With the coded-font model if a desired glyph does not map from a character in the coded character set, it cannot be in the coded font and the glyph cannot be displayed or printed. For example, if the U+FB01 latin small ligature fi "
fi" character is not in the coded character set, then the "?" glyph is not available for display or printing. Consequently, the widespread use of the coded-font model resulted in pressure to include some glyphs into coded character sets. The other two font models, which can be implemented to do sophisticated glyph selection, do not require that all the glyphs in a font resource be coded as characters in the coded character set to print or display the glyphs.The coded font model is less suitable than the other two models for the more complex glyph-selection requirements of printing and publishing. For example, the Arabic script requires special processing in the coded-font model. If the input to the general layout process includes Arabic characters, the process also needs to convert the Arabic characters to the correct Arabic presentation forms.