From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Sun Mar 16 2008 - 11:03:35 CST
The problem with arbitrarily wide embellishments on text is how to
define scope. The mathematical horizontal brackets are used as
characters, but within a system that uses other mechanisms to assign
scope. Therefore, the character only encodes the type of the
embellishment to be used, not its beginning or end.
That's a methodology that has been used by the Unicode Standard for (at
least) two distinct notational systems: mathematical notation and
musical notation.
Applying this to ordinary text, however, is problematic, since the layer
that defines a notational convention is missing in the usual
implementations for ordinary text. Coding a bracketing pair of
characters, with the understanding that the text between these
characters is to be embellished would allow someone to express the
desired semantic, but introduces non-local scope into plain text,
something that should be avoided if at all possible. All text handling
software might be in a situation where it is exposed to the requirement
to keep track of that kind of scope, or else, ruin the desired effect.
And, of course, at least some implementations would have to support this
convention, or else the users would never see the intended text decoration.
Generically, such over and underlining schemes would seem to belong into
extensions to CSS or similar standards. To mimic the traditional
flexibility of TeX and LaTeX in these areas, CSS could adopt a
convention whereby one could specify a (horizontal) bracket character
(or an accent) to designate the desired effect, as is done for mathematics.
CSS (and HTML) would take care of the scope, and Unicode's role could be
limited to furnish any required building blocks, that is character codes
for any required additional shapes used in making such text decorations.
A./
This archive was generated by hypermail 2.1.5 : Sun Mar 16 2008 - 11:06:50 CST