[v4 beta.6] visual regression around (poor use of) \mid

[pkra]

This falls under bad authoring and we will probably change the source to \vert but I wanted to let you know that the following had visual changes (for the worse) when upgrading from beta.4 to beta.6. (I'm guessing it's really an SRE issue.)

\begin{equation*} d\big ((t,x,\lambda ),(s,y,\mu )\big )=\mid t-s \mid + \parallel x-y\parallel + d_{M}(\lambda ,\mu ). \end{equation*}

$A=\{x\in \mathbb{R}^2: 1\leqslant \mid x \mid \leqslant 2\}$

beta 4:

beta 6:

Edit: updated screenshots.

[pkra]

Sorry, the screenshots were wrong. I've added new ones (using STIX Two) - my "vanilla" v4 setup was failing me just now. I'll try to reproduce it there, too.

[dpvc]

I'm not able to reproduce this with a generic beta.6 using STIX2 fonts. This is what I see:

This seems in line with your beta.4 images, though mine has the proper spacing between the equal sign and the \left\vert in the displayed equation, which is not the case for yours. I'm wondering if there isn't something else affecting your output?

[dpvc]

Or perhaps the first one was supposed to use \mid rather than \left\vert and \right\vert?

[pkra]

Oh, sorry!! You're of course correct. I had paste the "fixed" example :pensive: no wonder I couldn't reproduce :rofl:

[pkra]

I've edited the top card.

[dpvc]

OK, the spacing between the equal and the \mid is now gone, but I don't get any extra space after the \mid in the first expression, and the second expression didn't change, so still is as in my image above. So I'm still not able to reproduce the problem spacing that you are seeing.

[pkra]

So I'm still not able to reproduce the problem spacing that you are seeing.

Ugh, that's unfortunate. I'm not doing anything new in the AMS setup. But as you know it's somewhat custom, server-side etc. I'll try to build a better test case.

[dpvc]

Perhaps it is browser-specific? Which browser are you using? I tested in Firefox, but will try Chrome.

[dpvc]

Nope, Chrome works for me, too.

[pkra]

I'm seeing this in a nodejs setting. I'll try to get you a MWE tomorrow.

[dpvc]

I am cleaning gout old issues, and noticed this one isn't complete. Is it still an issue for you?

[pkra]

Thanks for checking in. Yes, I'm still seeing this with the AMS's setup using mathjax beta.7 and sre beta.11. It's not a high priority (IIRC the source was changed).

Sorry for never getting you an MWE. I'll try to find time for that.