- From: Jonathan Kew <jonathan@jfkew.plus.com>
- Date: Tue, 22 Jun 2010 12:41:16 +0100
- To: John Daggett <jdaggett@mozilla.com>
- Cc: Bram Pitoyo <brampitoyo@gmail.com>, www-font <www-font@w3.org>
On 22 Jun 2010, at 07:11, John Daggett wrote: >>> […] proposed revised syntax: >>> font-feature-settings: cpsp, pkna(0); >> >> The proposed syntax would make the CSS behavior more consistent with >> the existing font-variant property draft. Do you know if >> dash-separated "friendly name" would be allowed here? >> >> So our proposed example could be typed longhand as: >> font-feature-settings: capital-spacing, proportional-kana(0) > > That would be nice but there's no standard for that and nothing > in the font data that contains a name like that. I think for > something like this to work CSS would need to define a table of > friendly names to feature tags. We already have (in the draft) the font-variant-* properties to specify features that are explicitly defined in CSS, and are expected to be mapped to any available rendering technology that supports them. The purpose of font-feature-settings is to allow authors access to *any* features that may happen to be available in their fonts, without requiring CSS to be aware of them. As such, it is essential to allow arbitrary tags to be specified and passed to the rendering system. It seems to me that the distinction between standardized properties defined in CSS (expressed via font-variant-* subproperties with friendly names) and low-level access to the "raw" font technology is an important one. Adding a mapping of friendly names to tags for use in font-feature-settings would tend to blur this distinction in a way that I don't think is helpful. JK
Received on Tuesday, 22 June 2010 11:42:02 UTC