FontSync has introduced the concept of a font reference in order to uniquely identify a font. The font reference is an opaque object containing enough information about a font to be described to the user and related to a font that the user may have. Font references can only be created for fonts that can be interpreted by Open Font Architecture and can be represented as font objects (that is, fonts for which there is a scaler). They are stored in documents and can be sent over a network and matched with actual fonts on a host system.
The data that is stored in a font reference is algorithmically derived from the font data. This distinguishes FontSync from the approach taken in
PANOSE
specifications:
PANOSE
numbers are subjectively assigned by the font designer. Some of the data contained in a font reference includes the QuickDraw font family name, the ATSUI-visible font name, the type of font, the font version, checksums of the data, and information from the font name table.
Because a font reference contains only a portion of the data contained in the actual font, there is a slight chance that when a match is reported, the fonts don't actually match. However, there is a high degree of confidence in a match because any difference in font data triggers a mismatch. On the other hand, spurious mismatches may also occasionally occur. For example, when glyphs are renumbered in a font, they trigger a mismatch, even though the font hasn't changed. Matching fonts of different technologies is a particular case of this. This is not currently supported in FontSync 1.0.