Re: [nmh-workers] Can't forward MIME-encoded message

[Ken Hornstein]

>> - If it was run on a message with a CTE or Mime-Version header, it would
>>   error out.
>...
>> Why that specific behavior?  No idea!  But it was very deliberate.
>
>I agree the earliest git commit checks both, but the RFC says C-T-E is
>optional.  I think the only mandatory MIME header is the version so
>can't we ditch the other test?  Otherwise it just leads to calls to
>extend the list.

Well, I wanted to provide some perspective and history, but I can't
justify it after a closer inspection.  I find it odd that it bails
out on a Mime-Version header AND a CTE, but it silently eats the
Content-Type.

Perhaps you could help me by filling out the following matrix?  Do you
want the desired behavior for each box to be "exit(0)", "exit(1)",
"Remove" (for remove header), or something else?.  I don't think mhbuild
can just ignore existing MIME headers.  This is what we have now:

+---------------------------+---------+----------+
| Header                    | -auto   | -noauto  |
+---------------------------+---------+----------+
| Content-Transfer-Encoding | exit(0) | exit(1)  |
+---------------------------+---------+----------+
| Mime-Version              | exit(0) | exit(1)  |
+---------------------------+---------+----------+
| Content-Type              | Remove  | Remove   |
+---------------------------+---------+----------+

--Ken