Target Audience: Manager, Software Engineer, Marketer

Session Level: Beginner, Intermediate

The Unicode Standard encodes characters, not glyphs. But for human readers, characters must be displayed or printed with glyphs, and this requires cooperation between Unicode implementors and font developers.

A common puzzle is graphical harmony. In a given font, or set of fonts intended to be used together, how closely should the glyphs in one Unicode block coordinate in "look" with those in another? As the range of Unicode blocks broadens, from ASCII to Extended Latin, and from Latin to non-Latin, and from alphabetic to ideographic, the potential complexity of the problem of harmonization increases.

Another common problem involves the alternation of variant glyphs for Unicode characters whose graphical expression may vary according to locale, culture, or typographic preference.

Yet another issue is the naming of characters. Some font formats identify characters by name as well as by Unicode, and font rendering software may expect particular names, even though they have not reached the degree of standardization of Unicode encodings.

Another question is: how should Unicode implementors and font developers respond to requests from software developers for "Unicode fonts"? Does that term simply mean a font whose glyphs are encoded as Unicode characters, as some implementors might say? Or does it mean a font that contains every character in the Unicode standard?, as some developers might expect? Or something in-between?

This tutorial illustrates these and other puzzles and discusses possible solutions, drawing on experience gained from developing Unicode compliant fonts for the various operating systems of Microsoft, Apple, Sun, and Lucent.