CSS interoperability with vertical-align

[dani31415]

The CSS vertical-align property should work as per specification with MathML (CSS spec). The resolution of this issue is perhaps just related to create some tests and does not affect the MathML specification.

Some examples are

<img height="40" src="https://www.w3.org/html/logo/img/mark-word-icon.png">
<math><mi style="vertical-align:top">24</mi></math>

with result

and

<math>
    <mn>4</mn>
    <mo>-</mo>  
    <mfenced>
        <mrow style="vertical-align:middle">
            <mfrac>
                <mfrac>
                    <mi>x</mi>
                    <mn>2</mn>
                </mfrac>
                <mn>2</mn>
            </mfrac>
        </mrow>
    </mfenced>
</math>

with result

[fred-wang]

I think this makes sense for the long term, IMHO it's way too early to consider this for the Core spec. Igalia is still working with Google to clarify the box and layout models, so I'd prefer to keep things simple and just ignore the vertical-align value for box layout for now.

[fred-wang]

I think vertical-align is not what we want. It is is supposed to apply to inline boxes.

The dominant-baseline property looks closer to what we want:
https://www.w3.org/TR/css-inline-3/#propdef-dominant-baseline

[NSoiffer]

I wasn't aware that there was a "dominant-baseline" property. Interestingly they mention a value of mathematical. They don't use the word "axis" in their description of it, but it seems like we can influence the language they used (it is marked with an issue that the section is being rewritten) and perhaps this is a way to get the notion of math axis into CSS...

[fred-wang]

This is mentioned in the current core spec:
https://mathml-refresh.github.io/mathml-core/#box-model

[NSoiffer]

What's left to do? Can this be closed or does some other issue need to be resolved before this can be closed?

[NSoiffer]

Need to write a test that shows setting vertical-align does nothing.

[rwlbuis]

I will try to add that test.