[SM-USERS] why can't squirrelmail deal with both header and body charsets?

[<ji...@ji...>]

Gentlemen, consider a message with

MIME-Version: 1.0
Content-Type: text/html; charset="big5"
Content-Transfer-Encoding: base64
Subject: =?utf-8?b?5YWE5byf...
From: =?utf-8?b?5bGx5Z+O6aK...

Note how the headers are in UTF-8, but the body in big5.

But squirrelmail 1.4.19 shows them both in on one web page frame.

Therefore, there needs to be some iconv'ing done to either the headers
or the body.

It seems our best strategy is to set
Language preferences
1.  Default Language       : en_US
2.  Default Charset        : utf8

and then hope that squirrelmail will say
"is this header not utf8? well let's iconv it first before sending it to
the browser". Same with the body.

Mutt can deal with this. Why not squirrelmail?

In squirrelmail one must adjust their browser once to read the body, and
then once again to make the headers sane again. It's like needing to
carry around two sets of glasses.

Sure, one can set the Default Language to big5, but then one cannot read
the headers. And what if someone then sends you a UTF8 message.