On Sun, Jun 28, 2026 at 17:25 +0200, Vincent Lefevre wrote:
If the 8th bit is interpreted as being Meta, then it is no longer
possible to have non-ASCII characters in input.
Sure, but not everybody needs non-ASCII characters.
I'm in the US, don't often need even accent marks, much less the Greek
alphabet. I used xterm and Emacs that way for years, with ASCII and
with Meta encoded as the 8th bit. Don't think I ever used ISO-8859-*.
That's in the past for me. Now I use the locale C.UTF-8, so xterm can
send Unicode, and the xterm setting metaSendsEscape, so Meta is
converted to an Escape prefix. Occasionally I use an X mechanism to
type a non-ASCII Unicode character, like this: é.
Of course, in the X window system, Meta is a genuine modifier, separate
from characters, so it works fine, including when Emacs is an X client.
Meta only becomes a problem when squeezed through a terminal emulator.
I get the impression the world is moving toward Unicode, but I don't
know much. Is that happening in Europe?
