RE: SVG Fonts inside of OpenType fonts? [Cross-post from www-font@w3.org]

On Thursday, June 30, 2011 7:00 PM Cameron McCormack wrote:
> 
> Levantovsky, Vladimir:
> > I am not sure what you mean by content. Would plain sequence of
> > Unicode codepoints be considered a content?
> 
> I think Alex means the content of the <glyph>, i.e. you can have
> 
>   <svg …>
>     <defs>
>       <path id="a" d="M …"/>
>     </defs>
>     <font>
>       <glyph>
>         <use xlink:href="#a"/>
>         …
>       </glyph>
>     </font>
>   </svg>
> 

Okay, thanks.

> 
> In cases where you have a complex script that isn’t supported directly
> by SVG Font’s ligature and Arabic form features, then you can use
> <altGlyph> to select an explicit glyph to use for a run of Unicode
> characters.  For example:
> 
>   <font>
>     <glyph id="complex">
>       …
>     </glyph>
>   </font>
>   <text>The <altGlyph xlink:href="#complex">xyzzy</altGlyph>
>     glyph.</text>
> 

But this would require:
a) that you know the text content upfront;
b) that you know the specific details of the language script in question and know how to shape it, and
c) that all shaping/layout decisions can only be done at the authoring time.

I can't see how this would work if your text is dynamic and comes from an outside source (e.g. via content aggregation or if it's simply a user comment on the post you made) formatted as a Unicode string. One way it would work is if you can use SVG glyph descriptions that are tied to OpenType glyph indices that would allow using the already existing shaping / layout engines, which is the answer to what

Alex Danilo wrote:
> I don't see that shoving SVG Fonts into an OpenType container does
> anything  more
> than force us to restrict them in arbitrary ways and so seems a bit
> silly.

Thank you and regards,
Vlad