Re: Unicode request for CSS property to control emoji style [I18N-ACTION-500]

On Fri, Mar 25, 2016 at 6:07 AM, Tab Atkins Jr. <jackalmage@gmail.com>
wrote:

> On Thu, Mar 24, 2016 at 2:01 PM, Phillips, Addison <addison@lab126.com>
> wrote:
> > The Unicode Consortium, via its liaison to us (Mark Davis, copied),
> would like I18N to request a new CSS property to control emoji
> presentation. You can see the request in [1]. The I18N WG considered this
> request in our teleconference of 24 March [2] and I've been actioned to
> follow up with your WG. We would like to encourage Mark and/or the Unicode
> emoji subcommittee to work actively with you on this request.
> >
> > Their specific request can be found at [3]. To summarize briefly, emoji
> characters come in two flavors. Some characters most frequently are used as
> "normal" (non-emoji) textual characters, but sometimes are used as emoji
> also. These characters have a default display of 'text'. Some characters
> most frequently are used as emoji characters, but occasionally are used in
> a plain text context. These characters have a default display of 'emoji'.
> Additional details and illustration can be found in the proposed update to
> UTR#51 [4]
> >
> > Unicode is requesting a property that would allow sequences of emoji
> characters to be styled in one of three ways:
> >
> > default : style characters to use each character's default display
> > text : style characters to use their text display
> > emoji : style characters to use their emoji display
> >
> > If this request should be redirected or converted e.g. to a github
> issue, please let me know. Please note that the I18N WG is not itself
> tracking this as an issue at this time.
>
> This is definitely the right place.
>
> This was originally brought up back in 2014 by John Daggett, in
> <https://lists.w3.org/Archives/Public/www-style/2014Aug/0322.html>.
>
> The thread there drifted a bit to talk about multicolor fonts, but the
> initial mail from John was explicitly about selecting the
> "color"/"image" variant of an emoji character vs its "monochrome"
> version that was colored by the 'color' property.
>

What about calling it font-variant-emoji?

I agreed that it was useful at the time, and still agree with that
> now.  The important question is still hanging: what to do if a font
> has only an emoji version of a character?  Should we attempt to
> synthesize a monochrome variant, or just do fallback?  I guess
> "fallback" is the right answer, since we're essentially rendering a
> different variant character, right?
>

Should we do fallback here? I'd propose we just behave as if there is a
variation selector attached, and let the font do whatever it wants.

It reveals another question, what should we do if the text have already
explicitly included the variation selector? Should we always override it or
only do so for characters without the selector?

- Xidorn