On Wed, Mar 01, 2006 at 03:30:10PM +0100, Daniel Tiefnig wrote:
> > Either way, one "Received:" header should be displayed as one, long
> > line, without CRLFs,
>
> But what's "displayed" here? If I'm looking at the message source with
> my mail client, should headers allready be unfolded? Or does this only
> concern headerlines that are displayed rigth away with the rendered message?
If you view the source, everything should be shown unmodified, although I
have seen the client evolution showing a "fixed up" message when looking
at the source without saying a word about it.
Many clients offer an option to see all header fields, not just the
ones important to most people. That should unfold headers, strip MIME
encoding and translate the character set to what your display uses.
That's where tabs look ugly.
RFC 2822 calls it a "representation":
For example, the header field:
Subject: This is a test
can be represented as:
Subject: This
is a test
It's still the same header, just represented/encoded differently,
which tells white space is part of the header content, but CRLF is not.
Software may re-fold headers just fine, but if it changes white space
between printable characters, it changes the message content.
I guess that's why removal of white space between MIME words is part
of MIME header decoding: That way you can fold a long line that does
not contain white space at the right points without changing its
MIME-decoded content.
E-mail is weird, isn't it?
Michael
