Re: [tlug] Email encoding for DoCoMo and AU (was: font/char set question)

["Stephen J. Turnbull" <stephen@example.com>]

Curt Sampson writes:

 > On Wed, 8 Aug 2007, Josh Glover wrote:

 > > Isn't UTF-8 supposed to be Base-64 encoded, per some RFC or
 > > other?

A transfer encoding was recommended in ancient RFCs (up to about RFC
1642 in 1994, which defines UTF-7), but any system that can handle
Japanese at all is unlikely to have trouble with UTF-8 because of the
high bits.  You just add a "Content-Transfer-Encoding: 8-bit" header.
(True binary, where you have to worry about line ending conversion and
such-like, is another matter.)

However, you cannot be sure about whether 8-bit material will arrive
in that format, because it's perfectly reasonable for a relaying MTA
to ask if the next-hop partner can handle 8-bit, or even assume that
eventually the message will pass through a munging gateway.  In such
case a relay may BASE64 the message body.

 > Not as far as I know. I see no more reason for base-64-encoding it than
 > any other 8-bit data, such as Shift_JIS or ISO-2022-JP.

*snicker*  You do have a bad memory.  ISO-2022-JP is old enough for
bytesize to matter (RFC 1468).

For the interested, RFCs can be accessed by number via the template
URL ftp://ftp.rfc-editor.org/in-notes/rfc${RFC#}.txt.  The full list
of RFCs is in rfc-index.txt at the same URL.