On Sun, Mar 31, 2002 at 08:15:43PM +0100, Philip Hazel wrote:
| On Sat, 30 Mar 2002, dman wrote:
|
| > Suppose this is the scenario :
| >
| > 1) a user agent creates a message with the high-order bit set on some
| > characters
| >
| > 2) the user agent pipes the mesasge to exim like that
| >
| > 3) when exim tries to hand off the message via SMTP, the remote host
| > doesn't advertise 8BITMIME
| >
| > WIll exim bounce the message as rfc1652 seems to require, or is that
| > considered a bug in the user agent that created the message?
|
| As documented, Exim is 8-bit clean. It will "just send 8". It doesn't
| even look at what the receiving host advertises.
Ok, it's considered a bug in the user agent. Perhaps it would be
useful (for local pipes) for exim to output a warning message on
stderr and exit with non-zero status if any characters are not 7-bit?
Of course, this behavior should be toggleable like the accept_8bitmime
option.
| The justification for this is given in my book. (Executive summary: this
| action is more likely to "do what the user wanted" than anything else.)
I have no problem with simply being 8-bit clean rather than
pedantically following the "SMTP is 7-bit" definition by stripping the
high-order bit.
I was just wondering what the effect would be if a user agent whose
output wasn't 7-bit tried to send Tony a message with exim as the
system's mta. Apparently that's not a good thing to do.
HAND,
-D
--
Your beauty should not come from outward adornment, such as braided hair
and the wearing of gold jewelry and fine clothes. Instead, it should be
that of your inner self, the unfading beauty of a gentle and quiet
spirit, which is of GREAT WORTH in God's sight. For this is the way the
holy women of the past used to make themselves beautiful.
I Peter 3:3-5
