W3C home > Mailing lists > Public > www-style@w3.org > February 2011

Re: [CSS3-UI] text-overflow:ellipsis (freshly rewritten/expanded and incorporated into editor's draft)

From: João Eiras <joao-c-eiras@telecom.pt>
Date: Mon, 28 Feb 2011 16:58:38 +0000
To: "www-style@w3.org" <www-style@w3.org>
Message-ID: <201102281658.39083.joao-c-eiras@telecom.pt>

> > I strongly suggest including <string> values in the specification. Else, it is not multi lingual.
> Thanks for your suggestion.
> Absent data I am considering this a theoretical request.

Oh, there is plenty of data. I've shown three examples on the previous email. e.g. latin ellipsis is different in thai, laotian or chinese scripts. Not theoretical at all.

And given that I do not understand languages that use those scripts, I cannot copy paste an example from real life websites, but I can give you this reference. http://en.wikipedia.org/wiki/Ellipsis#Computer_representations

> But no one implemented <string> values. For over 7 years.
> That's why it got dropped when the feature was moved to CSS3 UI.
> Before adding this feature to CSS4 UI, you will need to justify
> how/why implementations will consider it differently than they have
> for the past 7+ years (they ignored it). Otherwise it is pointless to
> spend the time writing it up.

Although lack of implementations sometimes drives specifications, but serious accessibility or internationalization features need to be keep and forced upon the implementors.

e.g.: the acid tests did a good job, and test suites over all do, of asking user agents to align better with certain specifications.

> > Many non latin scripts use something else than 3 dots for the same effect as an ellipsis.
> The current editor's draft allows implementations to use something
> other than 3 dots:
> "Implementations may substitute a more language/script-appropriate
> ellipsis character."
> from http://dev.w3.org/csswg/css3-ui/#text-overflow

If you want to specify the language detection algorithm, fine. But it would be error prone or if miscontextualized, would be useless (e.g. mixing different scripts). But usually, authors know better. So detecting language would remain just a recommendation.

> FAQ recorded:
> http://wiki.csswg.org/spec/css3-ui#text-overflow-for-non-latin-scripts
> If you have suggestions of specific non-latin scripts and their
> respective something else other than 3 dots for the same effect as an
> ellipsis, please provide them with citations to style guides for those
> non-latin scripts so that we may provide more precise implementation
> guidance.

Above (and mentioned in the previoous e-mail).

> > A single 3-dot ellipsis is also not a bullet proof solution. The author might wish to use "(...)" for instance in citations, or use the vertical ellipsis in maths.
> Recorded as theoretical use-cases.
> http://wiki.csswg.org/spec/css4-ui#text-overflow-string
> Please provide URLs of real world examples on the public web of
> "(...)" in citations, or vertical ellipsis in maths in order to
> upgrade this from theoretical to real-world based.

Not theoretical. The proper way to omit parts of quotes is to use "(...)" or "[...]". Can't quickly find a grammar book online to give you that evidence.

> Note that this suggested syntax is an abuse of HTML5 data-*
> attributes, as those are supposed to be site/page specific and NOT
> used in any standards, formats etc.

data-* would be site specific, used only as example. text-overflow and attr() are standard css constructs.
Received on Monday, 28 February 2011 16:59:37 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 16:28:23 UTC