Re: On Unicode

[Ted Harding]

On 19-Mar-99 address@hidden wrote:
> 
> I think lout could be adapted to handle this input quite easily. The
> real work would be in the PostScript backend. How does it handle large
> character sets? Is it unicode-based?

PostScript has its own methods quite independent of character encoding in
the calling software. Basically, a given PS device knows of named "fonts"
either built-in or downloaded; also each "glyph" in each "font" has a
name and a corresponding code, either built-in or defined by downloaded
code. Although any glyph in any font has at any time a code from 0-255,
this arrangement allows arbitrarily many glyphs to be accessed in
PostScript. It's up to the calling software to tell PostScript what it
wants, using PostScript language to do so. So calling software can use
Unicode and call PostScript (which knows nothing about Unicode).

When software (like Lout) has its internals closely tied to PostScript,
there could be complications in adapting it to Unicode which would be
less of a problem, or none, with software where the conversion to
PostScript was done by post-filtering a device-independent output stream.

> Also Unicode doesn't define
> character codes for ligatures, which would make support of them
> diffucult.

This strikes me as a possible defect in Unicode, though it doesn't stop
ligatures being implemented: the software needs to be able to recognise
sequences such as "fi" and "ffi" and wrap them up into a call to print
the ligature (PostScript, of course, already has some ligature glyphs in
the standard fonts).

The major complication with this approach is that different languages
have different conventions about ligatures. In some languages it's
necessity rather than convention: in Turkish, for instance, there is an
"i" with a dot and an "i" with no dot. Therefore, since "fi" (where "i"
has dot) would be indistinguishable in ligature from "fi" (dotless "i"),
the "fi" ligature is not used. Also, even in English you will find that
dictionaries usually avoid ligatures, not only for entries but even in
explanatory text.

Can of worms, this one!

Ted.

--------------------------------------------------------------------
E-Mail: (Ted Harding) <address@hidden>
Date: 19-Mar-99                                       Time: 11:41:02
------------------------------ XFMail ------------------------------