[Mailman-Users] member names vs. us-ascii/iso-8859-1

[Lauren Weinstein]

Greetings.  I was recently stung by the situation where somebody
apparently subscribed to my us-ascii configured system with an
extended Latin-1 char in their "real name" (probably via e-mail,
though I'm not certain).  This resulted in mail distributions failing
and being shunted upon hitting that name due to an "out of range"
error. 

I cured this quickly by switching the mailman definition
over to iso-8859-1 to get the mail out, but this has side-effects
I'd prefer not to have continue (including ordinary ascii Subject
lines being iso-encoded unnecessarily).

Is there a simple change I can make to simply prevent the member
name(s) in question from shunting the entire distribution?
I've been looking over the code but I don't know it well enough
to be sure that I won't cause unintended side-effects.  

Thanks much.

--Lauren--