🌐 US-Proxy
class="logged-out env-production page-responsive" style="word-wrap: break-word;" >
Skip to content

[css-text] Consider removing distribute value for text-justify from the spec #7322

Closed

Description

@nt1m

Currently there is this note in this section:

For legacy reasons, UAs must also support the alternate keyword distribute which must compute to inter-character, thus having the exact same meaning and behavior.

Looking at browser support for text-justify:

  • Chrome only supports the property behind the experimental web features flag
  • IE (which used to support this value) is now retired
  • Safari has never shipped text-justify

...which leaves Firefox being the only browser supporting the CSS property.

I suggest that we remove this note from the spec, since there is no clear win in web compatibility in supporting this.

Activity

  1. nt1m commented on May 27, 2022

    @nt1m
    MemberAuthor

    cc @emilio @dbaron Wdyt?

  2. nt1m commented on May 27, 2022

    @nt1m
    MemberAuthor
  3. fantasai commented on May 31, 2022

    @fantasai
    Collaborator

    Note that justification doesn't break pages in the way that, say, a layout bug would, so the compat impact of not having this value is going to be less than breaking/dropping any kind of layout behavior. But it's also very cheap to implement, being just an alias.

    This value was in a CR-level spec in the past and supported by the most dominant browser of the time, so I'm hesitant to drop it if there's not either a compelling reason or substantial proof that it's not used by some long tail of existing CJK content. https://www.w3.org/TR/2003/CR-css3-text-20030514/#justification-prop

  4. fantasai commented on May 31, 2022

    @fantasai
    Collaborator
  5. karlcow commented on May 31, 2022

    @karlcow
    Member

    Following the discussion in #6156

  6. nt1m commented on May 31, 2022

    @nt1m
    MemberAuthor

    Note that justification doesn't break pages in the way that, say, a layout bug would, so the compat impact of not having this value is going to be less than breaking/dropping any kind of layout behavior. But it's also very cheap to implement, being just an alias.

    The spec (and relevant WPT) specifies it as a parse-time alias, which while not hard to implement, complicates the code unnecessarily, since WebKit does not have such a mechanism (you need to create a custom parsing method, and can't re-use the fast paths). Compute-time is fine though I don't see the benefit of supporting it. Fwiw, Chrome only implements it compute-time.

  7. kojiishi commented on May 31, 2022

    @kojiishi
    Contributor

    Edge has degraded this when it moved to Chromium. If Edge is fine, I think it's a good sign we can remove this safely. /cc @dlibby-

  8. 19 remaining items

  9. frivoal commented on Dec 28, 2022

    @frivoal
    Collaborator

    Hi @nt1m
    We resolved to keep the distribute value in the spec, but it can be implemented either by computing to inter-character (which Chrome) or by treating it as a parse-time alias (as Gecko does). Would appreciate if you can confirm this is acceptable to you.

  10. nt1m commented on Dec 28, 2022

    @nt1m
    MemberAuthor

    Should probably be fine to keep it.

  11. frivoal commented on Dec 29, 2022

    @frivoal
    Collaborator
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions