The qpopper list archive ending on 4 May 2001
Topics covered in this issue include:
1. Re: Qpopper 4.0.1 (final) available
Qpopper Bugs <qpopper-bugs at qualcomm dot com>
Wed, 25 Apr 2001 10:55:09 -0700
2. Re: Qpopper 4.0.1 (final) available
Steven Champeon <schampeo at hesketh dot com>
Wed, 25 Apr 2001 14:16:57 -0400
3. Re: FTP site via broswer
"Kenneth Porter" <shiva at well dot com>
Wed, 25 Apr 2001 12:04:24 -0700
4. Re: Qpopper 4.0.1 (final) available
"Kenneth Porter" <shiva at well dot com>
Wed, 25 Apr 2001 12:07:20 -0700
5. Re: Qpopper 4.0.1 (final) available
Steven Champeon <schampeo at hesketh dot com>
Wed, 25 Apr 2001 15:32:06 -0400
6. RPM's
"Kenneth Porter" <shiva at well dot com>
Wed, 25 Apr 2001 13:01:49 -0700
7. Re: RPM's
Steven Champeon <schampeo at hesketh dot com>
Wed, 25 Apr 2001 16:37:41 -0400
8. I/O Error
"Lance Rasmussen" <lance at cdesoftware dot com>
Wed, 25 Apr 2001 13:48:40 -0700
9. Re: Qpopper 4.0.1 (final) available
Qpopper Bugs <qpopper-bugs at qualcomm dot com>
Wed, 25 Apr 2001 14:11:58 -0700
10. ftp.qualcomm.com is empty
"Kenneth Porter" <shiva at well dot com>
Wed, 25 Apr 2001 16:29:31 -0700
11. XTND XMIT (4.0: send on POP broken?)
Jacques Distler <distler at golem.ph.utexas dot edu>
Wed, 25 Apr 2001 23:15:18 -0700
12. Re: XTND XMIT (4.0: send on POP broken?)
Eric Krichbaum <eric at telicsolutions dot net>
Thu, 26 Apr 2001 08:39:14 -0400
13. Re: XTND XMIT (4.0: send on POP broken?)
Jacques Distler <distler at golem.ph.utexas dot edu>
Thu, 26 Apr 2001 10:24:23 -0700
14. Re: XTND XMIT (4.0: send on POP broken?)
"Lance Rasmussen" <lance at cdesoftware dot com>
Thu, 26 Apr 2001 11:58:52 -0700
15. Re: XTND XMIT (4.0: send on POP broken?)
Jacques Distler <distler at golem.ph.utexas dot edu>
Thu, 26 Apr 2001 12:48:36 -0700
16. Error: Unable to get spool name
"Matthew Koch" <mattkoch at genevaonline dot com>
Fri, 27 Apr 2001 15:08:37 -0500
17. Qpopper Client for APOP and TLS/SSL
"Frank P. Troy" <ftroy at pop dot net>
Sat, 28 Apr 2001 14:13:01 -0400
18. Re: Qpopper Client for APOP and TLS/SSL
Chuck Yerkes <chuck+QPopper at yerkes dot com>
Sun, 29 Apr 2001 01:05:32 -0700
19. Invalid cross-device link
Ted George <ted at kcnet dot com>
Mon, 30 Apr 2001 07:44:21 -0500
20. reading from /var/spool/mail AND $HOMEdir?
Peter Rose <peter at cyberscreen dot com>
Mon, 30 Apr 2001 13:58:06 +0100
21. Qpopper/qmail
trini0 <trini0 at optonline dot net>
Mon, 30 Apr 2001 13:54:03 -0400
22. Unable to obtain socket and address of client
John Hogan <jhogan2112 at yahoo dot com>
Mon, 30 Apr 2001 13:55:26 -0500
23. Re: Qpopper Client for APOP and TLS/SSL
"Kenneth Porter" <shiva at well dot com>
Mon, 30 Apr 2001 13:52:53 -0700
24. Re: RPM's
James Ralston <qralston+ml.qpopper at andrew.cmu dot edu>
Mon, 30 Apr 2001 18:50:39 -0400 (EDT)
25. Re: RPM's
"Kenneth Porter" <shiva at well dot com>
Mon, 30 Apr 2001 16:20:20 -0700
26. TLS shutdown errors with MS Outlook clients (4.0.1 LX)
Scott McDermott <mcdermot at questra dot com>
Tue, 1 May 2001 17:46:15 -0400
27. Server mode based on shell
Robert Brewer <rbrewer at lava dot net>
Tue, 01 May 2001 16:59:53 -1000
28. Problems compiling/installing qpopper 4.0.1
Bryce Newall <data at dreamhaven dot org>
Wed, 2 May 2001 12:02:16 -0700 (PDT)
29. Re: Invalid cross-device link
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 12:37:05 -0700
30. Re: Server mode based on shell
Scott McDermott <mcdermot at questra dot com>
Wed, 2 May 2001 15:48:48 -0400
31. RE: Problems compiling/installing qpopper 4.0.1
"Matthew Koch" <mattkoch at genevaonline dot com>
Wed, 2 May 2001 16:00:21 -0500
32. make-install doesn't work for 4.0.1
"Martin, Rich" <rich.martin at reden-anders dot com>
Wed, 2 May 2001 16:14:00 -0500
33. Re: reading from /var/spool/mail AND $HOMEdir?
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 14:53:04 -0700
34. Re: Unable to obtain socket and address of client
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 14:58:18 -0700
35. Re: TLS shutdown errors with MS Outlook clients (4.0.1 LX)
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 15:01:58 -0700
36. Re: Qpopper/qmail
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 14:55:28 -0700
37. Re: Server mode based on shell
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 15:15:06 -0700
38. Re: Hash Directories
Chuck Yerkes <chuck+qpopper at yerkes dot com>
Wed, 2 May 2001 15:59:37 -0700
39. RE: make-install doesn't work for 4.0.1
"Martin, Rich" <rich.martin at reden-anders dot com>
Wed, 2 May 2001 20:08:25 -0500
40. RE: make-install doesn't work for 4.0.1
Randall Gellens <randy at qualcomm dot com>
Wed, 2 May 2001 18:37:29 -0700
41. Re: Server mode based on shell
Scott McDermott <mcdermot at questra dot com>
Wed, 2 May 2001 23:55:17 -0400
42. Re: Problems compiling/installing qpopper 4.0.1
Bryce Newall <data at dreamhaven dot org>
Thu, 3 May 2001 11:59:59 -0700 (PDT)
43. Re: Server mode based on shell
Randall Gellens <randy at qualcomm dot com>
Thu, 3 May 2001 12:58:58 -0700
44. Re: Server mode based on shell
Scott McDermott <mcdermot at questra dot com>
Thu, 3 May 2001 16:03:15 -0400
45. DON'T DELETE ....
"Genghis Rios Kruger" <grios at pucp.edu dot pe>
Thu, 3 May 2001 22:18:25 -0500
46. AW: DON'T DELETE ....
"Philipp Gimm" <philipp at problemchen dot de>
Fri, 4 May 2001 05:35:01 +0200
47. For help on PAM using samba authentication
"Rogerwoo" <rogerwoo7508 at sina.com dot cn>
Fri, 4 May 2001 13:44:08 +0800
48. Re: For help on PAM using samba authentication
"Kenneth Porter" <shiva at well dot com>
Thu, 03 May 2001 23:51:48 -0700
49. Re: For help on PAM using samba authentication
Joseph S D Yao <jsdy at cospo.osis dot gov>
Fri, 4 May 2001 12:18:42 -0400
50. Re: AW: DON'T DELETE ....
Robert Brewer <rbrewer at lava dot net>
Fri, 04 May 2001 11:34:56 -1000
Date: Wed, 25 Apr 2001 10:55:09 -0700
From: Qpopper Bugs <qpopper-bugs at qualcomm dot com>
Subject: Re: Qpopper 4.0.1 (final) available
At 10:05 AM -0400 4/25/01, Steven Champeon wrote:
> I'd also like to request that the kind folks at qpopper pay closer
> attention to this list, where complete patches often fix problems with
> the distros that your subsequent releases ignore. For example: there
> is no rule to make install for poppassd in 4.0.1; there is no definition
> of INSTALL in the Makefile for poppassd; there is still a bug in the
> build rule for poppassd that ignores the settings for ${LIBS} (so that
> if you define -lcrypt it gets added to ${LIBS} but ${LIBS} is not used
> to build the binary, only ${NETWORK_LIBS}; etc.
>
> Of course, you're free to ignore us, but please bear in mind that it is
> your users who often find the gruesome holes in your software and supply
> patches weeks before anyone else does, and if we keep getting ignored,
> we may as well stop submitting patches. Understand that I don't really
> care that my patch was ignored or that I was not given any credit for
> publicly posting the patch. What irks me is that I have to mention that
> the distro still has most of the same bugs it had a week and a half ago.
I apologize for overlooking your patch. Certainly I would have
preferred that it be included in 4.0.1.
--Randy
Date: Wed, 25 Apr 2001 14:16:57 -0400
From: Steven Champeon <schampeo at hesketh dot com>
Subject: Re: Qpopper 4.0.1 (final) available
on Wed, Apr 25, 2001 at 10:55:09AM -0700, Qpopper Bugs wrote:
> I apologize for overlooking your patch. Certainly I would have
> preferred that it be included in 4.0.1.
Just so we're all clear on this, what is the procedure for submitting
patches? Should we submit them to both qpopper-bugs and the qpopper
list? What form of patch should we submit, and what arguments to diff
should be used to create them? I looked at the Web site, but the item
that discusses bug reports:
http://www.eudora.com/qpopper_general/#FEEDBACK
refers the reader to the FAQ:
http://www.eudora.com/qpopper/faq.html
which has no information regarding how to submit a proper bug report
or patch. Might be useful to make that information more public. Next
time (unless directed otherwise) I will submit the patch to both the
bugs list and the discussion list.
Thanks,
Steve
PS: can we expect a 4.0.2? ;)
--
"Euphoria is not a business strategy" -- Louis Rossetto
From: "Kenneth Porter" <shiva at well dot com>
Date: Wed, 25 Apr 2001 12:04:24 -0700
Subject: Re: FTP site via broswer
On Wed, 25 Apr 2001 12:15:01 -0400, Mark wrote:
>Is anyone else having a problem seeing the FTP site via a browser?
>Qualcomm, can you fix this please. Makes it easy to check for updates as opposed to FTPing in
It's not just browsers. I just used the command line client from Linux
and couldn't get a listing of any directory. I just get a 226 Transfer
Complete.
The site banner is properly displayed, though.
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
From: "Kenneth Porter" <shiva at well dot com>
Date: Wed, 25 Apr 2001 12:07:20 -0700
Subject: Re: Qpopper 4.0.1 (final) available
On Wed, 25 Apr 2001 10:05:41 -0400, Steven Champeon wrote:
>This is a polite request to everyone using popper - if you have handy
>patches to qpopper, such as those apparently sent in by both myself
>and Michael C Tiernan to fix the Makefile problems when building
>poppassd, please share them with the list
Is anyone using RPM's? I've been maintaining a spec file but Red Hat
has been having problems with their incoming server for contrib, so I
haven't uploaded a new RPM in quite a while.
As soon as I can get 4.0.1 off the FTP site (currently broken) I can
make a new SRPM available.
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
Date: Wed, 25 Apr 2001 15:32:06 -0400
From: Steven Champeon <schampeo at hesketh dot com>
Subject: Re: Qpopper 4.0.1 (final) available
on Wed, Apr 25, 2001 at 12:07:20PM -0700, Kenneth Porter wrote:
> On Wed, 25 Apr 2001 10:05:41 -0400, Steven Champeon wrote:
>
> >This is a polite request to everyone using popper - if you have handy
> >patches to qpopper, such as those apparently sent in by both myself
> >and Michael C Tiernan to fix the Makefile problems when building
> >poppassd, please share them with the list
>
> Is anyone using RPM's? I've been maintaining a spec file but Red Hat
> has been having problems with their incoming server for contrib, so I
> haven't uploaded a new RPM in quite a while.
I use RPMs for all the non-critical stuff, and SRPMs for critical
stuff where possible, but generally build from source distros when the
build has many options that aren't usually set up the same way I need
them to be (I was running DRAC for a long time, way before qpopper had
DRAC support, for example).
Just another data point,
Steve
--
"Euphoria is not a business strategy" -- Louis Rossetto
From: "Kenneth Porter" <shiva at well dot com>
Date: Wed, 25 Apr 2001 13:01:49 -0700
Subject: RPM's
On Wed, 25 Apr 2001 15:32:06 -0400, Steven Champeon wrote:
>I use RPMs for all the non-critical stuff, and SRPMs for critical
>stuff where possible, but generally build from source distros when the
>build has many options that aren't usually set up the same way I need
>them to be (I was running DRAC for a long time, way before qpopper had
>DRAC support, for example).
I use SRPM's exclusively, mostly so I'll have source on hand if Things
Go Wrong (tm) ;-). I just tweak spec files and add my initials to a new
release number if an SRPM isn't exactly the way I want it.
My current qpopper SRPM enables DRAC support, and I've got another SRPM
for DRAC (package name is "dracd", available in Red Hat contrib), on
which the qpopper SRPM depends.
Does anyone have a good working xinetd config file for pop3? I haven't
yet installed xinetd on my RH6.1 system, but I'd like to provide the
config file in my SRPM for those that use it.
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
Date: Wed, 25 Apr 2001 16:37:41 -0400
From: Steven Champeon <schampeo at hesketh dot com>
Subject: Re: RPM's
on Wed, Apr 25, 2001 at 01:01:49PM -0700, Kenneth Porter wrote:
> I use SRPM's exclusively, mostly so I'll have source on hand if Things
> Go Wrong (tm) ;-). I just tweak spec files and add my initials to a new
> release number if an SRPM isn't exactly the way I want it.
Yup - that's a good way to go.
> My current qpopper SRPM enables DRAC support, and I've got another SRPM
> for DRAC (package name is "dracd", available in Red Hat contrib), on
> which the qpopper SRPM depends.
OK, cool. We're finally moving to SMTP-AUTH, so we (hopefully) won't need
DRAC anymore, but it's good to know it's still around if we need it.
> Does anyone have a good working xinetd config file for pop3? I haven't
> yet installed xinetd on my RH6.1 system, but I'd like to provide the
> config file in my SRPM for those that use it.
Here's what I use:
defaults
{
instances = 25
log_type = SYSLOG daemon info
log_on_success = HOST PID DURATION EXIT
log_on_failure = HOST RECORD
per_source = 20
}
service pop-3
{
flags = REUSE NAMEINARGS
socket_type = stream
protocol = tcp
wait = no
user = root
server = /usr/sbin/in.qpopper
server_args = in.qpopper -s
}
--
"Euphoria is not a business strategy" -- Louis Rossetto
Date: Wed, 25 Apr 2001 13:48:40 -0700
From: "Lance Rasmussen" <lance at cdesoftware dot com>
Subject: I/O Error
--=====_98823172030676=_
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable
Relatively new to FreeBSD.
I installed the new 4.3 release of FreeBSD yesterday and installed the
Qpopper 4.0 package.
Set up everything and got permissions, etc set up.
I am getting the following error
I/O error flushing output to client at (ip) Operation not permitted (1).
The message comes frequently on the screen making things a pain when I am
working on the server.
We have a Window's based CSM program that pulls and sends mail. I set up
the FreeBSD box to act
as a mail server and web server. The CSM program polls for mail every 5
min. There are 3 boxes that
it pulls from.
Suggestions or guidance would be appreciated.
Nothing custom or special has been done other than using the -R flag in the
Inetd.conf for QPopper to get
rid of the reverse DNS messages.
Lance Rasmussen
CDE Software
--=====_98823172030676=_
Content-Type: text/html; charset="us-ascii"
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2462.0" name=GENERATOR></HEAD>
<BODY style="FONT-FAMILY: Arial" text=#000000 bgColor=#ffffff>
<DIV><FONT size=2>Relatively new to FreeBSD.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>I installed the new 4.3 release of FreeBSD yesterday and
installed the Qpopper 4.0 package.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Set up everything and got permissions, etc set
up.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>I am getting the following error</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>I/O error flushing output to client at (ip) Operation
not permitted (1).</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>The message comes frequently on the screen making
things a pain when I am working on the server.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>We have a Window's based CSM program that pulls and sends
mail. I set up the FreeBSD box to act</FONT></DIV>
<DIV><FONT size=2>as a mail server and web server. The CSM program polls
for mail every 5 min. There are 3 boxes that</FONT></DIV>
<DIV><FONT size=2>it pulls from.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Suggestions or guidance would be appreciated.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Nothing custom or special has been done other than using the
-R flag in the Inetd.conf for QPopper to get</FONT></DIV>
<DIV><FONT size=2>rid of the reverse DNS messages.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Lance Rasmussen</FONT></DIV>
<DIV><FONT size=2>CDE Software</FONT></DIV></BODY></HTML>
--=====_98823172030676=_--
Date: Wed, 25 Apr 2001 14:11:58 -0700
From: Qpopper Bugs <qpopper-bugs at qualcomm dot com>
Subject: Re: Qpopper 4.0.1 (final) available
On Wednesday, April 25, 2001, at 11:16 AM, Steven Champeon wrote:
> Just so we're all clear on this, what is the procedure for submitting
> patches? Should we submit them to both qpopper-bugs and the qpopper
> list?
If you're just fixing a bug, send to <qpopper-bugs at qualcomm dot com>. If
the patch does anything else (adds new features, for example), send to
<qpopper-patches at qualcomm dot com>.
From: "Kenneth Porter" <shiva at well dot com>
Date: Wed, 25 Apr 2001 16:29:31 -0700
Subject: ftp.qualcomm.com is empty
I'm trying to pull down the latest qpopper and I'm finding that I can't
get a file listing of any directory in the tree. I was able to guess
the name of the file
(ftp://ftp.qualcomm.com/eudora/servers/unix/popper/qpopper4.0.1.tar.gz)
and download it, though. Looks like it's just directory listings that
are failing.
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
Date: Wed, 25 Apr 2001 23:15:18 -0700
From: Jacques Distler <distler at golem.ph.utexas dot edu>
Subject: XTND XMIT (4.0: send on POP broken?)
-----BEGIN PGP SIGNED MESSAGE-----
"Roy C. Bixler" <rcb at press-gopher.uchicago dot edu> wrote:
I tried to upgrade from qpopper 3.12 to 4.0 on a Linux server, but found
that I had to revert when I got complaints from users of Eudora 3.0 who
use send on POP feature. Their messages were not going out as usual
because the headers of their sent messages suddenly appeared all on one
line. I plan to have those users upgrade to a later Eudora version and
use authenticated SMTP instead, but wonder if the apparent send on POP
breakage is a known issue.
I see the following comment in popper/pop_xmit.c:
* 06/11/00 [rg]
* - Applied patch by Clifton Royston to translate network EOL
* (CRLF) to local EOL ('\n').
* - Now only initial ".\n' is recognized as end-of-msg (instead
* of any line which starts with a "." and ends with a ".").
Could that have anything to do with the problem?
Sorta.
XTND XMIT is, indeed (still) broken (as of 4.0.1). The buffer is written to
the temp file with no '\n' to terminate the line(s) (Doh!).
The following patch fixes it:
*** popper/pop_xmit.c.orig Tue Feb 20 19:15:21 2001
- --- popper/pop_xmit.c Thu Apr 26 00:43:34 2001
***************
*** 165,172 ****
- --- 166,175 ----
fputs ( &buffer[1], tmp );
else
fputs ( buffer, tmp );
+ fputc ( '\n', tmp );
} else
fputs ( buffer, tmp );
+ fputc ( '\n', tmp );
return 0;
}
Jacques
-----BEGIN PGP SIGNATURE-----
Version: PGP
Comment: Public Key - http://golem.ph.utexas.edu/~distler/distler.asc
iQCVAwUBOue9EaIBi34rsX+ZAQGC3QP+Jb5eWn90zp4hNU4jDpYvLYVi2c4fEDc1
EACjC6iQbdbEGgstLjQvtbIDfZtRM7SxNg5Sut3bkbe8GDuI/sFeI+YzAFHdeSOM
YA/54n1fcaABuRGanMIMBVVEWvKzP6RZCCphsFmJwk030JmI8gwD5LHi8i6ZmC2a
6RNPEPA/aSw
=U9aO
-----END PGP SIGNATURE-----
Date: Thu, 26 Apr 2001 08:39:14 -0400
From: Eric Krichbaum <eric at telicsolutions dot net>
Subject: Re: XTND XMIT (4.0: send on POP broken?)
Doesn't this break the if...then... else logic? I didn't/can't test this
because we don't use send on POP but shouldn't this be:
if ( buffer[0] == '.' ) {
/*
* Exit on end of message
*/
if ( buffer[1] == '\0' ) {
DEBUG_LOG0 ( p, "Received end of msg" );
fclose ( tmp );
p->xmitting = FALSE;
Pop ( &(p->InProcess) );
pop_xmit_exec ( p );
}
/*
* sendmail will not remove escaped ..
*/
else {
if ( buffer[1] == '.' )
fputs ( &buffer[1], tmp );
else {
fputs ( buffer, tmp );
fputc ( '\n', tmp );
} else {
fputs ( buffer, tmp );
fputc ( '\n', tmp );
}
return 0;
}
At 11:15 PM 4/25/2001 -0700, Jacques Distler wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>
>"Roy C. Bixler" <rcb at press-gopher.uchicago dot edu> wrote:
>
>I tried to upgrade from qpopper 3.12 to 4.0 on a Linux server, but found
>that I had to revert when I got complaints from users of Eudora 3.0 who
>use send on POP feature. Their messages were not going out as usual
>because the headers of their sent messages suddenly appeared all on one
>line. I plan to have those users upgrade to a later Eudora version and
>use authenticated SMTP instead, but wonder if the apparent send on POP
>breakage is a known issue.
>
>I see the following comment in popper/pop_xmit.c:
>
> * 06/11/00 [rg]
> * - Applied patch by Clifton Royston to translate network EOL
> * (CRLF) to local EOL ('\n').
> * - Now only initial ".\n' is recognized as end-of-msg (instead
> * of any line which starts with a "." and ends with a ".").
>
>Could that have anything to do with the problem?
>
>
>Sorta.
>
>XTND XMIT is, indeed (still) broken (as of 4.0.1). The buffer is written to
>the temp file with no '\n' to terminate the line(s) (Doh!).
>
>The following patch fixes it:
>
>*** popper/pop_xmit.c.orig Tue Feb 20 19:15:21 2001
>- --- popper/pop_xmit.c Thu Apr 26 00:43:34 2001
>***************
>*** 165,172 ****
>- --- 166,175 ----
> fputs ( &buffer[1], tmp );
> else
> fputs ( buffer, tmp );
>+ fputc ( '\n', tmp );
> } else
> fputs ( buffer, tmp );
>+ fputc ( '\n', tmp );
> return 0;
> }
>
>Jacques
>
>-----BEGIN PGP SIGNATURE-----
>Version: PGP
>Comment: Public Key - http://golem.ph.utexas.edu/~distler/distler.asc
>
>iQCVAwUBOue9EaIBi34rsX+ZAQGC3QP+Jb5eWn90zp4hNU4jDpYvLYVi2c4fEDc1
>EACjC6iQbdbEGgstLjQvtbIDfZtRM7SxNg5Sut3bkbe8GDuI/sFeI+YzAFHdeSOM
>YA/54n1fcaABuRGanMIMBVVEWvKzP6RZCCphsFmJwk030JmI8gwD5LHi8i6ZmC2a
>6RNPEPA/aSw
>=U9aO
>-----END PGP SIGNATURE-----
--------------------------------------------------------------------------------------------------
New software available at http://www.telicsolutions.net/soft
If you have ICQ you can message me at ICQ#:9736582
Eric Krichbaum, MCSE, MCP+IS, ASE, CCNA, A+, CNA, developer
--------------------------------------------------------------------------------------------------
Date: Thu, 26 Apr 2001 10:24:23 -0700
From: Jacques Distler <distler at golem.ph.utexas dot edu>
Subject: Re: XTND XMIT (4.0: send on POP broken?)
-----BEGIN PGP SIGNED MESSAGE-----
At 8:39 AM -0400 4/26/01, Eric Krichbaum wrote:
>Doesn't this break the if...then... else logic? I didn't/can't
>test this because we don't use send on POP but shouldn't this be:
> [snip]
I tested the previous patch, and it did work (actually, I left off a
"fputc(...)" which swallows a newline for an "escaped" '.'),
but the pagination of that section
of code was bad enough that it sure didn't look right.
Anyway, here, properly paginated, is what that section of code should
look like, followed by the patch. I have tested this (PGP signatures
verify, including for messages with escaped '.'s).
Jacques
%%%%%%%%%%%%% the code %%%%%%%%%%%%%%%%%%
int
pop_xmit_recv ( p, buffer )
POP *p;
char *buffer;
{
/*
* Data written to a file should follow UNIX or local EOL format ('\n').
* sendmail and some MTAs will ignore superfluous CRs, but not all will.
*/
/*
* Look for initial period
*/
DEBUG_LOG2 ( p, "Receiving (%d): \"%.128s\"", strlen(buffer), buffer );
if ( buffer[0] == '.' ) {
/*
* Exit on end of message
*/
if ( buffer[1] == '\0' ) {
DEBUG_LOG0 ( p, "Received end of msg" );
fclose ( tmp );
p->xmitting = FALSE;
Pop ( &(p->InProcess) );
pop_xmit_exec ( p );
}
/*
* sendmail will not remove escaped ..
*/
else {
if ( buffer[1] == '.' ) {
fputs ( &buffer[1], tmp );
fputc ( '\n', tmp );
}
else {
fputs ( buffer, tmp );
fputc ( '\n', tmp );
}
}
}
else {
fputs ( buffer, tmp );
fputc ( '\n', tmp );
}
return 0;
}
%%%%%%%%%%%%%% the patch %%%%%%%%%%%%%%
*** popper/pop_xmit.c.orig Tue Feb 20 19:15:21 2001
- --- popper/pop_xmit.c Thu Apr 26 11:59:55 2001
***************
*** 161,172 ****
/*
* sendmail will not remove escaped ..
*/
! else if ( buffer[1] == '.' )
! fputs ( &buffer[1], tmp );
! else
! fputs ( buffer, tmp );
! } else
fputs ( buffer, tmp );
return 0;
}
- --- 162,182 ----
/*
* sendmail will not remove escaped ..
*/
! else {
! if ( buffer[1] == '.' ) {
! fputs ( &buffer[1], tmp );
! fputc ( '\n', tmp );
! }
! else {
! fputs ( buffer, tmp );
! fputc ( '\n', tmp );
! }
! }
! }
! else {
fputs ( buffer, tmp );
+ fputc ( '\n', tmp );
+ }
return 0;
}
-----BEGIN PGP SIGNATURE-----
Version: PGP
Comment: Public Key - http://golem.ph.utexas.edu/~distler/distler.asc
iQCVAwUBOuhZ4aIBi34rsX+ZAQEavQQAzO14iuLtgyEVE1TrNoFIINSHu9iUaOOg
/Wx8eX+S4x7D1pS7YIZgYDALqY4jJRSENpPW0LgQwCYT0tcLekbC13m53RyaVuRx
Hcn7AsmrjLKSDi7zSodDfjytO5dF8gDhsmf7dIl2H0rhTsK1wnKqql5yOc5p28bR
wkGqZCjVNno
=r3PC
-----END PGP SIGNATURE-----
Date: Thu, 26 Apr 2001 11:58:52 -0700
From: "Lance Rasmussen" <lance at cdesoftware dot com>
Subject: Re: XTND XMIT (4.0: send on POP broken?)
I tried and this seemed to take care of an issue I was having.
Lance Rasmussen
*********** REPLY SEPARATOR ***********
On 4/25/2001 at 11:15 PM Jacques Distler wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>
>"Roy C. Bixler" <rcb at press-gopher.uchicago dot edu> wrote:
>
>I tried to upgrade from qpopper 3.12 to 4.0 on a Linux server, but found
>that I had to revert when I got complaints from users of Eudora 3.0 who
>use send on POP feature. Their messages were not going out as usual
>because the headers of their sent messages suddenly appeared all on one
>line. I plan to have those users upgrade to a later Eudora version and
>use authenticated SMTP instead, but wonder if the apparent send on POP
>breakage is a known issue.
>
>I see the following comment in popper/pop_xmit.c:
>
> * 06/11/00 [rg]
> * - Applied patch by Clifton Royston to translate network EOL
> * (CRLF) to local EOL ('\n').
> * - Now only initial ".\n' is recognized as end-of-msg
(instead
> * of any line which starts with a "." and ends with a ".").
>
>Could that have anything to do with the problem?
>
>
>Sorta.
>
>XTND XMIT is, indeed (still) broken (as of 4.0.1). The buffer is written
to
>the temp file with no '\n' to terminate the line(s) (Doh!).
>
>The following patch fixes it:
>
>*** popper/pop_xmit.c.orig Tue Feb 20 19:15:21 2001
>- --- popper/pop_xmit.c Thu Apr 26 00:43:34 2001
>***************
>*** 165,172 ****
>- --- 166,175 ----
> fputs ( &buffer[1], tmp );
> else
> fputs ( buffer, tmp );
>+ fputc ( '\n', tmp );
> } else
> fputs ( buffer, tmp );
>+ fputc ( '\n', tmp );
> return 0;
> }
>
>Jacques
>
>-----BEGIN PGP SIGNATURE-----
>Version: PGP
>Comment: Public Key - http://golem.ph.utexas.edu/~distler/distler.asc
>
>iQCVAwUBOue9EaIBi34rsX+ZAQGC3QP+Jb5eWn90zp4hNU4jDpYvLYVi2c4fEDc1
>EACjC6iQbdbEGgstLjQvtbIDfZtRM7SxNg5Sut3bkbe8GDuI/sFeI+YzAFHdeSOM
>YA/54n1fcaABuRGanMIMBVVEWvKzP6RZCCphsFmJwk030JmI8gwD5LHi8i6ZmC2a
>6RNPEPA/aSw
>=U9aO
>-----END PGP SIGNATURE-----
Date: Thu, 26 Apr 2001 12:48:36 -0700
From: Jacques Distler <distler at golem.ph.utexas dot edu>
Subject: Re: XTND XMIT (4.0: send on POP broken?)
-----BEGIN PGP SIGNED MESSAGE-----
On Thu, 26 Apr 2001, "Lance Rasmussen" <lance at cdesoftware dot com> wrote:
>I tried and this seemed to take care of an issue I was having.
>
>Lance Rasmussen
Please note that my first attempt at a patch did not handle escaped
'.'s correctly (ate the subsequent '\n'). The second one included the
fputc ( '\n', tmp );
for that case too.
Cheers,
Jacques
-----BEGIN PGP SIGNATURE-----
Version: PGP
Comment: Public Key - http://golem.ph.utexas.edu/~distler/distler.asc
iQCVAwUBOuh7oqIBi34rsX+ZAQGzjAQA0ePkgL+stKL+jNO84VnXOBRjNje1vxkY
nrhySnTZsyQNH/DcE7wBrM7xhKM74d0E4kM/IbEjfHtimSQFW1ccTz8EtK6xnzMx
1OfH6/rAF7Mt8bJogEG++k89mg01o6qcdrvxRvd0oGYZEzEppJBaFilShbqP+bN7
JAZ1Yym1WK4
=KHSH
-----END PGP SIGNATURE-----
From: "Matthew Koch" <mattkoch at genevaonline dot com>
Subject: Error: Unable to get spool name
Date: Fri, 27 Apr 2001 15:08:37 -0500
I am getting the following error(s) when trying to check
mail.
Apr 27 15:00:02 somehost ./popper[35811]: Insufficient room
to generate path for user mildew; need more than 163; have
only 64
Apr 27 15:00:02 somehost ./popper[35811]: mildew at
somehost.com (111.111.111.111): -ERR [SYS/TEMP] Unable to
get spool name: /home/mildew/
My configure for this was:
./configure --enable-home-dir-mail=.mail --enable-standalone
--enable-server-mode
I also added to the logging in pop_dropcopy.c to get the
spool name logged. I'm running FreeBSD 4.2, procmail 3.15.1,
and sendmail 8.11.3. I have compiled procmail to deliver to
${HOME}/.mail (which it does correctly). I just can't seem
to get qpopper to read this mail spool correctly. Any
suggestions? (I'm going through all this to get mailquotas
working as best I can. I hope to have pop temp files on
another filesystem as well)
Matthew Koch
Geneva On-Line Inc.
Date: Sat, 28 Apr 2001 14:13:01 -0400
From: "Frank P. Troy" <ftroy at pop dot net>
Subject: Qpopper Client for APOP and TLS/SSL
Hi,
What is the best windows pop client to use for QPopper compiled with
APOP and TLS/SSL?
Thanks
Frank
From: Chuck Yerkes <chuck+QPopper at yerkes dot com>
Subject: Re: Qpopper Client for APOP and TLS/SSL
Date: Sun, 29 Apr 2001 01:05:32 -0700
Er, Eudora might to well. Outlook gets weekly viruses and is
likely one of the biggest hazards to business that we have out.
Netscape doesn't support APOP. Dunno if QPopper supports
using personal Certs AS the auth mechanism (yet), but that
makes passing passwords of any kind moot (and opens up the
use of smart cards).
Quoting Frank P. Troy (ftroy at pop dot net):
> What is the best windows pop client to use for QPopper
> compiled with APOP and TLS/SSL?
Date: Mon, 30 Apr 2001 07:44:21 -0500
From: Ted George <ted at kcnet dot com>
Subject: Invalid cross-device link
we are running a redhat 6.2 box as a pop mail server with qpopper in
server mode. we have been using qpopper 3.1.2 and several versions
before that. recently tried 4.0 and everything looked ok except these
errors started showing up in the logs.
Apr 29 11:45:54 mail popper[18125]: Unable to move
/var/spool/poptemp/.username.pop
to /var/spool/mail/username: Invalid cross-device link (18)
i built the binary with
--enable-servermode
--enable-temp-drop-dir=/var/spool/poptemp
--enable-cache-dir=/var/spool/popcache
did not use the enable-cache before with 3.1.2. all other options the
same as before. the only other thing we did different with 4.0 is using
the -F option from the command line, popper -sF. also we run a quota on
/var/spoo/mail, and no quota on the other 2 dirs (the other dirs have
plenty of available space). does 4.0 now need more space
/var/spool/mail even though it was successfully using the poptemp and
popcache dirs, we verified files were being created there.
does anyone have any information on this error, we have downgraded back
to 3.1.2 until we can find out what caused this
thanks in advance
Date: Mon, 30 Apr 2001 13:58:06 +0100
From: Peter Rose <peter at cyberscreen dot com>
Subject: reading from /var/spool/mail AND $HOMEdir?
Hi -
I am just upgrading to 4.0 from 3.1.1 on RedHat 6.1
I am using sendmail 8.9.3 with procmail as local mailer.
Up to now, my system by default has delivered mail to
/var/spool/mail/user,
but now I want to reset the system so mail is appended to a file in the
users' home dir.
I can put a .procmailrc file in each home dir and the mail gets written
there OK, and I have just built qpopper 4.0 with the --enable-home-
dir=.mail.
On testing, qpopper finds the mail in the $HOME/.mail files OK, but it
doesn't seem to be reading the spool for accounts that still have mail
left there.
I thought qpopper checked both by default once it was configured to read
mailfiles in the home dir?
Or will I have to set a changeover date and copy all the mailfiles from
/var/spool over to their corresponding $HOME dir/.mail files?
What am I missing? Is it a case of qpopper only looks in one or the
other?
TIA
Peter Rose
Date: Mon, 30 Apr 2001 13:54:03 -0400
From: trini0 <trini0 at optonline dot net>
Subject: Qpopper/qmail
Hey all. I just setup qpopper4 to work with qmail using Mailbox format.
I saw a patch for qpopper 2.53, would that still work for ver 4, or does
it already support it, or is there another patch??
Thanks
trini0
Date: Mon, 30 Apr 2001 13:55:26 -0500
From: John Hogan <jhogan2112 at yahoo dot com>
Subject: Unable to obtain socket and address of client
new install of qpopper to match new install of qmail...
on:
/usr/local/sbin/popper -s -d qpopper
i get:
Apr 30 13:40:14 baywave /usr/local/sbin/popper[13474]: Unable to obtain socket and address of client: Socket operation on non-socket (88)
what does this mean and how can i fix it?
- hogan
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
From: "Kenneth Porter" <shiva at well dot com>
Date: Mon, 30 Apr 2001 13:52:53 -0700
Subject: Re: Qpopper Client for APOP and TLS/SSL
On Sat, 28 Apr 2001 14:13:01 -0400, Frank P. Troy wrote:
>What is the best windows pop client to use for QPopper compiled with
>APOP and TLS/SSL?
I haven't tried this, but I understand you can set up stunnel, or some
other ssh tunneling program, on your client. You have it act as a local
POP server and pass all traffic to Qpopper's secure port at the remote
host. You then have any ordinary mail client connect to the local end
of the tunnel as if it were a POP server.
With this kind of secure connection, you no longer need APOP so much,
because the whole channel is encrypted.
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
Date: Mon, 30 Apr 2001 18:50:39 -0400 (EDT)
From: James Ralston <qralston+ml.qpopper at andrew.cmu dot edu>
Subject: Re: RPM's
On Wed, 25 Apr 2001, Steven Champeon wrote:
> service pop-3
> {
> flags = REUSE NAMEINARGS
> socket_type = stream
> protocol = tcp
> wait = no
> user = root
> server = /usr/sbin/in.qpopper
> server_args = in.qpopper -s
> }
The official service name for version 3 of the Post Office Protocol is
"pop3":
http://www.isi.edu/in-notes/iana/assignments/port-numbers
The name "pop-3" is an unofficial alias, and is not guaranteed to be
listed in a machine's services file. If it isn't, the definition you
have used will break.
I suggest you use "pop3" instead of "pop-3".
--
James Ralston, Information Technology
Software Engineering Institute
Carnegie Mellon University, Pittsburgh, PA, USA
From: "Kenneth Porter" <shiva at well dot com>
Date: Mon, 30 Apr 2001 16:20:20 -0700
Subject: Re: RPM's
On Mon, 30 Apr 2001 18:50:39 -0400 (EDT), James Ralston wrote:
>I suggest you use "pop3" instead of "pop-3".
Will do. That's already what I use for my PAM file. Didn't realize that
IANA had made the name official. (The RFC for the SRV DNS record
suggests that there's no standard for the name.)
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
Date: Tue, 1 May 2001 17:46:15 -0400
From: Scott McDermott <mcdermot at questra dot com>
Subject: TLS shutdown errors with MS Outlook clients (4.0.1 LX)
I am getting TLS shutdown errors with every single MS Outlook POP
retrieval from a couple of test users I'm serving pop3s to. Looks like
this:
May 1 17:26:36 hades xinetd[31233]: START: pop3s_roc pid=14021 from=10.20.11.15
May 1 17:26:36 hades sqpopper[14021]: (v4.0.1) TLSv1/SSLv3 handshake with client at ro1-cjanel.ddns.roc.questra.com (10.20.11.15); new session-id; cipher: RC4-MD5 (RC4-MD5 SSLv3 Kx=RSA Au=RSA Enc=RC4(128) Mac=MD5 ), 128 bits [pop_tls_openssl.c:510]
May 1 17:26:36 hades sqpopper[14021]: (v4.0.1) POP login by user "cjanel" at (ro1-cjanel.ddns.roc.questra.com) 10.20.11.15 [pop_log.c:244]
May 1 17:26:36 hades sqpopper[14021]: Stats: cjanel 0 0 0 0 ro1-cjanel.ddns.roc.questra.com 10.20.11.15 [pop_updt.c:296]
May 1 17:26:36 hades sqpopper[14021]: TLS shutdown Error [pop_tls_openssl.c:784]
May 1 17:26:36 hades sqpopper[14021]: (v4.0.1) Timing for cjanel at ro1-cjanel.ddns.roc.questra dot com (normal) auth=0 init=0 clean=0 [popper.c:371]
May 1 17:26:36 hades xinetd[31233]: EXIT: pop3s_roc status=0 pid=14021 duration=0(sec)
They all look like that in repeated tests. I want to be sure this is
normal behavior before I start offering this to the masses. Mail does
seem to retrieve just fine without any problems, and snooping/traces
reveal full encryption. My guess is Outlook isn't following the
spec and just disconnects without shutdown and QUIT...
I'm getting some more testing in tomorrow with different clients, but
wanted to get this message off as well to see if anyone else has already
found this to be normal behavior.
One other question: is standalone mode a lot faster for serving with TLS
encryption (presumably because it does session key init infrequently) ?
We make heavy use of concurrency and rate limiting feature of xinetd, so
standalone mode isn't an option, but perhaps an external session cache
would be of help here...I might try implementing something like this if
it's a lot faster, since our mail server doesn't have the most
horsepower in the world :)
Date: Tue, 01 May 2001 16:59:53 -1000
From: Robert Brewer <rbrewer at lava dot net>
Subject: Server mode based on shell
I'm upgrading our qpopper from 3.0 to 4.0 in an attempt to improve
performance. In particular, I'd like to be able to turn on server mode.
However, some users on our system have shell accounts, and some do not.
Therefore, I have to make sure that server mode is not turned on for shell
users.
Qpopper 4.0 has a variety of options for choosing whether or not to do server
mode: groups, spool config files, and user directory config files. However,
what I really want is a direct check: if the user has a valid shell (i.e.
listed in /etc/shells) then don't use server mode, otherwise do use server
mode. Or the inverse would be OK too: any user with a particular shell like
"/bin/false" gets server mode, and everyone else defaults to non server mode.
It appears that qpopper already has some support for shell checking via the
CHECK_SHELL macro, but that is for enabling or disabling authentication and
not server mode.
For now, I'll probably just use groups and make sure that when a user's shell
is enabled or disabled that group membership follows in lock step. But please
consider this a suggestion for future development. Mahalo.
Date: Wed, 2 May 2001 12:02:16 -0700 (PDT)
From: Bryce Newall <data at dreamhaven dot org>
Subject: Problems compiling/installing qpopper 4.0.1
Hi everyone,
I'm running a FreeBSD 4.3-STABLE system and am attempting to upgrade from
Qpopper 3.1.2 to 4.0.1. I've run into a few snags, all but one of which
I've managed to work around.
I used the following configure command:
/configure --enable-servermode --enable-bulletins
--enable-home-dir-mail=Mailbox --enable-log-login --enable-poppassd
Then, upon compiling, popper itself compiled fine, but when it got to
compiling poppassd...
gcc -o poppassd auth_user.o poppassd.o ../common/libcommon.a
auth_user.o: In function `auth_user':
/usr/local/src/qpopper4.0.1/password/auth_user.c(.text+0x18): undefined
reference to `crypt'
*** Error code 1
I found that in the poppasswd Makefile (./password/Makefile), there was no
reference to the LIBS variable in the following section:
poppassd: ${OBJS} ${common_dir}/libcommon.a
${CC} -o poppassd ${OBJS} ${NETWORK_LIBS} \
${common_dir}/libcommon.a
So I added it to the last line:
${common_dir}/libcommon.a ${LIBS}
make clean; make; everything went fine there. However, there appears to
be no "install:" target for poppassd. When I did "make install", popper
installed fine, but then...
Installed popper as /usr/local/sbin/popper
if [ "xpoppassd" != "x" ]; then cd ../password && make install ; fi
make: don't know how to make install. Stop
*** Error code 2
Looks like a few adjustments to password/Makefile.inc are in order. :) I
decided to start over again, this time without the --enable-poppassd
configure option. make realclean; configure (with the rest of the command
line arguments up above); make; make install. Everything was fine there.
But then after installing, when I tried to connect to my POP server, I got
the following messages in my system logs:
May 2 11:05:08 ds9 popper[33072]: Insufficient room to generate path for
user data; need more than 956; have only 64
May 2 11:05:08 ds9 popper[33072]: data at localhost (127.0.0.1): -ERR
[SYS/TEMP] Unable to get spool name
So at that point, I was forced to revert back to 3.1.2. Any suggestions
on the above problems?
Thanks very much in advance,
*********************************************************
* Bryce Newall * Email: data at dreamhaven dot org *
* www.dreamhaven.org/~data *
* "Computers make very fast, very accurate mistakes." *
*********************************************************
Date: Wed, 2 May 2001 12:37:05 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: Invalid cross-device link
At 7:44 AM -0500 4/30/01, Ted George wrote:
> we are running a redhat 6.2 box as a pop mail server with qpopper in
> server mode. we have been using qpopper 3.1.2 and several versions
> before that. recently tried 4.0 and everything looked ok except these
> errors started showing up in the logs.
>
> Apr 29 11:45:54 mail popper[18125]: Unable to move
> /var/spool/poptemp/.username.pop
> to /var/spool/mail/username: Invalid cross-device link (18)
>
> i built the binary with
>
> --enable-servermode
> --enable-temp-drop-dir=/var/spool/poptemp
> --enable-cache-dir=/var/spool/popcache
>
> did not use the enable-cache before with 3.1.2. all other options the
> same as before. the only other thing we did different with 4.0 is using
> the -F option from the command line, popper -sF. also we run a quota on
> /var/spoo/mail, and no quota on the other 2 dirs (the other dirs have
> plenty of available space). does 4.0 now need more space
> /var/spool/mail even though it was successfully using the poptemp and
> popcache dirs, we verified files were being created there.
>
> does anyone have any information on this error, we have downgraded back
> to 3.1.2 until we can find out what caused this
>
> thanks in advance
There is no need to fall back to 3.1.2. All that is happening is
that you told Qpopper 4.0 to move instead of copy the spool during
the end-of-session update; when Qpopper tried to do so it got an
error, so it let you know (and of course fell back on copying).
So you can continue using Qpopper 4, and either (a) ignore the
errors, (b) turn off the '-F' option, or (c) make sure that your
temp spool directory is on the same file system as your spool
directory.
Date: Wed, 2 May 2001 15:48:48 -0400
From: Scott McDermott <mcdermot at questra dot com>
Subject: Re: Server mode based on shell
Robert Brewer on Tue 1/05 16:59 -1000:
> I'm upgrading our qpopper from 3.0 to 4.0 in an attempt to improve
> performance. In particular, I'd like to be able to turn on server
> mode. However, some users on our system have shell accounts, and some
> do not. Therefore, I have to make sure that server mode is not turned
> on for shell users.
I don't know why the LX manual says continually that shell users
shouldn't be allowed access to server mode and the like...what they
should really say is, "don't let shell users have access to programs
that modify the mail spool and don't respect F_SETLK and/or dotlocks."
If you have been sure to compile all your programs to cooperate with
respec to fcntl()/flock() (which Qpopper does properly) then I don't see
where the problem would arise.
FWIW, any sane program from the last decade works appropriately with
locking (and they even work over NFS as long as you have NLM4 running).
If you're unsure, use `nm' to see if the symbols are there, and `strace'
to verify that the locks are being tested for, set where appropriate,
and honored.
From: "Matthew Koch" <mattkoch at genevaonline dot com>
Subject: RE: Problems compiling/installing qpopper 4.0.1
Date: Wed, 2 May 2001 16:00:21 -0500
-----Original Message-----
From: Bryce Newall [mailto:data at dreamhaven dot org]
Sent: Wednesday, May 02, 2001 3:31 PM
To: mildew at genevaonline dot com
Subject: SPAM: Problems compiling/installing qpopper 4.0.1
> Everything was fine there.
> But then after installing, when I tried to connect to my
POP server, I got
> the following messages in my system logs:
>
> May 2 11:05:08 ds9 popper[33072]: Insufficient room to
generate path for
> user data; need more than 956; have only 64
> May 2 11:05:08 ds9 popper[33072]: data at localhost
(127.0.0.1): -ERR
> [SYS/TEMP] Unable to get spool name
>
> So at that point, I was forced to revert back to 3.1.2.
Any suggestions
> on the above problems?
>
>Thanks very much in advance,
I had a similar problem when using home-dir-mail. After some
fooling
around in the source code for genpath.c I found that 956
seems to be
the length qpopper thinks the users home directory string
is. That is,
/usr/home/someuser would be 18.
I'm no C guru, but I think when genpath goes to read the
path for a
users home dir it's falling off the end of the string and
grabbing
too much data. Then it goes to do a string copy. There's a
check to
make sure the string copy didn't truncate the string. When
it sees
that the old string was 956 and the new one was only 64, it
assumes
it truncated the home dir, and aborts. If you're brave
(and/or stupid)
you can comment out that check. I rewrote that section to
look like
this:
/*
if ( len1 > iDropLen ) {
pop_log ( p,
POP_PRIORITY,
HERE,
"Insufficient room for len1 to generate
path for user %.100s"
"; need more than %i; have only %i",
pszUser,
MAX ( len1, len2 ),
iDropLen );
return -1;
}
*/
if ( len2 > iDropLen ) {
pop_log ( p,
POP_PRIORITY,
HERE,
"Insufficient room for len2 to generate
path for user %.100s"
"; need more than %i; have only %i",
pszUser,
MAX ( len1, len2 ),
iDropLen );
return -1;
}
if ( len3 > iDropLen ) {
pop_log ( p,
POP_PRIORITY,
HERE,
"Insufficient room for len3 to generate
path for user %.100s"
"; need more than %i; have only %i",
pszUser,
MAX ( len1, len2 ),
iDropLen );
return -1;
}
It works. Just make sure you don't give someone a home dir
that'll get truncated.
If I knew more C code I'd write a patch and submit it, but
I'll leave that to the
experts. Has anyone gotten home-dir-mail to work on FreeBSD?
Matthew Koch
Geneva On-Line Inc.
From: "Martin, Rich" <rich.martin at reden-anders dot com>
Subject: make-install doesn't work for 4.0.1
Date: Wed, 2 May 2001 16:14:00 -0500
I'm trying to install on a plain-vanilla HP/UX 11.00 box...
./configure --enable-shy --enable-log-login --enable-standalone
Have tried with and without --enable-standalone.
No errors are reported by ./configure, or by make. However, when I go to
make install...
/da01/qpopper4.0.1> make install
./install-sh -c -m 0644 -o root ./man/popper.8 /usr/local/man/man8/
Make: Cannot load ./install-sh. Stop.
*** Error exit code 1
Stop.
/da01/qpopper4.0.1>
I am logged on as root for the installation. Does anybody have any
suggestions? Thanks in advance for your time and consideration!
-Rich
Date: Wed, 2 May 2001 14:53:04 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: reading from /var/spool/mail AND $HOMEdir?
At 1:58 PM +0100 4/30/01, Peter Rose wrote:
> Is it a case of qpopper only looks in one or the
> other?
That's correct.
Date: Wed, 2 May 2001 14:58:18 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: Unable to obtain socket and address of client
At 1:55 PM -0500 4/30/01, John Hogan wrote:
> Unable to obtain socket and address of client: Socket operation
> on non-socket (88)
>
> what does this mean and how can i fix it?
>
Usually this means you compiled a normnal (inetd) Qpopper and tried
to run it from a command line. If this is the case, either run
Qpopper from inetd or compile a stand-alone version. See the
Administrator's Guide for more information.
Date: Wed, 2 May 2001 15:01:58 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: TLS shutdown errors with MS Outlook clients (4.0.1 LX)
At 5:46 PM -0400 5/1/01, Scott McDermott wrote:
> One other question: is standalone mode a lot faster for serving with TLS
> encryption (presumably because it does session key init infrequently) ?
>
> We make heavy use of concurrency and rate limiting feature of xinetd, so
> standalone mode isn't an option, but perhaps an external session cache
> would be of help here...I might try implementing something like this if
> it's a lot faster, since our mail server doesn't have the most
> horsepower in the world :)
Qpopper 4 doesn't take advantage of several performance
improvements that could be done in standalone mode, such as doing a
lot of the initialization only once, and using a session cache.
Feel free to add this :-).
Date: Wed, 2 May 2001 14:55:28 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: Qpopper/qmail
At 1:54 PM -0400 4/30/01, trini0 wrote:
> Hey all. I just setup qpopper4 to work with qmail using Mailbox format.
> I saw a patch for qpopper 2.53, would that still work for ver 4,
> or does it already support it, or is there another patch??
> Thanks
> trini0
You probably want to use '--enable-home-dir-mail' with ./configure
(after a 'make realclean') or 'set home-dir-mail' in a config file.
See the Administrator's Guide for more information.
Date: Wed, 2 May 2001 15:15:06 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: Server mode based on shell
At 3:48 PM -0400 5/2/01, Scott McDermott wrote:
> I don't know why the LX manual says continually that shell users
> shouldn't be allowed access to server mode and the like...what they
> should really say is, "don't let shell users have access to programs
> that modify the mail spool and don't respect F_SETLK and/or dotlocks."
> If you have been sure to compile all your programs to cooperate with
> respec to fcntl()/flock() (which Qpopper does properly) then I don't see
> where the problem would arise.
Qpopper does not keep the spool locked for the duration of the
session. Qpopper only locks the spool at the beginning and end of
sessions. In non-server mode, this is fine, as Qpopper makes no
assumptions about the state of the spool between times. But in
server mode, you are asserting that the only modifications to the
spool are done by a final delivery agent, which only appends. If a
shell user edits the spool in any other way during a POP session,
and Qpopper is in server mode, you've got spool corruption.
Date: Wed, 2 May 2001 15:59:37 -0700
From: Chuck Yerkes <chuck+qpopper at yerkes dot com>
Subject: Re: Hash Directories
Have modified the hash directory stuff a bit to use a different
hash (basically, look the name up in a .db file to find it).
I have to presume that their's stuff available to DELIVERY to
hashed directories. But I find no docs with popper on it. I
use sendmail. Must I hack mail.local myself or has someone
got patches??
Thanks,
chuck
From: "Martin, Rich" <rich.martin at reden-anders dot com>
Subject: RE: make-install doesn't work for 4.0.1
Date: Wed, 2 May 2001 20:08:25 -0500
Actually, /bin/sh does exist; for the sake of argument I changed it to
/sbin/sh. No dice. I tried the GNU "make install" command, and that came
back with:
/da01/qpopper4.0.1> /usr/local/bin/make install
./install-sh -c -m 0644 -o root ./man/popper.8 /usr/local/man/man8/
make: execvp: ./install-sh: Permission denied
make: *** [install] Error 127
/da01/qpopper4.0.1>
Don't know what's giving me the "permission denied" errors; all the file
ownerships and perms in the qpopper directory look fine. Unfortunately, I'm
not a C or C++ programmer, or this would be less mystifying.
My boss told me a couple minutes ago that the "make" environment on HP/UX is
different than on most x-IXes; he's had problems like this with similar
utilities he's tried to compile. Looking around on the web, I'm noticing
that the last version of qpopper people using HP/UX seem to be successfully
using is somewhere right around 2.14 or 2.53.
If I can't get this to work, I will, of course, solicit suggestions for an
alternative...
-----Original Message-----
From: Bryce Newall [mailto:data at dreamhaven dot org]
Sent: Wednesday, May 02, 2001 2:24 PM
To: Martin, Rich
Subject: Re: make-install doesn't work for 4.0.1
On Wed, 2 May 2001, Martin, Rich wrote:
> /da01/qpopper4.0.1> make install
> ./install-sh -c -m 0644 -o root ./man/popper.8
/usr/local/man/man8/
> Make: Cannot load ./install-sh. Stop.
> *** Error exit code 1
This may be elementary, but did you check the first line of the install-sh
script to see what shell it's pointing to? If, for example, it's looking
for /bin/sh and /bin/sh doesn't exist on your system, you'll get that
exact error message. If that's the problem, you can either edit the
script, or make a symlink to the shell so that the script will be able to
find it where it's looking.
*********************************************************
* Bryce Newall * Email: data at dreamhaven dot org *
* www.dreamhaven.org/~data *
* "Computers make very fast, very accurate mistakes." *
*********************************************************
Date: Wed, 2 May 2001 18:37:29 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: RE: make-install doesn't work for 4.0.1
At 8:08 PM -0500 5/2/01, Rich Martin wrote:
> Don't know what's giving me the "permission denied" errors; all the file
> ownerships and perms in the qpopper directory look fine. Unfortunately, I'm
> not a C or C++ programmer, or this would be less mystifying.
>
> My boss told me a couple minutes ago that the "make" environment on HP/UX is
> different than on most x-IXes; he's had problems like this with similar
> utilities he's tried to compile. Looking around on the web, I'm noticing
> that the last version of qpopper people using HP/UX seem to be successfully
> using is somewhere right around 2.14 or 2.53.
Does 'ls -l install-sh' show it has execute permission? If not,
try 'chmod u+x' on it.
If you can't get 'make install' to work, just install Qpopper
manually. The Administrator's Guide has detailed information on
how to do this. Until recently, it was the expected way to install
Qpopper. It's just a few copies.
It's certainly not worth falling back to an old version of Qpopper
over. There's been far too many code improvements to even think of
that, as I see it.
Date: Wed, 2 May 2001 23:55:17 -0400
From: Scott McDermott <mcdermot at questra dot com>
Subject: Re: Server mode based on shell
Randall Gellens on Wed 2/05 15:15 -0700:
> Qpopper does not keep the spool locked for the duration of the
> session. Qpopper only locks the spool at the beginning and end of
> sessions. In non-server mode, this is fine, as Qpopper makes no
> assumptions about the state of the spool between times. But in server
> mode, you are asserting that the only modifications to the spool are
> done by a final delivery agent, which only appends. If a shell user
> edits the spool in any other way during a POP session, and Qpopper is
> in server mode, you've got spool corruption.
Ok, I've looked at the code...it looks like in server mode the cache is
created on pop_updt() and left around after process exit. Then with
each new session, during pop_dropcopy() the cache is read if it exists
(and init_dropinfo() is bypassed) and used to fill in the MsgInfoList
without having to parse the spool:
else { /* SERVER_MODE */
/*$
* Save the temporary drop FILE and fid values
*/
p->hold = p->drop;
p->drop = fdopen ( mfd, "r+" );
...
/*
* If we have a usable cache file, that saves us from having to
* process the spool.
*/
rslt = cache_read ( p, fileno(p->drop), &p->mlp );
if ( rslt == POP_FAILURE ) {
rslt = init_dropinfo ( p, p->drop_name, p->drop, time_locked );
if ( rslt != POP_SUCCESS )
goto bailout;
/*
* If the cache file is smaller than the spool, we need to
* process the new messages.
*/
}
}
Looks like the comment at the end indicates that corruption problems
between sessions are prevented by testing in cache_read():
if ( header.spool_end > stbuf.st_size ) {
pop_log ( p, POP_PRIORITY, HERE,
"spool smaller than indicated in cache file %s "
"(expected %ld; found %ld)",
cache_name, (long) header.spool_end, (long) stbuf.st_size );
close ( cache_fd );
return POP_FAILURE;
}
/*
* For now, exit if the spool has been modified or is larger than
* indicated in the cache. This is temporary.
*/
if ( header.spool_end != stbuf.st_size ||
header.modtime != stbuf.st_mtime ) {
DEBUG_LOG5 ( p, "spool modified or larger than indicated in "
"cache: %s (%ld %ld vs %ld %ld)",
cache_name, (long) header.spool_end, header.modtime,
(long) stbuf.st_size, stbuf.st_mtime );
close ( cache_fd );
return POP_FAILURE;
}
/*
* Now we know that the cache information is probably compatible
* with the spool. Note that the local delivery agent may have
* added messages to the spool since the cache was written,
* in which case the new messages need to be processed normally.
*/
in which case init_dropinfo() is called anew in pop_dropcopy() for
safety. Now we have the lock on the file this whole time, but why do we
run
if ( p->server_mode )
flock ( mfd, LOCK_UN );
next, allowing the spool to change before pop_updt() is called?
It shouldn't really be a performance hit to move the unlock to quitting
time instead of unlocking after pop_dropcopy() does its thing, leaving
the window open during the session, making the assumption in server
mode, and then, locking again before the update. While the lock is
held and the POP session is active, MDAs should do the right thing and
keep trying. If not or if the MDA times out waiting for a lock the MTA
is responsible for handling temporary failure codes returned from the
MDA. And a user with a shell MUA modifying the spool directly will be
expected to have a client that respects the locks as well. This allows
the advantages of the cache file without disallowing shell access and
cooperative use of the spool.
Am I missing something? If not it should be easy to make a quick patch
which surrounds the session instead of just the dropcopy and the update.
Date: Thu, 3 May 2001 11:59:59 -0700 (PDT)
From: Bryce Newall <data at dreamhaven dot org>
Subject: Re: Problems compiling/installing qpopper 4.0.1
On Wed, 2 May 2001, Gary Schrock wrote:
> I'd suggest trying the port, and just adding the configure options you want
> to the makefile for the port. I never did get qpopper to properly compile
> when trying to do it myself, but when the port got updated to 4.0.1 I had
> no trouble at all.
Thanks, that worked... for the most part. (I had to update my ports
collection first, though.) I still can't use poppassd, though, since the
same error in poppassd's Makefile apparently exists in the port as in the
tarball I originally downloaded from Qualcomm. But if I don't include
that option, when I install the new qpopper, everything works as it
should. Thanks again for the help!
*********************************************************
* Bryce Newall * Email: data at dreamhaven dot org *
* www.dreamhaven.org/~data *
* "Computers make very fast, very accurate mistakes." *
*********************************************************
Date: Thu, 3 May 2001 12:58:58 -0700
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: Server mode based on shell
The issue isn't the state of the spool between sessions, it's
changes to the spool during a session. Qpopper unlocks the spool
during a session to allow delivery agents to modify it.
Date: Thu, 3 May 2001 16:03:15 -0400
From: Scott McDermott <mcdermot at questra dot com>
Subject: Re: Server mode based on shell
Randall Gellens on Thu 3/05 12:58 -0700:
> The issue isn't the state of the spool between sessions, it's changes
> to the spool during a session. Qpopper unlocks the spool during a
> session to allow delivery agents to modify it.
Right, but why is that necessary? If you went on to read towards the
end of my message I made plenty of arguments for why it's unnecessary to
unlock during the middle of the run. Why not simply get the lock before
pop_dropcopy() and release it after pop_updt() ? You then get the
advantages of the cache and the rest of server mode without alienating
shell users.
From: "Genghis Rios Kruger" <grios at pucp.edu dot pe>
Subject: DON'T DELETE ....
Date: Thu, 3 May 2001 22:18:25 -0500
Hello !
Is there any way to use Qpopper in conjuction with WU's IMAP and NOT get the
"DON'T DELETE THIS MESSAGE -- FOLDER INTERNAL DATA" message.
This message is generated for the pine program on the mail server. I get
these all of the time on a RedHat 6.2 and 7.0 Linux server.
Thanks!
Genghis Ríos Kruger
Dirección de Informática PUCP
email: grios at pucp.edu dot pe
From: "Philipp Gimm" <philipp at problemchen dot de>
Subject: AW: DON'T DELETE ....
Date: Fri, 4 May 2001 05:35:01 +0200
Hi!
at configure time, use --without-uw-kludge, or --diasble-uw-kludge
...or similiar (don't remember ;)) look in the manual...
/p
> Hello !
>
> Is there any way to use Qpopper in conjuction with WU's IMAP and
> NOT get the
> "DON'T DELETE THIS MESSAGE -- FOLDER INTERNAL DATA" message.
> This message is generated for the pine program on the mail server. I get
> these all of the time on a RedHat 6.2 and 7.0 Linux server.
>
> Thanks!
>
>
> Genghis Ríos Kruger
> Dirección de Informática PUCP
> email: grios at pucp.edu dot pe
>
>
From: "Rogerwoo" <rogerwoo7508 at sina.com dot cn>
Subject: For help on PAM using samba authentication
Date: Fri, 4 May 2001 13:44:08 +0800
This is a multi-part message in MIME format.
------=_NextPart_000_063F_01C0D4A0.4B2C3C60
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Hi, everybody.
Our company have migrate from NT 4.0 to Linux, We are now using samba
service to provide share and domain login authentication to windows
wokstation users.
So that is natural to replace Exchange Server5.5 to sendmail to provide
mail service to users, we found Qpopper is a good pop3 service and
planed to use Qpopper 4.0 and pam-smb PAM authentication to use samba
password information. But I tried more than twice all failed following
pam-smb README . When I tried from window 98 workstation error message
said module not found but the pam-smb module in the directory.
Any help will be appreciated.
------=_NextPart_000_063F_01C0D4A0.4B2C3C60
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">
<HTML>
<HEAD>
<META content=text/html;charset=iso-8859-1
http-equiv=Content-Type>
<META content='"MSHTML 4.72.3110.7"' name=GENERATOR>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT size=2>Hi, everybody.</FONT></DIV>
<DIV><FONT size=2>Our company have migrate from NT 4.0 to Linux, We
are now
using samba service to provide share and domain login authentication to
windows
wokstation users.</FONT></DIV>
<DIV><FONT size=2>So that is natural to replace Exchange
Server5.5 to
sendmail to provide mail service to users, we found Qpopper is a good
pop3
service and planed to use Qpopper </FONT><FONT size=2>4.0 and pam-smb
PAM
authentication to use samba password information. But I tried more than
twice
all failed following pam-smb README . When I tried from window 98
workstation
error message said module </FONT><FONT size=2>not found but the
pam-smb
module in the directory. </FONT><FONT
size=2> </FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Any help will be appreciated.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV></BODY></HTML>
------=_NextPart_000_063F_01C0D4A0.4B2C3C60--
From: "Kenneth Porter" <shiva at well dot com>
Date: Thu, 03 May 2001 23:51:48 -0700
Subject: Re: For help on PAM using samba authentication
On Fri, 4 May 2001 13:44:08 +0800, Rogerwoo wrote:
>When I tried from window 98 workstation error message said module not found but the pam-smb module in the directory.
What are the exact error messages you get, and where are they from (ie.
which log file)?
I'd suspect either PAM (in which case check the PAM mailing list) or
Samba (in which case check the comp.protocols.smb newsgroup). Where
does pam-smb come from? I don't have that in my Samba distribution
(v2.0.7).
Ken
mailto:shiva at well dot com
http://www.sewingwitch.com/ken/
[If answering a mailing list posting, please don't cc me your reply. I'll take my answer on the list.]
Date: Fri, 4 May 2001 12:18:42 -0400
From: Joseph S D Yao <jsdy at cospo.osis dot gov>
Subject: Re: For help on PAM using samba authentication
I'm not sure, but you might ask on pam-list at redhat dot com. I think this
has come up before on that forum.
List-Subscribe: <https://listman.redhat.com/mailman/listinfo/pam-list>,
<mailto:pam-list-request at redhat dot com?subject=subscribe>
List-Archive: <https://listman.redhat.com/mailman/private/pam-list/>
--
Joe Yao jsdy at cospo.osis dot gov - Joseph S. D. Yao
OSIS/COSPO Computer Support EMT-B
-----------------------------------------------------------------------
This message is not an official statement of COSPO policies.
Date: Fri, 04 May 2001 11:34:56 -1000
From: Robert Brewer <rbrewer at lava dot net>
Subject: Re: AW: DON'T DELETE ....
--On Friday, May 4, 2001 5:35 AM +0200 Philipp Gimm <philipp at problemchen dot de>
wrote:
> Hi!
>
> at configure time, use --without-uw-kludge, or --diasble-uw-kludge
>
> ...or similiar (don't remember ;)) look in the manual...
>
> /p
I think he wants the "--enable-uw-kludge" configure option, not
disable-uw-kludge. :)
Genghis Rios Kruger <grios at pucp.edu dot pe> wrote:
>> Hello !
>>
>> Is there any way to use Qpopper in conjuction with WU's IMAP and
>> NOT get the
>> "DON'T DELETE THIS MESSAGE -- FOLDER INTERNAL DATA" message.
>> This message is generated for the pine program on the mail server. I get
>> these all of the time on a RedHat 6.2 and 7.0 Linux server.
>>
>> Thanks!
>>
>>
>> Genghis Ríos Kruger
>> Dirección de Informática PUCP
>> email: grios at pucp.edu dot pe
Last updated on 4 May 2001 by Pensive Mailing List Admin