RE: Emoji Unicode

Thanks so much for such an informative reply Chris! I must reread it a few more times.

Adam Twardoch's proposal looks both intriguing and informed. (where are Cliff Notes when you need them?) The references on Unicode and emoji are very welcome.

I could imagine the glyph definitions of emoji becoming more relevant/operative/normative than the font definition, since the semantics of the Unicode things seem a bit distracting or vague: "green round wiggling face (though it need not be either green nor round nor wiggling and, btw, face-ness is optional and culturally relative)".

DD> the canonical set of
DD> semantic primitives might not consist of monochromatic path
DD> geometry alone, but rather of richer colored, textured and even
DD> animated objects.

CL>The semantic primitives are characters.
CL>In a visual representation, characters are represented by glyphs.

I see what you're saying, though it comes in and out of focus for me a bit. Maybe, I'm imagining that if we think of semantic primitives not as "characters" but as gestures, odors or essences, then my tendency (and I suspect I'm not the first) to confuse them with geometry may fade a bit. The term "character" brings with it the perhaps unfortunate legacy of geometry. If we could indeed define the canonical semantic primitives, then would they be just 100 [1] extra characters in a Unicode alphabet of 2^16 or 2^32? Perhaps, instead, the character definitions of those 100, and the resulting fonts, would produce projections from semantic geometry into the visual world and concurrently into other media as well. Characters, were, in the first case, recognizable geometries built to convey semantics, and because of the limitations of clay, ink and stylus, they were no more than geometry + representation = meaning. The user-interface of early writing systems was compromised because of limitations in the technology.

At any rate, I share your interest in knowing if anyone is working on SVG glyphs for an emoji font. There was a very intriguing group of presenters at SVG Open 2007 [2] who were looking at an  existing set of already "universally recognized symbols" and determining the degree to which those symbols might form a semantically complete system of expression. It was very cool.


[2] (SVuGy - Exploring the Space between Procedural and Declarative Graphics
Authors:  Prof. Dr. Martin J. Dürst, Associate Professor, Aoyama Gakuin University
Makoto Fujimori, Aoyama Gakuin University
Dr. Kazunari Ito, Assitant professor, Aoyama Gakuin University
Takeshi Maemura, Aoyama Gakuin University
Tohru Koga, Aomaya Gakuin University

-----Original Message-----
From: [] On Behalf Of Chris Lilley
Sent: Monday, October 03, 2011 3:04 PM
To: David Dailey
Cc: 'www-svg'
Subject: Re: Emoji Unicode

On Sunday, October 2, 2011, 2:49:49 AM, David wrote:

DD> Does anyone know how to get the Unicode version of Emoji
DD> characters (see for example or
DD> ) to actually display? Is there a font I’d have to install to see them?

As with all newly standardized characters, you need a font that supports them, at the correct code positions, to see them. Otherwise you will see the 'missing glyph'.

This faq on emoji and dingbats may be helpful:

specifically "there is no way based on character code alone to tell whether a character should be presented using an “emoji” style; that decision depends on context". ie styling.

DD> Are these characters available anywhere as an SVG font? TTF?

Apple has a (static, bitmapped) color emoji font in OS X Lion,
and this is gradually being supported by tools,eg

Microsoft has emoji font support in Windows Phone Mango (the new mobile OS).
See also this typophile thread

DD> Are the colors and gradients a part of the Unicode definition, or
DD> does Unicode merely encode path geometry (like WOFF)?


Unicode does not describe the glyph at all (for any character). They standardise the code position and the character properties (like whether it is upper case,whether it is a numeral, whether it is rtl or ltr or neutral directionality, etc).

The code charts do illustrate these with sample glyphs, but these are not normative. As an example, there is a sample glyph for the latin lowercase "a" but the actual glyph can be the curly one with a handle or the round one without a handle (or indeed whatever other formis appropriate).

in particular
and for the character properties, see

in UnicodeData.txt for example you will see


and the meaning of that is documented here

specifically, Someans it is "a symbol of other type",the "0" means that its combining class is "Not_Reordered: Spacing and enclosing marks; also many vowel and consonant signs, even if nonspacing",and so on. (the ;is a field separator,and the absence of a value means that the character has the default value).

DD> The Emoji character sets distributed through mobile phones appear
DD> to come with gradient definitions, implying perhaps another use case for SVG fonts over WOFF?

Or indeed SVG glyphs inside Opentype, served as WOFF. See Adam Twardoch's proposal at

DD> the canonical set of
DD> semantic primitives might not consist of monochromatic path
DD> geometry alone, but rather of richer colored, textured and even
DD> animated objects.

The semantic primitives are characters.

In a visual representation, characters are represented by glyphs.

DD>  That that animation should be borne
DD> declaratively and proximal to the object rather than as a style or
DD> script seems self-evident,  if so. 

Yes, I would expect that good glyphs for emoji would use a font technology that provides multicolour (preferably, parameterisable colour rather than hard-coded) and animating glyphs.

I would be interested to see SVG glyphs for an emoji font. Anyone working on that?

 Chris Lilley   Technical Director, Interaction Domain                 
 W3C Graphics Activity Lead, Fonts Activity Lead
 Co-Chair, W3C Hypertext CG
 Member, CSS, WebFonts, SVG Working Groups

Received on Tuesday, 4 October 2011 02:28:26 UTC