Thanks Chris. I'm attaching a testcase for contenteditable MathML.
Indeed, the formula is properly rendered. It can also be edited and this
will likely lead to "invalid" MathML markup (e.g. fraction ending up
with only one child) but Firefox and WebKit don't perform integrity
check either. And I guess it's fine, probably someone should really rely
on JS to extend the native editor.
Morten was actually only mentioning the case when table is within a
multicol so probably that excludes the Wikipedia use case (I'm going to
do more checks there).
Printing would still be an issue.
So yeah, I personally don't think that should block shipping MathML but
there are remaining issues with legacy layout that we need to be aware of.
On 23/06/2022 17:54, Chris Harrelson wrote:
EditingNG has already shipped. Only table fragmentation and printing
are still to ship, and hopefully they will both be done by the end of
the year.
On Thu, Jun 23, 2022 at 4:39 AM Frédéric Wang <fw...@igalia.com> wrote:
Hi Morten,
Thanks for raising this issue, I forgot to mention it. We
experimented
this in the past and IIRC we found issues with:
- Printing: indeed printing pages with math is likely to happen!
- Multicol: Wikipedia uses that for the "References" section at the
bottom of articles and some pages actually do use math inside!
- SVG: definitely needed for math in graphics. This is no longer an
issue now that foreignObject was implemented in LayoutNG.
- Editing: I'm not sure what's the status of EditingNG but I
believe we
wanted to disable editability on MathML anyway due to complexity for
editing math, which will require some tree fixup. This is better
handled
by other tools.
I don't remember what was the recent status, but we definitely
need to
check again. I believe we have crash tests in WPT and another to
check
correct layout for math-in-svg.
On 23/06/2022 13:06, Morten Stenshorne wrote:
> Exciting times!
>
> Since the layout implementation of MathML is pure NG (i.e. it has no
> legacy engine counterpart), you should be aware that this is
somewhat
> uncharted territory. We *do* have mechanisms that kick in when
we need
> to fall back to legacy layout (e.g. when printing, or when there are
> tables inside multicol), and have NG-only content inside (such as
> MathML), but no pure NG features have shipped yet, so we don't
really
> know how well this works on the world wild web.
>
> That said, we're also about to ship container queries, which has the
> same problem - so you're not alone. I was hoping that we'd be
rid of the
> legacy engine completely before we started shipping pure NG
features,
> because of this risk. I'm not saying that we definitely need to
wait for
> the legacy engine to be completely gone, but you need to be
aware of the
> risk, at least. Maybe clusterfuzz will find something
interesting once
> this feature is switched to "stable". :)
>
> By the way, I just tested, and MathML inside legacy layout
doesn't seem
> to work. I filed
> https://bugs.chromium.org/p/chromium/issues/detail?id=1338882
>
> Frédéric Wang <fw...@igalia.com> writes:
>
>> Contact emails
>>
>> fw...@chromium.org, rb...@chromium.org, bkard...@igalia.com
>>
>> Explainer
>>
>>
https://github.com/mathml-refresh/mathml-core/blob/master/docs/explainer.md
>> https://people.igalia.com/fwang/explainer-font-family-math
>>
>> Specification
>>
>> https://mathml-refresh.github.io/mathml-core
>>
>> Design docs
>>
>> https://drafts.csswg.org/css-fonts-4/#valdef-font-family-math
>> https://w3c.github.io/mathml-aam
>>
https://docs.google.com/document/d/1biGEaWN8ThNTDtAbT1M5GIf6N5uQLWdxh2QhrG9uN5c
>>
https://docs.google.com/document/d/1bvY_Npe2zLW_705KXdmecsH6P9I9wBMpHRsZ9CxWNOI/edit#heading=h.u9hwm9tp8nuy
>>
>> Summary
>>
>> Adds native DOM/IDL, accessibility support and CSS-compatible
OpenType-based layout of mathematical formulas. The reference
>> specifications are MathML Core (which describes in extensive
details a fundamental subset of the MathML 3 recommendation),
MathML AAM
>> (which describes mapping to platform accessibility APIs). This
includes the "math" font-family as well as features included in
section "CSS
>> Extensions for Math Layout".
>>
>> Blink component
>>
>> Blink
>>
>> Search tags
>>
>> mathml
>>
>> TAG review
>>
>> https://github.com/w3ctag/design-reviews/issues/438
https://github.com/w3ctag/design-reviews/issues/313#issuecomment-460523527
>>
>> TAG review status
>>
>> Issues addressed
>>
>> Risks
>>
>> Interoperability and Compatibility
>>
>> - This is not implementing the complete list of MathML3,
OpenType MATH or accessibility features that one can find in other
browsers.
>>
>> - New enhancements (e.g. clarification regarding CSS
interoperability) may not be implemented in other browsers yet.
>>
>> - MathML Core and MathML AAM are still working drafts.
>>
>> Gecko: Shipped/Shipping
>>
>> WebKit: Shipped/Shipping
>>
>> Web developers: Positive
(https://twitter.com/search?q=mathml%20chrome&f=live
<https://twitter.com/search?q=mathml%20chrome&f=live>) Users have
recently been excited about it,
>> experimented exiting implementation under a flag and are eager
to see it shipped soon. It is currently the 5th most starred Blink
issue, see
>>
https://bugs.chromium.org/p/chromium/issues/list?can=2&q=component:Blink&sort=-stars&colspec=ID%20Stars%20Pri%20Status%20Component%20Opened%20Summary
<https://bugs.chromium.org/p/chromium/issues/list?can=2&q=component:Blink&sort=-stars&colspec=ID%20Stars%20Pri%20Status%20Component%20Opened%20Summary>
>>
>>
>> Other signals:
>>
>> Ergonomics
>>
>> In the past, MathML trees were not treated specially at all and
this was fast ! Now, more work is needed for pages that do math
rendering
>> e.g. accessing data from the MATH table, handling operator
dictionary, using ink text bounds, performing low-level shaping of
stretchy
>> operators, loading and applying CSS UA sheet, or exposing an
accessibility tree. However, the only performance report we
received so far
>> was issue #1073760, which was discarded after further analysis.
In any case, performance is definitely better than existing non-native
>> technologies. Additionally, pages will likely need loading WOFF
fonts on operating systems that don't ship math fonts. Work is in
progress
>> here, see
https://frederic-wang.fr/update-on-open-type-math-fonts.html for
recent status.
>>
>> Activation
>>
>> - Editing MathML source is difficult but special authoring
tools & converters do exist.
>>
>> - Implementation may not be as complete as expected by users
but the spec tries to allow extensibility and the Math WG provides
polyfills
>> at https://github.com/mathml-refresh/mathml-polyfills
>>
>> - Users may require WOFF fonts to properly render formulas (see
above).
>>
>> Security
>>
>> There is a risk due to the new attack surface:
>>
>> - Relatively large amount of new code.
>>
>> - Newly exposed APIs.
>>
>> - Interacting with several parts of the web platform.
>>
>> However, risks related to violation of layout algorithm are
reduced now that the specification clarifies how to perform it in
a CSS-compatible
>> way. We've fixed 17 issues reported during 2.5 years of
development and there are currently 0 open security issues (last
report was in
>> March).
>>
>> Finally, similar security & privacy considerations as existing
layout specifications apply, see the relevant sections from the
MathML Core
>> spec:
https://w3c.github.io/mathml-core/#security-considerations
https://w3c.github.io/mathml-core/#privacy-considerations
>>
>> WebView application risks
>>
>> Does this intent deprecate or change behavior of existing APIs,
such that it has potentially high risk for Android WebView-based
>> applications?
>>
>> No.
>>
>> Debuggability
>>
>> - Nothing particular expected besides what is already provided
for SVG and HTML.
>>
>> - Existing devtools API was extended to handle "font-family:
math", see
>>
https://chromedevtools.github.io/devtools-protocol/tot/Page/#method-setFontFamilies
>>
>> Will this feature be supported on all six Blink platforms
(Windows, Mac, Linux, Chrome OS, Android, and Android WebView)?
>>
>> Yes
>>
>> Is this feature fully tested by web-platform-tests?
>>
>> Yes
>>
>> Flag name
>>
>> MathMLCore
>>
>> Requires code in //chrome?
>>
>> True (this is only to implement support for “font-family:
math”, which has a user preference menu for math font, see
>>
https://docs.google.com/document/d/1bvY_Npe2zLW_705KXdmecsH6P9I9wBMpHRsZ9CxWNOI/edit#heading=h.u9hwm9tp8nuy).
>>
>> Tracking bug
>>
>> http://crbug.com/6606
>>
>> Sample links
>>
>>
https://people.igalia.com/fwang/2022-06-igalia-week-shipping-mathml-in-chromium/#/12
>> https://fred-wang.github.io/MathFonts
>>
>> Estimated milestones
>>
>> DevTrial on desktop 103
>>
>> DevTrial on Android 103
>>
>> Anticipated spec changes
>>
>> Open questions about a feature may be a source of future web
compat or interop issues. Please list open issues (e.g. links to
known github
>> issues in the project for the feature specification) whose
resolution may introduce web compat/interop risk (e.g., changing
to naming or
>> structure of the API in a non-backward-compatible way).
>>
>> https://github.com/w3c/csswg-drafts/issues/5866
>> https://github.com/w3c/mathml-core/issues/74
>> https://github.com/w3c/mathml-core/issues/75
>> https://github.com/w3c/mathml-core/issues/76
>> https://github.com/w3c/mathml-core/issues/77
>>
>> Link to entry on the Chrome Platform Status
>>
>> https://chromestatus.com/feature/5240822173794304
>>
>> Links to previous Intent discussions
>>
>> Intent to prototype:
https://groups.google.com/a/chromium.org/g/blink-dev/c/OOZIrtSPLeM
>>
>> This intent message was generated by Chrome Platform Status.
>> --
>> Frédéric Wang
--
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 blink-dev+unsubscr...@chromium.org
<mailto:blink-dev%2bunsubscr...@chromium.org>.
To view this discussion on the web visit
https://groups.google.com/a/chromium.org/d/msgid/blink-dev/37ca5f43-60a8-54a6-e2e6-f781fe8c71bd%40igalia.com.
--
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 blink-dev+unsubscr...@chromium.org.
To view this discussion on the web visit
https://groups.google.com/a/chromium.org/d/msgid/blink-dev/354fba5f-de49-7e9a-dc91-440108e17908%40igalia.com.
Title: Contenteditable MathML
