# Seems to be a (minor) bug with displaying mathjax

Hardly noticeable but I still decided to report it - just noticed: at Find the number of boolean functions of n variable that satisfy the following condition one sees this

The second $k$ below is obviously spurious...

PS Added the link as suggested by quid; the bug is not always there however, right now it reappeared after I reloaded that page.

PPS On Windows 10 pro I see this in Waterfox 44.0.3 and Chrome 48.0.2564.116 m (64-bit), have not tried any other browsers

Later - seen another instance of the phenomenon in the "Related" column of Is this a model for $K$-theory of a triangulated category? which might shed some light on the nature of it. Result looks like

while the actual content of the garbled part should be "Proof that the homotopy category of a stable $\infty$-category is triangulated"

Thus the whole thing is broken into columns in a wrong way somehow.

Also as pointed out by Martin Sleziak in a comment below this also happens on math.SE and physics.SE. At links provided by him there is a suggested solution by Davide Cervone who seems to be one of the developers of MathJax

• It may make sense in the related column of which question you had that title; if not it will be hard to see if it can be reproduced. – user9072 Feb 29 '16 at 10:14
• I've seen this happen every now and then, too. – Emil Jeřábek supports Monica Feb 29 '16 at 11:01
• @quid You are right, I should add this, the question was mathoverflow.net/q/232311/41291. I visited it again now and the thing was gone; however when I reloaded the page it appeared again. – მამუკა ჯიბლაძე Feb 29 '16 at 11:22
• Added, mentioning that it is not always there – მამუკა ჯიბლაძე Feb 29 '16 at 11:25
• This is very likely caused by something intefering with the CSS of the output from the AssitiveMML extension (which embeds visually hidden MathML for assistive technology). You can disable the extension via the MathJax menu (under math settings => math renderer). If you run into this again, it would be helpful if you could use the FF or Chrome developer tools to see what CSS is interfering. (Unfortunately, I can't reproduce it myself.) – Peter Krautzberger Feb 29 '16 at 12:44
• @PeterKrautzberger Sorry I don't know enough to figure out myself. When I inspect it, the stray $k$ is inside something called <span style="font-family: MathJax_Math; font-style: italic;" id="MathJax-Span-8" class="mi">k</span> which itself is inside <span id="MathJax-Span-7" class="mrow"> which is inside <span style="position: absolute;"> which is inside <span style="display: inline-block; position: relative; width: 0em; height: 0px; font-size: 120%;"> which is inside <span role="math" id="MathJax-Span-6" class="math"> which also holds the correct piece (with correct $k$) – მამუკა ჯიბლაძე Feb 29 '16 at 13:12
• What is involved seems to be called inline:10 from .MathJax span but I have no idea how to get any further information – მამუკა ჯიბლაძე Feb 29 '16 at 13:14
• Thanks for taking a look. What I'm looking for is information in the panel names "styles" on the right-hand-side here developer.chrome.com/devtools#dom-and-styles. For the element with class="MJX_Assistive_MathML" I'm hoping to find out if those styles (a section with the name of the class) are being overruled, in particular heigh, width, clipping etc. They would be struck out in the list and some other section in that panel would indicate what stylesheet is overriding them. – Peter Krautzberger Feb 29 '16 at 13:46
• @PeterKrautzberger There are indeed several lines overruled under <span role="presentation" class="MJX_Assistive_MathML">, of about 16 lines, 5 towards the end, 4 of those marked with "Invalid property value", namely -webkit-touch-callout: none;, -khtml-user-select: none;, -ms-user-select: none; and user-select: none;. It is fourth in the list, the first being element with no entries, the second .MathJax span with some lines overruled too, the third .MathJax * with two of five lines also overruled and marked "Invalid property value", then some other stuff... What must I seek? – მამუკა ჯიბლაძე Feb 29 '16 at 15:54
• Thanks for digging deep! Maybe we should move this to chat (though my day is coming to an end). Could you share a screenshot perhaps? I know the style panels can be rather long and it's difficult to know what's important. – Peter Krautzberger Feb 29 '16 at 16:37
• @PeterKrautzberger Sure. I've got a screenshot (although not everything fits in it obviously, and I also have to leave soon). If you tell me how to arrange for chat and/or how to send you the screenshot I'll do it. Or we can postpone this if you prefer. Email is also an option. – მამუკა ჯიბლაძე Feb 29 '16 at 16:43
• Email sounds good. Can you ping me at peter.krautzberger@mathjax.org? – Peter Krautzberger Feb 29 '16 at 17:30
• @PeterKrautzberger Only saw this now, sorry. I've sent to the gmail account that I found on your page, now forwarded here too. – მამუკა ჯიბლაძე Feb 29 '16 at 21:17
• Isn't this the same problem as here and here? – Martin Sleziak Apr 25 '16 at 7:32
• In fact, if you believe this comment, the problem should be fixed soon: "Thanks for reporting. Fix is already in repo waiting for build. Should be live in < 24h." – Martin Sleziak Apr 25 '16 at 8:16