Re: '.' at the beginning of summary buffer lines

[Nuutti Kotivuori <nuutti.kotivuori@smarttrust.com>]

Simon Josefsson wrote:
> Nuutti Kotivuori <nuutti.kotivuori@smarttrust.com> writes:
> 
>>> Anyone know how to insert the U+2018 and U+2019 characters in
>>> emacs?
>>
>> Assuming MULE-UCS - insert-ucs-character. #x2018 works in XEmacs,
>> ?\x2018 in Emacs, I believe - the function takes a simple number
>> from the minibuffer as input. However, the documentation is in
>> MULE-UCS package, you can refer to it there.
> 
> So ****this**** should look nice to everyone that manages UTF-8?

It didn't :) I replaced a whole load of crap with stars so the message
does not get split into 5 charset parts when sending ;)

> It looks nice to me, with a vanilla Emacs 21 and vanilla RedHat
> fonts.

After sending?

> Err, Gnus won't let me post this mail.  QP encoding fails.  It seems
> to think the (emacs) character set for this buffer should be
> `eight-bit-control' instead of `mule-unicode-0100-24ff'.

Hmm. Maybe mm-mime-mule-charset-alist might be in error. I think that
one uses unicode-a, unicode-b, ... for charsets, not
mule-unicode-something.

And another thing is in mm-find-mime-charset-region - it uses a rather
dumb logic in picking utf-8 as the encoding.

> Doesn't Gnus handle the Unicode stuff in Emacs 21?

No idea - I have gotten it to work in XEmacs with MULE-UCS and a small
hack to Gnus code.

[...]

> (I'm manually forcing this mail to be sent anyway, if you edit the
> charset=eight-bit-control into charset=utf-8 maybe it works.)

I got an error in QP decoding :)

>> Character set       ‘   ’
>> --------------------------
>> Macintosh Roman    212 213
>> Macintosh Cyrillic 212 213
>> ISO 8859-7:1987    161 162
>>
>> If you want to get the ISO 8859-7 character, you can say:
>>
>> (make-char 'greek-iso8859-7 161)
>> (make-char 'greek-iso8859-7 162)
> 
> Doesn't seem to work for me, the characters look like a upside down
> !  and a cents character with a vertical line.

Well, those two characters in the ISO 8859-7:1987 are defined to be:

Hex | Dec | Chr | Code | ISO/IEC 10646-1:2000 Character Name
----+-----+-----+------+----------------------------
 A1 | 161 |  ‘  | 8216 | LEFT SINGLE QUOTATION MARK  
 A2 | 162 |  ’  | 8217 | RIGHT SINGLE QUOTATION MARK

How did you see the characters I put in the message? (after Character
set, up there). I put them in the buffer with insert-ucs-character. If
I am not entirely wrong, something bugs if you don't see them as
quotation marks - or can't insert them as such. The sending charset
for the message appears to be ISO-8859-7.

-- Naked