[kronolith] Incompatible Invitations (was: Reminders not conforming to RFC 2822)

Otto Stolz Otto.Stolz at uni-konstanz.de
Fri Apr 7 07:57:41 PDT 2006


Hello,

Jan Schneider schrieb:
> You are probably talking about invitations sent to meeting attendees.  
> Neither IMP 3, nor Thunderbird can make any use of these invitations,  
> you need a mail client that does, like IMP 4.

After having solved the sendmail problem (cf. my recent message),
my messages comply with RFC 2822, viz. their lines are delimited
with CRLF (not LF, as with the -- still faulty -- SMTP driver).
However, that issue turned out to be a red herring: Still, the
"attachment" is not displayed by conventional e-mail clients.

A closer look at the message source has revealed the true issue:
the MIME content-type is given as "multipart/alternative".
This means that there is no attachment, at all. Rather, according
to <http://rfc-ref.org/RFC-TEXTS/2046/chapter5.html#sub1sub4>,
each of the body parts is an "alternative" version of the same
information.

However, the Kronolith invitation messages do not comply with
thise RFC-2046 semantic. Rather, the first part essentially
refers the reader to the second (alternative) part for the
details of the event, as if it were an attachment -- but the
two alternative parts never are displayed together. So, the
user of any mail client that chooses to display the first part
is fooled.

To mend the situation,
- either the first (text/plain) alternative part must be
   enhanced to contain the same information as the second
   (text/calendar) one,
- or the MIME type of the whole message must be altered
   into "multipart/mixed".

Now, my questions are:
- Is there a configuration item to achieve one of these
   goals?
- If not so, would it be wiser to enhance the text/plain
   part, or to make the whole message multipart/mixed?
   Particularly, if I would do the latter, would iCalender-
   capable clients still work with the text/calendar part?
- Has anybody mended this issue, already, and is willing
   to share their solution?

Thank you, in advance, for your hints.

Best wishes,
   Otto Stolz





More information about the kronolith mailing list