Intent to Ship: CSS Automatic Hyphenation for Windows/Linux/ChromeOS

[Koji Ishii]

Contact emails

ko...@chromium.org

Specification

https://drafts.csswg.org/css-text-3/#propdef-hyphens

Summary

Blink supports `manual` and `none` values of the CSS `hyphens` property since M55, but `auto`, which enables the automatic hyphenation, is supported only on Android and Mac. This feature enables `hyphens: auto` on Windows, Linux, and ChromeOS by using dictionaries used in Android Open Source Project. The previous feature entry for M55 is https://www.chromestatus.com/features/5642121184804864

Blink component

Blink

TAG review

None

TAG review status

Not applicable

Risks

Interoperability and Compatibility

I don't think there are interoperability and compatibility risks because Blink is already shipping in 2 platforms, and also other browsers are already shipping.

Gecko: Shipped/Shipping
WebKit: Shipped/Shipping
Web developers: Strongly positive, the tracking bug has 339 stars.

Will this feature be supported on all six Blink platforms (Windows, Mac, Linux, Chrome OS, Android, and Android WebView)?

Yes. Android and Mac are already shipping in M55.

Is this feature fully tested by web-platform-tests?

Yes

Tracking bug

https://bugs.chromium.org/p/chromium/issues/detail?id=652964

Launch bug

https://bugs.chromium.org/p/chromium/issues/detail?id=652964

Link to entry on the Chrome Platform Status

https://www.chromestatus.com/feature/5672891947417600

[Yoav Weiss]

LGTM1

--
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+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACQRE%2BSLsj8W9YAu_7%3DxJUqH%2BzYpPM0tAx4dqKmGa4QdPKJ%2B3A%40mail.gmail.com.

[Thomas Steiner]

Non-owner LGTM. Thanks for finally shipping this useful feature! 

--
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+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACQRE%2BSLsj8W9YAu_7%3DxJUqH%2BzYpPM0tAx4dqKmGa4QdPKJ%2B3A%40mail.gmail.com.

--

Thomas Steiner, PhD—Developer Advocate (https://blog.tomayac.com, https://twitter.com/tomayac)

Google Germany GmbH, ABC-Str. 19, 20354 Hamburg, Germany
Geschäftsführer: Paul Manicle, Halimah DeLaine Prado
Registergericht und -nummer: Hamburg, HRB 86891

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.1.23 (GNU/Linux)

iFy0uwAntT0bE3xtRa5AfeCheCkthAtTh3reSabiGbl0ck0fjumBl3DCharaCTersAttH3b0ttom.hTtPs://xKcd.cOm/1181/
-----END PGP SIGNATURE-----

[Mike West]

LGTM2. Aligning our behavior across platforms is a net positive for developers; doubly so when we're also aligning with other vendors.

-mike

[Daniel Bratell]

LGTM3

/Daniel

[Matt Menke]

Are different dictionaries used based on user language/locale, or is there a single universal dictionary?  Just wondering if this is yet another way for sites to deduce information about a user's localization settings.

[Koji Ishii]

Are different dictionaries used based on user language/locale, or is there a single universal dictionary?  Just wondering if this is yet another way for sites to deduce information about a user's localization settings.

Different dictionaries based on the content locale. We will use the same engine and dictionaries as Android Open Source Project.

[fantasai]

> <https://android.googlesource.com/platform/external/hyphenation-patterns/+/refs/heads/master>.

Are you enabling hyphenation on untagged content, or only on language-tagged
content (per spec)?

~fantasai

[Koji Ishii]

Are you enabling hyphenation on untagged content, or only on language-tagged
content (per spec)?

There's an issue filed for that, please see crbug.com/676270.

[fantasai]

> <http://crbug.com/676270>.

I see. Then, I don't think this should be shipped until this bug is fixed.

It is not only a spec violation but
* It creates bugwards-compat pressure when it does.
* It exacerbates existing non-interop among browsers.
* It can cause readability problems if the heuristic
doesn't match the actual content language.
* Because it depends on the system configuration,
it also causes interop problems between users using
the *same browser* on differently-configured systems!
* Lastly, using the system language in such cases adds
additional fingerprinting surface.

~fantasai

[PhistucK]

While I do not see the point of blocking shipping expanded support for this feature due to this bug (since most users are probably on Android anyway, it would not matter much), I do think it should be highly prioritized.

☆PhistucK

--
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+...@chromium.org.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/3f3e4c73-fc50-ac08-8e66-b787415d1415%40inkedblade.net.

[Yoav Weiss]

I agree this should not be a blocker for this intent, which is merely extending platform support for the feature.

At the same time, it would be good to fix such compatibility issues.

[j.j.]

Am Donnerstag, 22. Oktober 2020 10:43:48 UTC+2 schrieb Yoav Weiss:

I agree this should not be a blocker for this intent, which is merely extending platform support for the feature.

I think crbug.com/676270 should block shipping. Author may start using this feature for desktop sites and have expected results without a lang attribute in Chromium browsers (only). When that  bug is fixed, they have to change their code.

j.j.

 

On Thu, Oct 22, 2020 at 10:23 AM PhistucK <phis...@gmail.com> wrote:

While I do not see the point of blocking shipping expanded support for this feature due to this bug (since most users are probably on Android anyway, it would not matter much), I do think it should be highly prioritized.

☆PhistucK

On Thu, Oct 22, 2020 at 6:00 AM fantasai <fantas...@inkedblade.net> wrote:

On 10/19/20 11:43 PM, Koji Ishii wrote:
 > > Are you enabling hyphenation on untagged content, or only on language-tagged
 > > content (per spec)?
 >
 > There's an issue filed for that, please see crbug.com/676270
 > <http://crbug.com/676270>.

I see. Then, I don't think this should be shipped until this bug is fixed.

It is not only a spec violation but
  * It creates bugwards-compat pressure when it does.
  * It exacerbates existing non-interop among browsers.
  * It can cause readability problems if the heuristic
     doesn't match the actual content language.
  * Because it depends on the system configuration,
     it also causes interop problems between users using
     the *same browser* on differently-configured systems!
  * Lastly, using the system language in such cases adds
     additional fingerprinting surface.

~fantasai

--
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 blin...@chromium.org.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/3f3e4c73-fc50-ac08-8e66-b787415d1415%40inkedblade.net.

To unsubscribe from this group and stop receiving emails from it, send an email to blin...@chromium.org.

[Koji Ishii]

I think we'll look at it soon.

Blink and WebKit can apply incorrect dictionaries for the lack of the feature today, but Gecko doesn't do so. I agree it's a good feature. Thank you for helping to make Chromium better.

2020年10月23日(金) 2:28 j.j. <fri...@jeka.info>:

To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/53bbdc1b-77ad-4092-92cc-7dcd7e60b744o%40chromium.org.

[Koji Ishii]

We've fixed crbug.com/676270 for M88, the fix is available in Canary 88.0.4313.0. Thank you for the feedback.

2020年10月23日(金) 22:57 Koji Ishii <ko...@chromium.org>: