This section was constructed from a posting by Johannes Schmidt-Fischer in Jun 1993.
All PostScript Type 1 fonts should contain a UniqueId. This is a number which should be, as the name suggests, unique (at least among the fonts that you download to the printer at any given time).
There are many PostScript fonts on the 'Net which have identical UniqeIds. If two of these fonts are downloaded to the same printer at the same time, attempts to use either font may cause the wrong characters to be printed.
In a nutshell, the reason that the wrong characters may be printed is that the printer may be storing the rendered glyphs in its font cache, addressed by UniqueID. So, if two fonts, /Foo and /Bar, both have UniqueID=5 and /Foo's 10pt ``A'' is currently in the cache, a request for /Bar's 10pt ``A'' will cause the wrong character to be printed. Rather than rendering /Bar's ``A'' from its (correct and unambiguous) outline, the printer will note that the cache contains a 10pt ``A'' for font 5 and will copy it from the cache (resulting in /Foo's ``A'' printing for /Bar).
Adobe's ``Red Book'' contains a detailed discussion of this topic.