Re: Unicode Devanagari Font in Mozilla]

From: James Kass (
Date: Sun Jul 14 2002 - 21:39:06 EDT

Prabhat Hedge wrote,
> >
> >>* Indian language web-sites use mis-use charset tag "x-user-defined".
> >>
> >
> >So do some non-Indian language web sites.
> >
> But most of them do not need CTL (Complex Text Layout) support as is
> essential for Indian scripts.

There are different reasons that people use the "x-user-defined"
character sets. One reason is that, if the material is in HTML
numeric character references (example क = क), at least
one browser will have some display differences between whether
the page is *displayed* under UTF-8 or "User Defined". In some
cases forcing automatic "User Defined" encoding makes a better

Another reason that people might use "x-user-defined" is to
avoid the complex text layout requirement and Unicode encoding
altogether. In this case, the page author would be using a custom
(proprietary) font encoding. X-user-defined is not a valid character
set, but, at least calling a page x-user-defined for a custom encoding
scheme is not wrong. (There is no valid character set for these
custom encoding fonts.)

> Which means that either a glyph-set or font-encoding standard is
> required or that all platforms that
> need this support need to have OT/Intelligent font support.

The best way to render a Devanagari page is with Unicode encoding
and smart font technology. With an up-to-date version of the
Uniscribe software installed, Devanagari can be properly displayed
even on Win 9x, as long as the browser uses the Uniscribe engine.
Of course, for input and editing of Unicode Devanagari, the best
(Windows) method is to use Win NT-and-up, which has built-in
essential tools like keyboard drivers.

There are "work-arounds" for handling Unicode Devanagari on
platforms which don't use smart font technology. One way of
doing this is to accept Unicode data from the internet, but
convert it into a custom font encoding for display on older
systems. (And, reverse this process when the user is sending
data, convert the custom encoded material back into Unicode
before it is transmitted.) There are various ways of handling
this and somebody has to "do the work" of setting up conversion
data tables. Because there are so many different custom font
encodings for Devanagari, this can be a lot of work. Fortunately,
people like Mark Leisher have already done a lot of work in this
regard. If you are interested in this approach, please visit Mark
Leisher's web site to learn more about using PERL scripts for this
kind of conversion.

Best regards,

James Kass.

This archive was generated by hypermail 2.1.2 : Sun Jul 14 2002 - 20:55:16 EDT