On Tue, 2008-11-04 at 13:36 -0800, Mark Sapiro wrote:
tct13 wrote:
I am Romanian and want MM to use utf-8 encoding because iso-8859-2 (current MM charset for Romanian) is not commonly used due its wrong display of s and t letters.
I would really appreciate if you would tell us what changes you have made to your MM instance so German can have utf-8 encoding.
I have tried changing only the line in mm_cfg.py/Defaults.py add langauge ("Romanian", ....... charset="utf-8") but that's not enough.
You also have to use iconv or something similar to transliterate messages/ro/LC-MESSAGES/mailman.po and templates/ro/* from iso-8859-2 to utf-8 and change the charset= parameter on the Content-Type: header in messages/ro/LC-MESSAGES/mailman.po and then use msgfmt or Mailman's bin/msgfmt.py to update messages/ro/LC-MESSAGES/mailman.mo from messages/ro/LC-MESSAGES/mailman.po.
After I make this change I end up with stuck emails (MM doesn't deliver them any more).
If you check Mailman's error log, you'll probably find uncaught exceptions due to trying to process iso-8859-2 templates and/or messages as utf-8.
Thank you, it worked! Now we have utf-8 encoding and it's cool. ;-)
We would like to write a minitutor for changing the encoding in MM and a patch which can change this in the current dev version. I don't know a single Romanian who still uses iso-8859-2 standard.
I'll come back with references.
Thank you, Tibi