If you try to print the testcase, the "+" symbol disappear. See https://github.com/mathjax/MathJax/issues/137

Is this mathml-specific? I seem to recall there being general issues with @font-face and printing...

MathJax devs reported another issue: https://github.com/mathjax/MathJax/issues/172 which corresponds to that Bugzilla entry: https://bugzilla.mozilla.org/show_bug.cgi?id=468568 but web-fonts do not seem to be involved in the present bug. I would say this bug has something to do with the <mo> operators, which are treated in a specific way.

I ran out of time yesterday but I was beginning to think there was a connection with characters that were centred (+, -, &ge; &le;) as they were the _single_ characters that I could get to fail. [In particular I was looking at those mentioned at http://mxr.mozilla.org/mozilla-central/source/layout/mathml/nsMathMLmoFrame.cpp#220 ]

(In reply to John P Baker from comment #3) > I ran out of time yesterday but I was beginning to think there was a > connection with characters that were centred (+, -, &ge; &le;) as they were > the _single_ characters that I could get to fail. > > [In particular I was looking at those mentioned at > http://mxr.mozilla.org/mozilla-central/source/layout/mathml/nsMathMLmoFrame. > cpp#220 ] I'm able to reproduce the problem with "[" too. It disappears when I use for example <mtext> instead of <mo> or if I use character outside the operator dictionary ("a" for example) I think it's an issue with <mo> operators for which nsMathMLmoFrame::UseMathMLChar returns true i.e. those drawn in nsMathMLChar.

This bug may depend on bug 736010.

I still see the bug even after bug 468568 is fixed.

Oddly, I have 2 Windows 7 64-bit laptops. I see the issue on one and not the other. Trying to figure out what the difference is.

I no longer see this bug on Linux, with Firefox 18. Bill, are you able to reproduce the bug on one of your Windows machine?

(In reply to Frédéric Wang (:fredw) from comment #8) > I no longer see this bug on Linux, with Firefox 18. Bill, are you able to > reproduce the bug on one of your Windows machine? Printing of the attached testcase works correctly on all of my windows machines using Firefox 18.0.1.