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

RE: [css3-fonts] humane 'unicode-range'

From: Phillips, Addison <addison@lab126.com>
Date: Sun, 1 May 2011 11:15:31 -0700
To: Koji Ishii <kojiishi@gluesoft.co.jp>, John Daggett <jdaggett@mozilla.com>, Christoph Päper <christoph.paeper@crissov.de>
CC: CSS WWW Style <www-style@w3.org>, "CJK discussion (public-i18n-cjk@w3.org)" <public-i18n-cjk@w3.org>
Message-ID: <131F80DEA635F044946897AFDA9AC3476A92A8B718@EX-SEA31-D.ant.amazon.com>
Hi Koji,

I don't think adding this to "unicode-range" makes sense. The 'unicode-range' selector is one specific, effective mechanism for one specific use--selecting font based on a range of Unicode code points.

If we want to provide for selection according various Unicode properties, we should provide a property selector.

Addison

Addison Phillips
Globalization Architect (Lab126)
Chair (W3C I18N WG)

Internationalization is not a feature.
It is an architecture.


> -----Original Message-----
> From: public-i18n-cjk-request@w3.org [mailto:public-i18n-cjk-request@w3.org]
> On Behalf Of Koji Ishii
> Sent: Sunday, May 01, 2011 4:35 AM
> To: John Daggett; Christoph Päper
> Cc: CSS WWW Style; CJK discussion (public-i18n-cjk@w3.org)
> Subject: RE: [css3-fonts] humane 'unicode-range'
> 
> This is in reply to a little old thread as this was marked as an issue in the
> spec[1].
> 
> One use case that came up in my mind is to switch fonts for UAX #11 [2]
> EAW=A code points.
> 
> EAW=A contains mostly punctuation and symbols that were unified. I think
> there are cases where authors want:
> * CJK fonts for EAW=A|F|H|W
> * Latin fonts for EAW=N|Na
> 
> One famous example for EAW=A is U+2026 HORIZONTAL ELLIPSIS; ellipsis are
> at baseline in Latin fonts while ellipsis at the vertical center in CJK fonts. If I
> were writing Japanese documents, I expect it be drawn at the vertical center.
> 
> It would be great if I can use a font like this:
> 
> @font-resource {
>   font-family: myfont;
>   src: local(CJK-font-name);
> }
> 
> @font-resource {
>   font-family: myfont;
>   src: local(Latin-font-name);
>   unicode-range: EAW=N|Na;
> }
> 
> Or to do this in opposite way:
> 
> @font-resource {
>   font-family: myfont;
>   src: local(Latin-font-name);
> }
> 
> @font-resource {
>   font-family: myfont;
>   src: local(CJK-font-name);
>   unicode-range: EAW=A|F|H|W, U+5C;
> }
> 
> Thoughts?
> 
> [1] http://dev.w3.org/csswg/css3-fonts/#unicode-range-desc

> [2] http://unicode.org/reports/tr11/

> 
> 
> Regards,
> Koji
> 
> -----Original Message-----
> From: www-style-request@w3.org [mailto:www-style-request@w3.org] On
> Behalf Of John Daggett
> Sent: Thursday, June 17, 2010 6:06 PM
> To: Christoph Päper
> Cc: CSS WWW Style
> Subject: Re: [css3-fonts] humane 'unicode-range'
> 
> 
> 
> Christoph Päper wrote:
> 
> > Would it make sense to add keywords to ‘unicode-range’ or to add
> > another font-decriptor with similar purpose to identify scripts or
> > writing systems (i.e. script + language) supported by (and requested
> > to be used from) the font resource?
> 
> There was a discussion of this last year:
> http://lists.w3.org/Archives/Public/www-style/2009May/0212.html

> 
> I think it boiled down to whether to use names from the Unicode database or
> not and if so, how specific the names should be.  I think the idea of named
> ranges for the 'unicode-range' descriptor is interesting.
> I don't like the idea of additional descriptors for this (e.g. script-coverage, etc.).
> 
> It would help to have a clearer idea of the use case you imagine for this to be
> able to judge whether named ranges are better than simple ranges.
> 
> One other use I can imagine is for easily dividing up large CJK fonts into defined
> character ranges.
> 
> Ex:
> 
>   unicode-range: jis-level-1, jis-level-2; /* alias for the set of codepoints in the
> JIS Level 1 and 2 ranges */
> 
> JIS Level 1 characters occur more commonly than JIS Level 2, etc., so this would
> be a convenient way for font vendors to package fonts so that fonts containing
> infrequently used characters were only downloaded in fallback situations.
> 
> No matter what the set of aliases, I think we can only practically reference
> ranges that are defined clearly and have a standard reference point such as the
> Unicode database.
> 
> Cheers,
> 
> John Daggett

Received on Sunday, 1 May 2011 18:16:02 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:40 GMT