Hello,

My previous email was generated from the information of chromestatus.com <http://chromestatus.com> ; I’d like to elaborate a bit more about the situation of MathML in Chromium. In 2019, we published a roadmap [1] which we basically implemented downstream the same year. As of today, we consider this roadmap complete upstream, modulo imperfect math layout of stretchy operators due to the limitation of the current LayoutNG architecture (the rendering team is aware of that and consensus is that it’s ok for a first implementation).

In addition, we have been receiving feedback from various people and tried to adjust our initial roadmap accordingly:

1. Some concerns were shared by some browser vendors, TAG members or
   some participants to the “Getting Math onto Web Pages CG” and are
   summarized on the official TAG’s feedback [2]. During these recent
   years, we’ve collaborated with the MathML and CSS working groups as
   well as browser vendors to publish specifications & tests that
   address these concerns. This also added many new tasks on the
   roadmap (e.g. new CSS properties, MathMLElement IDL, mixing with
   CSS/SVG, etc).
2. MathML supporters have also been asking a more complete support for
   MathML3 features, raising concerns about backward compatibility.
   Apple shares this concern about breaking content from their
   applications (e.g. Apple Books) and is generally reluctant to remove
   shipped features. Mozilla has been more positive about that and
   several patches landed to unship legacy features with low usage. We
   decided to solve that disagreement by publishing two specifications:
   MathML Core (intended to be implemented in browsers) and a superset
   called “MathML Full”. Polyfills for MathML features that are not
   part of the core spec have also been written. Incidentally, the
   layering/extensibility effort for (1) became instrumental to write
   such polyfills.
3. We also received feedback about making MathML accessible, so we
   added that in our plan. We implemented parity with Gecko/WebKit for
   macOS/NSAccessibility and Linux/ATK. We’ve also been working on a
   corresponding MathML AAM specification to standardize existing
   support and we wrote (currently internal) tests for that. Other
   platforms (Windows, ChromeOS, Android) don’t seem to have any
   special way to expose MathML nodes to assistive technologies, so
   instead we continue to rely on existing DOM-based approaches.
4. Finally, a lot of users have praised our work, tested existing
   support by enabling web platform features, reported bugs and they
   keep asking us when this is going to be officially shipped.
   Hopefully this intent-to-ship will allow to address that last
   feedback :-)

At that time, we believe the implementation is good enough for a first release. To be fully transparent these are known issues:

 * There are bugs to fix or more features users would like to see. We
   expect the layering/extensibility approach will make possible to
   work around current limitations.
 * Not all operating systems come with a pre-installed OpenType MATH
   font, which means poor math rendering on such systems by default.
   However, Windows has Cambria Math pre-installed and things seem to
   improve on the macOS & Android side [3]. The workaround for that is
   to use web fonts (on pages or as a web extension).
 * There are a few interactions with CSS properties (e.g. width/height,
   out-of-flow, etc) that we would like to clarify in the spec and/or
   the implementation and ensure we have interoperability tests for
   that. These are likely edge cases for MathML authors but can be
   important for polyfills.

That’s it, we are looking forward to hearing your feedback !

[1] https://mathml.igalia.com/project/
[2] https://github.com/w3ctag/design-reviews/issues/313#issuecomment-460523527
[3] https://frederic-wang.fr/update-on-open-type-math-fonts.html

--
Frédéric Wang

--
You received this message because you are subscribed to the Google Groups 
"blink-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/a/chromium.org/d/msgid/blink-dev/7af80b42-9838-d0cf-8e6c-aabf6a7fb2d5%40igalia.com.

Reply via email to