[Mailman-Users] Switching to UTF-8 and strange unicoded email accounts in member lists

[Ulf Dunkel]

Hello Mark et al.

Please let me note some remarks on my experiences while switching to 
UTF-8 for Mailman first:

I have now finally managed to switch my Mailman 2.1.9 to UTF-8. I found 
many helpful hints and comments in this mailing list and want to thank 
you all for your great work.

But now I ran into that miraculous email encoding error (emails listed 
with u'name at domain.tld') which can be detected using withlist as 
recommended by Mark, running the Python commands:

$ bin/withlist listname
 >>> for mem in m.getMembers():
...   print repr(m.getMemberCPAddress(mem))

This indeed listed some strange email addresses and I have NO idea why, 
except that I used to use different iso encodings for my own email 
script before.

I fixed all them by doing this:

a) mailto:listname-request at mydomain.tld who list-password
b) mass unsubscribed ALL members from the relevant list (quietly)
b) mass subscribed ALL members using the who reply list

Then all emails were listed properly when I re-checked everything. Now 
all lists run fine (again) and I wonder if the current stable 2.1.12 
does already have more "built-in" UTF-8 stuff. Can't wait to update my 
Debian Etch server to get my hands on the current Mailman version.

Maybe my notes can help some other Mailman users out there.

Best regards,
--Ulf Dunkel