Diacritics font

Robert Zydenbos zydenbos at BIGFOOT.DE
Sun Nov 28 20:55:13 UTC 1999

On Sun, 28 Nov 1999 17:49:59 +0000, Dominik Wujastyk wrote:

> On Wed, 24 Nov 1999, birgit kellner wrote:
> > It is my understanding however that CSX+-encoded fonts are not
> > particularly useful for mixed-language documents, that is, documents
> > which contain Sanskrit and European languages other than English (e.g.
> > German, French, Danish) that require in themselves diacritical
> > characters. [...]

> This is the opposite of the actual case.  The CS and CSX character sets
> were designed specifically to include as many European accented characters
> as could be accommodated, in addition to those essential for Sanskrit and
> Vedic.  The German and French scholars who sat round the table in Vienna,
> where CS/CSX was born, ensured that this was the case.

Actually, both of you are right / wrong, in different ways. The problem
is that the accented European letters bear certain ASCII codes
according to the ANSI coding scheme (used by Windows and Linux)
but other codes according to the DOS/OEM coding scheme (used by
the older PC-DOS and MS-DOS). The TrueType CSX+ fonts, as
found on Dr. Smith's FTP site in Cambridge, follow the DOS scheme.
This means that if we write something using Windows9x, or X
Windows under Linux, we can type in Umlaute etc. according to the
ANSI scheme, but when we change the font to one of the CSX+
fonts, all those Umlaute etc. will go wrong.

Because I like those fonts so much, I have worked on this problem
and solved it in the latest version of my own diacritical preprocessor.
What you have to do is save your text in RTF format (Rich Text
Format), which can be done when using word processors like
StarOffice and MS Word (and others; this means you can use
footnotes, italics, underlining, various paragraph formats, etc. etc.),
then run the text through my program, then open the transformed
text in the word processor, and you can change to one of the CSX+
fonts with all the accented letters as they ought to be. This sounds
more cumbersome than it really is; I demonstrated this at the
Oriental Research Institute here in Mysore, and within minutes
people of the institute were using the program correctly.

The Windows95/98 version is already available at:


I hope to put the Linux version there within a few days.


More information about the INDOLOGY mailing list