William_J_G Overington wrote:
Would it be a good idea to define a new block of characters within
Unicode/10646 such that characters would be encoded in pairs, possibly
with visible glyphs as context-specific markup brackets?
[...]
I am thinking that this would mean that where some applications use a
combination of Unicode/10646 characters, (sometimes including specific
circumstances characters such as Hieroglyphics characters) and markup,
that the fact that some of the stream of characters are used in a
markup context and some are not would be detectable from within the
character stream, in both forward and backward parsing, instead of
being designated in a possibly non-interoperably-notifiable manner
from outside of the character stream.
I do want to commend you for coming up with an idea to encode things
that would actually be characters in the sense that term is generally
understood.
I think in this case, the problem of identifying characters used in a
markup context, not with their normal plain-text meaning, is a
well-understood problem that has been successfully addressed by parsers
for half a century.
I also think the history of languages (programming, text markup, etc.)
that rely on obscure or difficult-to-type characters is not encouraging.
APL is a classic example. It's not that it would be hard to build a
specialized editor with macros to allow keyboard access to the special
characters, but that one could *only* use such an editor to write such
text, and not one's otherwise-preferred editor. We Unicode wonks tend to
forget that for many users, if it's not a visible key on the keyboard,
it may as well not exist.
I think a proposal to encode markup brackets would have to demonstrate a
real-world problem that isn't adequately solved with current solutions
(not just "it would be nice for some future use"). However, again, at
least these are characters, which is progress.
--
Doug Ewell | Thornton, Colorado, USA
http://www.ewellic.org | @DougEwell
