Adobe's "Unicode and Glyph Names" (WAS: RE: dotless j)

From: Christopher J. Fynn (cfynn@dircon.co.uk)
Date: Mon Jul 05 1999 - 17:55:48 EDT


Peter Constable asked:

> And how is this scheme to deal with reordering in Devanagari?

You should probably ask Adobe about this.

I'd guess you'd have to include composite ligature glyphs to represent
the various strings of characters which might otherwise have to reordered.
(Or for that matter, have been reordered) The naming scheme allows for
a single glyph to represent up to seven characters which should be
sufficient for Devanagri - though it is two or three characters short
of the requirements for some complex Tibetan stacks.
 
Seven characters is certainly sufficient for any combinations in Latin
script - moreover this scheme allows anyone with a copy of Fontographer,
or another Type1 font editor, to build fonts to work with Unicode.
No need to build more complex OT character to glyph GSUB tables.
Once user friendly tools are generally available for building OT fonts,
and there is more support for such fonts in operating systems,
it should be easy to have a utility parse the names in these and build
the equivalent OT tables.

There is already support for this naming convention in Adobe Type
Manager for Windows NT - and rumour has it that a version of ATM with
these features is built into Windows 2000. It should also be reasonably
straightforward to implement on other systems which already have a Type 1
font rasteriser and support Unicode characters. Certainly easier than
implementing full fledged support for OpenType fonts.

IMO this is a worthwhile (and workable) solution - at least until there
are tools generally available which make it much easier to build OpenType
 fonts *and* full support for OpenType (or something similar) is present
on the average user's desktop.

- Chris



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:48 EDT