pop server access?

pop server access?

[Norbert Koch]

Hi Lars, 

have you made any changes regarding the pop server access? I can't get
onto the server after the installation of 0.35.

Hmm, any point I can start debugging?

TIA, norbert.

-- 
Dr Norbert Koch, DELTA Industrie Informatik GmbH, Fellbach, Germany
A rattlesnake that doesn't bite teaches you nothing!

Re: pop server access?

[Lars Magne Ingebrigtsen]

Norbert Koch <n.koch@delta-ii.de> writes:

> have you made any changes regarding the pop server access? I can't get
> onto the server after the installation of 0.35.

This patch was applied to make things work under Emacs 20.3:

Index: pop3.el
===================================================================
RCS file: /local/lib/CVSROOT/gnus/lisp/pop3.el,v
retrieving revision 5.1
retrieving revision 5.2
diff -c -r5.1 -r5.2
*** pop3.el	1998/08/29 17:43:28	5.1
--- pop3.el	1998/10/19 22:25:09	5.2
***************
*** 111,124 ****
  Returns the process associated with the connection."
    (let ((process-buffer
  	 (get-buffer-create (format "trace of POP session to %s" mailhost)))
! 	(process))
      (save-excursion
        (set-buffer process-buffer)
!       (erase-buffer)
!       (setq pop3-read-point (point-min))
!       )
      (setq process
  	  (open-network-stream "POP" process-buffer mailhost port))
      (let ((response (pop3-read-response process t)))
        (setq pop3-timestamp
  	    (substring response (or (string-match "<" response) 0)
--- 111,126 ----
  Returns the process associated with the connection."
    (let ((process-buffer
  	 (get-buffer-create (format "trace of POP session to %s" mailhost)))
! 	(process)
!         (coding-system-for-read 'no-conversion)
!         (coding-system-for-write 'no-conversion)
!         )
      (save-excursion
        (set-buffer process-buffer)
!       (erase-buffer))
      (setq process
  	  (open-network-stream "POP" process-buffer mailhost port))
+     (setq pop3-read-point (point-min))
      (let ((response (pop3-read-response process t)))
        (setq pop3-timestamp
  	    (substring response (or (string-match "<" response) 0)

It looks a bit weird to me, now that I think about it.  Why has the
`pop3-read-point' setting moved downwards?  I'll move it back up.

Ratinox, I know you don't like Muletilations of pop3, but it is
necessary it it's going to work in that release.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: pop server access?

[Norbert Koch]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Norbert Koch <n.koch@delta-ii.de> writes:
> 
> > have you made any changes regarding the pop server access? I can't get
> > onto the server after the installation of 0.35.
>
> This patch was applied to make things work under Emacs 20.3:
> 

[...]


Hmm, the problem still exists after an update to 0.36. C-g comes up
with

Signaling: (quit)
  accept-process-output(#<network connection "POP" (110 . "mail.delta-ii.de") state:run> 3)
  pop3-read-response(#<network connection "POP" (110 . "mail.delta-ii.de") state:run> t)
  pop3-open-server("mail.delta-ii.de" 110)
  pop3-movemail("/zeus_home/norbert/mail/.gnus-crash-box")
  nnmail-pop3-movemail("po:nok_mail" "/zeus_home/norbert/mail/.gnus-crash-box")
  byte-code("..." [nnmail-movemail-program inbox tofile 0 result] 3)
  nnmail-move-inbox("po:nok_mail")
  nnmail-get-new-mail(nnml nnml-save-nov "/home/norbert/mail/" nil)
  nnml-request-scan(nil "nk")
  gnus-request-scan(nil (nnml "nk"))
  gnus-read-active-file(nil nil)
  gnus-setup-news(nil nil nil)
  byte-code("..." [dont-connect did-connect gnus-group-quit gnus-run-hooks gnus-startup-hook gnus-make-newsrc-file gnus-startup-file gnus-current-startup-file gnus-slave gnus-use-dribble-file gnus-dribble-read-file gnus-use-grouplens bbb-login add-hook gnus-summary-mode-hook gnus-grouplens-mode gnus-setup-news nil level gnus-setup-news-hook gnus-start-draft-setup gnus-group-list-groups gnus-group-first-unread-group gnus-configure-windows group gnus-group-set-mode-line gnus-started-hook] 4)
  gnus-1(nil nil nil)
  gnus(nil)
  call-interactively(gnus)


At this point the system seems to hang (well, I don't know for how
long, I'm sort of impatient :-).

Any leads? Tia, norbert.

-- 
Dr Norbert Koch, DELTA Industrie Informatik GmbH, Fellbach, Germany
A rattlesnake that doesn't bite teaches you nothing!

Re: pop server access?

[Norbert Koch]

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

> "viteno" == Norbert Koch <n.koch@delta-ii.de> writes:
> 
> viteno> Signaling: (quit)
> viteno>   accept-process-output(#<network connection "POP" (110
> viteno>   . "mail.delta-ii.de") state:run> 3)
> 
> What happens when, from a shell on the same machine as you are running
> Emacs, you type 'telnet mail.delta-ii.de 110'?

Hi,

odd enough this works like a charm. I'm also able to pop my mail using
earlier, ie pre 0.35, versions from pgnus. All the fuzz has begun
happening using 0.35/6, damn :-(

/norbert.

-- 
Dr Norbert Koch, DELTA Industrie Informatik GmbH, Fellbach, Germany
A rattlesnake that doesn't bite teaches you nothing!

Re: pop server access?

[Norbert Koch]

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

> "viteno" == Norbert Koch <n.koch@delta-ii.de> writes:
> 
> >> What happens when, from a shell on the same machine as you are running
> >> Emacs, you type 'telnet mail.delta-ii.de 110'?
> 
> viteno> Hi,
> 
> viteno> odd enough this works like a charm.
> 
> I need to know what happens when you telnet to your pop server directly,
> from the machine that is having problems, or I am never going to be able to
> figure out what is failing.

Hi ratinox,

first of all, thanks for the time you spend on helping me solving this 
problem (it still exists on 0.39)

So, here we go. I've written a simple test mail and retrieved it by
telnet.

(apollonius 3:3)$ telnet mail 110
Trying 192.9.200.11...
Connected to mail.delta-ii.de.
Escape character is '^]'.
+OK QPOP (version 2.2) at apollonius.delta-ii.de starting.  
user nok_mail
+OK Password required for nok_mail.
pass <you don't need to know my password, do you :-)>
+OK nok_mail has 1 message (415 octets).
list
+OK 1 messages (415 octets)
1 415
.
retr 1
+OK 415 octets
Return-Path: <norbert>
Received: by mail.delta-ii.de (Smail3.2.0.101/lists.delta-ii.de)
        via Delta-II
        id m0zXOcX-000AAcC; Sun, 25 Oct 1998 12:46:05 +0100 (MET)
Message-Id: <m0zXOcX-000AAcC@mail.delta-ii.de>
Date: Sun, 25 Oct 1998 12:46:05 +0100 (MET)
From: norbert (Norbert Koch)
To: nok_mail@delta-ii.de
subject: pop access test
X-UIDL: a734e2360b4080d73ed014c285d1c5a4

simple mail for pop access.

.
dele 1
+OK Message 1 has been deleted.
quit
+OK Pop server at apollonius.delta-ii.de signing off.
Connection closed by foreign host.
(apollonius 3:4)$ 


All in all, this looks just fine to me. 

Tia, norbert.

-- 
DELTA Industrie Informatik GmbH, Fellbach, Germany (+49.711.5715137)
A rattlesnake that doesn't bite teaches you nothing!

Re: pop server access?

[Norbert Koch]

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

> "viteno" == Norbert Koch <n.koch@delta-ii.de> writes:

> Though you might want to update your qpopper to 2.52 RSN.
> 
> You might also wish to try the pop3.el that is with pgnus 0.39.  If that
> does not work, then there is a problem much deeper in FSF Emacs.

I've now updated to qpopper-2.52 and pgnus-0.40 and still run into the
same troubles.

The one thing that strikes me is that I'm the only one who has this
problem. Therefore, I guess, the real culprit must lie somewhere else
in the system configuration, although I can't remember of any other
things I could have changed in the meantime.

Oh what the hell, I postpone the issue for a couple of days. Maybe,
(the) something slips back into my mind :-)

Cheers, norbert.

-- 
Cursed, cursed creator! Why did I live? Why, in that instant, did I
not extinguish the spark of existance that you had so wantonly
bestowed?

Re: pop server access?

[Norbert Koch]

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

> "viteno" == Norbert Koch <n.koch@delta-ii.de> writes:
> 
> viteno> I've now updated to qpopper-2.52 and pgnus-0.40 and still
> viteno> run into the same troubles.
> 
> Does it work if you start emacs with --unibyte?

Yup, no help with it.

<offtopic>
since you reminded me ... I had to switch back to the old qpopper
version, since nobody here was able to retrieve mail with the new
version installed. It didn't accept any passwords. I have shadow
password enabled. Do you know about any problems with this (not
shadowed passwords, fetched via NIS, seem to work)?
</offtopic>

L8er, norbert.

-- 
Cursed, cursed creator! Why did I live? Why, in that instant, did I
not extinguish the spark of existance that you had so wantonly
bestowed?

Re: pop server access?

[Shenghuo ZHU]

>>>>> "RLP" == Stainless Steel Rat <ratinox@peorth.gweep.net> writes:
User-Agent: Gnus/5.07004 (Pterodactyl Gnus v0.40) Emacs/20.3

>>>>> "UE" == Urban Engberg <ue@cci.dk> writes:
User-Agent: Gnus/5.070035 (Pterodactyl Gnus v0.35) Emacs/20.3

>> "viteno" == Norbert Koch <n.koch@delta-ii.de> writes:
User-Agent: Gnus/5.07004 (Pterodactyl Gnus v0.40) XEmacs/21.2(beta3) (Aglaia)

viteno> I've now updated to qpopper-2.52 and pgnus-0.40 and still run into the
viteno> same troubles.

RLP> Does it work if you start emacs with --unibyte?

viteno> Yup, no help with it.

UE> Could it be that you run Emacs in unibyte-mode, then?  It was I who
UE> made the bug report, after detecting that pop3.el did not work in
UE> multibyte-mode (CRLF pairs from the pop server were translated into
UE> mere CR's).  Switching to unibyte made it work for me before the fix. 

This bug is because of MULE. Stainless Steel Rat fixed it in FSF Emacs
20 by adding 
	(coding-system-for-read 'no-conversion)   ;; because FSF Emacs 20
	(coding-system-for-write 'no-conversion)  ;; is st00pid

But Norbert Koch's pop3 does not work because of XEmacs 21.2.

My solution is using "binary" as the coding system. "binary" is alias
of "no-conversion" in FSF Emacs 20, "nil" in XEmacs 20 (hope it works
in XEmacs 21). The patch is attached.

-- 
Shenghuo

:- cut -----
--- ChangeLog	1998/11/04 07:24:09	1.4
+++ ChangeLog	1998/11/04 07:25:23
@@ -1,3 +1,8 @@
+Wed Nov  4 02:24:39 1998  Shenghuo ZHU  <zsh@cs.rochester.edu>
+
+	* pop3.el (pop3-open-server): Use "binary" instead of
+ 	"no-conversion".
+
 Sun Nov  1 01:26:42 1998  Shenghuo ZHU  <zsh@cs.rochester.edu>
 
 	* gnus-srvr.el (gnus-browse-foreign-server): Set

--- pop3.el	1998/11/04 07:13:24	1.1
+++ pop3.el	1998/11/04 07:16:54
@@ -112,8 +112,8 @@
   (let ((process-buffer
 	 (get-buffer-create (format "trace of POP session to %s" mailhost)))
 	(process)
-	(coding-system-for-read 'no-conversion)   ;; because FSF Emacs 20
-	(coding-system-for-write 'no-conversion)  ;; is st00pid
+	(coding-system-for-read 'binary)   ;; because FSF Emacs 20 and
+	(coding-system-for-write 'binary)  ;; XEmacs 20/1 are st00pid 
     )
     (save-excursion
       (set-buffer process-buffer)

Re: pop server access?

[Jan Vroonhof]

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

> I dunnow which is worse, FSF Emacs for being st00pid, or XEmacs for being
> st00pid in a diametrically opposed fashion.

[I think in this case FSF for being a changing target to stay
compatible with and ignoring the prior mule implementation in Xemacs.
However it does't really matter.]

Sigh, it just that
there is

binary    -- Fully transparant
raw-text  -- just EOL conversion

These work in both Emacsen.

On FSF Emacs no-conversion == binary and the documentation talks about 
no-conversion in that context.

On FSF XEmacs no-conversion == raw-text and the documentation talks about 
no-conversion in that context (minibuffer also shows the raw text
coding system as NoConv).

So if we could just get both .texi files to just talk about binary and 
raw-text and let no-conversion die that would be very nice.

Jan

P.S. To add to your pain there is also EOL conversion in non-mule
Emacsen (for DOS/Win support). 

Re: pop server access?

[Urban Engberg]

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

> Lars> Ratinox, I know you don't like Muletilations of pop3, but it is
> Lars> necessary it it's going to work in that release.
> 
> I must deny statement.  I have been using an un-muletilated version of
> pop3.el (1.3l) with FSF Emacs 20.3 and pgnus since about the time of the
> first pgnus release, and have not had any problems with it (none that I
> have noticed, anyway).

Could it be that you run Emacs in unibyte-mode, then?  It was I who
made the bug report, after detecting that pop3.el did not work in
multibyte-mode (CRLF pairs from the pop server were translated into
mere CR's).  Switching to unibyte made it work for me before the fix. 

  Urban

-- 
ue@cci.dk

Re: pop server access?

[Urban Engberg]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> It looks a bit weird to me, now that I think about it.  Why has the
> `pop3-read-point' setting moved downwards?  I'll move it back up.

Probably because I submitted a bugreport and patch to rms, which was
based on the pop3.el of Emacs 20.3, where the setting is positioned
down there.  I don't know why.

-- 
ue@cci.dk