0
$\begingroup$

I started getting [Math Processing Error] in the comments under this answer after WimC posted the comment that begins with "@joriki: Err, actually I started with the pdf". The errors occur roughly half the time when I refresh the page, whether hard or soft, with a hard refresh apparently having a higher probability of toggling the state; the other half, everything looks fine. I cleared my browser cache and checked that the about box says MathJax 2.1 is being used. I'm using Safari Version 5.1.7 (6534.57.2) under Mac OS 10.6.8. I went through all bug reports that mention [Math Processing Error]; the most relevant one seemed to be this one, but none seemed to be identical. The only unusual thing I can find in the $\TeX$ code in the comments is the use of \tfrac.

Here are screen shots of the erroneous display

erroneous display

and of the correct display

enter image description here

$\endgroup$
11
  • $\begingroup$ Renders fine in Chrome and Opera on Windows 7. $\endgroup$
    – Bill Dubuque
    Nov 5, 2012 at 5:47
  • $\begingroup$ 'Say "intermittent fault" to any programmer and watch their face fall.' -- Simon Tatham, "How to Report Bugs Effectively" (link) $\endgroup$
    – Douglas S. Stones
    Nov 5, 2012 at 6:56
  • $\begingroup$ When you get the error messages, would you type javascript:alert(MathJax.Hub.lastError.message) in the URL typein and let me know what you get. Also try javascript:alert(MathJax.Hub.lastPrepError.message), though that might not produce anything. Thanks $\endgroup$
    – Davide Cervone
    Nov 5, 2012 at 11:52
  • $\begingroup$ @Davide: I get "'undefined' is not an object" for the first one, nothing for the second one. Feel free to provide more instructions. $\endgroup$
    – joriki
    Nov 5, 2012 at 17:46
  • $\begingroup$ @joriki, Two possibilities: (1) did you do this when the processing error messages are present, and (2) did you get the capitalization correct in the command? (Did you copy and paste it, or type it by hand?) $\endgroup$
    – Davide Cervone
    Nov 5, 2012 at 21:19
  • $\begingroup$ @Davide: (1) yes and (2) yes (copied and pasted). The problem isn't that the javascript isn't being processed, it is; e.g. when I leave off .message, I get "TypeError: 'undefined' is not an object" instead, and when I set the value with javascript:MathJax.Hub.lastError.message="0" and then display it again, I do get "0"; so it seems that MathJax.Hub.lastPrepError.message really does have the value "'undefined' is not an object". $\endgroup$
    – joriki
    Nov 6, 2012 at 0:29
  • $\begingroup$ @Davide: I entered the debugger and printed MathJax.Hub.lastError.message -- here it says "'undefined' is not an object (evaluating 'd.prefilterMath')" $\endgroup$
    – joriki
    Nov 6, 2012 at 0:36
  • $\begingroup$ @Davide: Interesting -- when I do that when the display is correct, I get an error object with "message" property "'undefined' is not an object (evaluating 'MathJax.Hub.lastError.message')" -- so it seems the two cases can only be distinguished in the debugger where the "evaluating ..." part is displayed. $\endgroup$
    – joriki
    Nov 6, 2012 at 0:45
  • $\begingroup$ @joriki, of course, the 'undefined' is not an object was the error message stored in MathJax.Hub.lastError! I was foolish not to realize that. Of course, the (evaluating 'd.prefilterMath') that was part of the message would have helped. I will check into it. In the meantime, are you running any extensions that could affect MathJax? $\endgroup$
    – Davide Cervone
    Nov 6, 2012 at 10:34
  • $\begingroup$ @Davide: "would have helped" -- that may be a misunderstanding -- (evaluating 'd.prefilterMath') wasn't part of the message in the alert box -- I only got that when I printed MathJax.Hub.lastError.message in the debugger. By the way, I can no longer reproduce the bug now; I've tried a couple dozen times. $\endgroup$
    – joriki
    Nov 6, 2012 at 13:42
  • $\begingroup$ @joriki, interesting both that the message from the javascript call was different from the console result, and that you can no longer reproduce it. I'm thinking we are not going to be able to get the bottom of it without that. Let me know if you get it again. $\endgroup$
    – Davide Cervone
    Nov 6, 2012 at 16:38

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .