Created attachment 206547 [details]
Here is a testcase. "blah" should align with the base of the msubsup.
See bug 95015 for a related change.
Created attachment 206600 [details]
WebKit rendering looks fine to me. The text is nicely aligned next to the <math> element. (see 2nd testcase)
Though FF renders it differently, but that's because they calculate the <math> element's height differently (they end up with a lot smaller box) which makes the vertical-align property almost non-relevant here.
(In reply to comment #0)
> Created an attachment (id=206547) [details]
> Here is a testcase. "blah" should align with the base of the msubsup.
> See bug 95015 for a related change.
so I believe 'vertical-align: baseline' should align with the base of the msubsup as opposed to 'bottom' (unless I miss something here)
I was considering vertical align for table cells which is interpreted differently than for inline elements:
I think my comment on bug 95015 was misleading. I found the align bug while trying to write reftests for bug 95015, but that's probably unrelated since vertical-align is not inherited.
Actually, perhaps the bug is not related to MathML at all. Bug 90335 is a similar issue with HTML-only elements. I'm guessing using a div as a workaround should work for the present bug too.
BTW, Gecko's height bug is: https://bugzilla.mozilla.org/show_bug.cgi?id=415413
(In reply to comment #4)
> BTW, Gecko's height bug is: https://bugzilla.mozilla.org/show_bug.cgi?id=415413
Sorry it is https://bugzilla.mozilla.org/show_bug.cgi?id=219873
*** This bug has been marked as a duplicate of bug 90335 ***