Re: [Evolution] displaying mail source bug?

[Dan Winship <danw ximian com>]

> You say, when viewing a mail as source, the charset value will not be
> evaluated. Which character set is used then?
>
> It feels like 7 bit ANSII and every character not fitting into those 7
> bit is replaced by a '?' char. Is that true?

Yes.

> The reason why I come up with that: As you said just some minutes
> before, the Euro symbol really is a 0xA4 byte. That byte and even the
> German chars äöü are not only _displayed_ as '?', they are _replaced_. I
> verified that by copying the same string to gedit.
>
>
> As you explained it above, the reason for viewing as source is
> displaying the headers, not the body. So there is nothing wrong with
> that. Just wondering, why chars apparently get substituted in the body
> string opposed to just let them untouched and as-is. Isn't that extra
> work?

GtkHTML only displays UTF8. The input message is not UTF8, so it has to
be converted *somehow*. The conversion we make is to strip out non-ASCII
characters, since without looking at the charset, there's no way to know
what they mean.

-- Dan