IndUni fonts

October 2010: a third upgrade of the fonts has been carried out. New characters have been added with a view to providing everything likely to be needed for transliteration of Arabic and of the Iranian language Wakhi. Note that the full set of characters required for the latter is only found in IndUni-T, which, as well as j-underdot (provided previously for Shina) now also contains j-underdot-caron, barred “u” and ezh.

April 2009: a second major upgrade of the fonts has been completed. Further new characters are included and some bugs are fixed (including the accidental omission of non-breaking space in the previous release). As well as a comprehensive set of “Indian” characters, all the letter forms required for Avestan and for the Pinyin representation of Chinese, a set of Cyrillic characters and a basic set of Greek letters, the fonts implement almost the whole of the Multilingual European Subset 1 of Unicode. The IndUni-T fonts also contain versions of j-underdot, used for some Dardic languages. All accented characters can be accessed in “composed” or “decomposed” form — e.g., a-macron is accessible both as U+0101 (a-macron) and as U+0061 + U+0304 (“a” + macron). This is a considerable advantage, given the different possible ways in which Unicode can be, and is, used.

January 2009: a major upgrade of the fonts has been completed. Many new characters are included, some bugs are fixed, and all accented characters are now accessible in both “composed” and “decomposed” Unicode forms. For details see the README files accompanying the fonts.

June 2007: at the request of Linotype, I have rebuilt the IndUni fonts under new names, to avoid clashes with names on which Linotype hold a trademark. The fonts that used to be called Courier IndUni are now called IndUni-C; those that used to be called Helvetica IndUni are now called IndUni-H; those that used to be called NCS IndUni are now called IndUni-N; those that used to be called Palatino IndUni are now called IndUni-P; those that used to be called Times IndUni are now called IndUni-T.

The OpenType fonts in this archive are designed to allow the representation of Indian-language (and similar) material in Roman script using the Unicode character set. They are based on fonts designed by

URW++ Design and Development Incorporated
Poppenbuetteler Bogen 29A
D-22399 Hamburg

and made available under the terms of the GNU General Public License, a copy of which is included in the font zipfiles themselves in the file COPYING. The chief provisions of the GPL are that software licensed under it may be freely redistributed provided the author's copyright is properly acknowledged. As permitted under Section 2 of the GPL, I have modified the fonts to implement the accented characters needed by Indologists and other such scholars, using mkt1font, a program I wrote to create arbitrarily accented Type 1 PostScript fonts. The fonts output by mkt1font were converted to OpenType using Adobe's program MakeOTF (part of their OpenType Font Development Kit). In developing them I have also made substantial use of George Williams's excellent program FontForge. The modified fonts, like the originals, are distributed under the GPL. However, the copyright remains with URW++ Design and Development Incorporated. The modified fonts were created on January 18, 2009.

The fonts work well under Windows XP, provided that Service Pack 2 has been installed and “complex scripts” enabled. They presumably work well with Windows Vista, of which I have no experience; I understand, though, that in Vista it is no longer necessary to “enable complex scripts”.They work well with Word 2003 and later. I have used them myself under Linux and under Mac OS X Tiger and Leopard.

To enable complex scripts, Windows XP users should open the Start menu and choose “Control panel”, then “Date, time, language, and regional options”, then “Add other languages”. Now check the box marked “Install files for complex script and right-to-left languages”, and then click “OK”.

The fonts are contained in the five .zip files: contains fonts based on URW++'s Courier-lookalike; contains fonts based on URW++'s Helvetica-lookalike; contains fonts based on URW++'s New Century Schoolbook-lookalike; contains fonts based on URW++'s Palatino-lookalike; contains fonts based on URW++'s Times-lookalike;

(I do not currently have access to any software permitting me to build reliable .dfont files for Mac OS X, so Mac users should simply install the four individual .otf files contained within each of the .zip archives.)

In addition, the file contains Courier-like fonts that are monospaced (this can only be achieved by omitting the double-width characters kh-underbar, Kh-underbar and KH-underbar). These fonts may be found useful with programs such as terminal emulators that require fixed-pitch fonts.

The file contains a Windows keyboard layout that gives access to a large proportion of the special characters from the keyboard (if you encounter problems installing it on a Vista or Windows 7 system, the README file provides a way round them). The file contains an equivalent keyboard layout for Mac OS X. Note that on XP systems (I cannot answer for Vista) and on some earlier releases of Mac OS X, Word follows the usual Microsoft-knows-best rule and ignores most of the keystrokes defined in these layouts. (As of Mac OS X Leopard 10.5.4, this issue has been resolved, and Word behaves properly.) If you encounter this problem with Word, it is best to use the accent macros available here; it is easy to assign your own keystrokes to these. (The same collection of macros contains methods to permit the conversion of Word documents using the old Norman and CSX+ encodings to Unicode.)

Many people may find or its Mac OS X port NeoOffice preferable to Word and the other Microsoft Office components. Both are freeware, both support Microsoft's file formats, and both work well.

The file IndUni.def contains a list of the special characters contained in the fonts.

In addition to the IndUni OpenType fonts, a set of virtual fonts is available implementing the same set of characters for Omega, the 16-bit Unicode-aware development from the typesetting system TeX. They reference the same URW++ fonts as have been used to build the OpenType fonts described above. They were built using vpl2ovp, a program I wrote to create arbitrarily accented virtual fonts for Omega.

Please email any problems to

Back to John Smith's home page