[Freefont-bugs] deleting Telugu, Oriya, Kannada

[Steve White]

I have deleted Telugu in CVS, on account of the existing range of
glyphs not representing a writing system for the language.
Some research needs to be done yet, but I think the same will happen
to Oriya and Kannada.

The reason this should be done is: many font display systems work
glyph-by-glyph.  They may find a glyph for a character in FreeFont,
and use it to display a character, but not find the next character in
FreeFont, and so pull it out of another font.  The result is a very
disappointing mess.

Generally, it is better for a font not to advertize support for a
writing system, than to advertize it and really contain only a subset
of the required functionality.

I tried to rescue Telugu.  But it was a waste of a couple of evenings.

In a check of FreeSerif, I found that Unicode-encoded Telugu text
displayed as a mess, missing lots of glyphs, and with many malformed.
Looking into it, I found that many Telugu consonants were simply
missing, and almost all the vowels, too (and the digits).  So I found
the source font, the old "Tikkana".  There I found many of the missing
glyphs.

But...then I read about how these languages are meant to be encoded in Unicode.
     http://www.unicode.org/versions/Unicode5.0.0/ch09.pdf
These languages require lots of special font features and
supplementary glyphs to work properly.  Looking at some working Telugu
fonts, I see it would be days of work for a person familiar with the
language to get this working.

The work is simply out of the question for me.  I could not do it justice.

There are several very nice free fonts supporting Telugu commonly
available.  These serve the community much better than a
partially-implemented range in FreeFont could.

If somebody is interested in taking this on, they are welcome.  (The
old glyphs are still in the CVS archives.)

Another option would be to import some existing Telugu Unicode font.
But that too would be starting from scratch.

Cheers!
Cheers!