death to qp

death to qp

[Karl Kleinpaste]

I am tired of finding Gnus generating quoted-printable "for" me.  I
never, ever, ever, ever, EVER in my WORST NIGHTMARE want to find that
Gnus has sent qp-encoded crap _anywhere_ -- not mail, not news, not
nothing.  To me, qp-encoding is an annoying fly buzzing around my
head, and I want to swat it with a Buick because I like annoying flies
really, REALLY flat.  What's the most effective personal configuration
twiddle for squishing qp flat for all future eternity?  8bit will do
just fine for me.  Honest.

Re: death to qp

[Lloyd Zusman]

Karl Kleinpaste <karl@charcoal.com> writes:

> I am tired of finding Gnus generating quoted-printable "for" me.  I
> never, ever, ever, ever, EVER in my WORST NIGHTMARE want to find that
> Gnus has sent qp-encoded crap _anywhere_ -- not mail, not news, not
> nothing.  To me, qp-encoding is an annoying fly buzzing around my
> head, and I want to swat it with a Buick because I like annoying flies
> really, REALLY flat.  What's the most effective personal configuration
> twiddle for squishing qp flat for all future eternity?  8bit will do
> just fine for me.  Honest.

But we're all friends here.  You don't need to mince words with us.
Don't hold back ... tell us what you really feel about
quoted-printable.

:)


-- 
 Lloyd Zusman
 ljz@asfast.com

Re: death to qp

[Karl Kleinpaste]

Yes, well, sometime back there was discussion on this sort of thing,
and the suggestion then was to modify the definition of this variable,
found in mm-encode.el:

(setq mm-content-transfer-encoding-defaults
      '(("text/x-patch" 8bit)
	("text/.*" 8bit)
	("image/.*" base64)
	("message/rfc822" 8bit)
	("application/emacs-lisp" 8bit)
	("application/x-patch" 8bit)
	(".*" qp-or-base64)))

This changes the default for "text/.*" from qp-or-base64 to 8bit, but
it doesn't seem to be having the right effect.  I'd like to know why.

Re: death to qp

[Florian Weimer]

Karl Kleinpaste <karl@charcoal.com> writes:

> This changes the default for "text/.*" from qp-or-base64 to 8bit, but
> it doesn't seem to be having the right effect.  I'd like to know why.

Have you looked at "gnus-group-posting-charset-alist"?

Re: death to qp

[Bjørn Mork]

Karl Kleinpaste <karl@charcoal.com> writes:

> Yes, well, sometime back there was discussion on this sort of thing,
> and the suggestion then was to modify the definition of this variable,
> found in mm-encode.el:
> 
> (setq mm-content-transfer-encoding-defaults
>       '(("text/x-patch" 8bit)
> 	("text/.*" 8bit)
> 	("image/.*" base64)
> 	("message/rfc822" 8bit)
> 	("application/emacs-lisp" 8bit)
> 	("application/x-patch" 8bit)
> 	(".*" qp-or-base64)))
> 
> This changes the default for "text/.*" from qp-or-base64 to 8bit, but
> it doesn't seem to be having the right effect.  I'd like to know why.

I think the variable gnus-group-posting-charset-alist is what you are
looking for.


Bjørn

Re: death to qp

[Karl Kleinpaste]

"Bjørn Mork" <bmork@dod.no> writes:
> I think the variable gnus-group-posting-charset-alist is what you are
> looking for.

Thanx to Bjørn and Florian for the mention of -posting-charset-alist.

However, I don't understand why Gnus is doing this to me.  I have the
default value for this variable, of course (since I didn't even know
it was there), and its docstring from gnus-msg.el is:

     "Alist of regexps and permitted unencoded charsets for posting.
     Each element of the alist has the form (TEST HEADER BODY-LIST), where
     TEST is either a regular expression matching the newsgroup header or a
     variable to query,
     HEADER is the charset which may be left unencoded in the header (nil
     means encode all charsets),
>>>  BODY-LIST is a list of charsets which may be encoded using 8bit
>>>  content-transfer encoding in the body, or one of the special values
>>>  nil (always encode using quoted-printable) or t (always use 8bit).

     Note that any value other tha nil for HEADER infringes some RFCs, so
     use this option with care."

Now, call me crazy if you must, but the last list element is
     (message-this-is-news nil t)
where the trailing t means "always use 8bit", yet when I posted to
news.software.nntp last evening and used superscripts (¹,²) in the
text, Gnus nonetheless hurt me by qp-encoding.

I'm still mystified, and would be grateful for enlightenment.

--karl

Re: death to qp

[Florian Weimer]

Karl Kleinpaste <karl@charcoal.com> writes:

> I'm still mystified, and would be grateful for enlightenment.

Let me guess: Your XEmacs doesn't have MULE support, does it? ;)

Re: death to qp

[Karl Kleinpaste]

Florian Weimer <fw@deneb.cygnus.argh.org> writes:
> Let me guess: Your XEmacs doesn't have MULE support, does it? ;)

XEmacs 21.2  (beta31) "Iris" [Lucid] (i686-pc-linux, Mule) of Wed Feb 23 2000 on mesquite.charcoal.com

Re: death to qp

[Karl EICHWALDER]

Florian Weimer <fw@deneb.cygnus.argh.org> writes:

> Have you looked at "gnus-group-posting-charset-alist"?

Okay, I switched back to the default.  8bit posting happens as wanted
(News), but these headers are used:

    MIME-Version: 1.0
    Content-Type: text/plain; charset=us-ascii
    Content-Transfer-Encoding: 8bit

Why not "charset=iso-8859-1"?

Another general question: I "customized"
gnus-group-posting-charset-alist some time ago; now the format of this
variable changed (note the "mismatch" flag):

gnus-group-posting-charset-alist: [Hide] 
'(("^no\\." iso-8859-1)
  (message-this-is-mail nil)
  ("de.*" nil)
  (".*" iso-8859-1)
  (message-this-is-news iso-8859-1))
   [State]: this option has been set and saved. (mismatch)
Alist of regexps and permitted unencoded charsets for posting. [More]

Is there a function to check all "customized" settings?

-- 
work : ke@suse.de                          |
     : http://www.suse.de/~ke/             |          ------    ,__o
home : ke@gnu.franken.de                   |         ------   _-\_<,
     : http://www.franken.de/users/gnu/ke/ |        ------   (*)/'(*)

Re: death to qp

[Florian Weimer]

Karl EICHWALDER <ke@gnu.franken.de> writes:

>     MIME-Version: 1.0
>     Content-Type: text/plain; charset=us-ascii
>     Content-Transfer-Encoding: 8bit
> 
> Why not "charset=iso-8859-1"?

Have you set `message-default-charset'? What does `M-x apropos-value
RET us-ascii RET' say?  (Beware, this command can take a few minutes
to execute.)

> gnus-group-posting-charset-alist: [Hide] 
> '(("^no\\." iso-8859-1)
>   (message-this-is-mail nil)
>   ("de.*" nil)
>   (".*" iso-8859-1)
>   (message-this-is-news iso-8859-1))
>    [State]: this option has been set and saved. (mismatch)
> Alist of regexps and permitted unencoded charsets for posting. [More]

The current default settings should do the correct thing.

Re: death to qp

[Kai Großjohann]

Karl EICHWALDER <ke@gnu.franken.de> writes:

> Why not "charset=iso-8859-1"?

When you post, Gnus looks at your message.  If it contains only ASCII
characters, Gnus will say charset=us-ascii.  Only if it contains
non-ASCII characters, Gnus will say charset=iso-8859-1 or something
else.

kai
-- 
Beware of flying birch trees.

Re: death to qp

[Daniel Neri]

Karl Kleinpaste <karl@charcoal.com> writes:

> (setq mm-content-transfer-encoding-defaults
>       '(("text/x-patch" 8bit)
> 	("text/.*" 8bit)
> 	("image/.*" base64)
> 	("message/rfc822" 8bit)
> 	("application/emacs-lisp" 8bit)
> 	("application/x-patch" 8bit)
> 	(".*" qp-or-base64)))
> 
> This changes the default for "text/.*" from qp-or-base64 to 8bit, but
> it doesn't seem to be having the right effect.  I'd like to know why.

I asked a similar question a while back; until a good answer appears,
I've settled for the following addition to my ~/.gnus.el:

,----
| ;; qp sux!
| (setq mm-body-charset-encoding-alist '((iso-8859-1 . 8bit)))
`----


Regards,
/Daniel

-- 
Daniel Neri
dne@mayonnaise.net

Re: death to qp

[Karl EICHWALDER]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> When you post, Gnus looks at your message.  If it contains only ASCII
> characters, Gnus will say charset=us-ascii.  Only if it contains
> non-ASCII characters, Gnus will say charset=iso-8859-1 or something
> else.

In theory, but there's a bug somehow somewhere left; I hope this example
will make it through the list:

X-From-Line: nobody Sun Apr 30 19:02:46 2000
Sender: ke@tux.gnu.franken.de
Newsgroups: knf.test
Subject: =?iso-8859-1?q?b=E4sser?= (Re: article unavailable)
References: <slrn8gie0u.3sp.uplawski@sulla.franken.de> <shem7nfs82.fsf@tux.gnu.franken.de> <shaeibfrqu.fsf@tux.gnu.franken.de>
Mail-copies-to: never
From: Karl Eichwalder <ke@gnu.franken.de>
Date: 30 Apr 2000 19:02:45 +0200
Message-ID: <sh66szfrnu.fsf_-_@tux.gnu.franken.de>
Organization: The Moon On Earth
User-Agent: Gnus/5.0805 (Gnus v5.8.5) Emacs/20.6
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 8bit
Lines: 12
Xref: tux.gnu.franken.de news.2000-04:12
X-Gnus-Article-Number: 12   Sun Apr 30 19:02:46 2000

Karl Eichwalder <ke@gnu.franken.de> writes:

> Karl Eichwalder <ke@gnu.franken.de> writes:
> 
> > Täßt
> 
> 567

-- 
Karl EICHWALDER ....................................... GNU/Linux 2.2.14

  7:02pm  up 4 days, 17:22,  7 users,  load average: 0.00, 0.00, 0.10

-- 
work : ke@suse.de                          |
     : http://www.suse.de/~ke/             |          ------    ,__o
home : ke@gnu.franken.de                   |         ------   _-\_<,
     : http://www.franken.de/users/gnu/ke/ |        ------   (*)/'(*)

Re: death to qp

[Karl EICHWALDER]

Florian Weimer <fw@deneb.cygnus.argh.org> writes:

> Have you set `message-default-charset'?

I didn't set it: "nil".  Maybe, it would be a good idea to derive it
from the call to `set-language-environment'?  I'm using:

    (set-language-environment "Latin-1")

> What does `M-x apropos-value RET us-ascii RET' say?  (Beware, this
> command can take a few minutes to execute.)

It doesn't take that long; the result is 1242 lines (some of the a
little bit longish) containting a lot of confidental info.  I cannot
post it.  If it's important, I'll try to prepare a clean Emacs session
;)

> > gnus-group-posting-charset-alist: [Hide] 
> > '(("^no\\." iso-8859-1)
> >   (message-this-is-mail nil)
> >   ("de.*" nil)
> >   (".*" iso-8859-1)
> >   (message-this-is-news iso-8859-1))
> >    [State]: this option has been set and saved. (mismatch)
> > Alist of regexps and permitted unencoded charsets for posting. [More]
> 
> The current default settings should do the correct thing.

Of course, I resetted this value before doing my tests.  Using the old
value Gnus produced valid messages but qp'ed.

Thanks for your investigations!

-- 
work : ke@suse.de                          |
     : http://www.suse.de/~ke/             |          ------    ,__o
home : ke@gnu.franken.de                   |         ------   _-\_<,
     : http://www.franken.de/users/gnu/ke/ |        ------   (*)/'(*)

Re: death to qp

[Florian Weimer]

Karl EICHWALDER <ke@gnu.franken.de> writes:

> Florian Weimer <fw@deneb.cygnus.argh.org> writes:
> 
> > Have you set `message-default-charset'?
> 
> I didn't set it: "nil".  Maybe, it would be a good idea to derive it
> from the call to `set-language-environment'?  I'm using:
> 
>     (set-language-environment "Latin-1")

I think this variable isn't used on MULE Emacsen anyway.
> 
> > What does `M-x apropos-value RET us-ascii RET' say?  (Beware, this
> > command can take a few minutes to execute.)
> 
> It doesn't take that long; the result is 1242 lines (some of the a
> little bit longish) containting a lot of confidental info.  I cannot
> post it.  If it's important, I'll try to prepare a clean Emacs session
> ;)

Hmm.  It would be interesting to know if you've set a strange config
option somewhere.  I'm running Emacs 20.6 as well and I don't have any
problems of this kind, that's why I think it's a configuration issue.

Re: death to qp

[Kai Großjohann]

Karl EICHWALDER <ke@gnu.franken.de> writes:

> In theory, but there's a bug somehow somewhere left; I hope this example
> will make it through the list:

This does not happen for me.

In my ~/.gnus, I have this:

(add-to-list 'gnus-group-posting-charset-alist
             '("^de\\." nil (iso-8859-1)))

Otherwise, I use the default setting of
gnus-group-posting-charset-alist.  I then started Gnus, saw your
message, and evaled the following in the scratch buffer:

(add-to-list 'gnus-group-posting-charset-alist
             '(message-this-is-mail iso-8859-1 t))

I then sent myself a mail with 8bit characters in the header and in
the body, and Gnus correctly said `charset=iso-8859-1'.

Strange that it doesn't work for you.  Hm.

kai
-- 
Beware of flying birch trees.

Re: death to qp

[Jari Aalto+mail.emacs]

* 2000-04-30 Karl EICHWALDER <ke@gnu.franken.de> list.emacs-ding
* Message-Id: <shya5vcpq2.fsf@tux.gnu.franken.de>
| Florian Weimer <fw@deneb.cygnus.argh.org> writes:
| 
| > Have you looked at "gnus-group-posting-charset-alist"?
| 
| Okay, I switched back to the default.  8bit posting happens as wanted
| (News), but these headers are used:
| 
|     MIME-Version: 1.0
| 
| Why not "charsetiso-8859-1"?
| 
| Another general question: I "customized"
| gnus-group-posting-charset-alist some time ago; now the format of this
| variable changed (note the "mismatch" flag):
| 
| gnus-group-posting-charset-alist: [Hide] 
| '(("^no\\." iso-8859-1)
|   (message-this-is-mail nil)
|   ("de.*" nil)
|   (".*" iso-8859-1)
|   (message-this-is-news iso-8859-1))
    =================================

Howabout changing this to:

   ((lambda (x) (message-news-p)) iso-8859-1))

Jari