pop3.el & multibyte problems

pop3.el & multibyte problems

[Jean-Yves Perrier]

In order to make some more tests with pgnus, I tried the following:

I remove my (standard-display-european t) in .emacs

Now when I start gnus: it hangs forever after

Loading pop3...done

and before asking me for my pop password

Note: I'm using pop3.el and Emacs 20.3; set-language-environment may be
set to Latin-1 or not: no difference.

Any idea?
-- 
Jean-Yves Perrier

Re: pop3.el & multibyte problems

[Dave Love]

>>>>> "Rat" == Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

 Rat> "JP" == Jean-Yves Perrier <perrier@nagra-kudelski.ch> writes:

 JP> Note: I'm using pop3.el and Emacs 20.3; set-language-environment
 JP> may be set to Latin-1 or not: no difference.

[I don't have the context since the sunsite server seems to have
zapped back articles.]

 Rat> Should not make any difference whatsoever, since pop3.el does
 Rat> nothing with with languages or character sets or coding systems
 Rat> or any of that crap.  As far as it is concerned, messages are
 Rat> data.

It currently uses `pop3-movemail-file-coding-system' when appending to
the crashbox but I'd have thought it should also specify binary for
the network connexion in the Brave New Multibyte World.

Re: pop3.el & multibyte problems

[Michael Welsh Duggan]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> "JP" == Jean-Yves Perrier <perrier@nagra-kudelski.ch> writes:
> 
> JP> Note: I'm using pop3.el and Emacs 20.3; set-language-environment may be
> JP> set to Latin-1 or not: no difference.
> 
> Should not make any difference whatsoever, since pop3.el does nothing with
> with languages or character sets or coding systems or any of that crap.  As
> far as it is concerned, messages are data.

Maybe so, but it can make a difference anyway.  I fixed any multibyte
problems when using pop3.el by binding coding-system-for-read to
'binary in pop3-open-server.  I also bound it in pop3-read-response,
but subsequent research shows that that is probably not necessary.

-- 
Michael Duggan
(md5i@cs.cmu.edu)
.

Re: pop3.el & multibyte problems

[Hrvoje Niksic]

Dave Love <d.love@dl.ac.uk> writes:

>  Rat> Should not make any difference whatsoever, since pop3.el does
>  Rat> nothing with with languages or character sets or coding systems
>  Rat> or any of that crap.  As far as it is concerned, messages are
>  Rat> data.
> 
> It currently uses `pop3-movemail-file-coding-system'

Not in `pop3.el' distributed with Gnus.

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
We are all just prisoners here of our own MAKEDEV.

Re: pop3.el & multibyte problems

[Jean-Yves Perrier]

Hrvoje Niksic <hniksic@srce.hr> writes:
> Not in `pop3.el' distributed with Gnus.

Does that mean that there is another pop3.el around? 

-- 
Jean-Yves

Re: pop3.el & multibyte problems

[Dave Love]

>>>>> "HN" == Hrvoje Niksic <hniksic@srce.hr> writes:

 >> It currently uses `pop3-movemail-file-coding-system'

 HN> Not in `pop3.el' distributed with Gnus.

Sorry.  It's only in my local copy.  I thought Lars had merged this
from the Emacs 20.3 version, which I reckoned might avoid grief.  I
don't know if it does avoid it, as I haven't had to use POP yet.

BTW, it looks as though `file-name-coding-system' still needs binding
too, whereever `path-name-coding-system' is.  (Morioka 1997-10-21 change.)

1997-08-28  Miyashita Hisashi  <himi@etl.go.jp>

	* pop3.el (pop3-movemail-file-coding-system): Append it for
	assigning a coding system to receive mail with pop3.
	(pop3-movemail): Modify for writing messages with
	pop3-movemail-file-coding-system.

*** pop3.el	1997/04/16 22:13:18	1.1
--- pop3.el	1997/08/28 04:51:02	1.2
***************
*** 60,65 ****
--- 60,68 ----
    "Timestamp returned when initially connected to the POP server.
  Used for APOP authentication.")
  
+ (defvar pop3-movemail-file-coding-system nil
+   "Crashbox made by pop3-movemail with this coding system.")
+ 
  (defvar pop3-read-point nil)
  (defvar pop3-debug nil)
  
***************
*** 85,91 ****
        (pop3-retr process n crashbuf)
        (save-excursion
  	(set-buffer crashbuf)
! 	(append-to-file (point-min) (point-max) crashbox)
  	(set-buffer (process-buffer process))
  	(while (> (buffer-size) 5000)
  	  (goto-char (point-min))
--- 88,95 ----
        (pop3-retr process n crashbuf)
        (save-excursion
  	(set-buffer crashbuf)
! 	(let ((coding-system-for-write pop3-movemail-file-coding-system))
! 	  (append-to-file (point-min) (point-max) crashbox))
  	(set-buffer (process-buffer process))
  	(while (> (buffer-size) 5000)
  	  (goto-char (point-min))

Re: pop3.el & multibyte problems

[Hrvoje Niksic]

Jean-Yves Perrier <perrier@nagra-kudelski.ch> writes:

> Hrvoje Niksic <hniksic@srce.hr> writes:
> > Not in `pop3.el' distributed with Gnus.
> 
> Does that mean that there is another pop3.el around?

The pop3.el is FSF Emacs distribution is definitely not the same as
pop3.el in standalone Gnus distribution.  The former is infested with
coding-systems, which Stailness Steel Rat disapproves of.

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
4.  Thou shalt not warlorde a sig if it bee the sig of Kibo, nor if
    it bee the sig of the Inner Circle.