3)Both the old and current versions of table sorely lack in terms of reliable reference.These are main reasons.There is no big deal,So do not rush yourself into it and take your time.Thank you.Rezameyqani (talk) 13:49, 19 August 2014 (UTC)
 * On the issue of flag icons and country links, the icon should correspond to whatever country existed at the time. So I agree with keeping USSR as the flag icon in most of the entries at the current protected version of the article.  To me the links to the Kingdom of Prussia make no sense at all.  All of the German born mathematicians were born in Germany, period.  As for Klaus Roth, the Kingdom of Prussia is equally inappropriate.  He was born Wrocław, which was then part of Germany (even though now it is part of Poland).   Sławomir Biały  (talk) 14:22, 19 August 2014 (UTC)
 * @ Sławomir Biały Thank you for your time. As you probably know,former USSR was consisted of several soviet republics.all of them were autonomous.So,I think when we say,for example Voevodsky was born in Russian SFSR,it clearly implies that he was born in USSR.about Prussia,one of these mathematicans were born in Free State of Prussia,Grothendieck and Roth.the problem was that I could not find the proper flagicon for Free State of Prussia.Rezameyqani (talk) 14:49, 19 August 2014 (UTC)
 * The Free State of Prussia was a state of the German Reich, that is, Germany. Likewise Russia was a state of the USSR.  We don't list states as birthplace generally.  Charlie Fefferman was born in Maryland, but we list his birthplace as USA.  Why should other geopolitical subdivisions be given special treatment, especially when there is no obvious benefit?  Sławomir Biały  (talk) 15:08, 19 August 2014 (UTC)
 * Your logic clearly convinced me.I will change birthplace flagicon of Grothendieck and Roth to German Reich and all mathematicians who were born in USSR,to flagicon of USSR.Would you find this change logical? Rezameyqani (talk) 15:15, 19 August 2014 (UTC)
 * So if you address the flag icon and country links, what differences remain between your version and the one appearing in the article?  Sławomir Biały  (talk) 20:48, 19 August 2014 (UTC)
 * @ Sławomir Biały The current table does not give any information about current institution of medalists. Also the information on the current table does not cite any sort of references.Rezameyqani (talk) 06:25, 20 August 2014 (UTC)
 * Ok, although the existing table does have information about the current institution, I prefer the way you have done it (with flag icons beside the institution names). Would it also make sense to include flagicons beside the Institution (at the time of receiving the medal) field?  Or would this just be too many icons?  As for the references, it's a little unclear to me what purpose the references serve here.  Are they meant to refer only to the "current/last institution" field of the table?  That's how I interpret them, and I think that this column of the table probably has the greatest need for references.  Presumably the rest of table can for the most part be referenced as a whole to some other source such as http://www.mathunion.org/general/prizes/fields/prizewinners.  You might want to make this explicit somehow.   Sławomir Biały  (talk) 11:32, 20 August 2014 (UTC)
 * @ Sławomir Biały Your interpretation about the use of references are true.I put references to clarify the current affiliation. Also,Yes, putting more flags in the table would make it a messy.Rezameyqani (talk) 12:42, 20 August 2014 (UTC)
 * Ok, then just to summarize: in the birthplace/ICM fields, only include the flag icon of the country that existed at the time. You might consider including an explanatory statement about the references, and a general reference that covers the rest of the content of the table.  Otherwise, the new table is fine by me.   Sławomir Biały  (talk) 13:03, 20 August 2014 (UTC)
 * Error: The flag icon for Columbia University appears to be Russia, but it should be the USA.  Sławomir Biały  (talk) 13:05, 20 August 2014 (UTC)
 * With regard to birthplace column,I think it would be better for countries that have been formed after dissolution of USSR,We mention the new country's flag.Same goes for Weimar republic or West Germany.Thank YouRezameyqani (talk) 14:33, 20 August 2014 (UTC)
 * I thought we had already discussed this. The country of birth should obviously listed as whatever country existed at the time.  Any other designation would be an ahistorical anachronism.  Sławomir Biały  (talk) 14:57, 20 August 2014 (UTC)
 * @ Sławomir Biały You're right.I looked at some of the medalist's CVs,and they mentioned their birthplace exactly as what you said.Would you mind correcting this error yourself?Thank YouRezameyqani (talk) 15:03, 20 August 2014 (UTC)
 * I wanted to add, but had an edit conflict: a firm position on this also helps to ensure a neutral point of view. Nationality designations are notoriously prickly on Wikipedia.  Sławomir Biały  (talk) 15:10, 20 August 2014 (UTC)
 * And If it is possible for you,correct these flags on the table which exists on Fields Medal talk page,OK?Thank buddy. I just made corrections and updated the table.hope you approve it.ThanksRezameyqani (talk) 15:05, 20 August 2014 (UTC)
 * The version that appears above now looks very well done to me. I have no further comments at the moment.   Sławomir Biały  (talk) 11:10, 21 August 2014 (UTC)

Why are there flags at all? By the guideline MOS:FLAG none are needed. They should only appear when the subject represents a country – a national team for example, or the country itself. They should also include the county name as it's often non-obvious from the icon what the country is. So these are both unnecessary and incomplete. None are needed; if someone reading doesn't know and wants to know what country Oslo is in, or the University of Helsinki, or Harvard University they can click on those links. See also MOS:FLAGS – the above table is far worse than the example they give as excessively cluttered and redundant icons. They also result in massive overlinking, giving e.g. dozens of links to United States. There are much better ways to indicate the number of medals per country, such as has been done at Fields Medal.-- JohnBlackburne wordsdeeds 11:43, 21 August 2014 (UTC)


 * It would have been better to have this discussion at Talk:Fields Medal. I, for one, am only now seeing it. Check out at my latest suggestion there and see what you think. I am not married to flags, as some of my comments there make clear, but I do see a potential benfit to having flags in front of the institution names, as I have alluded to over there. As for the separate "Landmarks" by-country table, I think it gives more emphasis to the countries than just having flags in the one table does. But that's just my opinion... In any case, please discuss this table on the talk page of the article where the table belongs. - dcljr (talk) 12:29, 21 August 2014 (UTC)

letter to the Foundation
Not exactly math, but maybe of interest for some here that followed the recent struggle of the foundation with parts of the community and/or dislike the curret decision making process or the media viewer:

m:Letter to Wikimedia Foundation: Superprotect and Media Viewer/en

--Kmhkmh (talk) 11:41, 21 August 2014 (UTC)


 * This is related, of course, to the ongoing discussion in the above thread.  Sławomir Biały  (talk) 13:19, 22 August 2014 (UTC)

Time Cube
Is Time Cube within the scope of WikiProject Mathematics? See diff. Johnuniq (talk) 02:33, 23 August 2014 (UTC)
 * It should be very obvious that it is not. The article should be deleted promptly. YohanN7 (talk) 03:11, 23 August 2014 (UTC)
 * Why? The web site is nonsense, of course, but that doesn't prevent it from being notable. And are mathematical cranks not something we should concern ourselves with here? —David Eppstein (talk) 04:31, 23 August 2014 (UTC)
 * What then is a notable mathematical crank? The article on Myron Evans is gone. He, if anyone, is a notable crank. YohanN7 (talk) 05:09, 23 August 2014 (UTC)
 * Mathematical crankery, on such topics as angle trisection, is certainly notable, and the subject has been covered in serious books from De Morgan to Underwood Dudley. It should also be within the purview of this project.  It follows that individual cranks or crankeries might well be notable simply for their crankishness, if they or their crankeries are covered in independent reliable sources such as Dudley's books.  Indeed, some cranks are notable for other reasons as well.  Incidentally, please be sure to distinguish mathematical crank (person) from crank (mathematics)  and crank conjecture from the crank conjecture!  Deltahedron (talk) 06:12, 23 August 2014 (UTC)
 * In answer to Johnuniq's question, I'd say that it is clearly not within scope, considering the site's content. David, let's also not confuse the person with the site: the site is the subject of the article ( so we'd have to rename the article to Otis Eugene Ray to justify not deleting it ). —Quondum 15:50, 23 August 2014 (UTC)
 * Not a maths crank so not in the purview of this project. Could possibly be considered a phyics nutter though. Dmcq (talk) 16:23, 23 August 2014 (UTC)

Problematic redirect: Cantor's paradise
Currently, Cantor's paradise redirects to set theory.

Not meaning to offend anyone, this does not appear to me to make much sense. Cantor's set theory was famously not axiomatic (whether it was "naive" depends on just what you mean by that, and that's another whole hornet's nest that needs to be addressed, particularly at Russell's paradox, but that's not what I came to talk to you about today).

It seems to me that there are about three things that could be done with it.


 * 1) It could be deleted.  That actually seems like a reasonably good option to me, compared to the alternatives.
 * 2) A brief article could be written on David Hilbert's famous quote, which introduced the phrase.
 * 3) An article could be written on the conceptual basis of set theory, and this term redirected there.  (No, that article is not naive set theory, which is a seriously problematic article, largely because it's not clear which of the several interpretations of that phrase the article is supposed to be about.)

Thoughts? --Trovatore (talk) 02:53, 21 August 2014 (UTC)


 * The alternatives are not at all bad. As for me, (2) is better than (1). About (3), it is definitely better than (2) provided that you are ready to write such article! Boris Tsirelson (talk) 05:30, 21 August 2014 (UTC)
 * (2) [maybe call it (2 minimal)] could be simply a redirect to David Hilbert and mention the phrase there. If enough material accumulates, one could then split it to an article. JMP EAX (talk) 07:14, 21 August 2014 (UTC)
 * That's not bad. --Trovatore (talk) 07:15, 21 August 2014 (UTC)
 * Regardless of where the paradise redirects to, the Hilbert quote should certainly be included at set theory! Tkuvho (talk) 07:24, 21 August 2014 (UTC)
 * Some possible sources with varying degrees of praise and criticism:     JMP EAX (talk) 07:31, 21 August 2014 (UTC)
 * Based on the various commentary in 2nd-ary sources it seems to me the phrase relates more to Cantor's concept of infinity rather than his concept of set. Although some do interpret it in that sense too. JMP EAX (talk) 07:36, 21 August 2014 (UTC)
 * I have seen (I think) Von Neumann universe being referred to as Cantor's paradise. It makes some sense. YohanN7 (talk) 11:35, 21 August 2014 (UTC)
 * Following what JMP EAX said, see Tav (number) and Absolute Infinite. JRSpriggs (talk) 13:57, 21 August 2014 (UTC)
 * What do reliable sources say? One possibility is .  Deltahedron (talk) 14:20, 21 August 2014 (UTC)
 * The book by Ferreiros seems carefully researched and would be a good basis for a short article on Cantor's paradise. One could mention that Cantor's paradise is an expression used by Hilbert in describing set theory and infinite cardinals developed by Cantor. The context of Hilbert's comment was his opposition to what he saw as Brouwer's reductive attempts to circumscribe what kind of mathematics is acceptable. Here one could provide a link to the "Brouwer-Hilbert controversy" page which is unfortunately in sad shape. Anyway the main purpose of the page would be to provide some historical context for the term and also to provide links to pages like Hilbert and Brouwer and set theory. Any objections? Tkuvho (talk) 09:10, 24 August 2014 (UTC)

Decidability of the first-order theory of the real numbers
Decidability of the first-order theory of the real numbers could use some work. Michael Hardy (talk) 12:32, 24 August 2014 (UTC)
 * According to the current contents this should be "decidability of first-order theories", without the definite article. Tkuvho (talk) 12:44, 24 August 2014 (UTC)
 * IMO this article must be merged with Tarski–Seidenberg theorem, which is about the proof of this decidability. D.Lazard (talk) 14:16, 24 August 2014 (UTC)

Issue in Fréchet mean
Please see Hermann Karcher's unanswered complaints from a month ago in Talk:Fréchet mean (and see also Karcher's arXiv preprint, linked there, for more detail). Karcher seems to object to the terminology "Karcher mean", preferring an older name "Riemannian center of mass". But it is unclear to me exactly what scope these names are supposed to have (metric spaces or Riemannian metric spaces?), whether Karcher's work really does concern the local-vs-global difference from the Fréchet mean that the article now ascribes to him (Grove and Karcher 1973 seems to be about uniqueness of minima in certain Riemannian spaces), or what the right name should be. Karcher recently contacted me by email asking for help getting this resolved. Perhaps someone with more expertise than me in these areas can help clarify this. —David Eppstein (talk) 23:34, 24 August 2014 (UTC)
 * I've cited a couple of papers that use the term, and I suspect that many more could be found. But I don't know what that implies under Wikipedia policy. So we could still use some help there. Mgnbar (talk) 00:17, 25 August 2014 (UTC)

It is perhaps also worth noting that the Definition section of the article is in incredibly poor shape and badly needs to be rewritten for clarity of writing, punctuation, mathematical English, and clarity of notation. (I am on mobile so am not able to easily do it myself.) --JBL (talk) 01:50, 25 August 2014 (UTC)

List of things named after members of the Bernoulli family
The new List of things named after members of the Bernoulli family could probably use some work. Which member of the family was each thing named after? If it were a long list, it could be divided into secctions by subject matter; with a list as short as this one, maybe a brief comment on each item could be included. Michael Hardy (talk) 15:33, 25 August 2014 (UTC)
 * In its current form, I don't see how this isn't redundant to the Mathematics subsection at the Bernoulli disambiguation page. -- Kinu  t/c 18:09, 25 August 2014 (UTC)
 * I would suggest merging it into Bernoulli family, which already basically duplicates the same list. bd2412  T 18:16, 25 August 2014 (UTC)
 * One difference from the dab page is that a list (as a separate article or in the Bernoulli family) can contain entries that would not reasonably be the target for a wikilink to "Bernoulli". These entries should not be included in a disambiguation page. Some possible such links include Euler–Bernoulli beam theory, Poly-Bernoulli number, Time-inhomogeneous hidden Bernoulli model, etc. —David Eppstein (talk) 18:53, 25 August 2014 (UTC)

Variation (combinatorics) proposed for deletion
Is the article titled Variation (combinatorics) worth keeping? And should it say that the chosen elements are distinct? Michael Hardy (talk) 13:23, 24 August 2014 (UTC)
 * I PRODed the article because I think that it is essentially WP:OR. The editor who created the page said (Talk:Combination) that until he created the page some Russian wikipedians were unable to find an English equivalent to the term they were using. I think that this is essentially correct, there is no common English term used to simultaneously refer to k-tuples and k-permutations (partial permutations), at least not in my extensive set of references on combinatorics. I believe that the term was made up to fill this hole in translation, and so, would not have any reliable secondary sources, hence the PROD. As an aside, if you make a statement about the distinctness of the elements you would lose one of the two meanings of the "proposed" term. Bill Cherowitzo (talk) 16:07, 24 August 2014 (UTC)


 * This book shows historical use of the term variation for m-permutations. Given the term exists, a redirect would be preferable to a deletion. --Mark viking (talk) 17:02, 24 August 2014 (UTC)


 * I can't really speak to use of the term in English but in German in it is common/standard term (see de:Variation (Kombinatorik). Should the term occasionally appear in (older) English literature as well, a redirect or rather a link from the disambiguation page for variation might be appropriate.--Kmhkmh (talk) 17:25, 24 August 2014 (UTC)


 * Thanks for that reference Mark. While it still doesn't give us an early English source (I don't think Bernoulli wrote in English ) it does change my opinion about this page. A redirect is slightly problematic since there is no single topic to point this to, so I'd prefer the link from the disambiguation page for variation. This link would permit splitting off variations with repetition from variations without repetition. I think this should be done even without any proof of archaic English usage, since the term is currently used in this way in German (my sole German combinatorics book does not use the term, so I missed this). Bill Cherowitzo (talk) 04:23, 25 August 2014 (UTC)


 * Done, as a redirect to a disambiguation page (may the Disambiguation Gods forgive me). Bill Cherowitzo (talk) 19:46, 25 August 2014 (UTC)

Commutative property
A good article (or even appropriately titled)? Discuss. I should mention that I lambasted this wikipearl on Jimbo's talk page. JMP EAX (talk) 01:58, 26 August 2014 (UTC)

I should note that the same anti-*ivity campaign was apparently carried out on another page given how an old version looked like. JMP EAX (talk) 02:13, 26 August 2014 (UTC)

Disambiguation help needed with Antisymmetric
Expert attention is needed in figuring out about a dozen and a half incoming links to Antisymmetric, which has recently been made into a disambiguation page. I am wondering whether it would be possible to write a general concept article on antisymmetry in mathematics to which links could generally be pointed. Note that Skew-symmetric and Skew symmetric also pointed here, but I retargeted both of those to Skew-symmetric matrix, as it is the only article that is a title match to these terms. Hopefully that was the right answer to that problem. Cheers! bd2412 T 02:50, 25 August 2014 (UTC)


 * I think you overlooked the existence of Skew-symmetric graph, Skew-symmetric form (redirect to Bilinear form) and Skew-symmetric tensor (redirect to Antisymmetric tensor).  I would suggest Skew-symmetric and Skew symmetric be re-redirected to Antisymmetric, as should Skew symmetry.  Deltahedron (talk) 08:03, 25 August 2014 (UTC)
 * In addition, the heading of Antisymmetry should link to Antisymmetric, or possibly move Antisymmetry to Antisymmetry (linguistic) and move and redirect Antisymmetric to Antisymmetry as the main disambiguation page. Deltahedron (talk) 08:32, 25 August 2014 (UTC)


 * I agree that antisymmetry should probably be moved to antisymmetry (linguistics).  Sławomir Biały  (talk) 11:21, 25 August 2014 (UTC)
 * @Deltahedron, note that Antisymmetric and Skew-symmetric may be synonyms for some purposes, but are not homonyms. If Skew-symmetric has multiple meanings itself then Skew-symmetric should be either its own article (if the meanings are related) or its own disambiguation page (if the meanings are unrelated). bd2412  T 12:42, 25 August 2014 (UTC)
 * I have no problem with a separate disambiguation page for skew-symmetry. Deltahedron (talk) 12:51, 25 August 2014 (UTC)
 * The question still remains whether there is a general concept of Skew-symmetry. Does the term "skew-symmetric" mean something different and unrelated in Skew-symmetric graph, Skew-symmetric form, and Skew-symmetric tensor, or is it possible to describe a single general concept of skew-symmetry that explains all of them? Also, if these are going to be disambiguation pages, the links will need to be fixed. bd2412  T 13:40, 25 August 2014 (UTC)
 * I think there is: skew-symmetry is a property of structures that have some notion of sign, and an involution which changes the sign. For example, a skew-symmetric graph has an involutions that reverses all arrows; a skew-symmetric matrix has an involution (transposition) which changes the ± sign on all the matrix entries; a skew-symmetric form is a function of two variables which changes sign on interchange of the variables.  Anti-symmetry is vaguer, but one might describe it as a property of structures with a Boolean set of values which reverse under the involution.  For example, an anti-symmetric relation is one which if is true for the terms in one order is false when the terms are interchanged.  However this is all a personal opinion.  Deltahedron (talk) 13:53, 25 August 2014 (UTC)
 * There must be sources describing these things (after all, there must be some basis for the development of your opinion on the matter). Note that NIST defines skew symmetry as: "The property that the flow is the same amount, but reversed direction, starting from either vertex of every edge of a flow network. More formally, for an edge e=(v,w), f(v,w) = -f(w,v), where f(a,b) is the flow from a to b". bd2412  T 15:07, 25 August 2014 (UTC)
 * That would seem like a rather specialized use of the term, not in accord with the way it is used in much of mathematics. So I don't think that's much of a foundation on which to build an article on the general concept.   Sławomir Biały  (talk) 21:32, 25 August 2014 (UTC)
 * Frankly, I am having a hard time finding much about it as a concept. Suppose we have an article on the general concept of Antisymmetry (mathematics) and discuss the distinction between antisymmetry and skew symmetry there, to the extent that there is a difference that can be sourced? That would also solve the problem of incoming links, since that all seem to be math-related. bd2412  T 22:28, 25 August 2014 (UTC)
 * Is there a difference? As far as I know, these are synonyms.  What exactly is being proposed here?  If we can't find an authoritative source that treats various notions of skew-symmetry (or antisymmetry) as a whole, then probably there shouldn't be an article on that subject.  We shouldn't have an article just in order to have something for ambiguous links to point to.    Sławomir Biały  (talk) 23:05, 25 August 2014 (UTC)
 * I don't want to force a square peg into a round hole here (sorry, topology joke), but it seems that the antisymmetry in effect in an antisymmetric matrix is the same kind of antisymmetry in effect in an antisymmetric relation or an antisymmetric tensor. If not, then the many math articles linking to antisymmetric need those links fixed (and the same for skew-symmetric cases), because readers following those links will have a hard time finding the right option on a disambiguation page. If they are related, then the relationship itself can be described both in terms of its functionality and in the history of its discovery and development. I rely on the experts here to determine the optimum solution. bd2412  T 03:05, 26 August 2014 (UTC)
 * @Sławomir Biały: but we do have such articles for example Self-adjoint, center (algebra) or centralizer. It think this is what WP:SIAs are for. JMP EAX (talk) 06:46, 26 August 2014 (UTC)
 * But I don't think these are analogous. For instance, basically all uses of "self-adjoint" come from the usage in star algebras, so that is the primary meaning.  An SIA is fine as a list in this case, but I really don't think there is enough here to write an article on.  Any article in this case would almost certainly be very bad.   Sławomir Biały  (talk) 10:35, 26 August 2014 (UTC)

Bad redirect
Super linear algebra is now a redirect to linear algebra. Anyone looking for information on super-anything will only be frustrated by such a redirect. It would be better if it was just a red link. YohanN7 (talk) 17:05, 29 August 2014 (UTC)


 * I made it into a redirect to Super vector space. Better than nothing. JRSpriggs (talk) 18:01, 29 August 2014 (UTC)


 * Our article superspace should really include a disambiguation note (indicating the other common usage of superspace as the opposite of subspace). I wonder also if there is a possibility of confusion with "super vector space" or (worse) "vector superspace".   Sławomir Biały  (talk) 18:28, 29 August 2014 (UTC)

Mathematics under Flow
I have been trying out mw:Talk:Sandbox to see how mathematics renders inside Flow. It doesn't seem to be working to well for me: has anyone else tried it recently? Reports at Wikipedia talk:Flow, not here, I suggest. Deltahedron (talk) 19:42, 29 August 2014 (UTC)

Mathematical Reviews and Zentralblatt MATH
There is a discussion taking place, regarding a proposal, which I think might be of interest to this project and community.

The discussion is taking place on  the talk page of Notability (academic journals). ---Steve Quinn (talk) 04:18, 29 August 2014 (UTC)


 * The talk page you've linked was last edited more than 1 year ago; the particular section that you've linked seems to be 3 years old. Did you mean to link something else?  JBL (talk) 06:08, 29 August 2014 (UTC)


 * Probably Wikipedia_talk:Notability_(academic_journals). Deltahedron (talk) 06:13, 29 August 2014 (UTC)

The context is Articles for deletion/Mathematics and mechanics of complex systems. —David Eppstein (talk) 06:51, 29 August 2014 (UTC)
 * Sorry about the wrong link - I was probably tired when I posted it. Thanks to the two editors who posted the corrections. -- Steve Quinn (talk) 05:53, 30 August 2014 (UTC)

A response from WMF
Assiduous readers here will recall that, following a challenge from Jimmy Wales, there was a discussion here about what we might want from WMF in the way of mathematics rendering and editing. I sent a summary of that discussion to Jimbo Wales, who responded that he had "copied this text to the Board wiki and emailed the board (and Lila) asking them to read it" and would "personally recommend that we allocate resources to this". I have just had a discussion with Rachel diCerbo, Director of Community Engagement (Product) at WMF and have to report that she has told me that "It's one of those things that is "on the radar" but for now I don't imagine that we would be able to put it on the roadmap for the foreseeable future". It's very disappointing, especially since the proposal had Jimmy Wales's backing, and I'm sorry not to be able to deliver better news. However, I suppose that at least we know where we stand now -- WMF are not going to allocate resources to improving mathematics editing and rendering software in the near future, not are they going to include it in their planning. Rachel mentions James Forrester as "the Product Manager responsible for this area" but without resources it seems hard to see what there might be to discuss with him. In short, it does not appear that mathematics has a future here. Deltahedron (talk) 16:38, 19 August 2014 (UTC)


 * This is misleading, to put it mildly. Mathematics coverage continues to have a strong future at Wikimedia, but the strength of that future continues to be based on what the community of interested individuals wishes to put in. It would be good to see some people engaging with and helping the other excellent, hard-working volunteers who have toiled away to give you the tools you have today, as well as those to come. Instead, so far mw:Extension talk:Math/Roadmap merely has a copy-pasted list of passively-given complaints, without any sign that people want to actively help make changes and improve things. Code is not written by magic fairies. It's written by real people. It'd be wonderful to have some positive engagement here. Jdforrester (WMF) (talk) 16:54, 19 August 2014 (UTC)


 * "Misleading" is a strong term, I reject it, and you should withdraw it. I have done my best to represent the discussion as I see it and have given a link to the conversation on meta for the interested.  The result of this discussion is that mathematics rendering and editing has no place in the WMF roadmap, and the WMF with over 100 employees and income in the tens of millions of dollars has not resources to allocate it, nor does it propose to include mathematics editing in its planning process (nor does it currently have a mecanism to do so).  The content of the encyclopaedia is not written by magic fairies either.  It's written by volunteers, many of them experts, for free.  For a paid employee to come here and tell people who are contributing to the project, for free, by writing the encyclopaedia, that if they wish to be able to continue to do so, the way to do that is that they have to become unpaid developers as well, is -- unhelpful.  Deltahedron (talk) 17:02, 19 August 2014 (UTC)


 * Having said which, let me pick up a couple of points to see if there's a way forward. Assuming that these volunteer developers do come forward, where is the locus for the discussion between what the editors and readers might want or need, and what those volunteers might want or be able to work on?  Is it mw:Extension talk:Math/Roadmap?  James is rather scornful of the "copy-pasted list", which I might mention was the compilation of discussions at this and similar pages, and the same list that Jimmy Wales was so enthusiastic about, as noted above.  What was wrong with it?  Is that the right place?  If not, where is?  Do WMF staff want to involve themselves in the conversation there, or is it solely between one group of vounteers and another?  Is that where WMF staff will alert us to impending changes in other bits of software that might affect us, or will they use some other channel?  What is the way forward, according to WMF?  Deltahedron (talk) 18:13, 19 August 2014 (UTC)


 * I believe that there are slightly fewer than 60 software engineers at WMF, out of 208 staff overall (I could be wrong – seer the list on Foundation wiki). My team of six(-ish) people are the ones working on tools for editing – everything from formulæ to VisualEditor to hieroglyphics to citations to sheet music to… The rest work on areas like performance, discussions ("Flow"), multimedia, mobile phone, tablet and native app support, languages, and stopping the site from falling over. You're welcome to lobby senior management to increase the number of people working on editing tools so that we could dedicate some paid developer time to mathematical support, but characterising it as a straight choice between supporting and not supporting your particular editing requirement isn't entirely fair. :-)
 * On how to engage with fellow volunteers, there's nothing 'wrong' with copy-pasting a complaint, but it's unlikely to get things to happen compared to actually talking these things through with the people that can help you. I find that an isolated discussion on one wiki – even a wiki as important as enwiki – is often not very useful in getting stuff done. Jdforrester (WMF) (talk) 18:59, 19 August 2014 (UTC)


 * I am sorry to hear that WMF has such a small paid staff of engineers at its disposal. We have precisely zero paid staff here at WPM.  This discussion is precisely because I did lobby senior management for more support for mathematics editing, and the result was, as described above, disappointing, if not exactly a huge surprise.  Of course everyone here realises that things have to be prioritised, I really do not think you need to explain that.  You have described what is not very useful to get things done.  What, then, in your opinion, should we do to increase the probability of getting the thing we think need to be done?  Who are these people who can help us, and where do we go to talk these things through with them?  Deltahedron (talk) 19:08, 19 August 2014 (UTC)


 * Additional. You describe this as a "complaint".  That's rather disimissive.  It's a summary of a constructive discussion across three languages, in response to a specific request by Jimmy Wales and subsequently described by him as "a very helpful and concise statement of the issues and concerns".  The fact that you dismiss it as a complaint suggests that perhaps you did not actually read it carefully?  Deltahedron (talk) 19:13, 19 August 2014 (UTC)


 * I'm talking about this copy-and-paste, FWIW. I don't see "complaint" as a "dismissive" term, and I'm again disappointed that you try to twist my words like this; it's just an observation that giving people a list of demands (do this, then do this, finally do this) is forum-shopping, not real engagement with people.
 * There is no "why" in what you pasted, no actual discussion of what was on the roadmap to which you were posting, it was clearly written for another audience (you do realise that you posted a list of demands to WMF on a page aimed at volunteer developers, right? Surely you can imagine how demoralising/marginalising that would feel to those volunteers?)
 * As to the best way to engage with fellow volunteers, I can't speak for them as to what works best, I can just advise on what often doesn't work. Have you asked them how they would like you to engage with them? Jdforrester (WMF) (talk) 20:09, 19 August 2014 (UTC)
 * I am unable to regard this as a helpful answer. It is quite astonishing that the Product Manager, VisualEditor team, with a remit "to ensure that his "team understands what the community wants and needs, is focussed on the things that matter, and is engaging with and understood by the community" is quite ignorant of how and where to effectively engage with a group of volunteers, and is unable to suggest anyone who might be able to help.  Let me just quote my question and your answer together
 * "Who are these people who can help us, and where do we go to talk these things through with them?"
 * "Have you asked them how they would like you to engage with them?"
 * This may strike you as amusing but it is unworthy of someone in your position. Now please may we have a considered answer to the question?  Deltahedron (talk) 20:22, 19 August 2014 (UTC)


 * For the tenth or eleventh time, the Roadmap page to which I directed you last month is probably the best venue. For which volunteers are best to talk to, I don't know for sure, but looking at this history page suggests mw:User:Physikerwelt and mw:User:Schubi87. Your continued grossly bad faith responses like this make me sad. Jdforrester (WMF) (talk) 20:28, 19 August 2014 (UTC)


 * ping User:Physikerwelt so he knows about this discussion. (I guess cross-wiki pings like the link to mediawikiwiki above don't work). --Jeremyb (talk) 06:35, 22 August 2014 (UTC)


 * pong. I'm waiting for code review here . I think we are on a good way here. I'll come back once the change is merged and the new Math rendering can be tested at beta-cluster. (I can be tested at the moment but does not work). I proposed to establish a mentioning system for volunteers (i.e. extension maintainers) to Quim Gil. I'm waiting for a reply here. I think it's essential that there is exactly one fixed point of contact on the WMF side for every extension used in production. --Physikerwelt (talk) 15:42, 22 August 2014 (UTC)
 * Good news, and thanks for all the effort you've been putting in on this. Deltahedron (talk) 16:00, 22 August 2014 (UTC)


 * "Probably"? You mean you don't know where or how volunteer effort is coordinated?  Deltahedron (talk) 20:35, 19 August 2014 (UTC)


 * I don't know for sure, no. I'm not in a position to order volunteers around, and that includes telling them where and how to congregate. I can merely observe and intuit based on what looks sensible. Do you think instead I should try to tell volunteers that they can only participate in venues I run? That doesn't seem very wiki like or respectful. Jdforrester (WMF) (talk) 21:19, 19 August 2014 (UTC)


 * I would expect that you would know where and how volunteer effort is coordinated and conducted, so that you may effectively advise volunteers about what is and is not a good use of their time and energy, and so that once work is done, you can most efficiently integrate it into the code base. Restricting yourself to observing seems an ineffective way of doing your job of engaging with the community.  The dichotomy posed is fallacious, but I only mention this because you asked.  Deltahedron (talk) 21:32, 19 August 2014 (UTC)

Deltahedron's aggressive stance on this matter notwithstanding, the state of mathematics rendering on Wikipedia is disgraceful. It has not meaningfully advanced in the years that I have been editing, despite massive improvements in the available pool of open source solutions. If Wikipedia is to continue to be used as a serious resource for people studying technical fields, improvement of the support for LaTeX needs to be made an immediate priority. The WMF's brush-off response is also unacceptable. Foundation representatives need to get their heads out of the sand and address this pronto. Sławomir Biały (talk) 20:46, 19 August 2014 (UTC)


 * Again, I disagree that everything is WMF's responsibility. Wikimedia's software needs are and have always been met by a collective of editors, developers and sysadmins, some paid, most volunteers. The WMF is not the communities' mother, and isn't and hasn't been solely responsible for much of the software you use daily; in many cases, it's almost entirely uninvolved.
 * Differing areas of the software have differing levels of staff vs. volunteer involvement. For instance, I believe that all of the maths-related technology you're currently disparaging as "disgraceful" was created by volunteers, not by paid developers, and almost all of it back in the early 2000s (and you're right, its age is showing, as is the case with much of the software for Wikimedia wikis). In the other direction, despite long term requests for a rich/visual editor starting in 2003, no volunteers felt ready to take on the task, so WMF has funded that from 2011 onwards (though volunteer coders have contributed significant amounts there too).
 * That said, I'd be keen to understand what in your view would be us "get[ting] [our] heads out of the sand". Jdforrester (WMF) (talk) 21:19, 19 August 2014 (UTC)
 * WMF decided to fund visual editor, a solution for a non-existent problem, but not proper LaTeX support, a problem that has been known about for ten years? Obviously the social media aspect of the encyclopedia is a bigger priority for the foundation than the encyclopedia aspect.  So I guess we have our answer.  Thanks.  Sławomir Biały  (talk) 21:29, 19 August 2014 (UTC)
 * Indeed, we did our best, we put up our best case, with the best support we could muster, and we were turned down. That is the answer.  Mathematics development from now on depends on volunteers appearing from somewhere (the magic pixie solution, I believe someone called it) and the staff being able to spare the time to help them integrate any code they come up with.  James believes this is a "strong future" and while I'm willing to look for a way forward, if someone can find it, I also think we need to consider whether the time has come.  WMF projects are clearly at a major change point, the old ways are no longer viable, and the old guard may find themselves no longer part of that strong future the WMF have planned.  Deltahedron (talk) 21:43, 19 August 2014 (UTC)
 * I think I'm being quite restrained, given that James has seen fit to accuse me of "misleading, to put it mildly", "you try to twist my words", "Your continued grossly bad faith responses" and has described this community's constructive proposals to WMF as "complaints" and "demands". However, I am still looking for a way forward, and if, as James thinks, mw:Extension:Math/Roadmap is the place to go then, then we shall just have to go there.  Deltahedron (talk) 20:56, 19 August 2014 (UTC)
 * Indeed. I think I would be more blunt.  Anyone giving you the run-around on this issue should be seriously worried about their possible future at the foundation.   Sławomir Biały  (talk) 21:09, 19 August 2014 (UTC)


 * I've read a couple of these threads, and I don't understand exactly what is happening. Is the worry that an upcoming visual editor will make math editing much less pleasant than the status quo? Or is the worry that Wikipedia's math support is stagnating, while the rest of the world progresses? In other words, are we worried about things getting worse, or just not getting better? Mgnbar (talk) 21:48, 19 August 2014 (UTC)


 * I am not normally a pessimist, but I expect WMF to begin actively driving out mathematics (and computer science, physics, and engineering) editors before too long. Our articles are a small fraction of Wikipedia, and unlike lists of Pokémon they require extensive development effort, effort that could go to shiny toys like WP:Flow and WP:VisualEditor.  Removing mathematics support would vastly simplify their lives and still satisfy 99% of their audience, so I doubt if it will be long before it happens.  Ozob (talk) 02:02, 20 August 2014 (UTC)


 * I have no intention of removing the existing mathematics support the community have worked so long to build. Worry not! Jdforrester (WMF) (talk) 02:11, 20 August 2014 (UTC)


 * But if a high-priority project is blocked because it would require months of work on mathematics rendering, are you really willing to fight back? Where would your engineers find time?  Look at their schedules: can one of them carve out three months (longer, if you account for their other responsibilities)?  I'm sure that WMF is willing to keep mathematics support as long as it doesn't conflict with other priorities, but years of experience have shown that it is clearly unwilling to do mathematics development.  Furthermore, because mathematics support is complicated, few people in the community have the technical skills, time, and interest to volunteer to lead such a large project.  Sure, it's conceivable; but if I were the one responsible for mathematics rendering, then, given the other demands I have on my time, deliverable code would take years.  With no support in sight, the obvious conclusion is that some day, mathematics will disappear from Wikipedia.
 * Put more simply: WMF doesn't care. Ozob (talk) 04:19, 20 August 2014 (UTC)


 * We had something like this a year or so ago when it became clear that there were issues about interaction of mathematics markup, Visual Editor and Flow. I'm far from saying that WMF were deliberately trying to break mathematics editing, but it was clearly not a high priority, even after they were made aware of the issues.  The mathematics component of VE was a GSOC student project: this is the best we may now reasonably expect -- not that it's removed, but that it simply ceases to function in newly released software because there is no plan to do continue it, no effort available to support it, and that it it survives, it's on the basis of ad hoc unplanned unsupported uncoordinated volunteer effort.  That is the "strong future" we have to look forward to.  Deltahedron (talk) 06:36, 20 August 2014 (UTC)


 * Just to be very clear, there was never even a remote possibility that VisualEditor or Flow would not support formulæ. There wasn't ever an "incompatibility"; such an issue existed purely in the minds of people raising the (understandable, if entirely misplaced) concern that something might not work. Repeating the unfounded claim now, however, and especially doing so whilst painting yourself into the position of heroically saving the wiki from the evil WMF who just doesn't listen unless you scream from the roof-tops, is just FUD (or I could use less charitable terms). Please don't let yourself be tricked into saying false things like this. Jdforrester (WMF) (talk) 18:08, 21 August 2014 (UTC)
 * My statement was "there were issues about interaction of mathematics markup, Visual Editor and Flow". Are you saying that is false?  If you wish to use the word "false", quote an assertion of mine; it is a waste of your, my, and everyone else's time to make statements that no-one ever made and then proclaim them to be false.  Here's an example.  You said "there was never even a remote possibility that VisualEditor or Flow would not support formulæ"  That is incorrect, and to show you that it is, I refer you to, where Brandon Harris says "I cannot promise that there will be mathematics markeup in normal discussion comments".  In other words, there was a time, namely 23:42 on 17 August 2013, when a senior member of WMF staff expressed his view that mathematics markup would not work in normal Flow discussions.  You also stated that there was "never an incompatibility".  That is incrrect, and it was resolved by using Parsoid, but not without bugs: see Wikipedia_talk:Flow/Archive_6.  So on 19 October 2013 there was an incompatibility.  Please review your comments in the light of the evidence, and then withdraw your words "unfounded" and "false".  Deltahedron (talk) 18:29, 21 August 2014 (UTC)
 * Additional comment on the claim "there was never even a remote possibility that VisualEditor or Flow would not support formulæ". I just tried it again at mw:Talk:Sandbox and mathematics does not work under Flow, a year after the discussions I mentioned.  Deltahedron (talk) 19:50, 29 August 2014 (UTC)
 * Deltahedron wrote a summary of the proposal here. Speaking for myself, there is an element of both concerns.  It is a definite concern that adequate support for editing mathematics will be removed from VisualEditor, since inclusion seems to be rather an afterthought entirely dependent upon the valiant efforts of one or two volunteers.  But at the end of the day, I don't really think that will happen; the blow-back would be Wikipedia's own doom.  The other element is the solid lack of improvement.  We have for years been screaming for proper mathjax support (as summarized as the first "Specific" point of Deltahedron's summary).  But instead of funding this important but mundane feature, the WMF throws its resources into developing VisualEditor, Flow, and other questionable and unimportant bits of flashy gadgetry.  This clearly demonstrates the perversity of WMF's engineering priorities.   Sławomir Biały  (talk) 11:43, 20 August 2014 (UTC)

It is important to remember that James is not in a position to be able to set the goals of the organisation. As such, he could not reassign engineers even if he wanted to. I would recommend following James's advice to contribute to the roadmap. Failing that, you could make your case directly to Erik Möller, the Vice President of Engineering and Product Development. --Dan Garry, Wikimedia Foundation (talk) 02:31, 20 August 2014 (UTC)
 * Dan, we know that, and did not ask or expect him to. If you follow the links in my first posting in this section, you'll see that we have already made our case, at Jimmy Wales's suggestion, and he passed it on to the Board and to Lila.  The answer was, as I quoted above, "It's one of those things that is "on the radar" but for now I don't imagine that we would be able to put it on the roadmap for the foreseeable future".  So your advice, though doubtless well-meant, is somewhat behind the curve.  We tried every element of what you suggest, and it failed.  The question we are discussing now is, how we can move forward to sustain mathematics rendering and editing in the absence of any material support or guidance from WMF.  That may not be easy -- indeed, I believe it calls into question the sustainability of mathematics content in Wikipedia.  Deltahedron (talk) 06:16, 20 August 2014 (UTC)


 * I'd read that summary. (Thanks anyway.) What I really needed to read were the linked Visual Editor and Flow articles. The former says that classic wikitext/LaTeX editing will be available. So things won't get worse than status quo? The latter is a new discussion system. Does it use Visual Editor? With or without the classic mode?
 * I still don't understand the basis for comments such as "it does not appear that mathematics has a future here" or "Is it time for mathematicians to leave Wikipedia?" (made by the same editor here). Mgnbar (talk) 15:20, 20 August 2014 (UTC)
 * Those comments were made at a time when there was definite reason to believe that mathematics markup would not be compatible with VE and Flow, or at least, that they might have to be edited in a different and less convenient way than ordinary text. (It is my private belief that the incompatibility had not been consciously recognised by VE/Flow developers until it was raised by WPM members, but that may be incorrect.)  After some discussion, it became clear that that mathematics markup editing would be available in VE, because a mathematics component for VEwas written, as a student summer project (ie, not by WMF staff).  Deltahedron (talk) 16:48, 20 August 2014 (UTC)
 * Although the current system of displaying equations in Wikipedia is marginally better than the output of an old-school latex2html script popular in the 1990s, it has not been substantially improved essentially from the very early days. This is in marked contrast to the rest of the internet, for which there are open source solutions that are a vast improvement over what we have here.  Largely due to volunteer efforts of some developers, we now have a MathJax system that kind of works.  There remain significant performance issues that need to be fixed.   Sławomir Biały  (talk) 15:49, 20 August 2014 (UTC)
 * (ec) Mgnbar, technically it sounds as though previous functionality will not be deliberately broken. However, what we have may be described as limping along: I have stopped using MathJax because it is horribly slow and frequently fails to render. I have tolerated the uncomfortable truce between PNG and HTML in articles in the expectation that this would be resolved in the not-too-distant future. Already, the LaTeX supported here requires a number of inventive workarounds using overlaying of symbols to create new symbols, or creating a mosaic of PNGs. Watching the WMF's responses, it is evident that I should now not expect any fixes, only a slow degradation as the remainder of WP becomes progressively more incompatible with the existing LaTeX support. The WMF's attitude comes across as "it's not my problem". It is already clear that several of the top mathematical WP contributors are taking umbrage to this attitude.
 * I'm with Deltahedron on this. If we can find a place where a group of people with mathematical and scientific interests can collaborate while actually feeling welcome, I'll be there like a shot. And if this perception is widespread enough, the comments you refer to might have relevance. —Quondum 16:10, 20 August 2014 (UTC)
 * There is the Right to fork. The database for en.wikipedia articles currently stands at around 10Gb compressed, 44Gb uncompressed for the articles, according to Database_download.   Articles in mathematics categories are currently about 1% of the total number of 4,585,263 articles, as far as I know (I don't have a reference to hand), so presumably about 1Gb compressed, 5 Gb uncompressed (in pther words, one memory stick).  One possibility would be to ask Encyclopedia of Mathematics whether they could host the articles in some holding area and then transfer them into the EoM mainspace.  Another would be to find someone with a server to spare...  Deltahedron (talk) 17:13, 20 August 2014 (UTC)
 * Portal:Mathematics states 30,412 mathematics articles currently. So just under 1%.  Deltahedron (talk) 20:50, 20 August 2014 (UTC)
 * I've just done a scan of and old database dump from March and then there were 34,809 articles with some latex mathematical formula in them. 27% had just one equation, 50% had 5 or fewer and 25% has 17 or more. The article with the most equations was Propositional calculus with 758 formula. The total number of equation is 545,870 in all mainspace articles.--Salix alba (talk): 19:02, 21 August 2014 (UTC)
 * The is not only the EoM as an already existing alternative but also PlanetMath and possible Citizendium, though personally I prefer Wikipedia to them so far for various reasons.--Kmhkmh (talk) 00:19, 22 August 2014 (UTC)
 * I just want to point out a small fact that EoM uses MediaWiki with mathjax. In other words, "in theory", it cannot have a better math rendering support than Wikipedia. To me, Deltahedron seems too pessimistic. For instance, not every math article needs to have formulae and thus latex, and at least things are not broken (yet). I do find the wanting on the part of the foundation, though. Why can't they just deliver what editors want instead of delivering what we didn't ask, the latest example being image viewer. What I want see is not the continuing display of drama (korean soap is enough for me) but just simple very reasonable technological upgrade that puts Wikipedia on par with other sites. -- Taku (talk) 18:16, 20 August 2014 (UTC)
 * I cannot say why WMF have decided not to "just deliver what editors want", but that they have chosen this path is undeniable. There's a very interesting personal posting by WMF Board chair Jan-Bart de Vreede here.  He makes it clear that WMF projects are at a crossroads and that WMF Board has selected Lila to take the projects in a particular direction.  Bullet points describing that new direction include:
 * I hear a lot from the few and the angry. There is an argument I hear a lot: “We are the community, without us the projects would be nothing. We are the ones who got us here.” That is true, to a degree. But at the same time… we don’t want to be here…. We want to be much further along the road.
 * We want to attract new editors. They don’t have to become heavy editors, they could even contribute once in a while, as long as we get lots of them. We have to make it easy enough for anyone to contribute so that people once again feel that “anyone can edit.”
 * We need to move faster than ever before. This means we need to be tolerant of things we may not like and let experimentation happen. We also need to remove things we are attached to that don’t have wide adoption.
 * We need to act as one community, not 1,000. This means we cannot enact the wishes of a few hundred, but have to build processes that support the successes of millions.
 * All of this is going to require change, change that might not be acceptable to some of you. I hope that all of you will be a part of this next step in our evolution. But I understand that if you decide to take a wiki-break, that might be the way things have to be. Even so, you have to let the Foundation do its work and allow us all to take that next step when needed. I can only hope that your break is temporary, and that you will return when the time is right.
 * All of this sounds to me like a fair warning to those of us who have helped to build the encyclopaedia that unless we move in the WMF's chosen direction, which involves tolerating whatever the WMF decide we should have, we can and should leave; attracting new editors means shiny new editor experiences which may or may not support mathematics; somehow the "we cannot enact the wishes of a few hundred" (while I presume that it primarily refers to the various community votes against VE and MV) suggests that the (few and angry?) mathematics editors are going to carry even less weight, if possible, than before. So in view of this strategic overview, coupled with the WMF's rejection of the proposals submitted via Jimmy Wales, and the absence of any engagement between WMF staff and volunteer developers in mathematics, it seems not unreasonable to prepare for a graceful and productive exit as an option that may be inevitable in the not-too-distant future.  Deltahedron (talk) 18:38, 20 August 2014 (UTC)
 * This story on the WMF's disinterest in maintaining a different important feature (the ability to find the source page of an image) is perhaps instructive about their attitude towards the preferences of existing editors and their ability to continue editing Wikipedia. —David Eppstein (talk) 19:14, 20 August 2014 (UTC)


 * Thanks for the link. (An analogous discussion on English Wikipedia is interesting.) When you say "find the source page", you mean the "File:..." page? I can get to it from the Media Viewer with one click. So that's two clicks total. It used to be one click. So getting to the "File:" page is less convenient than it used to be. On the other hand, it doesn't really affect my "ability to continue editing Wikipedia".
 * Similarly, the stated goals of Visual Editor and Flow seem noble and prudent for Wikipedia's future. (Whether the software fulfills those goals is another issue.) I've edited Wikipedia with PNGed LaTeX for 9 years. I can go another 5 years, unless it's clear that VE is going to break the PNGed LaTeX before then.
 * By the way, it's great that editors such as Deltahedron are engaging with the hierarchy, to make sure it stays aware of math typesetting issues. Mgnbar (talk) 03:46, 21 August 2014 (UTC)
 * Yes, you can get to it with one click, from a tiny little icon on the viewer display that their user studies demonstrated that users were unable to discover easily. But the link also mentions a proposed change to eliminate frames on images, and the "zoom" icon within the frame, which would make it impossible to get to the File: page for image maps (because clicking on the image itself would do something else). —David Eppstein (talk) 03:57, 21 August 2014 (UTC)
 * I've been following this on and off and have seen a lot of heated words but I can't see at all what over. I've seen no reports of problems with formulae and formula editing, or at least none that haven't been promptly and appropriately addressed. Nor are there changes on the horizon that will break formulae editing. Flow has been mentioned but that's months if not years from wide deployment. It cannot be judged now based on an early test version.


 * Formula support has been significantly enhanced in recent years. The addition off MathJax to MW is a welcome addition, and addressed the main complaint about the previous system, it's poor graphics quality. The Visual Editor's formula editor with live preview improves greatly on source editing. But though I don't have stats I suspect neither has replaced the previous system: editing source to generate PNGs still works, perfectly well for many if not most editors. It's not as if there was or is some pressing problem that needs addressing. There are things it would be nice to have, such as an entirely visual formula editor, but this cannot be a high priority as it would be a lot of work, nothing's broken now, and maths is only a small part of the encyclopaedia.-- JohnBlackburne wordsdeeds 19:23, 21 August 2014 (UTC)
 * If I'm not mistaken is PNG rendering still the default setting and still looks ugly and out of place when used within text. Although Mathjax might work great for some, it is still experimental and afaik not without issues. Insufficient math/latex support of VE and Flow is imho a big issue as well, at least when they become implemented as default options. A system for talk pages of math/sciences/engineering articles that has no proper latex support is imho a no-go, then I rather use the current talk page as it is.


 * More generally speaking I'm getting more and more the impression that the WMF develops a lot of software that large parts of the community don't want or don't care for while important requests from the community seem to be put on the back burner. In addition board member statements like this one(m:User_talk:LilaTretikov) make you wonder, whether they have their priorities and the project goals straight. They seem to primarily measure Wikipedia's success in terms of traffic rankings, in comparison to commercial internet companies and fancy guis. Such a mindset loses sight of the fact that Wikipedia is primarily about free encyclopedic high quality content and the success should be measured by the degree to which this is achieved. Traffic rankings are merely a proxy to assess that (high quality presumably leads to high traffic) and a particular stylish or fancy interface is at best an afterthought here. The notion that a state of the art gui will bring us much needed new editors willing to provide man hours and quality work seems a bit of a pipe dream and given that WMF is increasingly pissing off its current editors providing the core content is actually almost a bad joke.--Kmhkmh (talk) 22:12, 21 August 2014 (UTC)
 * The VE maths editor has improved since I last looked. They have addressed the problem with the edit box obscuring the equation, now it always appears just below the displayed equation. The box for the equation is a bigger, it does still require some scrolling for the more complicated expressions. Whilst it does not respect my preference for MathJax while editing, always using PNG mode, it does now gracefully go back to displaying MathJax after saving. An annoyance is that if while editing an equation there is a temporary syntax error (say if your start a \begin{align} but have not yet added the closing \end{align}) the displayed equation will show a red "failed to parse" message. This can cause the page to jump about a lot flipping between a good display and the red error message.--Salix alba (talk): 22:29, 21 August 2014 (UTC)


 * For what it's worth, one of the team in their spare time proposed a change in May to show MathJax rather than PNGs for formulæ in VisualEditor, and editing to MathJax rather than server-provided PNGs, but it needs some work. Jdforrester (WMF) (talk) 23:14, 21 August 2014 (UTC)
 * Just posting to note my agreement with JohnBlackburne. --JBL (talk) 01:43, 22 August 2014 (UTC)


 * Despite my earlier statements in this thread, I feel compelled to defend WMF a little here. You and I may be comfortable with wikitext and with how discussion on WP is done, but many people—most especially potential new editors—aren't.  I've lost count of the number of times I've seen IP editors start a badly formatted thread in the middle of a talk page.  They don't do it on purpose; but they don't know how to edit, and they don't know how talk pages work, and they make good faith mistakes.  I am absolutely certain that there have been many others who made it to the talk page but gave up without editing because they were overwhelmed and intimidated.  Talk pages are an ad hoc collection of conventions structured around tools meant for editing articles.  They are—completely objectively—terrible.
 * That doesn't mean that I think that VE and Flow are good ideas or will be a success, and I'm not convinced that they're a good use of development resources. But I'm also willing to accept that the Foundation might be right to try.  They've invested a lot more time thinking about this than I have (I prefer to think about math), and maybe I would come to the same conclusions as them if I had spent that much time thinking about these problems.
 * What I worry about is what WMF will do about support for less mainstream features. That includes not only mathematics (my biggest personal concern) but also features like those that Jdforrester mentioned above (e.g., music and hieroglyphics).  I am not convinced that WMF is in a position to accurately evaluate the needs of the editors who work on specialist topics, and I'm still convinced that WMF doesn't care about us.  Ozob (talk) 01:45, 22 August 2014 (UTC)

I should note that MathJAX in Wikipedia (besides being opt-in) uses client-side rendering which is pretty slow in practice, something that is well known: https://www.mediawiki.org/wiki/Extension:Math#Discussion_of_pros_and_cons JMP EAX (talk) 00:19, 22 August 2014 (UTC)


 * Why are we not doing server-side rendering? -- Taku (talk) 22:02, 22 August 2014 (UTC)


 * Server-side rendering is, as I understand it, the plan for the future that Physikerwelt is working on. From m:Extension:Math: "Beginning from Math 1.2 // MW 1.23+ you can use a Mathoid server that uses MathJax to convert texvc input on the server side to MathML+SVG rendering. Mathoid is the rendering mode that is going to be used on future Wikipedia." -- Jitse Niesen (talk) 15:00, 23 August 2014 (UTC)
 * Looking at MathJax Output formats the SVG mode is only supported by IE >=9 not all andriod devices support it. Its also limited in the character is can display so some obscure character might not work. MathML mode only works as standard on firefox and with a pluging on IE. So both server-side options will have some holes in support. The other problem with server-side is that it can't always respond well to client side changes, such as changing the page width, zooming or using different locl font size. Server-side rendering will work for a lot of people but it not the perfect solution, for that we will have to wait until the other browsers all adopt MathML (which is taking a little longer that waiting for the WMF).--Salix alba (talk): 20:32, 24 August 2014 (UTC)


 * For anyone interested in community participation in software development across all wiki projects, a discussion has been started at Community Engagement (Product)/Process ideas. Deltahedron (talk) 19:28, 22 August 2014 (UTC)


 * Math 2.0 is now live in betlabs. Thanks to User:Jdforrester (WMF) for the support. --Physikerwelt (talk) 21:29, 22 August 2014 (UTC)
 * Page takes forever to load in Google Chrome, i.e. nothing gets really displayed. JMP EAX (talk) 16:54, 23 August 2014 (UTC)
 * Same for me in FF 31.0. Deltahedron (talk) 17:12, 23 August 2014 (UTC)


 * Page is loading fine for me (FF 31/win8.1). However, first test page with MathML enabled produced "(Failed to parse (MathML with SVG fallback (experimental): Invalid response ("&lt;p&gt;Error fetching URL: couldn't connect to host...&lt;/p&gt;") from server "http://localhost:10042":)" error messages instead of equations.TR 11:49, 25 August 2014 (UTC)
 * The beta labs server seems to be working fine now and does give me MathML. -- Jitse Niesen (talk) 13:55, 30 August 2014 (UTC)

dvisvgm
I have used this to covert some tikz stuff I made to SVG (though noting for Wikipedia yet.) It seems it can also convert equations to SVG via htlatex at least, but that requires some htlatex hackery/config:. Beware that unlike dvipng, dvisvgm needs ghostscript to actually work (at least with tikz; that's because it relies on gs as interpreter for PS/PDF DVI specials. There's a direct SVG driver in tikz that can—in theory—produce SVG directly, but it has some major bugs. Also, that driver probably won't help with equations, which tikz itself doesn't handle.) JMP EAX (talk) 22:06, 24 August 2014 (UTC)

There are some production-quality SVG TeX backends like BaKoMa (which can even produce a bit of SVG+SMIL, i.e. declaratively animated SVGs), but they are all commercial. The htlatex/dvisvgm hackery seems the only free one, besides MathJax, which is honestly even more silly on the server side (via Mathoid) as it basically runs a fully fledged web browser process to render stuff and MathJax re-implements TeX in Javascript (and not terribly well). JMP EAX (talk) 22:10, 24 August 2014 (UTC)

mini-rant about MathJax
For some backgrond info: MathJax was basically made by Design Science (company), the makers of MathType. Before you think that's awesome, MathType was the tech behind the equation editor before Office 2007. M$ ripped out that (or rather made it legacy) after they implemented their own editor which uses the new OpenType Math fonts (and standard) that M$ designed inhouse, which is way, way better, coming close to TeX quality. Now MathType (Design Science) was probably going to go the Y&Y way [of the dodo] after that, but they managed to reinvent themselves as a web services company. Their actual tech is quite meeeeh, but it's the only free web thingy for math basically, given that browsers have started to drop [direct] support for MathML last year. MathJax is basically a ghetto monopoly now. JMP EAX (talk) 22:34, 24 August 2014 (UTC)

MathJax does not use actual OpenType Math font features (i.e. the superset of TeX math font metrics that M$ put into OpenType Math and recently into the ISO standard thereof). MathJax uses the STIX fonts, which themselves don't yet have that. Most glyph positioning is done by hackery in MathJax. The XITS fork of STIX (and Asana Math) have actual OpenType Math support in the free fonts world. JMP EAX (talk) 22:48, 24 August 2014 (UTC)

FYI, here's the WMF-ish plan:. JMP EAX (talk) 05:06, 25 August 2014 (UTC)

And just so you don't think I'm just gratuitously ranting about this, I've attached a sample problem with MathJax rendering in Google Chrome, where MathJax actually has to do some rendering/positioning (as opposed to just translating a subset of LaTeX to MathML, which is what it does in Firefox.) It's not an optical illusion, the enlarged version also has baseline alignment problems between sub-parts of the aa-1 formula. JMP EAX (talk) 07:11, 28 August 2014 (UTC)


 * So, what do you suggest for us? It's all very well to rant about MathJax, but without alternative suggestions that won't improve matters. From your description, it looks like it will be a lot of effort to get the dvisvgm route to work and it is not clear that it will be better. -- Jitse Niesen (talk) 08:26, 28 August 2014 (UTC)


 * No one has an illusion that there is a perfect solution. Given that the web doesn't support math natively (mathml was once promising but isn't getting anywhere), there will remain some degree of hackery in a foreseeable future. What we're is asking some "modest" performance and appearance improvement that puts Wikipedia "on par" with other sites. This is a feasible cheap goal. By the way, although Deltahedron likes to frame this in a math-editing issue, this is actually a larger issue: Wikipedia as a website looks old. For example, the main page is another part of Wikipedia that has remained unchanged for "ages". WMF clearly understands this and that's why they are trying to put stuff like VE and media viewer to modernize the website Wikipedia. (Ah, this reads like a rant.) -- Taku (talk) 12:27, 28 August 2014 (UTC)


 * But Deltahedron did hit a nail in the sense that the fundamental problem is no one, as it seems, is accountable for math rendering; that is, it would be nice if there is some math-tech guy who fixes the math implementation here whenever it breaks and also plans to a further development (not someone "oh, I don't know it's not part of my job description.) -- Taku (talk) 12:39, 28 August 2014 (UTC)
 * Since you guys are probably (directly or indirectly) paying AMS dues, it might make some sense to try talk to whoever is paying for MathJax at AMS to convince them that money is better spent on native MathML support in browsers. None of the MathML implementers (in any browser [engine]) got paid anything for their MathML implementation work. Instead AMS decide to sink money a polyfill (MathJax) that will always suck in the long run. But given the disaster that the STIX fonts were, I'm not terribly surprised AMS made another bad engineering decision... JMP EAX (talk) 13:05, 28 August 2014 (UTC)
 * You think Microsoft will implement MathML support for some cash? And then a decade later we can finally enjoy MathML in the majority of common browser versions? ;) --Unverbluemt (talk) 12:02, 31 August 2014 (UTC)
 * FWIW, I've asked someone I know (former prof of mine) who is family closely associated with the inner circle of ACM SIGACT and they don't have funds for something like this, nor do they see it as much within their remit. JMP EAX (talk) 18:26, 28 August 2014 (UTC)
 * I am not a member of the AMS, but will explore this within the UK community. Email me if interested, as this is clearly no longer a WP or even WMF issue.  Deltahedron (talk) 18:33, 29 August 2014 (UTC)

Alignment of PNGs with text
The following might be of interest http://tex.stackexchange.com/questions/44486/pixel-perfect-vertical-alignment-of-image-rendered-tex-snippets JMP EAX (talk) 02:28, 25 August 2014 (UTC)

MathJax and texvc don't support the same subset of LaTeX
I just realized this when I kept seeing the flashing "failed to parse [blah, blah]". Turning MathJax off makes that error permanent. Basically MathJax supports  (a fundamental TeX command) but texvc chokes on it. JMP EAX (talk) 17:00, 28 August 2014 (UTC)
 * I believe the current plans are to run all equations through texvc first. Basically this means the subset of LaTeX wikipedia maths supports is defined as those parts supported by texvc. There are a few other incompatibilities documented at Help:Formula.--Salix alba (talk): 17:17, 28 August 2014 (UTC)

Maltsev or Mal'cev or Malcev varieties or conditions
Do we really not have page for these? I tried various permutations, but I still can't find them on wiki. Nor are they mentioned in Anatoly Maltsev. JMP EAX (talk) 15:58, 31 August 2014 (UTC)


 * There is a brief mention in Group extension and we have Malcev algebra and Malcev-admissible algebra, but I could find nothing else. If you would like to create an article for Malcev variety, go for it! --Mark viking (talk) 16:34, 31 August 2014 (UTC)