Variants include typeface variations (e.g., italic, sans serif), and font encodings (e.g., Adobe standard, TeX text).
A fontname may require multiple variants. To resolve the worst ambiguities, we specify that any encoding variant (‘7’, ‘8’, or ‘9’, see below) come last and any other numeral variant come first (to avoid confusion with a design size). We recommend but do not require that the other variants be given in alphabetical order. (It’s not required because it’s too painful to implement the sorting in TeX, and many existing names already have non-alphabetized variants.)
The letterspacing possibilities introduced by
virtual fonts have not yet become sufficiently widespread to make
standardization beneficial. Likewise for the many possible ways to
generate small caps fonts.
Notes on specific variants, both old and new:
These variants (‘semisans’, ‘bright’, ‘semiserif’, and ‘fax’) were used only for single typefaces, so to conserve variant abbreviations, they now have typeface abbreviations instead. Also, for ‘b’, to avoid too-long names; and for the others, to avoid variant vs. design-size ambiguities.
Since these variants were not actually used in any font names, they are now obsolete and may be reassigned in the future.
These characters begin a two-character encoding variant. Generally, ‘7’ is for 7-bit encodings, ‘8’ is for 8-bit encodings, and ‘9’ is for expertised encodings with or without oldstyle digits (see the ‘x’ item below), but this is not an absolute rule. Also, ‘5’ is presently used for phonetic encodings and ‘6’ for Cyrillic encodings. The ‘9s’ (‘SuperFont’) variant contain all of Adobe, Latin 1, and Expert glyphs, perhaps with slightly different metrics than the original fonts.
For a font to be named with a certain encoding variant, it’s not
necessary that all the characters appear in precisely the same positions
as in the encoding definition. It’s enough that the usual TeX macros
work. In practice, this means that it’s ok for a font to be labelled
‘7t’ if the only difference from Computer Modern is that the
ligatures and the ‘lslash’ and ‘Lslash’ characters are
different, since it’s impossible to access or change the ligature table
from TeX. Standard PostScript fonts don’t have an ‘lslash’
character the way Computer Modern does, but they do have the
‘Lslash’ and ‘lslash’ characters themselves, so the usual
\Lslash macros can be made to work via
ligatures. See the file T1.etx file in the
distribution for details.
If a name does not contain a specific encoding variant, its encoding is unspecified. For example, some of the fonts distributed with Dvips(k) have names like ‘ptmr’ for ‘Times-Roman’; they use the Dvips encoding (see dvips), which is close to (but not the same as) the TeX text encoding (as in Computer Modern Roman). Similarly, the TFM files distributed with Dvilj(k) for the builtin LaserJet 4 fonts have names like ‘cunm’, since these fonts, while compatible with TeX text, contain many additional characters.
These items are needed only because ‘x’ (and possibly ‘j’) followed by a two-character regular variant makes some names too long. That is, ‘9t’ is equivalent in meaning to ‘x7t’, and ‘9d’ is equivalent to ‘jx7t’. (This may not be true for all ‘9x’ encodings, though.)
These variants (indicating Greek and Cyrillic fonts, respectively) are obsolete; future fonts in different scripts should be assigned an appropriate encoding abbreviation. These are not currently standardized, due to ignorance of the appropriate encodings. Please send suggestions for abbreviations to the address in Introduction.
These math-related variants remain for the sake of typeface-specific math encodings, e.g., Lucida. Fonts that use the Computer Modern math encodings should use ‘7m’ (see texmital), ‘7v’ (see texmext), and ‘7y’ (see texmsym).
This is used for several different (but very rare) variants: only the Stone typeface has an “informal” variant, and only a few Monotype fonts have a “schoolbook” variant, with different ‘a’, ‘g’, and ‘y’ shapes.
Included only if no other variants, including encodings, apply, and either the width (see Widths) is not ‘r’ or a design size is present. I.e., ‘r’ is only used as a placeholder. When the normal version of the typeface is sans serif (e.g., Helvetica), ‘r’ should be used when necessary, not ‘s’. Use ‘s’ only when the typeface family has both serif and sans serif variants.
Mittelbach in TUGboat 13(1) suggests that these variants (for ‘sans’ and ‘typewriter’) should be identified as part of the typeface name, because there are few typeface families with these variants. I feel that since they are logically variants, it’s best to name them that way. But ‘LucidaSans’ (see Typefaces) and a few others are exceptions, to avoid too-long names.
‘8x’ indicates a font in the ‘Expert’ encoding itself. ‘x’ indicates an expertised font, i.e., a composite (virtual) font that includes characters from an ‘8x’ font. And in fact an ‘xee’ sequence is replaced by ‘9e’, to save characters.
Fontname 1 assignments are shown in brackets in the following table, from the file variant.map. It is organized alphabetically by abbreviation. Each line consists of an abbreviation and either any parts of a PostScript ‘FontName’ which use that abbreviation or the PostScript ‘Encoding’ name.
0 inferior 1 superior 2 proportional digits, not tabular obsolete [3=>7f] Fraction obsolete [4 fax; now typeface ‘lx’, Lucida Fax] 5 escape for (presently) phonetic encodings 6 escape for (mostly) Cyrillic encodings [was ‘SemiSerif’] 7 escape for (mostly) 7-bit encodings 8 escape for (mostly) 8-bit encodings 9 escape for (presently) expert encodings [was oldstyle digits] a Alt Arrows Alternative [was alternate encoding] obsolete [b bright; now typeface lh, Lucida Bright] c SmallCaps d Display Titling Caption Headline TallCaps SwashCaps LombardicCaps Festive e Engraved Copperplate Elite f Fraktur Gothic OldEnglish Handtooled (‘gothic’ can also be sans) g SmallText lc only, or designed for small sizes [was grooved, as in the IBM logo] h Shadow i Italic Kursiv Ital text italic j old-style digits [was invisible] k Greek obsolete l Outline OpenFace Blanks m math italic n Informal Fashion Schlbk for Stone o Oblique Obl slanted p Ornaments obsolete [q=>8t Cork (TeX extended) encoding] r roman or sans; often omitted, see text s Gothic sans serif t Monospace fixed-width typewriter u underline or unslanted italic v MathExtension w Script Handwritten Swash Calligraphy Cursive Tango Ligature x built with Adobe expert encoding [was expert-encoded] y MathSymbol z Cyrillic font-dependent Cyrillic 5a PhoneticAlternate 5i PhoneticIPA 5s sil-IPA 5t TeX-IPA Fukui Rei, LaTeX T3 5w TeXAfricanLatin wsuipa fonts, LaTeX OT3 5z user 6a T2A 6b Cyrillic part of ISO 8859-5, seven bits 6c T2C 6d Cyrillic CP866 encoding 6g LGR Greek font encoding 6i ISO 8859-5 6k Cyrillic KOI8-R encoding 6m Cyrillic Macintosh encoding 6s Storm extra encoding 6t T2B 6w Cyrillic CP1251 encoding 6x X2 6y LCY 6z user 7a A alternate characters only 7c Dfr Fraktur 7d OsF OSF oldstyle digit encoding 7f Fraction 7k OT2Cyrillic 7m TeXMathItalicEncoding see texmital 7t TeX text encoding (as in Computer Modern Roman) 7v TeXMathExtensionEncoding see texmext 7y TeXMathSymbolEncoding see texmsym 7z user 82 GreekKeys 83 Ibycus1 84 Ibycus2 8a StandardEncoding Adobe standard encoding, see 8a 8c TeXTextCompanion LaTeX TS1 8e CE Adobe CE 8f TeXAfricanLatin LaTeX T4 8g groff 8i TS0 Intersection of TS1/Adobe Standard 8m Macintosh standard encoding 8n LM1 Textures 8q encqxoosix QX, from GUST 8r TeXBase1Encoding (see 8r) 8t ECEncoding CorkEncoding (see ec), aka tex256.enc 8u XT2Encoding cmtt + Latin 2, see xl2.enc 8v TeXVietnamese T5 8w Windows 3.1 ANSI encoding 8x Expert expert encoding 8y LY1 texnansi 8z XL2Encoding cmr + Latin 2, see xl2.enc 9c expert + Text companion 9d expert + oldstyle digits + Cork 9e expert + Cork 9i TS0X Intersection of TS1/Standard/Expert 9o expert + oldstyle digits + TeX text 9s SF SuperFont 9t expert + TeX text 9u Unicode-compatible 9x TeXnANSIEncodingX texnansx, texnansi without repeats 9z user - songti for mnm