Re: content encoding etc. (updated)

Hatuka*nezumi - IKEDA Soji writes:

> I understood.  I worried that shortcuts such as `11.1% algorithm' 
> would break backward compatibility on implementation for any
> charsets (for example, an ISO-8859-1 subject line including a bit 
> many accented characters).
> 
> At least charsets below allow both methods (while some 
> implementations prefer quoted-printable for encoding):
>   RFC1947 (ISO-8859-7) Greek
>   RFC1555 (ISO-8859-8) Hebrew
>   TIS-620 (ISO-8859-11) Thai
> 
> Perhaps more recently-coming (sometimes non-Latin) charsets allow 
> both methods... is that right?

_Any_ character set may be used with either quoted-printable or base64.  The 
content transfer encoding is completely independent of the text character 
set.  In fact, sqwebmail can simply set Content-Transfer-Encoding: 8bit for 
any character set.  The only time the message must be converted to some 
7-bit format (either quoted-printable or base64) is when the message is 
PGP-signed.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQBAsnT9x9p3GYHlUOIRAm+7AJ4v48fZphxUTedg7ojEgI10oVrGaACcDsfJ
cxCY+lqUK4Z0gSrQEHKxfAk=
=LZIP
-----END PGP SIGNATURE-----