The following reply was made to PR misc/34995; it has been noted by GNATS.

From: Sergey Svishchev <svs+pr@grep.ru>
To: Onno van der Linden <o.vd.linden@quicknet.nl>
Cc: gnats-bugs@netbsd.org, wiz@netbsd.org
Subject: Re: misc/34995
Date: Thu, 15 Feb 2007 22:41:34 +0300

 On Sun, Feb 11, 2007 at 08:57:40PM +0100, Onno van der Linden wrote:
 >
 >This PR (misc/34995) breaks the display of alternate charsets
 >as used for example in mutt or slrn.
 
 Indeed.  I managed to submit wrong definition (twice) :(
 
 >It's caused by wsvt using tc=vt220-8 instead of tc=vt220.
 >This changes the ae and as capabilities to ^O and ^N, which
 >aren't handled in wsemul_vt100.c
 
 They are, actually.  grep ASCII_SI and ASCII_SO.
 
 >Try the following with a -current that has termcap rev 1.98
 >TERM=vt220
 >/usr/bin/tput as;echo "tq";/usr/bin/tput ae
 >TERM=wsvt25
 >/usr/bin/tput as;echo "tq";/usr/bin/tput ae
 >
 >It should give some graphic lines in both cases, but it doesn't.
 
 That's because the alternate charset is identical to primary.  My
 working copy (in $HOME/.termcap) had ti= and te= sequences that define
 and reset alternate charset, but I've managed to forget about that...
 
 Here it is, please test.  (I've verified that mutt displays line
 drawing characters.)
 
 wsvt25|NetBSD wscons in 25 line DEC VT220 mode:\
 	:co#80:li#25:vb@:xn@:if@:Co#8:pa#64:NC#2:it#8:ut:\
 	:k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:k;=\E[21~:\
 	:AB=\E[4%dm:AF=\E[3%dm:op=\E[m:is=\E[!p:rs=\Ec:ti=\E)0:te=\E)B:\
 	:ac=``aaffggjjkkllmmnnqqttuuvvwwxxyyzz{{}}~~:as=\016:ae=\017:\
 	:@7=\E[8~:kh=\E[7~:kH=\E[8~:tc=vt220-8:
 
 -- 
 Sergey Svishchev