Re: [css3-fonts] truetype collection

Dear colleagues,

I think that we can agree on the following things:

- It would be nice if TTC can be supported by CSS Fonts.

- On top of a reference to a TTC, we have to specify a font name for
    locating a specific font in the given TTC.

- Fragment identifiers look promising as a means for specifying
  font names.

- Font names as fragment identifiers should be documented
  as part of the registration of the media type for OpenType.  (See RFC 4288)

The registration of a media type for OpenType has been discussed for
years and nothing has been registered.  Several private media types have been
widely used.  I do not know if there is a plan (rather than a hope) for
registration.

I don't know when CSS Font Level 3 will become a CR, but I think
it should should explain the current status and stop there.

Regards,
Makoto

2012/5/14 Koji Ishii <kojiishi@gluesoft.co.jp>:
>> From: John Hudson [mailto:tiro@tiro.com]
>> The typical way that software identifies one of the fonts in a given TTC is by font name
>> table entries, i.e. the same way software identifies individual TTF or OTF fonts. So, for
>> example, an installed TTC such as that which ships as part of the Microsoft Cambria family
>> displays in font menus as two distinct fonts based on the distinct name tables for Cambria
>> Regular and Cambria Math.
>>
>> I can see how this will require appropriate media type for CSS, but once that is in place does
>> one need a 'fragment identifier' beyond the name table entries?
>
> Fragment is needed because src:url() doesn't have syntax to specify by name.
>
> If we were doing this, I support John Daggett's original idea to specify by ordinals. We could use names in fragment, but it will require UAs to scan all fonts in the TTC.
>
>
> Regards,
> Koji
>



-- 

Praying for the victims of the Japan Tohoku earthquake

Makoto

Received on Friday, 1 June 2012 23:52:58 UTC