And wouldn't it be safe enough to check for "ISO-8859-*" as they all
have a similar structure, with the control characters in the same
place?

+ Kim


| From:    Kimmo Suominen <kim@tac.nyc.ny.us>
| Date:    Mon, 09 Sep 2002 09:16:21 -0400
|
| I like this idea.  For euc-jp you'd get exactly the behaviour you get
| with the old finger, but all the European users would be happier.
|
| Any objections to going forward with this plan?  I'll put in comments
| to remind what to rip out when there finally is a protocol extension
| to finger.
|
| + Kim
|
|
| | From:    Martin Husemann <martin@duskware.de>
| | Date:    Mon, 09 Sep 2002 14:11:30 +0200
| |
| | > It actually does, in this part of the world at least. Other OSes passes
| | > 8-bit without question (Solaris, Linux) which causes lot of complaints
| | > about NetBSD not being able to do it.
| |
| | That's why I suggested to just declare "ISO-8859-1" as the official charset
| | our fingerd sends - and hardcode that in finger too. (Note that this is
| | different to the original approach which started this discussion.) One
| | possible implementation in finger is to force output to ASCII if the user
| | locale settings indicate anything different to ISO-8859-1, which would make
| | itojuns output (in euc-jp) safe, since no shift/escape sequences could be
| | triggered.
| |
| | It won't change much, make NetBSD interoperate better with other OSes and
| | not prevent any steps towards a real solution. It is no real solution in
| | itself though.
| |
| | Martin
| |
|