Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

LaTeX: Math render delay

Bottom of Page

1 to 16 of 16

  1.  
    • CommentAuthorgeraldedgar
    • CommentTimeMar 13th 2012 edited
     

    I see a great math render delay today, unlike yesterday. (Safari 5.1.4 just out, is that related?) I open a page, a short question and two answers. Yesterday, rendering took a couple of seconds. Today (I timed it) 90 seconds. At least it is better than math.stackexchange, where math doesn't render at all...
    I tried it with an old Safari (3.0.4) and it has the same delay, so perhaps it is not the fault of the new version. Also tried Google Chrome, Firefox, Opera. All have the delay.

  2.  

    Relevant.

    Later addition: It seems that they have solved the problem.

  3.  

    That's weird, because we don't use the MathJax CDN; we host our own copy.

    • CommentAuthorEmil J
    • CommentTimeMar 14th 2012
     

    I got:

    A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.

    Script: http://latex.mathoverflow.net/mathjax1.1/MathJax.js?config=MOconfig:29

    • CommentAuthorEmil J
    • CommentTimeMar 14th 2012
     

    I see, MathJax.js consists of a banner comment and a single 42KB line number 29 which contains all of its actual source code. Very nice.

  6.  

    Whatever the cause, it is all better now.

  7.  

    Interesting. I'd forgotten that we host our own copy of the MathJax library. This actually seems like a bad idea --- even considering the recent episode, I suspect the CDN is more reliable than my web hosting (latex.mathoverflow.net is run off my dreamhost account).

    • CommentAuthorMariano
    • CommentTimeMar 19th 2012
     

    The CDN copy is also payed by Someone Else, no? (And probably updated also by Someone Else!)

  9.  

    Isn't MathJax 2.0 out already? Is it possible to upgrade the one on the site?

    One reason to suggest this upgrade is that with MathJax 1.1 both MO and MSE would hang sometime and completely overload the CPU until Firefox (or otherwise) ask me whether to stop the script or continue.

    With MathJax 2.0 on MSE I don't recall this problem anymore, while on MO it still bugs me from time to time.

  10.  

    I've just updated to MathJax 2.0, but not to the CDN (it doesn't seem to work immediately ... I'll debug later).

    I'm pretty sure I handled the fonts issue, so fonts should load quickly. Please let me know if there are any problems.

  11.  

    Anton, when MSE upgraded the MathJax it was noted that there is an option which allows mobile phone clients to have a 'smaller cache' and to render more often; which makes long pages render into a (partially) readable state faster.

    Is it possible to make the same configuration here? I visited several heavily TeX'd pages from my iPhone and it took longer than I'd like to load the pages completely.

  12.  
    Anton: I have a new issue with MathJax (just the past couple of days) and I wonder if its onset coincides with your upgrade:

    In Firefox, LaTex'ed expressions frequently overflow the space allotted for them so that they overlap surrounding text. Here, for example, is a screen grab of one of my own recent answers:



    Could this be upgrade-related?
  13.  

    For reference, Steve's answer pictured above is at http://mathoverflow.net/questions/91246/other-homology-theories-still-count-holes/91253#91253. Works fine for me, beta Chrome, Firefox 9.0.1, or Safari 5.1.4 on OS X.

    What's your Firefox version and OS? Perhaps this should be reported in the MathJax bug tracker <https://github.com/mathjax/MathJax/issues>. It would be nice if there were an easy hack to switch the MathJax version, but I can't think of one.

  14.  

    I have the same issue as Steve regarding spacing, in firefox 3.6.17 and 10.0.1 and 11.0 (but not firefox 3.5.3), all on linux. I have not been able to reproduce this on any other site which uses MathJax.

  15.  
    I'm having the problem in Firefox 3.6.12 under Windows Vista, which, for idiosyncratic but powerful reasons, is my much-preferred browser.
  16.  
    I've seen the phenomenon SteveLandsburg mentioned on Math SE using Chrome, but it usually goes away by simply reloading the page.

1 to 16 of 16

  1.  
Back to Discussions Top of Page