W3C home > Mailing lists > Public > www-style@w3.org > November 2013

Re: [css-fonts] Stylistic alternates

From: Chris Lilley <chris@w3.org>
Date: Sun, 10 Nov 2013 04:42:26 +0100
Message-ID: <1091870587.20131110044226@w3.org>
To: James Clark <jjc@jclark.com>
CC: "www-style@w3.org" <www-style@w3.org>
Hello James,

Sunday, October 27, 2013, 4:51:24 AM, you wrote:

> I don't understand how these are supposed to work in the current draft.


> At the font level, there's a "salt" feature, implemented by a type
> 3 (alternate substitution) lookup. For each glyph id, this will
> result in an ordered list of alternate glyph ids.

Right.

>  The usual
> behavior in an editing environment is that the UI allows the user
> manually to select the appropriate alternate.

Yes, agreed.

> An important point is that alternate N for glyph X does not
> necessarily have any stylistic relationship to alternate N for glyph
> Y: the effect of a particular alternate index is glyph-specific.

That is also correct.


> So the kind of thing I would want to do is to define stylistic
> feature value names such as:

> - horizontal-base: use alternate #1 for "f" and  #1 for "r"  
> - curve-base: use alternate #2 for "i", #1 for "j" and #3 for "r"

Right, you can't do that; for an entire paragraph or set of
paragraphs, to set what you want would require a style rule with a
large set of character-to-glyph-alternate mappings.

But you would also, quite often, not want to use the same alternate
all the time. On a case by case basis you would use one alternate on
one word and a different alternate for the same letter in a different
word. Its difficult to see how that would be done just in a style
rule; it really needs to be tied to the document structure.

An editing environment would likely persist those sort of case by case
choices by generating child elements around groups of letters (span
elements for example, in html) with a style rule which applies just to
that span, or all spans of that class.

> If I have understood the current draft correctly, it only allows
> the definition of a stylistic feature value name by using a
> particular alternate index uniformly for all glyphs. This doesn't
> seem to me to be quite what is needed.

Its not suitable for large stretches of text, certainly. The font
family would be set on the larger elements and stylistic alternates
set on smaller runs of text (even on single letters, for the finest
control) as required.




-- 
Best regards,
 Chris                            mailto:chris@w3.org
Received on Sunday, 10 November 2013 03:42:23 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:36 UTC