Kazu,
At 6:24 AM 12/14/94, Kazuhiko Yamamoto
=?ISO-2022-JP?B?GyRCOzNLXE9CSScbKEI=?= >This time I'd like to ask about a
CTE: issue. When I send a letter
whose body-part is "CT: text/plain; charset=iso-2022-jp", which CTE:
should I use? How about ISO-8859-1?
The distinction between Content-type and Content-transfer-encoding is a
source of continuing confusion, even for people who have worked with Mime
for quite some time. Hence, your query provides a good opportunity for
some review:
Content-type specifies the "canonical" representation of data. It pertains
to the syntax and semantics of the real user data, albeit in an
interoperable representation, such as GIF, JPEG, ASCII, etc.
Content-transfer-encoding specifies the modifications that have been made
to the canonical user data in order to get it through cantakerous transport
environments (e.g., most of the email transport world. sigh.)
For Content-types of text, the charset= parameter permits specification of
some detail about the encoded, data, but this pertains to the canonicalized
user data and not to transfer issues. Of the range of choices for charset,
ISO-2022-jp and ISO-8859-1 are two choices. There can be only one value
selected; hence they are mutually exclusive.
Content-transfer-encoding really only specifies two bits of information
about the data: Whether the 8th bit is one and whether 'line
canonicalization' has been performed on the data. Line canonicalization
means that CR-NUL/CR-LF sequences are maintained and that the number of
characters between CR-LF pairs is relatively short (1K bytes, as I recall.)
In today's transport environment, I believe that CTE's of 7-bit and 8-bit
are the most useful choices; both specify line canonicalization and differ
only in the matter of the high bit of data bytes.
However verbosely, I hope this answers your question.
d/
--------------------
Dave Crocker
Brandenburg Consulting +1 408 246 8253
675 Spruce Dr. fax: +1 408 249 6205
Sunnyvale, CA 94086
dcrocker(_at_)mordor(_dot_)stanford(_dot_)edu