> Thanks. This helps a good deal. I'm mystified as to why 166 (aka 'Broken
> bar') is displayed as a minute formless squiggle and 214 is displayed as a
> broken bar but I can survive without that knowledge
> >
> >This means that software that is converting files between Latin-1 (or
> >UNICODE, or
> >Microsoft "ANSI") and PC-DOS code page 850 ought to perform a conversion on
> >these characters.
>
> Yes. It performs an unwanted one :-). It looks like a problem between Java
> and the DOS commandline. What particularly worried me was that simple Java
> code using 'char' translated this character into 65446, which presumably
> has a completely different meaning in Unicode. IOW there is a danger that
> corruptions could take place.
This must be a bug. 65446 = FFA6, but I figure that 166=00A6 which is suspiciously
close. FFA6 is a naughty Korean character, so I guess someone has
programmed wrong.
I dont know whay 214 = D6 is displayed as a broken bar. Have a look in
the keycaps application whether 214= D6 is indeed a broken bar in the
font you are using. (It is also quite possible for a font designer to
decide to use a broken bar glyph where a single bar is wanted, and vice
versa. If that is the case, change the font to one that isnt broken.)
Rick Jelliffe