Thanks, I've fixed that, too.

The 8th bit got stripped for single-line files a bit too well due to
how char gets converted to int...

The code uses toascii to strip the 8th bit of unprintable characters,
which is a bit different from what the net.c code is doing.  I wonder if
these should use the same approach for uniformity.  In any case, this
will prevent the M- from being printed.

+ Kim


| From:    itojun@iijlab.net
| Date:    Sun, 11 Aug 2002 11:16:19 +0900
|
| >The client is the one running isprint() and isspace(), so it reflects
| >the settings of the person running the finger client.  If the received
| >character is printable on his/her screen it should be ok to print it,
| >regardless of what the remote end thinks it is sending.  I don't see how
| >this could cause any problems.
| >
| >There is no standard for non-ASCII finger, so there is no mandate to
| >strip out the 8th bit.  It is very common to have ISO-8859-1 printable
| >characters in the plan files of Finnish users.
|
| 	hmm.  do you know there's another issue in finger/lprint.c:vputc()?
| 	it strips off 8th bit from .plan, pass it to vis(3) before printing.
| 	(so iso-8859-1 chars will become M-A and such)
|
| itojun
|