Intent to Ship: Unprefix CSS 'line-break' property

26 views
Skip to first unread message

Koji Ishii

unread,
Jun 28, 2017, 3:28:06 AM6/28/17
to blink-dev, Quinten Yearsley
ko...@chromium.org,qyearsley@chromium.org https://drafts.csswg.org/css-text-3/#line-break-property The CSS 'line-break' property specifies the strictness of line-breaking rules. The exact definition of the strictness depends on languages. Our implementation relies on ICU (see doc link) which supports East Asian writing systems and Finnish.
This is a long-demanded feature primarily from Japan, where words are not separated by spaces and different line-breaking rules are strongly desired depends on the types/styles of the documents. Web developers using such styles consider not using 'strict' rules is a bug in the product.

Originally shipped in IE5.5 in July 2000, defined in CSS Text Level 3 and EPUB 3, and recently Safari unprefixed.
None. Firefox: Public support Edge: Shipped Safari: Shipped Web developers: Strongly positive (crbug.com/240344, Japanese stackoverflow)
None.
None.
Yes.

The actual supported writing systems and levels of the strictness depend on the ICU data. Android, or some distributions, may chose to reduce the size of ICU data. See crbug.com/306730#c17 for more details.
https://crbug.com/735539 https://www.chromestatus.com/features/5137645088800768
Yes.

Apologies in advance, but we have shipped the unprefixed version unintentionally in M58, and that this is a post facto intent due to our errors. More details in style-dev@ discussion.

The UMA counts 0.5% (or 0.4% for -webkit- prefix.)

TAMURA, Kent

unread,
Jun 28, 2017, 4:30:53 AM6/28/17
to Koji Ishii, blink-dev, Quinten Yearsley
LGTM1.
It's unfortunate that it was accidentally shipped.  However, it's worth to ship, and it seems some CLs to prevent us from such accidental ship are on-going.

--
You received this message because you are subscribed to the Google Groups "blink-dev" group.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACQRE%2BTW6CSiQson09VCoeT-YuJNFzhd-5ffQaKChrhD7hG1gQ%40mail.gmail.com.



--
TAMURA Kent
Software Engineer, Google


Mike West

unread,
Jun 28, 2017, 5:46:18 AM6/28/17
to TAMURA, Kent, Koji Ishii, blink-dev, Quinten Yearsley
Whoops!

LGTM2. It looks like https://chromium-review.googlesource.com/c/551877/ and https://chromium-review.googlesource.com/c/551400 will prevent us from repeating this mistake in the future, and I agree with tkent@ that it's worth shipping.

Given that Edge and Safari have already shipped this unprefixed, is there a bug against Firefox that y'all could ping to update them on Blink's support as well?

-mike

Koji Ishii

unread,
Jun 28, 2017, 5:58:10 AM6/28/17
to Mike West, TAMURA, Kent, Koji Ishii, blink-dev, Quinten Yearsley
On Wed, Jun 28, 2017 at 6:45 PM, Mike West <mk...@chromium.org> wrote:

Given that Edge and Safari have already shipped this unprefixed, is there a bug against Firefox that y'all could ping to update them on Blink's support as well?

3 years old bug here.​ I'll leave comments on our status, thank you for checking.

Rick Byers

unread,
Jun 28, 2017, 9:36:20 AM6/28/17
to Koji Ishii, Mike West, TAMURA, Kent, blink-dev, Quinten Yearsley, Alan Cutter, sh...@chromium.org
Thanks for catching this and doing the post-hoc intent.  Also thanks to Darren and Alan for quickly adding protection against this class of mistake.

What's the status of web-platform-tests for this feature?

Regardless, LGTM3

--
You received this message because you are subscribed to the Google Groups "blink-dev" group.

Koji Ishii

unread,
Jun 28, 2017, 9:53:39 AM6/28/17
to Rick Byers, Koji Ishii, Mike West, TAMURA, Kent, blink-dev, Quinten Yearsley, Alan Cutter, sh...@chromium.org
On Wed, Jun 28, 2017 at 10:35 PM, Rick Byers <rby...@chromium.org> wrote:
Thanks for catching this and doing the post-hoc intent.  Also thanks to Darren and Alan for quickly adding protection against this class of mistake.

What's the status of web-platform-tests for this feature?

Thank you for asking this, I missed to ​check that.

wpt has 556 tests created by Richard Ishida@W3C i18n WG. TestExpectations has 46 failures among them. Last I checked, some of them were caused by 1px rendering differences, while some needed more investigations, but I didn't record exact numbers.

​The I18N WG site has results for each browsers, ​but Chrome result is before this property is supported. I'll work with Richard to update the results on his site with this property shipping.
 
Regardless, LGTM3

​Thank you as always.​

Rick Byers

unread,
Jun 28, 2017, 12:02:45 PM6/28/17
to Koji Ishii, Mike West, TAMURA, Kent, blink-dev, Quinten Yearsley, Alan Cutter, sh...@chromium.org
Sounds good, thanks~  In particular the top priority is probably ensuring that Chrome passes the tests which Edge and Safari are passing (we definitely don't want to be the ones holding back interop).  Hopefully we'll have a dashboard soon that helps us easily find these cases :-)   (w3.org is 503ing for me at the moment).
Reply all
Reply to author
Forward
0 new messages