That did it, thanks. Now I just need to tweak various scripts and aliases to use $VISUAL instead of /usr/bin/vi, which I probably should’ve done long ago, anyhow. Seems there’s always another yak that needs shaving.
Note that I had no luck just setting the print option in the stock /usr/bin/vi, which reports that it's version 1.79 (10/23/96) The CSRG, University of California, Berkeley. -DaveP On Mon, Oct 26, 2026, at 00:03, Hugo Villeneuve-Lapointe wrote: > On Sun, Oct 25, 2026 at 08:02:59PM -0600, Dave Polaschek wrote: >> I’ve got a machine running OpenBSD 8.0 with all the patches (as >> of a few minutes ago). >> >> I’ve set LC_CTYPE to en_US.UTF-8. printenv shows that it’s set. >> >> My terminal type when I ssh in is xterm-256color. >> >> The terminal I’m using on my Chromebook is hterm >> (https://hterm.org/x/ssh/faq ) which supports UTF-8, and I’m using >> Google’s “Noto Sans Mono” as my terminal font, which has the >> characters I’m testing. >> >> If I type a ¢ into it, for example, I get the hex-escapes displayed >> in /usr/bin/vi, but when I save the file and dump it using more or >> cat, the ¢ comes out correctly. >> >> Is there some setting I need to tweak to get vi to display UTF-8 >> characters correctly, or am I just running into a limitation of vi >> here? I don’t need anything more than ASCII most of the time, but >> it would be nice to be able to make sure I’ve typed the correct >> unichar without having to ^z out and cat the file to be sure. >> >> Thanks! >> >> -DaveP >> > > vi in OpenBSD does not support UTF-8 yet. > > You'll have to install an alternate vi from the ports. Either nvi > or vim will work. > > Personnaly, I like nvi+iconv. My ksh's .profile startup script has > this in it: > > LANG=en_US.UTF-8 > if [ -f /usr/local/bin/nvi ]; then > VISUAL=/usr/local/bin/nvi > alias ex=/usr/local/bin/nex > alias vi=/usr/local/bin/nvi > alias view=/usr/local/bin/nview > else > VISUAL=/usr/bin/vi > fi > export VISUAL LANG > set -o emacs > > > Note that for nvi, weird thing happens if you edit a file that is > not UTF-8 but has high 8bit characters. You might have to run > "LANG=C nvi file" to edit it. (nvi+iconv has also a > fileencoding/inputencoding settings.) > > Also, without a LANG set to a UTF-8 encoding, nvi will work like > the built-in OpenBSD's vi. (Showing \xc3\xa9 instead of é.) > > Good luck
