On 5/5/2012 3:02 PM, Doug Ewell wrote:
In other words, the characters in the top and bottom rows are unified
in Unicode, according to both the Microsoft-provided mappings for
CP950 and (for the four listed code points) the obsolete Unicode
mapping for Big5. One would probably need to provide a second source
to show that these glyphs really have the indicated differences in
appearance, plus evidence that they are used in contrast to one
another, to make the case for either disunification or variation
sequences.
That might be at once easier and more difficult than you might think.
Ken Lunde, in the second edition of his book, states that there are
holes in the Unicode of non-han Big 5 characters. So, getting this
report is not unexpected. However, box drawing characters are a strange
breed of symbols. They are rather rooted to terminals and terminal
emulators (or "character mode" full-screen interfaces). These have
largely fallen out of use. So it may be very difficult to document
current usage, beyond looking at lists of mappings.
The latter is of limited value. The mappings and the encoding would be
expected to be mutually consistent, even if the encoding blurred actual
differences in implementation. Any "de facto" unification in that area
does not necessarily imply a well-researched, well-reasoned, and
well-defensible unification by design. A unification by default, for
lack of evidence at the time is equally likely.
A./