Talk:Main Page/Archive 12

From ProofWiki
Jump to navigation Jump to search
Filing cabinet.png
This is an article of past discussions, from 22-Dec-2015 to 1-Aug-2017.
Do not edit the contents of this page.
If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Extending ProofWiki to "derivations" and visual explanations?

I'm proposing the extension of ProofWiki to "derivations" and visual explanations of mathematical concepts and expressions. Examples would include the convolution integral or e.g. the Hilbert transform.

An example explanation is how the convolution integral is presented here:

What do you think?

No reason why not. We already have plenty of visual enhancements of proofs. The only thing limiting us has been the ingenuity of the people putting such visuals in place.
If you have particular skills and interests in this area, then please feel free to contribute. Your work will be very welcome. --prime mover (talk) 07:47, 24 December 2015 (UTC)
I'm willing to contribute, but I don't understand how to edit the Wiki (if I even have the rights to set up those pages). --Mviljamaa (talk) 09:18, 24 December 2015 (UTC)
You should have sufficient rights to edit pages. You also have all sorts of Help and FAQ pages to read. Feel free to study these. You have links on your home Talk page, and you also have various links down the left hand side to explore. --prime mover (talk) 13:56, 24 December 2015 (UTC)

Joke namespace

Since it is of course of the utmost and pressing importance that we ensure that we have documented all the mathematical jokes in the world, we are going to need a namespace for them.

Can we therefore have a "Joke" namespace, the same way we have a "Mathematician" and "Book" namespace?

What does anyone else think? --prime mover (talk) 08:29, 3 January 2016 (UTC)

Meh. Use subpages until it becomes too unwieldy. It's better that they're all in one place, to be read in a single run. — Lord_Farin (talk) 22:02, 3 January 2016 (UTC)

Style problem

Something gone wrong with this page Category:Definitions/Fundamental Dimensions

Anyone able to sort it out? --prime mover (talk) 18:42, 23 January 2016 (UTC)

no worries, flushed my browser cache, all OK --prime mover (talk) 19:02, 23 January 2016 (UTC)

Problem with MathJax?

Is anyone else seeing this problem?


All the $\LaTeX$ strings are appearing with a following vertical line.

I first noticed this yesterday on a different machine from my usual one (Windows 7 vintage), but now I notice it on this (my trusty old Vista). On both machines I am using Google Chrome.

I just tried it on Firefux and it's fine (I don't fancy getting IE out this morning to check that too) -- so it appears it may be a problem with Google Chrome. --prime mover (talk) 07:28, 30 January 2016 (UTC)

I can't reproduce it. — Lord_Farin (talk) 09:14, 30 January 2016 (UTC)
Very odd. As I say, it exhibits this problem on both of the machines I have access to: work and home, and on Google Chrome only. I can't get IE to connect to it (but then I can't get IE to connect to anything, I never use it). --prime mover (talk) 10:12, 30 January 2016 (UTC)
This has been happing to me for the past few months on the Canary version of Chrome. I assumed it was a bug that would be fixed. I guess it has made its way into the stable branch. --Joe (talk) 16:56, 30 January 2016 (UTC)
According to Stack Overflow this has been fixed in MathJax. --Joe (talk) 16:58, 30 January 2016 (UTC)
Updated MathJax, should be fixed now. --Joe (talk) 16:59, 30 January 2016 (UTC)
Yep, after a page refresh, this is now looking okay. Although I may stick with Firefox, the version I currently have (and have been using all day) seems to run more easily than Google Chrome ... the only thing I don't like about it is the tab management paradigm. --prime mover (talk) 20:28, 30 January 2016 (UTC)


Some recent pages brought the need to support more authors in the BookReference template. I've expanded it to 6, and also included a whitespaced sample of the (for display reasons necessarily) cluttered wiki code producing it. This should help in making such templates more maintenance friendly. — Lord_Farin (talk) 14:52, 21 March 2016 (UTC)

I think something might have gone wrong with it -- check out the Book:Books page. Also, Template:Book needs a similar expansion. I'd work on it myself but I lack the patience tonight :-) --prime mover (talk) 20:49, 21 March 2016 (UTC)
Done. (And this time, I did check that it works on existing stuff ;).) — Lord_Farin (talk) 11:45, 25 March 2016 (UTC)
Nice one. Might not have been good in all browsers, though -- I found you have to use   instead of a space at the start of any string after a pipe delimiter that is being used as an "else", otherwise the space is not interpreted. --prime mover (talk) 12:30, 25 March 2016 (UTC)

Invitation for suggestions on improvements to presentation of oscillatory systems

There's a lot of new material that's been added in the Mechanics section, all of which is working towards a treatment of (so far as I've reached) resonance. The overall plan is that the general second order ODEs defining the underlying mathematics be established separately from the applications (whether they be moving carts attached to walls by springs or L-R-C circuits, or whatever) and hence to decouple the mathematics from the physics.

I'd welcome suggestions as to how to assemble the various categories, and also how to structure the pages and even to name the pages if they appear to be inadequate. Also, any definitions which appear loose or inaccurate (and there are plenty) I definitely want to improve.

I may well be changing direction after this, as I'm getting tired. --prime mover (talk) 13:01, 27 March 2016 (UTC)

Good work. Personally I find physical applications of ODEs boring, so I'll refrain from having an opinion on that front. As to the particular ODEs, I would probably go with the following scheme:
  • ODE LHS (basically TeX without backslashes and braces)
    • /Homogeneous
    • /Inhomogeneous
      • /Instance
Furthermore I think there is a need for clear overview pages of the calibre Trigonometric Identities (e.g. Linear Second-Order ODEs). As much as possible, these should have TeX'ed click-throughs. Naming convention close to TeX will help searchability.
If all becomes too much to maintain, we will eventually end up introducing the Example namespace we've kept out thus far. For now, however, subpages will be fine. (We might even consider sub-wikis or other cross-wiki collaborations, but this is at present unrealistic.) — Lord_Farin (talk) 18:00, 28 March 2016 (UTC)
Overview pages was something I started, but they got a little unwieldy and I reckoned it would be easier to go back to them, but I never did.
Another thought I had was to subcategorise the 2nd order ones further into the three categories: over-, critically and underdamped, and then resonant, but at the "pure" level rather than at the application level, but that came to me late as I got into the thread with the oscillating cart. And there is a lot in there which need to be pulled back into the abstract, but that also needs work which I'm not prepared to do at the moment, I'm going to need to let it rest a little before I can see it from a fresh angle.
All contributions welcome here. I've not seen an example before of such a structured categorised approach to arranging DE's into categories before, so as to exactly what it is we're trying to achieve, I confess I'm not sure where it's going. --prime mover (talk) 19:28, 28 March 2016 (UTC)
I think one reason why there is no good categorisation for ODEs is because unlike e.g. finite simple groups there is provable failure in classification: we cannot solve them all in "elementary" form. This also makes me less eager to attempt such a thing.
But then, admittedly I have been out of the theory of DEs and everything that goes with them for a long time, so maybe I'm not in a position to comment on this. — Lord_Farin (talk) 19:13, 30 March 2016 (UTC)

Template:explain spans too many lines. It makes bad reading experience.

Example to illustrate that template:explain is bad.

Two solutions:

1. Use template:clarify.Template:Clarify

2. Remove redundant sentences in template:explain.

Golopot (talk) 03:28, 7 April 2016 (UTC)

I disagree. --prime mover (talk) 05:43, 7 April 2016 (UTC)
I like the template as it is, but that page is a an eye sore. What I'd prefer to see is the explain template used once there and either it lists all the problems in one section like a summary, or to have the explain not mention any specifics and have a smaller cleaner template like clarify to each specific. Whatever the case, those template shouldn't exist on the page, so this really is a nit pick. --Ybab321 (talk) 11:27, 7 April 2016 (UTC)
My take on this is that "explain" templates should not be there in the first place. They are supposed to make the page ugly, so as to inspire people who know what they are talking about to put in whatever fixes are necessary so as to allow that invocation of that template to be removed. --prime mover (talk) 12:19, 7 April 2016 (UTC)

First of all, thanks for the feedback. This helps to improve the site. (NB. Naturally you are encouraged to make changes yourself -- it's a wiki after all.) However I feel a third option didn't receive sufficient attention:

3. Resolve the calls to explain by improving the article.

That's what really ought to be done. I agree that the page is ugly and tough to read now, but the solution lies not with prettifying the maintenance template(s). That being said, I think Template:Explain could say everything in need of being said on three lines. I'll see what I can do about both. — Lord_Farin (talk) 14:33, 7 April 2016 (UTC)

Here goes: Fundamental Theorem of Algebra. — Lord_Farin (talk) 15:09, 7 April 2016 (UTC)

Problem loading WebFonts

After I have freshly booted up my (elderly Vista) machine (running Google Chrome), when I open certain pages I get a message in the bottom left corner of the screen saying: "Loading web-font TeX/Math/Italic" and the page just sits there, taking some time to finish. Then it puts the message up (same place): "Web-Fonts not available - using image fonts instead".

Take for example Area under Arc of Cycloid.

Examining the generated html, it looks, for example, something like this:

<span class="mi" id="MathJax-Span-51" style="display: inline-block;">
    <img src="" 
        style="width: 14px; 
        height: 14px; 
        vertical-align: 0px; 
        margin-right: 0.003em;">

Take as a contrast, for example, an element of a page Brachistochrone is Cycloid which does not show this error message:

<span class="mrow" id="MathJax-Span-38">
    <span class="mi" id="MathJax-Span-39" style="font-family: MathJax_Main;">sin</span>
    <span class="mo" id="MathJax-Span-40"></span>
    <span class="mi" id="MathJax-Span-41" style="font-family: MathJax_Math-italic; padding-left: 0.165em;">α</span>

... which shows the Math-italic font being used as per normal.

The appearance of the mathematical text is different as well -- the one using images is heavier and slightly fuzzy.

Has anyone else noticed this, or does it look as though it is a problem with my machine? --prime mover (talk) 06:09, 13 April 2016 (UTC)

... but now get this.
I have just made a no-change edit ("Edit" then "Save page") on the troublesome page Area under Arc of Cycloid in question, and now it is behaving correctly.
But now it has suddenly started doing it on Brachistochrone is Cycloid.
Ugh. And now Brachistochrone is Cycloid is working fine again. WTF? --prime mover (talk) 06:15, 13 April 2016 (UTC)

Prioritising stub and refactor-requiring pages

There are 993 pages that are marked as stubs, there are 853 pages that are marked as currently being refactored (refactor required or does this mean that someone is on the job?). This number is huge. Also Definition:Thermodynamics (for example) isn't marked as a stub, but looks like a stub. Surely there's more that can be said about thermodynamics!

Just a quick point: $\mathsf{Pr} \infty \mathsf{fWiki}$ is intended to be more of a dictionary than an encyclopedia, hence its intention is to contain definitions rather than expositions (with the possible exception of the "Mathematician" namespace, where I confess to getting carried away sometimes). There is ample room for discursion on e.g. Wikipedia, with whom we are not and never will be in competition. Our intention is to state the facts of a definition and that is all.
Hence although it "looks like a stub", to the extent that it fulfils its purpose: it defines what thermodynamics is. Such pages (those that define a branch of mathematics or other science) are not intended to be large. --prime mover (talk) 06:31, 22 April 2016 (UTC)

Anyway! I (the newest of the current newbies) would like to propose a grading system for stubs and a change to the refactoring system, and have a grade for each which denotes "low hanging fruit" to make it easier for newcomers to get started and learn. I'd rather (for example) have a proof for "equivalence relations partition a set" rejected on style grounds than pour time and effort into a longer, much more involved proof and have that declined.

Pages don't get "approved" or "declined" -- if their style does not come up to scratch they get a "tidy" template appended to them and then get brought up to style by one of the admins (er, yeah, that's me) who (irony alert) delights in such work.
So it should be no big deal. Paste up what you see fit; once it's in (if it's not a repeat of something already there, or is wildly incorrect) it stays -- and the only changes are those of, for example, stylistic improvement, correction of grammar/spelling or expansion of the argument, or rewording for clarity and/or linguistic consistency. --prime mover (talk) 06:42, 22 April 2016 (UTC)

Lastly (to keep this short) grades can convey urgency, for example Carathéodory's Theorem (Measure Theory) is extremely important in Measure Theory where as Book:Christopher Clapham/The Concise Oxford Dictionary of Mathematics, lets be honest, it isn't that urgent. Alec (talk) 19:51, 21 April 2016 (UTC)

Actually, I quite like that idea (particularly for stubs, anyway). I can see room for the Stub template to be expanded to take a new optional parameter "importance", or however the parameter may be named (let's be jocular and call it "defcon") to take a value of e.g. 1 to 5, where 1 is "goodness, this needs to be fixed urgently, it's embarrassing" and where 5 is "meh", and thence to automatically assign such pages into an appropriately-configured subcategory of the unwieldy and overflowing "stubs" category.
This is an area of development where "he who smelt it dealt it", so I'd be happy enough for you to take that on, if you so desire.
I am not so concerned about the "refactor" category, and (as I say to all new contributors) I would prefer to have this category left well alone except to old $\mathsf{Pr} \infty \mathsf{fWiki}$ hands. --prime mover (talk) 06:38, 22 April 2016 (UTC)
These lakes of stubs/refactor pages/etc. are one of the biggest pain points this site has to deal with. Any ideas to break these categories into manageable chunks are most welcome.
One of the templates that might be (and at times, has been) used for this purpose is Template:Finish which signifies that while the mathematical content isn't too exciting, it's still some work to complete the page.
As to grading I'm not convinced yet it would work because of the inherent subjectivity involved. But there might be a case for something in between Template:Stub and ProofWiki:Wanted proofs list.
Do you have concrete suggestions on how to introduce such sifting? — Lord_Farin (talk) 22:31, 21 April 2016 (UTC)
Just so as to get a handle on the quantity of work here, I have cleared 6 refactor flags over the last couple of days. It took all day to restructure the Definition:Polygonal Number refactor flag. The work is ongoing, and it gets picked up as and when the thread of what I am studying passes through one such.
As for the stubs, they are always available for someone coming into $\mathsf{Pr} \infty \mathsf{fWiki}$ to see whether they want to try their hand. --prime mover (talk) 20:33, 27 April 2016 (UTC)

This old rubbish again

It has been once more pointed out that it is extremely bad form to name one's results after oneself. I understand that it displays an arrogance and egotism that have no place in the field of mathematics.

Once more there have been calls to take this result down again, on the grounds that it makes $\mathsf{Pr} \infty \mathsf{fWiki}$ look bad.

While I completely take on board the philosophy behind this (there should be zero tolerance for self-aggrandisement in mathematics), for various reasons I am reluctant. So I will hold off purging it from the database if I get enough (? five or ten) well-argued posts to the contrary.

Just leave it there. It's not hurting anyone. — Lord_Farin (talk) 16:06, 26 April 2016 (UTC)

Original research

Can I add topics from my original research book? (note, that it is currently published only on the Web, not peer reviewed) --VictorPorton (talk) 07:25, 3 July 2016 (UTC)

Yeah sure. Bear in mind they will get editied a lot to fit them into house style. Be welcome. --prime mover (talk) 07:37, 3 July 2016 (UTC)
But what if my proof refers to a theorem present only in my book? --VictorPorton (talk) 08:02, 3 July 2016 (UTC)
Then you will need to add to $\mathsf{Pr} \infty \mathsf{fWiki}$ that theorem which is present only in your book, at which point it becomes public domain under the CC-by-SA 3.0 licence and the GNU Free Documentation Licence. See ProofWiki:Copyrights. --prime mover (talk) 09:09, 3 July 2016 (UTC)

can't do it any more

I'm staring at it and staring at it but I can't motivate myself to get any work done. It's all I can do to just post up links. Can someone take over because I'm tired. --prime mover (talk) 15:56, 6 July 2016 (UTC)

Hm, sad to hear that, although I've been in similar situations in the past. Just take some time off and give yourself a rest. You've earned it. — Lord_Farin (talk) 21:48, 7 July 2016 (UTC)

sorry i completely lost the plot

I went and changed all those results on $a^x$ from discussing them as exponentials to referring to them as powers. Now i realise this was all wrong. Sorry, i've broken everything. Someone is going to have to mend it. Not going to be me because I don't want to, it's too much like hard work. --prime mover (talk) 12:32, 7 July 2016 (UTC)

I felt a little too timid to mention. I'll fix up, once I've brought up my existing pages in the category to house style :) --Keith.U (talk) 14:25, 13 July 2016 (UTC)

New wave of spammers

In order to penetrate our spam blocker, prospective spammers need to send us a CV.

Here are a couple of recent CVs which have proven to be bogus:

Hey , I am Neal Wrangler , working with Synopsis LLC as a Senior Project Consultant


Neal is Director, Software Architect, and Meme Wrangler at ThoughtWorks, a global IT consultancy with an exclusive focus on end-to-end software development and delivery. Before joining ThoughtWorks, Neal was the Chief Technology Officer at The DSW Group, Ltd., a nationally recognized training and development firm.
Neal has a degree in Computer Science from Georgia State University specializing in languages and compilers and a minor in mathematics specializing in statistical analysis. He is also the designer and developer of applications, instructional materials, magazine articles, and video presentations. He is also the author of 6 books, including the most recent Presentation Patterns and Functional Thinking. Given his degree, Neal is a bit of a language geek, with affections including but not limited to Ruby, Clojure, Java, Groovy, JavaScript, Scala and C#/.NET. His primary consulting focus is the design and construction of large-scale enterprise applications. Neal is an internationally acclaimed speaker, having spoken at over 300 developer conferences worldwide, delivering more than 2000 presentations. If you have an insatiable curiosity about Neal, visit his web site at He welcomes feedback and can be reached at [email protected].

I'm also pretty suspect of this guy:

My name is Zachary Kniebel and I am a full-stack Web Application Developer and Software Developer, currently living in Philadelphia, PA. I have a Bachelor of Science in Computer Science from Northeastern University, and my primary focus and inspiration for my studies is Web Development. In my free time, I study human computer interface and the psychology of human computer interaction. I am both driven and self-motivated, and I am constantly experimenting with new technologies and techniques. I am very passionate about Web Development, and strive to better myself as a developer, and the development community as a whole.

... although as the latter hasn't contributed yet it is impossible to say with any certainty.

The thing in common with all these CVs is that they say nothing about the user's commitment to mathematics as such.

I did specifically reject one user's application to join $\mathsf{Pr} \infty \mathsf{fWiki}$ recently (I was in a bad mood) as he presented a similar CV. We will never know whether I was correct in my assessment. --prime mover (talk) 07:30, 12 July 2016 (UTC)

Some people take any text box on the internet as a potential job application form, apparently. I say, no mercy. — Lord_Farin (talk) 15:44, 12 July 2016 (UTC)
Can someone take a second opinion on a bunch of 3 job applications currently sitting in our intray? Dismiss them out of hand, accept them with a view that they may be potential spammers, or reject with a polite email saying "Get lost, spammer!"? I tend towards the first. --prime mover (talk) 09:12, 24 July 2016 (UTC)
I rejected them all. I'm thinking it might be a good idea to tweak the account request form somewhat so as to make it less daunting (without a CV upload button, which I somehow suspect to be the cause of the problems we've been having). I'll ask Joe to update it. — Lord_Farin (talk) 10:08, 24 July 2016 (UTC)
Something like: "Please tell us what your interest is in mathematics, then (no less than 5 words)" should do it. Adding an apologetic note about how this is how we filter out the obvious spambots may also be a good idea. Apparently this aspect of $\mathsf{Pr} \infty \mathsf{fWiki}$ has been discussed not-completely-positively on e.g. Reddit (suggestions of elitism, and so on), although when I explained the spam problem to someone who subsequently joined us, the reaction was completely understanding and supportive. --prime mover (talk) 10:22, 24 July 2016 (UTC)
I've simplified the account request form. Now it's just username, email, and notes. We'll want to update the system messages to tweak some of the wording. --Joe (talk) 13:06, 24 July 2016 (UTC)
I've made some minor text edits, someone else may want improve them. --Joe (talk) 13:12, 24 July 2016 (UTC)
How easy would it be to get it to automatically email to the user: "Welcome to ProofWiki!" rather than making it so you have to type it in every time we press the "Accept" button? Apart from that, it looks fine. --prime mover (talk) 13:22, 24 July 2016 (UTC)

... Further research: in the "Confirm account requests " page, the "Biography" field is empty. Can we replace this with the contents of the "Additional Notes" field which is what the user has to type something into now? This could then also go into the user's home page. Maybe. --prime mover (talk) 17:33, 24 July 2016 (UTC)

I could change the field back to "Biography", however I wonder that will just deter people from signing up? --Joe (talk) 13:20, 25 July 2016 (UTC)
Well yes, that -- and besides, we don't want it to be a "biography", that's the whole point, people just post up their boring CV, and that makes it difficult to screen out the spambots. But an answer to "Why do you like mathematics?" demands something more creative and bot-proof. Till they get wind of it (here cometh the Singularity). --prime mover (talk) 18:42, 25 July 2016 (UTC)

How to cite ProofWiki?

How to cite a ProofWiki article in my work? Should I cite at all? --VictorPorton (talk) 18:10, 21 July 2016 (UTC)

Depends on the citation style you're using. I'd imagine you can cite it as you would any encyclopedia entry. Perhaps we could add a "Cite this" link at the bottom of each page one day? One that autoformats citations for MLA, APA, etc. Chandra (talk) 19:02, 21 July 2016 (UTC)
Never thought about it, never dreamed anybody would want to, we never thought we'd be citeworthy.
If anyone wishes to design and specify a citation style, the world is yours. --prime mover (talk) 20:40, 22 July 2016 (UTC)

Changing the "ProofWiki" first name use

When you share the ProofWiki main page anywhere, the LateX markup shows up for the first instance of "ProofWiki" on the page. Makes sharing it on Facebook, for example, rather ugly. I move to change the first instance of the site name with the lemniscate in it to just "ProofWiki" Chandra (talk) 19:02, 21 July 2016 (UTC)

Perhaps we could replace it with a graphic. --prime mover (talk) 19:05, 21 July 2016 (UTC)
How does that work? --prime mover (talk) 10:30, 24 July 2016 (UTC)
... no, I see it does not. Anyone care to have a better try? I'm limited as to what FakeBook will allow me to do, I've compromised the usability of my account by being rude to Brexit scoundrels. --prime mover (talk) 10:37, 24 July 2016 (UTC)
We already have a graphic right next to the text. We could just put "ProofWiki" in bold Chandra (talk) 12:47, 25 July 2016 (UTC)
Do we really even even to make it look special. I say we just replace the markup with "plaintext". --Joe (talk) 13:24, 25 July 2016 (UTC)
I think the bold would make it look nice when on the main site without making the shared page ugly. And not just for the first paragraph, but also for the header bar above it. Chandra (talk) 17:22, 25 July 2016 (UTC)

Math on my wiki

I have started a new wiki site:

Please explain which MediaWiki extension and which configuration should I use for displaying math on my wiki.

We use MathJax, via this MediaWiki extension. I don't know all the details, but it shouldn't be too hard.
NB. Please sign your posts so we know who said what and when. — Lord_Farin (talk) 19:48, 22 July 2016 (UTC)
I've installed MathJax on my site but rendering of math formulas is horribly slow: Please help to investigate. --VictorPorton (talk) 20:38, 22 July 2016 (UTC)
For a start, you may find you need to use dollar sign delimiters in a MathJax installation on a MediaWiki site, as "math" delimiters do not always render properly. When we moved to MathJax from MediaWiki $\LaTeX$ we had to go through every single page and change them all. Kept us busy for a week or two. --prime mover (talk) 20:43, 22 July 2016 (UTC)
With dollar signs has the same problem (renders too slowly, about 20 secs) --VictorPorton (talk) 20:45, 22 July 2016 (UTC)

It would appear that there is some built-in delay configured for MathJax; Chrome measurements show that after ~20 secs, the render command is triggered and subsequently completes in ~100ms. So please check if there's any configuration parameter which you have inadvertently set. Otherwise I don't know; you might try the MathJax mailing list or some other dedicated channel for further help. — Lord_Farin (talk) 20:55, 22 July 2016 (UTC)

Idea for categories/maintenance templates in user namespace

I just cleared out some user pages from Category:Tidy by putting the template calls in nowiki tags. But on second thought it might make more sense to enhance the various maintenance templates to exclude the User namespace (see this).

This will then exclude those pages from showing up in the categories, so they won't hamper us, but if e.g. a page is moved live from a sandbox (I've done this numerous times) it will automatically be included in the proper maintenance categories.

Thoughts? — Lord_Farin (talk) 11:19, 30 July 2016 (UTC)

Never realised how bad Category:Tidy had got again. Suppose I'd better get working on it. --prime mover (talk) 05:15, 1 August 2016 (UTC)
Seems like a good idea! --Joe (talk) 13:18, 1 August 2016 (UTC)
't Has been done, and it was quite easy in the end. There might be various caching mechanisms in place which persist the User and Talk pages in the respective categories, but over time, or given a null edit, I expect them to disappear. In any case, the inflow has stopped. — Lord_Farin (talk) 19:30, 1 August 2016 (UTC)
I'm trying to reduce the number of pages in the maintenance categories but I still have so much more to do. But I just get so tired nowadays. I'm doing my best but it's just not good enough. --prime mover (talk) 20:10, 1 August 2016 (UTC)

Recent maintenance upgrade -- problems

Ever since the maintenance upgrade of a couple of days ago, I've been having sporadic trouble where the response time is such that it causes the browser to time out (it's a few seconds or so). Two things here:

a) It used to be a lot longer than that
b) The result of this is that the page is lost.

It never used to do this. The timeout situation lasts a few minutes, but when you're doing work it can interrupt the flow. The last time it did this was at about 08:45 British Summer Time 13th Aug 2016, and it came back soon after that. It did it about an hour before that as well, and I had to abandon what I was doing and do something else instead.

Is there something systematic that may be going wrong, or is it just me being impatient? --prime mover (talk) 03:52, 13 August 2016 (EDT)

I've worked out what it is.
Whenever you type anything into the "Search" field at the top right of the page, which does a text match on the stuff you are typing in, it crashes. This is seriously going to compromise my productivity because I use this all the time. Is there any way this can be fixed, or is this just the way it works now? --prime mover (talk) 08:39, 13 August 2016 (EDT)
I'm not seeing this. What exactly crashes, the browser tab, the search box results list? --Joe (talk) 09:59, 15 August 2016 (EDT)
When I type something in the search box, it usually puts up an autoselector which populates itself with all the options that match what you type in. After you're typed in a few characters, and specially if you hesitate for a while and let it sit for a few seconds, and then continue, you notice that it has stopped updating the autoselector with your latest keypresses. And a few seconds after that (you may or may not have actually selected a page by this time), the entire tab goes blank, leaving you with: "This site can’t be reached", " took too long to respond." "ERR_CONNECTION_TIMED_OUT".
Other ProofWiki tabs that you have open then behave the same way (i.e. if select another page, any page, e.g. from a link on the left hand side menu bar, the same thing happens in that tab too). But only ProofWiki tabs. Although one of my tabs put the ProofWiki favicon on it. This page won't save, I'll cut and paste this text into a copy buffer and wait for it to recover. It usually does after a couple of minutes. --prime mover (talk) 13:25, 15 August 2016 (EDT)
What browser and version are you running? --Joe (talk) 15:07, 15 August 2016 (EDT)
I think you were being caught up in an overly aggressive firewall rate limiter. I've made some firewall changes, let know if this still persists. --Joe (talk) 16:43, 15 August 2016 (EDT)

Latency is appalling

Even after you purge your cache and press F5 on your keyboard on all relevant pages, it still takes quite some time before redlinks turn to blue, pages appear in their correct categories and "What links here" reflects an accurate account of what links there.

I noticed this getting worse after the latest software upgrade. Maybe MediaWiki designers need to be made aware of this. --prime mover (talk) 03:10, 4 September 2016 (EDT)

I experienced this a few times as well. Somehow the updates of the database seem to be delayed. Since it's intermittent (at least for me), I'll try to discover a pattern. — Lord_Farin (talk) 04:58, 4 September 2016 (EDT)
It's as if the database updates are batched, and only sent to update the database intermittently. --prime mover (talk) 17:59, 4 September 2016 (EDT)
I suspect the issue maybe be caused by some overly aggressive caching from the web server. I'll temporarily disable it to see if it is indeed the cause. --Joe (talk) 21:24, 4 September 2016 (EDT)
Change made, let me know if you're still seeing this issue. --Joe (talk) 21:27, 4 September 2016 (EDT)
There is still a noticeable delay between committing -- and purging -- and seeing the page appear in its appropriate category. I haven't tested how quickly "what links here" reacts to changes. --prime mover (talk) 01:22, 5 September 2016 (EDT)
The other thing I can think of is the job queue. It runs about once a minute. --Joe (talk) 10:34, 5 September 2016 (EDT)
That's sort of what it feels like. The odd thing is: how come it has never been noticed before? --prime mover (talk) 14:31, 5 September 2016 (EDT)
The MediaWiki job queue has changed in the latest update (we actually jumped two major versions). We're now doing the recommended thing and running the job queue via cron, previously we were running on user requests. Running on user requests is not recommended as it can be quite slow sometimes. Is one minute too unbarable? --Joe (talk) 10:02, 8 September 2016 (EDT)
When doing a refactoring job, and I move a page (with redirect) and then I do a "what links here" on one of the pages, and I get "No links", it's too easy to say "that's all right then" and delete the page before it's had a chance to update itself. And then there are so many links broken.
Just means I'm going to have to change my working style and leave big gaps between each operation I do. --prime mover (talk) 11:05, 8 September 2016 (EDT)

"Recent changes" does not completely work

I added a page "Definition:Conditional Preference" earlier today, but for some reason the "Recent changes" link does not pick up on it. --prime mover (talk) 07:18, 27 September 2016 (EDT)

Did it resolve itself after? --Joe (talk) 11:58, 30 September 2016 (EDT)
When I edited it subsequently, the change was registered -- but the original creation of that file never actually did appear on the Recent Changes log. It doesn't matter in itself, just an indication of things not going quite right when the fabric of space and time was tearing itself apart the other day. --prime mover (talk) 17:15, 30 September 2016 (EDT)

A general shout out to contributors

While it is accepted that not all results are perfect, and some need to be improved, it would be nice if contributors were to spend some time creating new pages in areas where we have so far made little progress (perhaps even to attack our colossal backlog of stubs) rather than to spend their time continually reworking existing material. Just a suggestion. --prime mover (talk) 03:08, 13 November 2016 (EST)


Once upon a time there was discussion about adding examples. For example, how to use Theorem X in solving a certain kind of problem. Did that idea come to fruition? --GFauxPas (talk) 12:41, 14 November 2016 (EST)

Plenty. Search for "Examples" and see what comes up. --prime mover (talk) 14:20, 14 November 2016 (EST)


There is no button to add signatures on sandbox pages, like on User:GFauxPas/Sandbox and User:Lord Farin/Sandbox and ProofWiki:Sandbox. --GFauxPas (talk) 09:28, 20 November 2016 (EST)

You can use 4 tildes (~), which on US keyboard layout can be typed as Shift-backtick (`). But it's strange nonetheless. — Lord_Farin (talk) 10:36, 20 November 2016 (EST)


After I upload new versions of images, the old images sometimes still show on the page invoking [[File:. Unless you click the image and go to the appropriate File: page. .. except when this doesn't work, either. But if you click on the image on the File: space and open the image in its full resolution, then the updated image is there. Is this a known issue of MediaWiki? How does one fix it? --GFauxPas (talk) 17:25, 29 December 2016 (EST)

Purge. --prime mover (talk) 17:43, 29 December 2016 (EST)
I suspect this is a caching issue. Does the issue go away when you purge the page as Prime.move suggested? --Joe (talk) 19:05, 29 December 2016 (EST)
Purging didn't seem to have helped. --GFauxPas (talk) 19:36, 29 December 2016 (EST)
Purge the file page *and* the page it's included on. Works for me. --prime mover (talk) 01:30, 30 December 2016 (EST)
I did that and it still didn't work. Then I found the problem: I needed to clear the cache of my browser (Chrome) as well. All good, thanks. --GFauxPas (talk) 10:25, 30 December 2016 (EST)

Recent visits

Would it be possible to show some statistics of this website? A simple as daily visits would add some motivation, while other numbers like the quantity of proofs would be interesting, which I have not been able to find except for those announced by administrators. Also, other stats like most popular categories, proofs, definitions or other articles in general would help to define new goals, a success thereof would again be reflected on daily visits' part. Or, maybe, there are third party tools that are being used for this purpose? Julius (talk) 14:08, 17 January 2017 (EST)

There is Google Analytics, which shows that we have ~8000 daily visits, a number that is quite stable the past three years, modulated mostly by academic year and weekdays. There used to be an overview of the number of proofs and the number of definitions. However, showing this metric every time is quite resource-intensive and was harming page load times, so we removed it from the main page. — Lord_Farin (talk) 05:07, 29 January 2017 (EST)

Literature coverage progress

What do you think of explicitly showing the progress on literature coverage. By coverage I mean all theorems, lemmas, corollaries and problems which exist in a given book should be presented here without any missing steps with sufficient rigour. For example, this could be done for each section separately, say in the personal page of each book next to the corresponding section or chapter (percentage, or ratio of completed/total). This information could be exploited even further. One could make an invitation on the main page for a visitor, who, say, is confused by what is being said in a given book, to check if we have covered that book. Then there would be an increased chance for him to understand what was said in the aforementioned book. Maybe this could reduce the bounce ration of this webpage. Of course, this would require some micromanagement, and there are more important things missing, like whole branches of mathematics not having any related articles. Julius (talk) 09:07, 20 January 2017 (EST)

At this stage of ProofWiki's mathematics coverage, I feel like investing in this would be a waste compared to writing more definitions and proofs. Also, what is to be covered and what not is surprisingly hard to decide sometimes -- do you want this parenthetical example covered or not? — Lord_Farin (talk) 05:07, 29 January 2017 (EST)
I am making more of an attempt to be completist than I used to be, although I am finding it hard to make a case for the inclusion of the following exercise from 1965: Seth Warner: Modern Algebra: "Let $A$ be the set of words occurring in the first sentence of Chapter 1, $B$ the set of words occurring in the first sentence of the second paragraph of Chapter 1. What is $A \cup B$? $A \cap B$? $A - B$? $B - A$?" I find it difficult imagining a universe in which somebody can have thought this was an appropriate question for a graduate-level textbook. --prime mover (talk) 05:31, 29 January 2017 (EST)


Some time ago we enforced a policy whereby we removed a large number of redirects, on some grounds that I can't remember now, I expect someone will remind me.

Now it seems that we have re-implemented that policy of having lots of redirects.

Can somebody clarify our position here? I'm really confused. --prime mover (talk) 03:16, 2 February 2017 (EST)

Some observations: As the search function does not recognize wrongly spelled words, it is easy to search for a page and not find it; e.g. because you wrote "Real Numbers" instead of "Real Number". I'm certainly not in favor of adding redirects for wrongly spelled words, but for names of mathematicians with accents I think exceptions can be made. Also, searching for a theorem can be tricky: if the title of the page uses a synonym of a word you used in your search query, chances are you won't find the page. --barto (talk) 12:16, 2 February 2017 (EST)

Name space for algorithms?

Has it ever been considered to use a separate namespace for algorithms? If yes, what were the pros and cons? Currently they're put in the Theorem (main) namespace, but as they require a particular treatment (algorithm statement, followed by proof of finiteness etc., a section/theorem addressing its complexity), and as they often make for long articles, I thought it might be worth to work out a framework for them. --barto (talk) 13:20, 3 February 2017 (EST)

I have no problem with the concept. I have not been happy with the way we present algorithms so far. But it would need Joe to set it up as I believe I do not have the authorisation. --prime mover (talk) 15:10, 3 February 2017 (EST)
I see one problem: how to name those pages? Do we go for the verbose Algorithm:Prim's Algorithm or the weird Algorithm:Prim's :s ? --barto (talk) 16:17, 3 February 2017 (EST)
The former, unless someone has better ideas. It is only by accident of evolution that we don't have "Theorem:Cauchy's Theorem". THe clumsiness is regrettable but I fear unavoidable if we are to continue with the stylistic design of this site.
Better suggestions are invited. --prime mover (talk) 17:18, 3 February 2017 (EST)
Two possibilities are either including part of context in the title, say "Prim's minimum spanning tree determination" (or similar) such that it provides with enough details to separate from others, or to leave out algorithm, but mention the relevant field in parentheses like "Prim's (Graph Theory)", so it does not end abruptly. Julius (talk) 18:51, 3 February 2017 (EST)
Not a fan of the first idea for the simple reason that this is inconsistent with the rest of the naming philosophy: we don't do for example "Fermat's result that there is no integer greater than 2 such that the sum of two numbers to that power equals another number to that power". Can lead to unaesthetic verbal circumlocutions. One of the advantages of preferring named theorems is their brevity. --prime mover (talk) 19:11, 3 February 2017 (EST)
I'm with prime mover. Actually I think Algorithm:Prim's Algorithm isn't so bad. --barto (talk) 01:38, 4 February 2017 (EST)
I do not see a problem with that as long as that's the only item under the name of the author. However, as more items appear under a given person's name, some additional information should reside in titles to make a clear distinction between them. Hopefully we will never get to the point where a situation like the one above becomes real and necessary, but one could speculate about making an abbreviation of a given equation or lemma just for this website, and then refer to it whenever needed. --Julius (talk) 07:53, 4 February 2017 (EST)
You may wish to familiarise yourself with how things have been done so far on this website.


Hi. I just joined and I added a proof for the double angle of sin (proof 3). This is my first try and I will like to get some feed back - I know it's not perfect.

Also, I will like to help adding more proofs in geoemtry. Is there any one already working on this?

Thanks, ElyashivH (talk) 09:32, 28 February 2017 (EST)

Welcome. It is a pleasure to see our team growing. I am curious how you found this website.
The presentation of your proof is not far from the required form. One thing that is a bit visible is spelling. Anybody can make such mistakes, but usually they become marked in red on the code editor of this page. That should help you with spelling. Another aspect is formality of the proof. It would be better to start with "Let" or "Consider" instead of "look". A neutral language is preferred over a personalised invitation. Finally, it would be great if you could add a source of this proof. This is not always possible, but if there is a possibility, that would be welcome.
I do not think that anyone is working just on geometry. In the recent past it does not seem to be the main focus of anyone, although some members have lots of experience in this field already. It can be that you will work in geometry completely on your own. Could you be more specific on which areas of geometry you would like to write about? --Julius (talk) 13:27, 28 February 2017 (EST)
Sorry about the spelling - I worked from a computer in the library that doesn't have the dictionary installed.
I don't know a source for this proof - I just tried to calculate the sin of a double angle and did it this way. The proof isn't complete - It doesn't handle cases where the angle is more than 180.
I know just Euclidean geometry, so I guess I will need to stick to this.
ElyashivH (talk) 15:21, 28 February 2017 (EST)
Very well. If you feel that the proof is incomplete, use one of the templates used on this website, like Stub or WIP to denote it as an unfinished problem. Also, if you have any more questions regarding that exact proof, it would be better to use the talk page on the page of the proof ( or any other article). --Julius (talk) 15:56, 28 February 2017 (EST)

Problem with images


I had a problem with the image File:Thinker.gif which throws a complicated message when generating a thumbnail:

Error creating thumbnail: /usr/bin/timeout: the monitored command dumped core /var/www/proofwiki/web/w/includes/ line 101: 15784 Aborted /usr/bin/timeout MW_WALL_CLOCK_LIMIT /bin/bash -c "1" 3>&- Error code: 134

Any chance to look at this? I am trying to create a new template for "Proof Wanted". --prime mover (talk) 05:54, 5 March 2017 (EST)

A quick search shows that it could be memory limitation for such processes. Then one of the suggestions, except for faulty commands, is to increase that limit, like it is explained in [1]. --Julius (talk) 15:52, 5 March 2017 (EST)
I'll let someone else do that. --prime mover (talk) 18:32, 5 March 2017 (EST)
Can you confirm that this is still happening? --Joe (talk) 10:43, 7 March 2017 (EST)
It sure does. See Numbers whose Cyclic Permutations of 3-Digit Multiples are Multiples for an example. --prime mover (talk) 12:06, 7 March 2017 (EST)

Help needed with Template

I am trying to fix the SubjectCategory template to allow a "context" parameter to be passed in, which will be passed through into the SubjectCategoryNodef template to be displayed as "in the context of (link to definition page of subject)". There is an example of it invoked on Category:Baire Spaces. But for some reason I can't get it to work and I can't see what I am doing wrong. All I want to do is pass the "context" parameter as and when it is not blank.

Anybody care to have a go to see what I'm missing? --prime mover (talk) 07:01, 12 March 2017 (EDT)

I think you were thinking too complex. I removed the #if construct, now it works. — Lord_Farin (talk) 13:38, 13 March 2017 (EDT)
There is now a subtlety I've learned: the technique of adding a pipe into a parameter. Now I know how to simplify several templates. Thanks. --prime mover (talk) 14:46, 13 March 2017 (EDT)

Progress on Problem with Images

Did we get anywhere with this?

I had a problem with the image File:Thinker.gif which throws a complicated message when generating a thumbnail:
Error creating thumbnail: /usr/bin/timeout: the monitored command dumped core /var/www/proofwiki/web/w/includes/ line 101: 15784 Aborted /usr/bin/timeout MW_WALL_CLOCK_LIMIT /bin/bash -c "1" 3>&- Error code: 134

See above. --prime mover (talk) 02:17, 3 April 2017 (EDT)

Does this happen with all uploads or just this one in particular. --Joe (talk) 09:58, 3 April 2017 (EDT)
It just did it also with File:PascalsTriangle.gif when I put |left|100px against it just now on Definition:Pascal's Triangle. So no, it's not just Thinker.gif. --prime mover (talk) 10:09, 3 April 2017 (EDT)

This should be fixed now. Please let me know if it's not. --Joe (talk) 11:06, 3 April 2017 (EDT)

Yes indeed, that's all good. Good job. Cheers --prime mover (talk) 15:54, 3 April 2017 (EDT)

Policy on keeping categories small

So if I understand well, the goal is to have ~ max. 100 pages per category. I'm guessing that the correct way to tag e.g. the subpages of Definition:Connected (Topology) is: only the main page in the category Topology, all subpages under Connectedness? What about Definition:Connected Component (Topology): in the main category as well? --barto (talk) 09:52, 27 April 2017 (EDT)

Defining new MathJax commands

This has probably been discussed before, but I don't know how to search talk pages. Is it possible to define new commands, such as \lcm so they can be used everywhere? Would this slow down page loading or not make any difference? Note: I don't want to push through my own set of commands, just talking about some very common ones, such as \lcm, \Hom, \ord, about which there is no discussion how they should be named. This way we don't have to use \operatorname{} every time --barto (talk) 08:53, 28 April 2017 (EDT)

Yes we have had this conversation before, it crops up regularly -- but we haven't found anyone both committed and knowledgeable enough, and with sufficient privileges and authorisations to be able to do it. We did have some considerable discussion on what those abbreviations would be, and there was also some juicy disagreements as to whether a 2-letter command was better than a 2-word command or not. So it all got left up in the air, as usually happens. --prime mover (talk) 11:24, 28 April 2017 (EDT)
Technically, I should be able to make this happen. --Joe (talk) 11:27, 28 April 2017 (EDT)
Well what I propose is (if possible) to create commands that are literally what the output should be, like \Hom (and not \hom) for $\operatorname{Hom}$, so there can be no discussion or obscurity. (So not \Domain or \DomainOfMapping but \Dom for $\operatorname{Dom}$ and \dom for $\operatorname{dom}$, just to give another example.) --barto (talk) 12:02, 28 April 2017 (EDT)
I don't want both \Dom and \dom -- we use one or the other. Consistency. --prime mover (talk) 12:27, 28 April 2017 (EDT)
Of course, I was just giving an example. --barto (talk) 12:46, 28 April 2017 (EDT)
Also no things like \divides for whatever notation we use to express divisibility. --barto (talk) 12:05, 28 April 2017 (EDT)
\divides is a good example. We want that. We can't use \div unfortunately because that is already in use for $\div$.
... and because \div would have been perfect for $\operatorname{div}$ (used for divergence of vector fields and divisors in algebraic geometry). One has to be very careful when creating abbreviations (not to say avoid them), which is also why I want to focus on 'literal text commands'. While \divides may be an arguably useful command, I suggest not to focus on such commands now. It's much less standard and may prevent us from getting this project off the ground. --barto (talk) 12:46, 28 April 2017 (EDT)
We could then have a page where new commands can be requested and discussed. --barto (talk) 12:10, 28 April 2017 (EDT)
But \ProofWiki for $\mathsf{Pr} \infty \mathsf{fWiki}$ seems acceptable :) --barto (talk) 12:15, 28 April 2017 (EDT)
We already have <code>{{ProofWiki}}</code> -- I would discourage unnecessary proliferation of such things when we already have a shortcut like this. --prime mover (talk) 12:27, 28 April 2017 (EDT)
Incidentally, I have just remembered a good reason why we have not used "\def" in the past to define commands -- and that is when the page is transcluded, the "\def" is not carried over into the transclusion. We need to make sure we don't fall foul of this. --prime mover (talk) 12:30, 28 April 2017 (EDT)
Never mind about \ProofWiki --barto (talk) 13:02, 28 April 2017 (EDT)

Joe, would you mind experimenting a bit, say to try to implement \lcm as a start? --barto (talk) 04:35, 30 April 2017 (EDT)

Will do! --Joe (talk) 10:45, 4 May 2017 (EDT)
Done. --Joe (talk) 11:24, 4 May 2017 (EDT)
Can confirm that $\lcm$ works well, and also in transcluded pages (where this technique was under suspicion). --prime mover (talk) 15:46, 4 May 2017 (EDT)
Awesome! What about creating a separate page where we create our list of newly introduced commands, aside from Symbols:LaTeX Commands? We could then discuss them there and have a nice overview of what is new versus what is standard included in MathJax. --barto (talk) 16:19, 4 May 2017 (EDT)
Don't know about "separate from" -- "transcluded into" I am up for: Symbols:LaTeX Commands/ProofWiki Specific is where we start. --prime mover (talk) 16:36, 4 May 2017 (EDT)

MathJax CDN discontinuation

I just noticed that MathJax will permanently shut down its CDN services on April 30 (tomorrow): Is ProofWiki hosting its own MathJax or does it use their CDN? --barto (talk) 12:11, 29 April 2017 (EDT)

Thanks for the heads up. We were aware of the CDN shutdown, but we have a local instance running (as you noticed the past days). — Lord_Farin (talk) 11:31, 2 May 2017 (EDT)

Restricting the Rename tool

Does anyone know if it is possible to restrict the "Move" tool (which effectively renames a page) to "Admin" users or other suitably authorised editors? It is so easy to just change page names according to whim and personal preference, and completely compromise the linking structure. --prime mover (talk) 02:35, 2 May 2017 (EDT)

Joe should be able to manage this; it's internal MediaWiki configuration. I think it makes sense, given that we nowadays have the "Trusted user" group in between. — Lord_Farin (talk) 11:31, 2 May 2017 (EDT)
I'll take care of this. --Joe (talk)
Should be limited to the "Trused user" group and above now. --Joe (talk) 11:21, 4 May 2017 (EDT)
Jep, confirmed :) --barto (talk) 16:20, 4 May 2017 (EDT)

Strange behaviour with editor

Has anyone else noticed the following strange (but inconsistent) behaviour of the wiki page editor?

You place the cursor at the place in the page where you want to add a macro-generated construct, e.g. the <code>[[Category:]]</code> MediaWiki function from below the edit screen.

But instead of appearing on the screen where you put your cursor, it appears on the line (or 2 line, or 3 lines, ...) above.

This not only affects the macros below the screen, but I have also noticed it on the "add header" button (the big uppercase A on the list of edit helper buttons above the edit page). However, it is inconsistent -- it does not always happen, and I have been unable to identify the conditions under which it does happen.

Note that I am using Google Chrome (I don't have the patience for any other) on a Windoze 10 machine (yes I know, how brain-dead is that).

It has only started doing it for the past week or so (I can't recall when it started). I first wondered whether it was a glitch, but now I wonder if it is more general than just my machine. --prime mover (talk) 02:40, 4 May 2017 (EDT)

Error on protected featured proof.

I find it odd that the proof 3 in this featured proof (Complex numbers cannot be totally ordered) would have a typo. Near the beginning it states.

Aiming for a contradiction, suppose suppose that (C,+,×) can be ordered.

Suppose, suppose... Feels philosophical to me.

The person responsible is to be fired. --prime mover (talk) 14:01, 7 June 2017 (EDT)

On the begging of forgiveness

(It's better than asking permission.)

As some time ago we lost sight of the number of proofs (for some reason it got removed), we haven't been able to get a quick count of the number of theorems.

To make up for this, I have added the category Category:Proven Results, invoked automatically when the <code>{{qed}}</code> template is included at the bottom of any proof. This category is populating itself as I type. --prime mover (talk) 02:34, 9 June 2017 (EDT)

It seems cleaner to add Category:Proven Results to Category:Hidden categories. Nobody will ever look at that category anyway; it serves just to count the proofs. --barto (talk) 02:46, 5 August 2017 (EDT)
You're sure of that why? --prime mover (talk) 05:30, 5 August 2017 (EDT)
Because Category:Proven Results is too large to scroll through. You're much more likely to find what you're looking for with a normal search. But I won't insist. --barto (talk) 05:42, 5 August 2017 (EDT)
It can be fun. Offering the opportunity for someone to browse through it in all its manic colossalness is better than not doing so. I do it myself sometimes. --prime mover (talk) 14:14, 5 August 2017 (EDT)

The following are equivalent

What about a template "TFAE" for The following are equivalent:

The following are [[Definition:Logical Equivalence|equivalent]]:
? --barto (talk) 05:43, 30 July 2017 (EDT)
Yeah, why not? On it ... --prime mover (talk) 05:46, 30 July 2017 (EDT)
Cool. Just that sentence or also some functionality to include the equivalent statements, with :$(1): \quad$ etc? Maybe that makes it too complicated to use. Not sure. May aso clash with some LaTeX symbols then. --barto (talk) 05:51, 30 July 2017 (EDT)
How's that? See it in use on the page Equivalence of Definitions of Abelian Group. --prime mover (talk) 05:59, 30 July 2017 (EDT)
It's fine. The clashing I was talking about was in the hypothetical situation where the TFAE template would include parameters for the equivalent statements, similar to the Equation templates. But the way it's now is good. I like it. --barto (talk) 06:08, 30 July 2017 (EDT)
I considered adding the stuff to populate the rest of the page, i.e. by automatically transcluding the definitions, or indeed the statements themselves, but deciding how to pass in the number of such equivalences is non-trivial (as far as I know there is no easy way of setting it up to handle varying numbers of things to be defined without going down the complicated route we went through with the Citation template etc. --prime mover (talk) 06:21, 30 July 2017 (EDT)

Does this mean we can now sort the pages in Category:Definition Equivalences using [[:Category:Definition Equivalences|{{{def}}}]] in the template? --barto (talk) 10:11, 1 August 2017 (EDT)

Yes, it seems :) --barto (talk) 16:58, 1 August 2017 (EDT)

I don't understand the question. All pages are sorted alphabetically. --prime mover (talk) 17:20, 1 August 2017 (EDT)
I mean, to ignore "Equivalence of Definitions of" in the alphabetic sorting, so that the pages will be organized in sections according to the starting letter. We talked about this at Category talk:Definition Equivalences once. --barto (talk) 17:25, 1 August 2017 (EDT)

Expanding the Help Section

Many things are not yet addressed in the Help Section, leading to discussions being brought up over and over, while a consensus has already been reached. To make $\mathsf{Pr} \infty \mathsf{fWiki}$ more attractive to contributors, I decided to expand the Help Section (which has not been done for years, it seems).

Topics include:

  • When to Redirect - partially addressed at Help:Redirects
  • Which categories to add - partially addressed at Help:Categories
  • The balance between accessibility and generalizations of elementary results.
  • When to enumerate definitions and when to separate them into sections.
  • How to treat pages with multiple definitions
  • New LaTeX commands
  • Requests for moving pages
  • Commutative Diagrams
  • etc...

There are many such things, and the discussions are scattered on talk pages. I've recently started writing, but only with the input of many users can the Help Section become really great. --barto (talk) 09:33, 30 July 2017 (EDT)

More topics: informal definitions, proof outlines, proof outlines in the case of multiple proofs, what to put in the also see section and what not, (over)using the AoC Template, stubs and similar templates, requesting new categories, example categories. --barto (talk) 10:08, 30 July 2017 (EDT)

More: Named theorems, named theorems and disambiguation, Template:Languages. --barto (talk) 13:25, 30 July 2017 (EDT)

Sounds good to me. --Joe (talk) 10:38, 31 July 2017 (EDT)
I've started by creating a skeleton for the missing topics. Fleshing it out now. You're welcome to change the wording or give me some advice if you're not happy with it. There are some sections I can't complete, because they have not yet been discussed, or I because can't find the discussions. When I find one (e.g. on the archives of the main talk), I copy the consensus to the Help Section. --barto (talk) 16:47, 1 August 2017 (EDT)

Commutative diagrams using AMScd

Next to XyJax, which I learned is no longer maintained, some commutative diagrams can be written using the AMScd package, using \require{AMScd} for every CD. According to the MathJax Documentation, we can make a small change in the MathJax configuration to avoid using \require{AMScd} every time. --barto (talk) 09:57, 2 August 2017 (EDT)

I can add this to our configuration later today or tomorrow. We should also need to remove/replace anything that's using XyJax. --Joe (talk) 10:00, 2 August 2017 (EDT)
Thanks. As long as XyJax works and there is no good alternative, I'd rather not remove the existing diagrams though. A TikZcd extension would be ideal, but all I could find is an ordinary PGFTikZ Extension: It creates images, not MathJax, has a cumbersome syntax and will not be compatible with our new LaTeX commands. --barto (talk) 06:41, 24 August 2017 (EDT)