[css-fonts] Clarify font-optical-sizing behaviour under size-adjust

[drott]

In
web-platform-tests/wpt#28399
I made a ref-test ensuring that font-optical-sizing: auto behaves according to the adjusted / resulting font-size under size-adjust, pending review.

Since the wording for size-adjust says:

However, the computed font-size (and thus any values that derive from it, such as em units, percentages in text-underline-offset, etc.) remains unaffected.

And font-optical-sizing: auto says:

The user agent may modify the shape of glyphs based on the font-size and the pixel density of the screen. For OpenType and TrueType fonts using font variations, this is often done by using the "opsz" font variation.

we may want to clarify that the applied optical sizing value should be based on the resulting/adjusted/displayed font-size, not on the computed font size, in order to avoid confusion?

I believe the behaviour as tested in the new test makes sense as the glyph size is changed by size-adjust, and the scaled glyphs should then be rendered with the scaled/new optical size value. Or in other words, the glyph is displayed larger/smaller, and so should its optical sizing value be adjusted accordingly.

[jfkthame]

I agree this makes sense, although I note that the current Gecko implementation fails this test, because it sets 'opsz' based on the computed font size -- so that'll need to be fixed.

(In practice, I think most use cases for size-adjust will involve fairly moderate scaling amounts, so the difference will be barely noticeable; but still, we should specify and do the right thing, not just fudge it!)

This also prompted me to wonder how optical sizing interacts with font-size-adjust; I believe that is similarly broken in Gecko at present. Maybe you could include a similar test for that?

[drott]

Thanks for your feedback, Jonathan.

This also prompted me to wonder how optical sizing interacts with font-size-adjust; I believe that is similarly broken in Gecko at present. Maybe you could include a similar test for that?

We're currently not shipping font-size-adjust: and if Gecko would fail it, I wouldn't have a way of verifying the test atm, which would make it a bit brittle to submit to WPT.

[css-meeting-bot]

The CSS Working Group just discussed [css-fonts] Clarify font-optical-sizing behaviour under size-adjust #6190, and agreed to the following:

• RESOLVED: font-optical-sizing: auto references the used font-size, accounting for 'size-adjust' and 'font-size-adjust'

The full IRC log of that discussion

<fantasai> Topic: [css-fonts] Clarify font-optical-sizing behaviour under size-adjust #6190
<fantasai> github: https://github.com//issues/6190
<Rossen__> github: https://github.com//issues/6190
<fantasai> chrishtr: Question of size-adjust interaction with font-optical-sizing
<fantasai> chrishtr: The proposed resolution is to use font-optical-sizing of the used font size
<fantasai> myles: of course
<fantasai> RESOLVED: font-optical-sizing: auto references the used font-size, accounting for 'size-adjust' and 'font-size-adjust'