W3C home > Mailing lists > Public > www-style@w3.org > February 2014

Re: [css-counter-styles] Using character as unit is ambiguous

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Mon, 24 Feb 2014 14:19:07 -0800
Message-ID: <CAAWBYDDh1mdno7x8AhMK2x8ebmyKijj3UhBYJdFzd9g1pVmTew@mail.gmail.com>
To: Xidorn Quan <quanxunzhen@gmail.com>
Cc: www-style list <www-style@w3.org>
On Fri, Feb 21, 2014 at 9:53 PM, Xidorn Quan <quanxunzhen@gmail.com> wrote:
> In the current draft, "character" is used as a unit to describe the
> length limitation user agents could apply for some systems. However,
> as Simon Sapin mentioned in another thread, "character" is an
> ambiguous unit, which could refer to either Unicode code points or
> grapheme clusters.
>
> Things could even get more complex when we start discussing the exact
> meaning of "character" here. I believe what you meant via this word
> was in fact grapheme clusters. But if it means grapheme clusters, the
> limitation will be completely useless, since a grapheme cluster could
> contain arbitrary number of Unicode code points. However, if the
> "character" here means Unicode code points, 20 code points might not
> be enough for some counter styles, as a symbol may contains some
> cluster extenders to describe one grapheme.
>
> I suggest to allow 60-100 Unicode code points instead.

Good catch.  Changed it to 60 codepoints.

~TJ
Received on Monday, 24 February 2014 22:20:00 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:19 UTC