[BRLTTY] The unknown-character sign (26)

[Jason White]

On Mon, Nov 24, 2008 at 01:46:59AM -0500, Dave Mielke wrote:
> Perhaps the correct thing to do is to remove the attempt to use the question 
> mark and just stick with the Unicode replacement character, which, if not 
> defined explicitly, will be represented by dots 12345678. Opinions?

This seems reasonable: at least the resulting characters can't be confused
with question marks.

The next step would be to define appropriate characters in all of the text
tables, to ensure that the default representation doesn't occur in any of the
supplied configurations (unless for some languages it is specifically
desired).

I still think question mark with a dot 7 or dot 8, defined in each table, is a
good policy, but that's only my opinion.