Developers of Sympa

[John Kirkland <address@concealed>]

Hi, developers,

I'm encountering this problem in v6.1.1.

I have a portuguese mailing list in Sympa that I'm trying to convert 
fully to UTF-8.  One message per day is sent to the list, and it is sent 
as plain text.  I have recently converted this plain text from 
iso-8859-1 to utf-8 encoding.  So far, my only complaint about messed up 
character encoding was from an old Eudora user... everything looks 
normal to me in Thunderbird.

To tell the mail programs about the UTF-8 messages, I have added 
"Content-Type: text/plain; charset=UTF-8" as a header to the email.

The only problem is my sympa mail archives.  The body of the archived 
message is all run together without any mhonarc-inserted, HTML 
formatting in it.  Before this conversion to UTF-8, the archive message 
was neatly formatted by mhonarc with <br> and <p> at the ends of the lines.

I've done a little bit of debugging, and if I remove the Content-Type 
header from the email, then the message contains the mhonarc-inserted 
html formatting, but the characters are no longer legible.  It's as if 
it converted the utf-8 characters from latin1 to utf-8 -- even though 
they were already utf-8.

I also ran mhonarc from the command line using the same command as from 
the sympa debug log.  Now, it gets ever weirder... it works fine if I 
run the command from the command line.

Any ideas?

BTW, hijacking my own thread here.  If I don't put the content type 
headers into the plain text message, then bulk.pl will crash when it 
attempts to perform merging (new 6.1 feature) because there is no 
encoding specified.  The problem was in Bulk.pm around lines 288 and 
304.  If $charset is undefined, then from_to will 'die'.

Many thanks,
John Kirkland