Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>:
MK> However, mere implementations of the Unicode bidi algorithm are far from
MK> what we need to really understand how to handle bidi text in xterm or
MK> other VT100/ISO 6429 emulators.
Before we can ever start wondering about this, we need to decide how
much is to be handled in the terminal emulator and how much in the
application. At one extreme, as far as xterm is concerned, all
characters are L2R, and the application is in charge of sending them
to the terminal in ``visual'' order; I understand this is commonly
done for Hebrew. At the other extreme, xterm implements the full
Unicode BIDI algorithm, and the semantics of most control sequences
needs to be clarified. Somewhere in between, the terminal has no
concept of intrinsic directionality of characters, but allows the
application to switch writing direction using control sequences;
again, the semantics of other control sequences need to be clarified.
My personal preference is the first solution, due to the simple fact
that an application already needs to do bidi processing in order to
handle cursor movement in an intuitive (which usually means
configurable, as my intuitions and yours probably differ) way. Second
best is the third solution, which allows an application either to do
its own bidi processing (by not using the direction-switching
commands).
I would be glad to know which model is mandated by ISO 6429 and which
is implemented by Kermit. (We might as well try to ensure that the
two most popular terminal emulators are compatible.)
J.
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:51 EDT