<<< Date Index >>>     <<< Thread Index >>>

Re: w3m can't show html mail with charset: gb2312



 On Wednesday, July 27, 2005 at 11:21:03 PM +0800, phyrster wrote:

> On 11:42 Wed 27 Jul     , Alain Bench wrote:
                      ^^^^
    The year in attribution seems to lack.


> While copying or loading messages, the encoding won't get converted
> into that of the system's locale?

    No: <copy-message> [C] is a verbatim copy. What do you mean
"loading"? <fetch-mail> [G]? Verbatim.


> The attachment-1 is a copy of mal-displayed message. [...] I don't
> have that mal-formed From: message right now.

    First: Unfortunately both mails had 7 bits US-Ascii "From:", of
course well displayed everywhere. Nothing can be infered for your
"From:" problem. Show us when you have one.

    Att-2 is a well formed UTF-8 mail. Good display of subject and body,
in index and pager.

    But there is a big problem with att-1 body and subject: MIME label
says UTF-8, but real content is GB2312. You wrote att-1, so in fact
problem is not broken display (unavoidable with such a mess), but broken
sending. Check your editor writes UTF-8 as per current locale. And try:

| set 
send_charset="us-ascii:iso-8859-1:iso-8859-15:windows-1252:iso-8859-2:windows-1250:gb2312:utf-8"


> In my muttrc, I have these two lines:
>| charset-hook ^utf-8$ utf-8
>| charset-hook .* gb2312

    If you really want or need them, fine. But such ultra-wide aliasing
seems to me like an heresy. It breaks so much things... Trying to
workaround broken mails is good. But sacrificing display of valid mails
for it is bad. If you asked me I would probably allow only the harmless:

| charset-hook ^us-ascii$ gb2312

    ...and deal with encountered problems one by one, either with a
narrow charset-hook where possible, or manually thru <edit-type>.


> you mentioned last time that [$assumed_charset] can affect Subject in
> Debian patched mutt and they do work well.

    More exactly $assumed_charset affects raw 8 bits unencoded
unlabelled headers. Subject, From, and others. The impact perimeter is
not Subject versus From, but raw against valid 2047 encoding.


Bye!    Alain.
-- 
When you want to reply to a mailing list, please avoid doing so with
Lotus Notes 5. This lacks necessary references and breaks threads.