The qpopper list archive ending on 2 Jan 2002


Topics covered in this issue include:

  1. Re: Qpopper loch mechanism
       Clifton Royston <cliftonr at lava dot net>
       Thu, 20 Dec 2001 17:17:31 -1000
  2. Re: Qpopper loch mechanism
       Mark <admin at asarian-host dot net>
       Fri, 21 Dec 2001 10:42:07 +0100
  3. Re: BLOCK_UID?
       Randall Gellens <randy at qualcomm dot com>
       Sat, 22 Dec 2001 10:42:40 -0500
  4. Virtual Mail hosting.
       "Mark-Nathaniel Weisman" <mweisman at gci dot net>
       Sat, 22 Dec 2001 10:17:08 -0900
  5. Re: force clean mail after reading for specific users?
       Kenneth Porter <shiva at well dot com>
       23 Dec 2001 01:17:57 -0800
  6. Re: force clean mail after reading for specific users?
       Mark <admin at asarian-host dot net>
       Sun, 23 Dec 2001 12:58:10 +0100
  7. problem with qpopper 4.04
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Sun, 23 Dec 2001 22:19:10 +0800
  8. qpopper problem
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Sun, 23 Dec 2001 22:49:31 +0800
  9. Re: problem with qpopper 4.04
       Kenneth Porter <shiva at well dot com>
       23 Dec 2001 16:59:19 -0800
 10. problem with qpopper 4.04
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Mon, 24 Dec 2001 10:21:58 +0800
 11. Re: problem with qpopper 4.04
       Kenneth Porter <shiva at well dot com>
       23 Dec 2001 18:27:15 -0800
 12. RE: problem with qpopper 4.04
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Mon, 24 Dec 2001 12:56:11 +0800
 13. 30 second delay with qpopper
       "Tech Support" <adam at homebusiness dot to>
       Mon, 24 Dec 2001 03:06:11 -0500
 14. Re: 30 second delay with qpopper
       Peter Evans <peter at gol dot com>
       Mon, 24 Dec 2001 18:38:40 +0900
 15. Re: 30 second delay with qpopper
       Gerhard Gonter <gonter at buddy.wu-wien.ac dot at>
       Mon, 24 Dec 2001 10:44:15 +0100 (MEZ)
 16. Re: 30 second delay with qpopper
       Syed Zaeem Hosain <Syed.Hosain at aeris dot net>
       Mon, 24 Dec 2001 08:49:41 -0800
 17. Re: 30 second delay with qpopper
       Daniel Senie <dts at senie dot com>
       Mon, 24 Dec 2001 12:38:14 -0500
 18. RE: problem with qpopper 4.04
       Daniel Senie <dts at senie dot com>
       Mon, 24 Dec 2001 12:35:17 -0500
 19. RE: problem with qpopper 4.04
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Tue, 25 Dec 2001 12:25:03 +0800
 20. Re: 30 second delay with qpopper
       Chuck Yerkes <chuck+qpopper at yerkes dot com>
       Tue, 25 Dec 2001 11:20:14 -0800
 21. Re: 30 second delay with qpopper
       Daniel Senie <dts at senie dot com>
       Tue, 25 Dec 2001 14:42:31 -0500
 22. RE: problem with qpopper 4.04
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Wed, 26 Dec 2001 05:34:25 +0800
 23. Re: 30 second delay with qpopper
       Chuck Yerkes <chuck+qpopper at yerkes dot com>
       Wed, 26 Dec 2001 09:35:41 -0800
 24. Re: 30 second delay with qpopper
       Daniel Senie <dts at senie dot com>
       Wed, 26 Dec 2001 13:02:40 -0500
 25. How to create POP-Boxes without creating REAL users?
       Newsletter <newsletter at spellweaver dot de>
       Wed, 26 Dec 2001 21:55:55 +0100
 26. FW: problem with qpopper 4.04
       "Peter Santiago" <petersantiago at bigfoot dot com>
       Thu, 27 Dec 2001 07:48:53 +0800
 27. Client support?
       Tim Tyler <tyler at beloit dot edu>
       Thu, 27 Dec 2001 17:00:34 -0600
 28. Re: Client support?
       Daniel Senie <dts at senie dot com>
       Thu, 27 Dec 2001 19:53:22 -0500
 29. Fwd: NDN: Re: Client support?
       Daniel Senie <dts at senie dot com>
       Thu, 27 Dec 2001 20:08:58 -0500
 30. qpopper, qmail and maildirs
       "Arne Schmitz" <arne.schmitz at mmweg.rwth-aachen dot de>
       28 Dec 2001 22:56:00 +0100
 31. Re: qpopper, qmail and maildirs
       Kenneth Porter <shiva at well dot com>
       28 Dec 2001 23:55:37 -0800
 32. Re: How to create POP-Boxes without creating REAL users?
       Randall Gellens <randy at qualcomm dot com>
       Sat, 29 Dec 2001 23:29:39 -0800
 33. RE: problem with qpopper 4.04
       Randall Gellens <randy at qualcomm dot com>
       Sat, 29 Dec 2001 23:26:57 -0800
 34. Configuring APOP - gdbm_open Error
       Kai Bleek <newsletter at spellweaver dot de>
       Sun, 30 Dec 2001 19:10:17 +0100
 35. Re: Configuring APOP - gdbm_open Error
       Kenneth Porter <shiva at well dot com>
       30 Dec 2001 10:52:58 -0800
 36. Re[2]: Configuring APOP - gdbm_open Error
       Kai Bleek <newsletter at spellweaver dot de>
       Sun, 30 Dec 2001 20:13:38 +0100
 37. Re: Configuring APOP - gdbm_open Error
       Kai Bleek <newsletter at spellweaver dot de>
       Sun, 30 Dec 2001 20:39:40 +0100
 38. qpopper upgrade XUIDL issue...
       Daniel Suen <ttdsuen at ln.edu dot hk>
       Mon, 31 Dec 2001 09:43:35 +0800
 39. Emails are retrieved twice (sometimes 3 times) from Qpopper vesion 3.1.2. 
       "Habib Abassi" <habassi at telephonyatwork dot com>
       Sun, 30 Dec 2001 21:09:19 -0800
 40. Canonical name error...
       "Fred G Guiliano" <fred.guiliano at DPS-inc dot com>
       Wed, 2 Jan 2002 11:03:03 -0500
 41. unsubsribe
       James McLaughlin <katana at montrose dot net>
       Wed, 2 Jan 2002 09:40:57 -0700
 42. Re: Canonical name error...
       Gregory Hicks <ghicks at cadence dot com>
       Wed, 2 Jan 2002 09:33:16 -0800 (PST)
 43. Re: Canonical name error...
       Daniel Senie <dts at senie dot com>
       Wed, 02 Jan 2002 12:51:42 -0500
 44. Re: Canonical name error...
       Gregory Hicks <ghicks at cadence dot com>
       Wed, 2 Jan 2002 10:09:44 -0800 (PST)
 45. RE: Canonical name error...
       "Fred G Guiliano" <fred.guiliano at DPS-inc dot com>
       Wed, 2 Jan 2002 13:03:24 -0500
 46. Re: Canonical name error...
       Daniel Senie <dts at senie dot com>
       Wed, 02 Jan 2002 14:17:57 -0500
 47. Re: Canonical name error...
       Clifton Royston <cliftonr at lava dot net>
       Wed, 2 Jan 2002 09:48:05 -1000
 48. Re: Canonical name error...
       Clifton Royston <cliftonr at lava dot net>
       Wed, 2 Jan 2002 09:51:04 -1000
 49. Re: Emails are retrieved twice (sometimes 3 times) from
       Randall Gellens <randy at qualcomm dot com>
       Wed, 2 Jan 2002 11:55:38 -0800
 50. Re: Canonical name error...
       Daniel Senie <dts at senie dot com>
       Wed, 02 Jan 2002 15:07:14 -0500

Date: Thu, 20 Dec 2001 17:17:31 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: Qpopper loch mechanism

On Thu, Dec 20, 2001 at 11:10:03PM +0100, Mark wrote:
> Dear folks,
> 
> I have a question about Qpopper's locking mechanism (3.x). I have a Perl LDA
> on my FreeBSD 4.1.1 box, in which I use an exclusive flock to lock a user
> mailbox before writing to it. I always thought qpopper honors the same
> flock, and will not truncate on the user's mailbox until it can acquire a
> lock. But since I suspect to have lost some mail, I want to be sure now.

Hi, Mark!

  The normal behavior for Qpopper, in accessing the user mailspool, is
to follow the default lock convention for mail delivery programs, which
for historical reasons is user.lock in the mail spool directory.  When
it needs exclusive access to the spool Qpopper will attempt to create
the user.lock file in an "atomic" fashion, then lock it exclusively
with flock, and write its own pid to it.  (This last is a bit more
extensive than what many lock methods will do, but in general, creating
user.lock in an exclusive way is the requirement.)


> I know that (additionally?) qpopper uses the .user.lock and user.lock file.

  For mail spool, unless you've changed it, qpopper should *only* be
using user.lock in the spool directory.

  So yes, it is quite possible you're losing mail if your approach is
what you describe.  Sorry. :-/  On the bright side, it sounds like
you've correctly identified the reason.


> Should I try and emulate that, or will an flock on mbox suffice? Basically,
> all I need to know is which one should I set to prevent qpopper from writing
> and/or truncating mbox.

  Create user.lock in an atomic way (e.g. by creating a different file
and relinking it to user.lock, or by opening it with O_CREAT|O_EXCL if
you trust them) and if feasible flock it too, then reverse those steps
at the end to unlock it.  "man open" and "man flock", also read
common/maillock.h in the qpopper sources.

  The "lockfile" utility included with procmail does a good job for
mail locks, and is also a good starting point for understanding the
locking mechanism.

  Since I'm making some very specific assertions here, that should draw
out any needed corrections on any point I've got wrong. :-) 

  -- Clifton

-- 
 Clifton Royston  --  LavaNet Systems Architect --  cliftonr at lava dot net
   WWJD?   "JWRTFM!" - Scott Dorsey (kludge)   "JWG" - Eddie Aikau

From: Mark <admin at asarian-host dot net>
Date: Fri, 21 Dec 2001 10:42:07 +0100
Subject: Re: Qpopper loch mechanism

----- Original Message -----
From: "Clifton Royston" <cliftonr at lava dot net>
To: "Mark" <admin at asarian-host dot net>
Cc: "Subscribers of Qpopper" <qpopper at lists.pensive dot org>
Sent: Friday, December 21, 2001 4:17 AM
Subject: Re: Qpopper loch mechanism

> On Thu, Dec 20, 2001 at 11:10:03PM +0100, Mark wrote:
>
> > Dear folks,
> >
> > I have a question about Qpopper's locking mechanism (3.x). I have a
> > Perl LDA on my FreeBSD 4.1.1 box, in which I use an exclusive flock to
> > lock a user mailbox before writing to it. I always thought qpopper
> > honors the same flock, and will not truncate on the user's mailbox
> > until it can acquire a lock. But since I suspect to have lost some
> > mail, I want to be sure now.
>
> Hi, Mark!


Hi Clifton!

I already saw your name pop up in the mailing list archives; ltns. :) Your
answer was most comprehensive, lucid and total. Thanks!

> Create user.lock in an atomic way (e.g. by creating a different file and
> relinking it to user.lock, or by opening it with O_CREAT|O_EXCL if you
> trust them) and if feasible flock it too, then reverse those steps at the
> end to unlock it. "man open" and "man flock", also read common/maillock.h
> in the qpopper sources.
>
> The "lockfile" utility included with procmail does a good job for mail
> locks, and is also a good starting point for understanding the locking
> mechanism.

Since I already have procmail for some accounts, I also have "lockfile" too.
I will take it apart and use its mechanism.

Again, thanks!

- Mark

        System Administrator Asarian-host.org

---
"If you were supposed to understand it,
we wouldn't call it code." - FedEx


Date: Sat, 22 Dec 2001 10:42:40 -0500
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: BLOCK_UID?

Yes, it would make sense for ./configure to allow this to be set. 
Should be a run-time configuration file option as well.

From: "Mark-Nathaniel Weisman" <mweisman at gci dot net>
Subject: Virtual Mail hosting.
Date: Sat, 22 Dec 2001 10:17:08 -0900

-----BEGIN PGP SIGNED MESSAGE-----

All,
  Just wrote a real quick Perl script that seems to do exactly what I
needed it to do. I appreciate all the assistance that the list
offered as far as the real way to set it up. If anyone is interested,
I'd be glad to publish the little patch I used. Seems to work fine on
multiple virtual addresses, even for identical addresses in multiple
domains. Again thanks.

Thank you,
Mark-Nathaniel Weisman A+, MCP, CNA, MOUS MI
Network Systems Administrator
Career Academy
Anchorage, AK



-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>

iQEVAwUBPCTcM8dDd3XswOzXAQGnPQf7BhI0YzKWnyIQFYGDQVSi9DYHdPk2CKF5
/XQAkK6igBFM+xaJzbnR2/iienOE8HytpItaf2jYkQA8tUJDejuPtA22hpxB5Jrm
8nqqrI5hTq7xlK2TagX1LbBJtpSMgByyNlQdx+eOEkUuxrO9guwbbdm7xCOMovXl
v52SpCqxF25ZfyV7hkSTQGTmiQL2CacJBkLw3gm8J504AdhrBwxX/4Hva5YfablC
vSY8hkJ8XknW6qjinAoRh9YAFnJU5TXt7e2Ys2I4ZOynKtBULiCAUVLGe8RVSlfe
9wAT1cQ4r+gkFHRLOH1lsfw3DE80m57F7cbw/pXLz4LM7hFngJATiA=
=/3BU
-----END PGP SIGNATURE-----



Subject: Re: force clean mail after reading for specific users?
From: Kenneth Porter <shiva at well dot com>
Date: 23 Dec 2001 01:17:57 -0800

On Wed, 2001-12-19 at 18:20, Stephen Mawdsley wrote:
> I run a Perl script called trimmail which is set by cron to run every 
> night.  It goes through every user's mailbox and trims out unread messages 
> that are older than 30 days and read messages older than 3 days.  I this 
> way, my users are happy and I get to keep my hard drive clean.  If you want 
> a copy, just send me a message.

This sounds good. I'd appreciate a copy.

Your script? If so, I can make it available on my website.


From: Mark <admin at asarian-host dot net>
Date: Sun, 23 Dec 2001 12:58:10 +0100
Subject: Re: force clean mail after reading for specific users?

Yeah, sounds good; I would like a copy too.

- Mark

        System Administrator Asarian-host.org

---
"If you were supposed to understand it,
we wouldn't call it code." - FedEx


----- Original Message -----
From: "Kenneth Porter" <shiva at well dot com>
To: "Subscribers of Qpopper" <qpopper at lists.pensive dot org>
Sent: Sunday, December 23, 2001 10:17 AM
Subject: Re: force clean mail after reading for specific users?


> On Wed, 2001-12-19 at 18:20, Stephen Mawdsley wrote:

> > I run a Perl script called trimmail which is set by cron to run every
> > night.  It goes through every user's mailbox and trims out unread
messages
> > that are older than 30 days and read messages older than 3 days.  I this
> > way, my users are happy and I get to keep my hard drive clean.  If you
want
> > a copy, just send me a message.
>
> This sounds good. I'd appreciate a copy.
>
> Your script? If so, I can make it available on my website.


From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: problem with qpopper 4.04
Date: Sun, 23 Dec 2001 22:19:10 +0800

I'm not sure what the problem is...
Dec 23 22:18:07 gateway xinetd[6290]: libwrap refused connection to pop3
from 192.168.1.100
What does this particular line mean?

I'm using exim as MTA......

Peter


From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: qpopper problem
Date: Sun, 23 Dec 2001 22:49:31 +0800

qpopper needs dracd right?

Error message:
Dec 23 22:43:46 gateway in.qpopper[6663]: Unable to obtain socket and
address
of client: Socket operation on non-socket (88) [pop_init.c:1050]

I have checked /etc/services it has the pop3 entry.
dracd.allow contains the ff:
255.255.255.255 127.0.0.1
255.255.255.0 192.168.1.1


Subject: Re: problem with qpopper 4.04
From: Kenneth Porter <shiva at well dot com>
Date: 23 Dec 2001 16:59:19 -0800

On Sun, 2001-12-23 at 06:19, Peter Santiago wrote:
> I'm not sure what the problem is...
> Dec 23 22:18:07 gateway xinetd[6290]: libwrap refused connection to pop3
> from 192.168.1.100
> What does this particular line mean?

libwrap is probably tcp_wrappers. Look at the man pages for hosts.allow
and hosts.deny.



From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: problem with qpopper 4.04
Date: Mon, 24 Dec 2001 10:21:58 +0800

qpopper's complaint..
Dec 23 22:43:46 gateway in.qpopper[6663]: Unable to obtain socket and
address
.... if I used gnupop3d....without changing anything... works fine... really
puzzled....

Peter


Subject: Re: problem with qpopper 4.04
From: Kenneth Porter <shiva at well dot com>
Date: 23 Dec 2001 18:27:15 -0800

On Sun, 2001-12-23 at 18:21, Peter Santiago wrote:
> qpopper's complaint..
> Dec 23 22:43:46 gateway in.qpopper[6663]: Unable to obtain socket and
> address
> .... if I used gnupop3d....without changing anything... works fine... really
> puzzled....

How are you launching it? What's your inetd/xinetd setup look like? Or
are you running it standalone?



From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: RE: problem with qpopper 4.04
Date: Mon, 24 Dec 2001 12:56:11 +0800

# qpopper config file for xinetd
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
        # use this line once you create your certificate to enable STLS
        # server_args = in.qpopper -s -l 1 -f /etc/mail/qpopper.conf
}

-----Original Message-----
From: Kenneth Porter [mailto:shiva at well dot com]
Sent: Monday, December 24, 2001 10:27 AM
To: Subscribers of Qpopper
Subject: Re: problem with qpopper 4.04


On Sun, 2001-12-23 at 18:21, Peter Santiago wrote:
> qpopper's complaint..
> Dec 23 22:43:46 gateway in.qpopper[6663]: Unable to obtain socket and
> address
> .... if I used gnupop3d....without changing anything... works fine...
really
> puzzled....

How are you launching it? What's your inetd/xinetd setup look like? Or
are you running it standalone?

through xinetd.... if I launch it as stand-alone... wouldn't connect to
socket...........

# qpopper config file for xinetd
service pop3
{
        flags       = REUSE NAMEINARGS
        socket_type = stream
        protocol    = tcp
        wait        = no
        user        = root
        server      = /usr/sbin/in.qpopper
        server_args = in.qpopper -s
        # use this line once you create your certificate to enable STLS
        # server_args = in.qpopper -s -l 1 -f /etc/mail/qpopper.conf
}


From: "Tech Support" <adam at homebusiness dot to>
Subject: 30 second delay with qpopper
Date: Mon, 24 Dec 2001 03:06:11 -0500

Hello,

I am experiencing a 30 second (exactly) wait for a client to connect to the
server.  I try to set the nice variable to -20 and everything inbetween, but
I am unable to get rid of the delay.  Any help would be much appreciate.  My
direct email address is adam at homebusiness dot to.  Thanks.


Date: Mon, 24 Dec 2001 18:38:40 +0900
From: Peter Evans <peter at gol dot com>
Subject: Re: 30 second delay with qpopper

Tech Support (adam at homebusiness dot to) wrote:
> I am experiencing a 30 second (exactly) wait for a client to connect to the
> server.  I try to set the nice variable to -20 and everything inbetween, but
> I am unable to get rid of the delay.  Any help would be much appreciate.  My
> direct email address is adam at homebusiness dot to.  Thanks.

	This doesnt sound like qpopper, look elsewhere.

	eg,

	identd / tcp wrappers / ...

	The very fact that playing with nice doesnt do anything should have
	given you hints as to where to look.


	P
	


-- 
New Snibbo! Combined Breakfast Cereal and Washing-Up Liquid!

From: Gerhard Gonter <gonter at buddy.wu-wien.ac dot at>
Subject: Re: 30 second delay with qpopper
Date: Mon, 24 Dec 2001 10:44:15 +0100 (MEZ)

According to Peter Evans:
> Tech Support (adam at homebusiness dot to) wrote:
> > I am experiencing a 30 second (exactly) wait for a client to connect to the
> [...]
> 
> 	This doesnt sound like qpopper, look elsewhere.
> 
> 	eg,
> 
> 	identd / tcp wrappers / ...

... or there might be a problem with reverse DNS.

+gg
 
-- 
Gerhard.Gonter at wu-wien.ac.at  Fax: +43/1/31336/702  g dot gonter at ieee dot org
Zentrum fuer Informatikdienste, Wirtschaftsuniversitaet Wien, Austria

Date: Mon, 24 Dec 2001 08:49:41 -0800
From: Syed Zaeem Hosain <Syed.Hosain at aeris dot net>
Subject: Re: 30 second delay with qpopper

Hi!

Tech Support wrote:
> I am experiencing a 30 second (exactly) wait for a client to connect to the
> server.  I try to set the nice variable to -20 and everything inbetween, but
> I am unable to get rid of the delay.  Any help would be much appreciate.  My
> direct email address is adam at homebusiness dot to.  Thanks.

This sounds like a DNS lookup delay. With its timeout of 30 seconds.

Date: Mon, 24 Dec 2001 12:38:14 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: 30 second delay with qpopper

At 03:06 AM 12/24/01, Tech Support wrote:
>Hello,
>
>I am experiencing a 30 second (exactly) wait for a client to connect to the
>server.  I try to set the nice variable to -20 and everything inbetween, but
>I am unable to get rid of the delay.  Any help would be much appreciate.  My
>direct email address is adam at homebusiness dot to.  Thanks.

I was dealing with the exact same thing. On my systems, this turned out to 
be an IDENT protocol thing, caused by the config of XINETD. If you're using 
XINETD, go into the config file you're using, and remove "USERID" from any 
logging lines, and that'll help.

I should also note that I've found and reported a bug in the xinetd 
RedHat's currently shipping, in which it'll sometimes start doing ident 
again, even if you've told it not to.
-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: Mon, 24 Dec 2001 12:35:17 -0500
From: Daniel Senie <dts at senie dot com>
Subject: RE: problem with qpopper 4.04

At 11:56 PM 12/23/01, you wrote:
># qpopper config file for xinetd
>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
>         # use this line once you create your certificate to enable STLS
>         # server_args = in.qpopper -s -l 1 -f /etc/mail/qpopper.conf
>}

This will result in TCP Wrappers being called (via libwrap).

You need to add an entry in your /etc/hosts.allow to permit qpopper to run. 
It'll probably be something like:

in.qpopper: ALL: ALLOW

TCP Wrappers works based on the program name of the daemon, which would 
account for why the ipop3d worked and qpopper doesn't, assuming your 
/etc/hosts.allow is other than an empty file.

[You might want to tell us what your /etc/hosts.allow looks like, if the 
above doesn't fix the problem]

>-----Original Message-----
>From: Kenneth Porter [mailto:shiva at well dot com]
>Sent: Monday, December 24, 2001 10:27 AM
>To: Subscribers of Qpopper
>Subject: Re: problem with qpopper 4.04
>
>
>On Sun, 2001-12-23 at 18:21, Peter Santiago wrote:
> > qpopper's complaint..
> > Dec 23 22:43:46 gateway in.qpopper[6663]: Unable to obtain socket and
> > address
> > .... if I used gnupop3d....without changing anything... works fine...
>really
> > puzzled....
>
>How are you launching it? What's your inetd/xinetd setup look like? Or
>are you running it standalone?
>
>through xinetd.... if I launch it as stand-alone... wouldn't connect to
>socket...........
>
># qpopper config file for xinetd
>service pop3
>{
>         flags       = REUSE NAMEINARGS
>         socket_type = stream
>         protocol    = tcp
>         wait        = no
>         user        = root
>         server      = /usr/sbin/in.qpopper
>         server_args = in.qpopper -s
>         # use this line once you create your certificate to enable STLS
>         # server_args = in.qpopper -s -l 1 -f /etc/mail/qpopper.conf
>}

-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: RE: problem with qpopper 4.04
Date: Tue, 25 Dec 2001 12:25:03 +0800

content of /etc/hosts.allow

sshd: ALL
exim: ALL
in.telnetd:192.168.1.
in.qpopper:192.168.1.

-----Original Message-----
From: Daniel Senie [mailto:dts at senie dot com]
Sent: Tuesday, December 25, 2001 1:35 AM
To: Subscribers of Qpopper
Subject: RE: problem with qpopper 4.04


At 11:56 PM 12/23/01, you wrote:
># qpopper config file for xinetd
>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
>         # use this line once you create your certificate to enable STLS
>         # server_args = in.qpopper -s -l 1 -f /etc/mail/qpopper.conf
>}

This will result in TCP Wrappers being called (via libwrap).

You need to add an entry in your /etc/hosts.allow to permit qpopper to run.
It'll probably be something like:

in.qpopper: ALL: ALLOW

TCP Wrappers works based on the program name of the daemon, which would
account for why the ipop3d worked and qpopper doesn't, assuming your
/etc/hosts.allow is other than an empty file.

[You might want to tell us what your /etc/hosts.allow looks like, if the
above doesn't fix the problem]

>-----Original Message-----
>From: Kenneth Porter [mailto:shiva at well dot com]
>Sent: Monday, December 24, 2001 10:27 AM
>To: Subscribers of Qpopper
>Subject: Re: problem with qpopper 4.04
>
>
>On Sun, 2001-12-23 at 18:21, Peter Santiago wrote:
> > qpopper's complaint..
> > Dec 23 22:43:46 gateway in.qpopper[6663]: Unable to obtain socket and
> > address
> > .... if I used gnupop3d....without changing anything... works fine...
>really
> > puzzled....
>
>How are you launching it? What's your inetd/xinetd setup look like? Or
>are you running it standalone?
>
>through xinetd.... if I launch it as stand-alone... wouldn't connect to
>socket...........
>


Date: Tue, 25 Dec 2001 11:20:14 -0800
From: Chuck Yerkes <chuck+qpopper at yerkes dot com>
Subject: Re: 30 second delay with qpopper

30 seconds is almost always a wait for a DNS lookup.

Quoting Tech Support (adam at homebusiness dot to):
> Hello,
> 
> I am experiencing a 30 second (exactly) wait for a client to connect to the
> server.  I try to set the nice variable to -20 and everything inbetween, but
> I am unable to get rid of the delay.  Any help would be much appreciate.  My
> direct email address is adam at homebusiness dot to.  Thanks.

Date: Tue, 25 Dec 2001 14:42:31 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: 30 second delay with qpopper

At 02:20 PM 12/25/01, Chuck Yerkes wrote:
>30 seconds is almost always a wait for a DNS lookup.

I've found that not to be the case. A 30 second delay will ONLY be DNS the 
first time. The DNS server near (or co-resident) with the server providing 
caching will cache either the positive or negative result. So, you might 
see a 30 second delay once, but if you then check your mail a second time 
and get another 30 second delay, that would point elsewhere.

I've found the 30 second delay to be a telltale sign of IDENT processing in 
XINETD. Indeed, in Xinetd, someone hard-coded 30 seconds as a timeout, and 
didn't even provide a way to alter that value. Removing USERID logging in 
your xinetd config will turn this off, give or take bugs in xinetd that 
seem to turn it back on.
-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: RE: problem with qpopper 4.04
Date: Wed, 26 Dec 2001 05:34:25 +0800

Dunno what's going on.... qpopper's working now.... there's another
problem... I have configured Exim MTA to deliver mail to users home
directory ($HOME/Mailbox)...Accordingly I have added set home-dir-mail 
Mailbox to /etc/qpopper.config.... I also added -d to in.qpopper
commandline.... after restarting xinetd, when I examined /var/log/maillog I
found out that qpopper is still reading from the /var/spool/mail... Neither
Outlook nor Eudora can pick up any POP mail.....


Dec 26 05:16:06 gateway in.qpopper[3169]: +OK ready
<3169.1009361766 at gateway dot lan> [popper.c:242]
Dec 26 05:16:06 gateway in.qpopper[3169]: Qpopper ready for input from
(null)at petersan [192.168.1.100] [popper.c:281]
Dec 26 05:16:06 gateway in.qpopper[3169]: Received (19): "USER
peter.santiago"
 [pop_get_command.c:105]
Dec 26 05:16:06 gateway in.qpopper[3169]: home (26):
'/home/users/peter.santiago' [pop_user.c:209]
Dec 26 05:32:54 gateway in.qpopper[3351]: No such group: "purgemail"
[check_gr
oup.c:37]
Dec 26 05:32:54 gateway in.qpopper[3351]: genpath Spool (1) [hash: 0; home:
NU
LL] for user peter.santiago returning /var/spool/mail/peter.santiago
[genpath.
c:229]
Dec 26 05:32:54 gateway in.qpopper[3351]: genpath .pop (2) [hash: 0; home:
NUL
L] for user peter.santiago returning /tmp/.peter.santiago.pop
[genpath.c:229]
Dec 26 05:32:54 gateway in.qpopper[3351]: Temporary maildrop name:
'/tmp/.pete
r.santiago.pop' [pop_dropcopy.c:1255]
Dec 26 05:32:54 gateway in.qpopper[3351]: Opened Bulletin database
'/var/spool
/mail/bulletins/bulldb' (checking user 'peter.santiago')
[pop_dropcopy.c:1301]
Dec 26 05:32:54 gateway in.qpopper[3351]: uid = 503, gid = 12, euid = 503,
egi
d = 12 [pop_dropcopy.c:1485]
Dec 26 05:32:54 gateway in.qpopper[3351]: Opened temp drop
/tmp/.peter.santiag
o.pop (5) [pop_dropcopy.c:1501]
Dec 26 05:32:54 gateway in.qpopper[3351]: Set p->drop to stream for 5
[pop_dro
pcopy.c:1585]
Dec 26 05:32:54 gateway in.qpopper[3351]: Getting mail lock
[pop_dropcopy.c:16
25]
Dec 26 05:32:54 gateway in.qpopper[3351]: successfully opened (exclusive)
lock
 /var/spool/mail/peter.santiago.lock [maillock.c:477]

-----Original Message-----
From: Daniel Senie [mailto:dts at senie dot com]
Sent: Wednesday, December 26, 2001 1:12 AM
To: Peter Santiago
Subject: RE: problem with qpopper 4.04


At 11:25 PM 12/24/01, you wrote:
>content of /etc/hosts.allow
>
>sshd: ALL
>exim: ALL
>in.telnetd:192.168.1.
>in.qpopper:192.168.1.

You might want to read the man pages.

While what you have may work (old syntax), the current/preferred method
would say:

sshd: ALL: ALLOW
exim: ALL: ALLOW
in.telnetd: 192.168.1.0/255.255.255.0: ALLOW
in.qpopper: 192.168.1.0/255.255.255.0: ALLOW

ALL: ALL: DENY

----

The log file /var/log/secure will tell you about the success or failure of
TCP Wrappers activity.

You might also test and see if telnet daemon is working, as it's using the
same mechanisms.

Other things: In your qpopper config for xinetd, you have "service pop-3".
Check your /etc/services file to be sure that's correct. On my system, I
have "service pop3".

Another thing to do: "netstat -an | more" and see if there's a line that
reads in part "0.0.0.0:110".




Date: Wed, 26 Dec 2001 09:35:41 -0800
From: Chuck Yerkes <chuck+qpopper at yerkes dot com>
Subject: Re: 30 second delay with qpopper

Quoting Daniel Senie (dts at senie dot com):
> At 02:20 PM 12/25/01, Chuck Yerkes wrote:
> >30 seconds is almost always a wait for a DNS lookup.
> 
> I've found that not to be the case. A 30 second delay will ONLY be DNS the 
> first time. The DNS server near (or co-resident) with the server providing 
> caching will cache either the positive or negative result. So, you might 
> see a 30 second delay once, but if you then check your mail a second time 
> and get another 30 second delay, that would point elsewhere.

Except when that DNS lookup fails.  Most DNS servers don't do
negative caching (storing failures).  When a connection comes
in, a lookup of the connecting IP address is done.  Unfortunately,
too much of the Internet is not run by the competent any more,
and those lookups fail.  My favorites are that 2.1.220.in-addr.arpa
(e.g.) will have NS servers that are unreachable.  This will cause
a 30 second delay each time.

> I've found the 30 second delay to be a telltale sign of IDENT processing in 
> XINETD. Indeed, in Xinetd, someone hard-coded 30 seconds as a timeout, and 
> didn't even provide a way to alter that value. Removing USERID logging in 
> your xinetd config will turn this off, give or take bugs in xinetd that 
> seem to turn it back on.

Dunno, don't really use Linux much.  When I do, I build from
source (a BSD habit).  I happily change the timeouts (it's the
point of using Open Source tools).

To make this note actually *useful*, I'll suggest that when
facing this sort of problem, a useful tool to debug is to run
tcpdump (or a tool on top of that) on the segment, filtering
for the connecting network or host.  You'll see that an IDENT
packet has gone out and, due to misconfigured firewalls, has
been dropped rather than rejected, causing delay.  (we found
THAT with sendmail a lot, leading to the current sendmail
default timeout of 2s rather than 30s for ident packets).

Sniff the line, see what's going on.  Run in debug mode and
watch.  Remove portions of the system (inetd, qpopper, use
different clients) and see where the delay is introduced.

Date: Wed, 26 Dec 2001 13:02:40 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: 30 second delay with qpopper

At 12:35 PM 12/26/01, Chuck Yerkes wrote:
>Quoting Daniel Senie (dts at senie dot com):
> > At 02:20 PM 12/25/01, Chuck Yerkes wrote:
> > >30 seconds is almost always a wait for a DNS lookup.
> >
> > I've found that not to be the case. A 30 second delay will ONLY be DNS the
> > first time. The DNS server near (or co-resident) with the server providing
> > caching will cache either the positive or negative result. So, you might
> > see a 30 second delay once, but if you then check your mail a second time
> > and get another 30 second delay, that would point elsewhere.
>
>Except when that DNS lookup fails.  Most DNS servers don't do
>negative caching (storing failures).

That's sure news. They are supposed to do negative caching. The timeouts on 
negatives tend to be short, but if you don't do this, the world tends to go 
to hell.

>   When a connection comes
>in, a lookup of the connecting IP address is done.  Unfortunately,
>too much of the Internet is not run by the competent any more,
>and those lookups fail.  My favorites are that 2.1.220.in-addr.arpa
>(e.g.) will have NS servers that are unreachable.  This will cause
>a 30 second delay each time.
>
> > I've found the 30 second delay to be a telltale sign of IDENT 
> processing in
> > XINETD. Indeed, in Xinetd, someone hard-coded 30 seconds as a timeout, and
> > didn't even provide a way to alter that value. Removing USERID logging in
> > your xinetd config will turn this off, give or take bugs in xinetd that
> > seem to turn it back on.
>
>Dunno, don't really use Linux much.  When I do, I build from
>source (a BSD habit).  I happily change the timeouts (it's the
>point of using Open Source tools).
>
>To make this note actually *useful*, I'll suggest that when
>facing this sort of problem, a useful tool to debug is to run
>tcpdump (or a tool on top of that) on the segment, filtering
>for the connecting network or host.  You'll see that an IDENT
>packet has gone out and, due to misconfigured firewalls, has
>been dropped rather than rejected, causing delay.  (we found
>THAT with sendmail a lot, leading to the current sendmail
>default timeout of 2s rather than 30s for ident packets).

Another useful tidbit...

Careful use of IPCHAINS or similar filtering can allow you to reject the 
outbound IDENT packet (reject == icmp response).

In sendmail, I recofigure the timeout to zero, which results in sendmail 
NOT doing Ident. Ident is just not useful in the email world at this point. 
It only hurts, never helps.


>Sniff the line, see what's going on.  Run in debug mode and
>watch.  Remove portions of the system (inetd, qpopper, use
>different clients) and see where the delay is introduced.

-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: Wed, 26 Dec 2001 21:55:55 +0100
From: Newsletter <newsletter at spellweaver dot de>
Subject: How to create POP-Boxes without creating REAL users?

Hi all,

hope I'm not 'off-topic', but my problem concerns both Procmail and
Qpopper, I guess.

I'm just wondering if there was any possibility to recieve/accept
mails when there doesn't exist a REAL user.
Since my system shall only serve as a 'Mailsever' I do not want to
create a real User with an home-directory for every POP-Box I add.

Furthermore I'm searching for an Admin-Frontend (best web-based) for
creating/maintaining POP-Boxes, setting up forwards or vacation
messages and so on.
It would be great if someone could give me some hint!!

Thanx in advance and best Christmas wishes,
Kai



From: "Peter Santiago" <petersantiago at bigfoot dot com>
Subject: FW: problem with qpopper 4.04
Date: Thu, 27 Dec 2001 07:48:53 +0800

Dunno what's going on.... qpopper's working now.... there's another
problem... I have configured Exim MTA to deliver mail to users home
directory ($HOME/Mailbox)...Accordingly I have added set home-dir-mail 
Mailbox to /etc/qpopper.config.... I also added -d to in.qpopper
commandline.... after restarting xinetd, when I examined /var/log/maillog I
found out that qpopper is still reading from the /var/spool/mail... Neither
Outlook nor Eudora can pick up any POP mail.....


Dec 26 05:16:06 gateway in.qpopper[3169]: +OK ready
<3169.1009361766 at gateway dot lan> [popper.c:242]
Dec 26 05:16:06 gateway in.qpopper[3169]: Qpopper ready for input from
(null)at petersan [192.168.1.100] [popper.c:281]
Dec 26 05:16:06 gateway in.qpopper[3169]: Received (19): "USER
peter.santiago"
 [pop_get_command.c:105]
Dec 26 05:16:06 gateway in.qpopper[3169]: home (26):
'/home/users/peter.santiago' [pop_user.c:209]
Dec 26 05:32:54 gateway in.qpopper[3351]: No such group: "purgemail"
[check_gr
oup.c:37]
Dec 26 05:32:54 gateway in.qpopper[3351]: genpath Spool (1) [hash: 0; home:
NU
LL] for user peter.santiago returning /var/spool/mail/peter.santiago
[genpath.
c:229]
Dec 26 05:32:54 gateway in.qpopper[3351]: genpath .pop (2) [hash: 0; home:
NUL
L] for user peter.santiago returning /tmp/.peter.santiago.pop
[genpath.c:229]
Dec 26 05:32:54 gateway in.qpopper[3351]: Temporary maildrop name:
'/tmp/.pete
r.santiago.pop' [pop_dropcopy.c:1255]
Dec 26 05:32:54 gateway in.qpopper[3351]: Opened Bulletin database
'/var/spool
/mail/bulletins/bulldb' (checking user 'peter.santiago')
[pop_dropcopy.c:1301]
Dec 26 05:32:54 gateway in.qpopper[3351]: uid = 503, gid = 12, euid = 503,
egi
d = 12 [pop_dropcopy.c:1485]
Dec 26 05:32:54 gateway in.qpopper[3351]: Opened temp drop
/tmp/.peter.santiag
o.pop (5) [pop_dropcopy.c:1501]
Dec 26 05:32:54 gateway in.qpopper[3351]: Set p->drop to stream for 5
[pop_dro
pcopy.c:1585]
Dec 26 05:32:54 gateway in.qpopper[3351]: Getting mail lock
[pop_dropcopy.c:16
25]
Dec 26 05:32:54 gateway in.qpopper[3351]: successfully opened (exclusive)
lock
 /var/spool/mail/peter.santiago.lock [maillock.c:477]



Date: Thu, 27 Dec 2001 17:00:34 -0600
From: Tim Tyler <tyler at beloit dot edu>
Subject: Client support?

Qpopper experts,
   I have successfully installed qpopper with openssl on our AIX4.3.3 
system and Eudora appears to work fine retrieving email.  However, I can't 
get Outlook to work with ssl enabled.  Are there any compatibility issues 
that I should be aware of with Outlook?  If not, what might my problem 
be?  Here are log results in case that helps:

Dec 27 09:28:24 stud2 qpopper[16404]: Set tls-support to STLS (2)
Dec 27 09:28:24 stud2 qpopper[16404]: Set tls-server-cert-file to 
"/etc/mail/cer
ts/cert.pem"
Dec 27 09:28:24 stud2 qpopper[16404]: (v4.0.3) Unable to get canonical name 
of c
lient 144.89.40.89: Unknown host (1)
Dec 27 09:28:24 stud2 qpopper[16404]: (null) at 144.89.40.89 
(144.89.40.89): -ER
R Unknown command: "\200j^A^C^A".

     Any thoughts are much appreciated! -tim



Tim Tyler
Network Engineer - Beloit College
tyler at beloit dot edu


Date: Thu, 27 Dec 2001 19:53:22 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: Client support?

At 06:00 PM 12/27/01, Tim Tyler wrote:
>Qpopper experts,
>   I have successfully installed qpopper with openssl on our AIX4.3.3 
> system and Eudora appears to work fine retrieving email.  However, I 
> can't get Outlook to work with ssl enabled.  Are there any compatibility 
> issues that I should be aware of with Outlook?  If not, what might my 
> problem be?  Here are log results in case that helps:
>
>Dec 27 09:28:24 stud2 qpopper[16404]: Set tls-support to STLS (2)
>Dec 27 09:28:24 stud2 qpopper[16404]: Set tls-server-cert-file to 
>"/etc/mail/cer
>ts/cert.pem"
>Dec 27 09:28:24 stud2 qpopper[16404]: (v4.0.3) Unable to get canonical 
>name of c
>lient 144.89.40.89: Unknown host (1)
>Dec 27 09:28:24 stud2 qpopper[16404]: (null) at 144.89.40.89 
>(144.89.40.89): -ER
>R Unknown command: "\200j^A^C^A".
>
>     Any thoughts are much appreciated! -tim

Outlook doesn't support STARTTLS. You need to set up alternate port service 
on port 995 with the tls support set to just run TLS without negotiation. 
Most folks wind up setting up two config files, one for port 110 with STLS 
and one for port 995 with the tls-support set to alternate-port.

-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: Thu, 27 Dec 2001 20:08:58 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Fwd: NDN: Re: Client support?

Can someone with moderator authority figure out who this person is, and 
kick them off the list?

It appears caisnet.com can't figure out how to send non-delivery 
notifications to the ENVELOPE sender of messages, and this user has 
apparently abandoned an account which is over-quota and subscribed to this 
list.

The message below is received in response to EVERY message sent to the 
qpopper list.


>Date: Thu, 27 Dec 2001 18:01:58 -0700
>Subject: NDN: Re: Client support?
>To: dts at senie dot com
>From: "Mailer-Daemon" <Mailer-Daemon at caisnet dot com>
>
>Sorry. Your message could not be delivered to:
>
>Chuck D. Pitre (Mailbox or Conference is full.)

-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: 28 Dec 2001 22:56:00 +0100
From: "Arne Schmitz" <arne.schmitz at mmweg.rwth-aachen dot de>
Subject: qpopper, qmail and maildirs

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Is there a way to use qmail and mailboxes in the maildir-format with qpopper 
4.0.3?

Arne

- -- 
"How do I love thee?  My accumulator overflows."

[--- PGP key available on http://www.root42.de/ ---]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE8LOp6zpP3gv0FvtcRArPOAJ4knzpLUbkVE5HkTAnDy79GTSYq5gCglVeu
SKMiA+ej1Ijg1AeE8Qu/8Oc
=TFW2
-----END PGP SIGNATURE-----

Subject: Re: qpopper, qmail and maildirs
From: Kenneth Porter <shiva at well dot com>
Date: 28 Dec 2001 23:55:37 -0800

On Fri, 2001-12-28 at 13:56, Arne Schmitz wrote:
> Is there a way to use qmail and mailboxes in the maildir-format with qpopper 
> 4.0.3?

It's been on the list of desired features for some time, just waiting
for some enterprising user to code up the patch. ;-)



Date: Sat, 29 Dec 2001 23:29:39 -0800
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: How to create POP-Boxes without creating REAL users?

At 9:55 PM +0100 12/26/01, Newsletter wrote:

>  Since my system shall only serve as a 'Mailsever' I do not want to
>  create a real User with an home-directory for every POP-Box I add.

Qpopper currently needs a real user account per user, but does not 
require a home directory.  You can also set an invalid shell to 
prevent logins.

Date: Sat, 29 Dec 2001 23:26:57 -0800
From: Randall Gellens <randy at qualcomm dot com>
Subject: RE: problem with qpopper 4.04

At 5:34 AM +0800 12/26/01, Peter Santiago wrote:

>  there's another
>  problem... I have configured Exim MTA to deliver mail to users home
>  directory ($HOME/Mailbox)...Accordingly I have added set home-dir-mail 
>  Mailbox to /etc/qpopper.config.... I also added -d to in.qpopper
>  commandline.... after restarting xinetd, when I examined /var/log/maillog I
>  found out that qpopper is still reading from the /var/spool/mail... Neither
>  Outlook nor Eudora can pick up any POP mail.....

Did you set the command line to process the config file (/etc/qpopper.config)?

Date: Sun, 30 Dec 2001 19:10:17 +0100
From: Kai Bleek <newsletter at spellweaver dot de>
Subject: Configuring APOP - gdbm_open Error

Hi,

I'm just trying to configure Qpopper 4.03 with Apop-Support on my Red
Hat 7.2 system.

While running
./configure --with-drac --enable-apop=/etc/pop.auth --enable-popuid=pop
I get the following error message:

checking which database manager to use ... checking gdbm ... checking for gdbm_open in -lgdbm... (cached) no
didn't find gdbm

By running make I get (I guess this follows from the above error):

pop_user.c: In function `pop_user':
pop_user.c:133: `DBM' undeclared (first use in this function)
pop_user.c:133: (Each undeclared identifier is reported only once
pop_user.c:133: for each function it appears in.)
pop_user.c:133: `db' undeclared (first use in this function)
pop_user.c:135: parse error before `int'
pop_user.c:137: `datum' undeclared (first use in this function)
pop_user.c:177: `user_name_len' undeclared (first use in this function)
pop_user.c:210: `pw' undeclared (first use in this function)
pop_user.c:265: `st' undeclared (first use in this function)
pop_user.c:281: `fid' undeclared (first use in this function)
pop_user.c:308: `key' undeclared (first use in this function)
pop_user.c:321: `value' undeclared (first use in this function)
pop_user.c:324: `i' undeclared (first use in this function)
pop_user.c:328: `bFoundUser' undeclared (first use in this function)
make[1]: *** [pop_user.o] Error 1
make[1]: Leaving directory `/root/qpopper4.0.3/popper'
make: *** [popper_server] Error 2

So 'make' stops with an error and qpopper didn't compile correctly.

I saw the same question in the mailing-list-archive somewhere around
the middle of 2001 (or was it 2000?) but it stayed unanswered.

Hope someone can help me out.
Thanks in advance and all the best for 2002!

Best wishes,
Kai



Subject: Re: Configuring APOP - gdbm_open Error
From: Kenneth Porter <shiva at well dot com>
Date: 30 Dec 2001 10:52:58 -0800

On Sun, 2001-12-30 at 10:10, Kai Bleek wrote:

> While running
> ./configure --with-drac --enable-apop=/etc/pop.auth --enable-popuid=pop
> I get the following error message:
> 
> checking which database manager to use ... checking gdbm ... checking for gdbm_open in -lgdbm... (cached) no
> didn't find gdbm

Sounds like you don't have /usr/lib/libgdbm*. Do you have the gdbm-devel
RPM package installed? That should include the needed library.


Date: Sun, 30 Dec 2001 20:13:38 +0100
From: Kai Bleek <newsletter at spellweaver dot de>
Subject: Re[2]: Configuring APOP - gdbm_open Error

KP> Sounds like you don't have /usr/lib/libgdbm*. Do you have the gdbm-devel
KP> RPM package installed? That should include the needed library.

Yes, gdbm-devel is installed properly - here's an exerpt from
/usr/lib:

[root@panic lib]# ls -la libgdbm*
-rw-r--r--    1 root     root        52594 Jun 26  2001 libgdbm.a
-rwxr-xr-x    1 root     root          692 Jun 26  2001 libgdbm.la
lrwxrwxrwx    1 root     root           16 Dec 30 19:59 libgdbm.so -> libgdbm.so.2.0.0
lrwxrwxrwx    1 root     root           16 Nov 14 23:05 libgdbm.so.2 -> libgdbm.so.2.0.0
-rwxr-xr-x    1 root     root        30114 Jun 26  2001 libgdbm.so.2.0.0

Thannks anyway for the help!
Kai



Date: Sun, 30 Dec 2001 20:39:40 +0100
From: Kai Bleek <newsletter at spellweaver dot de>
Subject: Re: Configuring APOP - gdbm_open Error


> You wiped the config cache and reran configure, right?

HA! Got me...
That was the little thing I DID forget! (how stupid...)

> You might also do the same check for installed RPMs for the gdbm stuff, but
> I think the db3 RPMs are supplying the equivalency routines now.

Credits and many thanx to Daniel Senie for rescuing me!!
All the best,
Kai



Date: Mon, 31 Dec 2001 09:43:35 +0800
From: Daniel Suen <ttdsuen at ln.edu dot hk>
Subject: qpopper upgrade XUIDL issue...

Hi All,

We have upgraded the Qpopper from 3.1.2 to 4.0.3 a couple days ago. 
However, we seem to experience some problems with XUIDL. In both versions, 
I compiled with the same set of options to the configure script, but after 
the change took place, some old messages left on server got downloaded 
again. Can anyone explain what happened? I read the release notes and there 
did not seem to have any changes on XUIDL since version 3.0. The following 
is my compile option:

./configure  --enable-bulletins=/usr/local/var/qpopper/bulls 
--enable-log-login --enable-group-bulls --enable-temp-drop 
dir=/usr/local/var/qpopper/poptemp --enable-uw-kludge --with-pam=pop3 
--enable-shy --enable-old-uidl --no-create --no-recursion

Any ideas?

-Daniel.


Subject: Emails are retrieved twice (sometimes 3 times) from Qpopper vesion 3.1.2.
Date: Sun, 30 Dec 2001 21:09:19 -0800
From: "Habib Abassi" <habassi at telephonyatwork dot com>

Hi All,
Emails are retrieved twice (sometimes 3 times) from Qpopper vesion 
3.1.2. 
I am using a netscape library to connect and retrieve emails. After 
retrieving email I mark it as deleted, it should be deleted after quit, 
but when I connect again the same email still in mailbox and I can 
retrieved again .... If any body kn ow why this is happen and if there's 
a way to know that the email is marked as deleted before retrieving it.
Thanks in advance


From: "Fred G Guiliano" <fred.guiliano at DPS-inc dot com>
Subject: Canonical name error...
Date: Wed, 2 Jan 2002 11:03:03 -0500

This is a multi-part message in MIME format.

------=_NextPart_000_0003_01C1937D.0E82E920
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

Hi All,

I see at the the qualcomm site this particular explanation of the following
error:

What does the error message Unable to get canonical name of client mean?
  This error message in your log means that a client connected in and
Qpopper got an error when it tried to do a reverse lookup on the IP address.
If the client is local, this may indicate a problem with its DNS entry
(specifically the PTR record for that IP address), or it may be a temporary
DNS problem.
  Qpopper allows you to turn off this check if you like.  Qpopper also
includes the IP address of the client in the log message, to make it easier
to check its DNS entries.

In this explanation, it says that Qpopper allows you to turn off this check.
That's nice, but my question is HOW do you turn this check off? I didn't see
anything in the manual on how to do this (It maybe there but I guess I'm not
seeing it). Any help would be appreciated!

Thanks!
Fred G Guiliano
Network Consultant
Dorset Professional Services



------=_NextPart_000_0003_01C1937D.0E82E920
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!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 5.00.3315.2870" name=GENERATOR></HEAD>
<BODY>
<DIV><FONT face=Arial size=2><SPAN class 1425615-02012002>Hi 
All,</SPAN></FONT></DIV>
<DIV><FONT face=Arial size=2><SPAN 
class 1425615-02012002></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2><SPAN class 1425615-02012002>I see 
at the the 
qualcomm site this particular explanation of the following 
error:</SPAN></FONT></DIV>
<DIV><FONT face=Arial size=2><SPAN 
class 1425615-02012002></SPAN></FONT>&nbsp;</DIV>
<DIV><SPAN class 1425615-02012002><FONT face=Arial 
size=2><B>What does the 
error message <I>Unable to get canonical name of client </I>mean? 
</B></FONT></DIV>
<BLOCKQUOTE><FONT face=Arial size=2>This error message in your log 
means that 
  a client connected in and Qpopper got an error when it tried to do a 
reverse 
  lookup on the IP address.&nbsp; If the client is local, this may 
indicate a 
  problem with its DNS entry (specifically the PTR record for that IP 
address), 
  or it may be a temporary DNS problem. </FONT>
  <P><FONT face=Arial size=2>Qpopper allows you to turn off this 
check if you 
  like. &nbsp;Qpopper also includes the IP address of the client in the 
log 
  message, to make it easier to check its DNS entries. 
</FONT></P></BLOCKQUOTE>
<DIV><FONT face=Arial size=2><SPAN class 1425615-02012002>In 
this explanation, 
it says that Qpopper allows you to turn off this check. That's nice, but 
my 
question is HOW do you turn this check off? I didn't see anything in the 
manual 
on how to do this (It maybe there but I guess I'm not seeing it). Any 
help would 
be appreciated!</SPAN></FONT></DIV>
<DIV><FONT face=Arial size=2><SPAN 
class 1425615-02012002></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2><SPAN 
class 1425615-02012002>Thanks!</SPAN></FONT></DIV></SPAN>
<P><I><FONT color=#000080 face="Comic Sans MS">Fred G 
Guiliano</FONT></I> 
<BR><FONT color=#000080 face="Comic Sans MS" size=2>Network 
Consultant</FONT> 
<BR><FONT color=#000080 face="Comic Sans MS" size=2>Dorset 
Professional 
Services</FONT> <BR></P></BODY></HTML>

------=_NextPart_000_0003_01C1937D.0E82E920--


From: James McLaughlin <katana at montrose dot net>
Subject: unsubsribe
Date: Wed, 2 Jan 2002 09:40:57 -0700



Date: Wed, 2 Jan 2002 09:33:16 -0800 (PST)
From: Gregory Hicks <ghicks at cadence dot com>
Subject: Re: Canonical name error...

> From: "Fred G Guiliano" <fred.guiliano at dps-inc dot com>
> Date: Wed, 2 Jan 2002 11:03:03 -0500
> 
> Hi All,
> 
> I see at the the qualcomm site this particular explanation of the 
> following error:
> 
> What does the error message Unable to get canonical name of client 
> mean?
[...snip...]
>   Qpopper allows you to turn off this check if you like.  Qpopper also
> includes the IP address of the client in the log message, to make it 
> easier to check its DNS entries.
> 
> In this explanation, it says that Qpopper allows you to turn off this 
> check.  That's nice, but my question is HOW do you turn this check 
> off? I didn't see anything in the manual on how to do this (It maybe 
> there but I guess I'm not seeing it). Any help would be appreciated!

Although qpopper may allow you to turn the check off (and no, I don't 
know the option, but it is in the docs or the config options), it seems 
to me - and this is a 'hostmaster' and 'postmaster' speaking - that if 
you cannot reverse look up one of your own hosts, you *might* have a DNS 
config issue...  That should be solved...

Just my own $0.02 worth...

Regards,
Gregory Hicks

---------------------------------------------------------------------
Gregory Hicks                           | Principal Systems Engineer
Cadence Design Systems                  | Direct:   408.576.3609
555 River Oaks Pkwy M/S 6B1             | Fax:      408.894.3479
San Jose, CA 95134                      | Internet: ghicks at cadence dot com

"There is no limit to what a man can do or how far he can go if he
doesn't mind who gets the credit." - Robert Woodruff

"The trouble with doing anything right the first time is that nobody
appreciates how difficult it was."

When a team of dedicated individuals makes a commitment to act as
one...  the sky's the limit.


Date: Wed, 02 Jan 2002 12:51:42 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: Canonical name error...

At 12:33 PM 1/2/02, Gregory Hicks wrote:
> > From: "Fred G Guiliano" <fred.guiliano at dps-inc dot com>
> > Date: Wed, 2 Jan 2002 11:03:03 -0500
> >
> > Hi All,
> >
> > I see at the the qualcomm site this particular explanation of the
> > following error:
> >
> > What does the error message Unable to get canonical name of client
> > mean?
>[...snip...]
> >   Qpopper allows you to turn off this check if you like.  Qpopper also
> > includes the IP address of the client in the log message, to make it
> > easier to check its DNS entries.
> >
> > In this explanation, it says that Qpopper allows you to turn off this
> > check.  That's nice, but my question is HOW do you turn this check
> > off? I didn't see anything in the manual on how to do this (It maybe
> > there but I guess I'm not seeing it). Any help would be appreciated!
>
>Although qpopper may allow you to turn the check off (and no, I don't
>know the option, but it is in the docs or the config options), it seems
>to me - and this is a 'hostmaster' and 'postmaster' speaking - that if
>you cannot reverse look up one of your own hosts, you *might* have a DNS
>config issue...  That should be solved...

Think "travelling user." You'll never have control over everything.

On every version of qpopper I install, I comment out the code that sprays 
the error message about DNS lookup. I don't want to turn the option off, 
though, since if the name DOES resolve, I want to find out what it is and 
log it.

Next time I get a chance to do some work on the qpopper code (will probably 
be soon) I think I'll take a look at this. Simply removing this message 
entirely may be the best answer.
-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: Wed, 2 Jan 2002 10:09:44 -0800 (PST)
From: Gregory Hicks <ghicks at cadence dot com>
Subject: Re: Canonical name error...

> Date: Wed, 02 Jan 2002 12:51:42 -0500
> From: Daniel Senie <dts at senie dot com>
> 
> At 12:33 PM 1/2/02, Gregory Hicks wrote:
> > > From: "Fred G Guiliano" <fred.guiliano at dps-inc dot com>
> > > Date: Wed, 2 Jan 2002 11:03:03 -0500
> > >
> > > Hi All,
> > >
> > > I see at the the qualcomm site this particular explanation of the
> > > following error:
[...snip...]

> >Although qpopper may allow you to turn the check off (and no, I don't
> >know the option, but it is in the docs or the config options), it 
> >seems to me - and this is a 'hostmaster' and 'postmaster' speaking - 
> >that if you cannot reverse look up one of your own hosts, you *might* 
> >have a DNS config issue...  That should be solved...
> 
> Think "travelling user." You'll never have control over everything.

Daniel:

Even a traveling user *should* reverse lookup - even if they come from
some unrelated ISP.  Or am I not understanding the"Travelling user"
issue?

If some hotel room doesn't reverse lookup then THEY have a DNS issue
that should be corrected.  I've made it a practice to tell these people
what problems I, and my users, are experiencing.  Sometimes, they fix
the issue, sometimes not.  Major chains seem to react well though.

Regards,
Gregory Hicks

[...snip...]

---------------------------------------------------------------------
Gregory Hicks                           | Principal Systems Engineer
Cadence Design Systems                  | Direct:   408.576.3609
555 River Oaks Pkwy M/S 6B1             | Fax:      408.894.3479
San Jose, CA 95134                      | Internet: ghicks at cadence dot com

"There is no limit to what a man can do or how far he can go if he
doesn't mind who gets the credit." - Robert Woodruff

"The trouble with doing anything right the first time is that nobody
appreciates how difficult it was."

When a team of dedicated individuals makes a commitment to act as
one...  the sky's the limit.


From: "Fred G Guiliano" <fred.guiliano at DPS-inc dot com>
Subject: RE: Canonical name error...
Date: Wed, 2 Jan 2002 13:03:24 -0500

Thanks! I could always create the reverse lookup zone but unfortunatly I
can't (too long of a story). I did, however,find the switch that would do
this (-R) which does work.

-----Original Message-----
From: Daniel Senie [mailto:dts at senie dot com]
Sent: Wednesday, January 02, 2002 12:52 PM
To: Gregory Hicks; qpopper at lists.pensive.org; fred dot guiliano at dps-inc dot com
Subject: Re: Canonical name error...


At 12:33 PM 1/2/02, Gregory Hicks wrote:
> > From: "Fred G Guiliano" <fred.guiliano at dps-inc dot com>
> > Date: Wed, 2 Jan 2002 11:03:03 -0500
> >
> > Hi All,
> >
> > I see at the the qualcomm site this particular explanation of the
> > following error:
> >
> > What does the error message Unable to get canonical name of client
> > mean?
>[...snip...]
> >   Qpopper allows you to turn off this check if you like.  Qpopper also
> > includes the IP address of the client in the log message, to make it
> > easier to check its DNS entries.
> >
> > In this explanation, it says that Qpopper allows you to turn off this
> > check.  That's nice, but my question is HOW do you turn this check
> > off? I didn't see anything in the manual on how to do this (It maybe
> > there but I guess I'm not seeing it). Any help would be appreciated!
>
>Although qpopper may allow you to turn the check off (and no, I don't
>know the option, but it is in the docs or the config options), it seems
>to me - and this is a 'hostmaster' and 'postmaster' speaking - that if
>you cannot reverse look up one of your own hosts, you *might* have a DNS
>config issue...  That should be solved...

Think "travelling user." You'll never have control over everything.

On every version of qpopper I install, I comment out the code that sprays
the error message about DNS lookup. I don't want to turn the option off,
though, since if the name DOES resolve, I want to find out what it is and
log it.

Next time I get a chance to do some work on the qpopper code (will probably
be soon) I think I'll take a look at this. Simply removing this message
entirely may be the best answer.
-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: Wed, 02 Jan 2002 14:17:57 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: Canonical name error...

At 01:09 PM 1/2/02, Gregory Hicks wrote:

> > Date: Wed, 02 Jan 2002 12:51:42 -0500
> > From: Daniel Senie <dts at senie dot com>
> >
> > At 12:33 PM 1/2/02, Gregory Hicks wrote:
> > > > From: "Fred G Guiliano" <fred.guiliano at dps-inc dot com>
> > > > Date: Wed, 2 Jan 2002 11:03:03 -0500
> > > >
> > > > Hi All,
> > > >
> > > > I see at the the qualcomm site this particular explanation of the
> > > > following error:
>[...snip...]
>
> > >Although qpopper may allow you to turn the check off (and no, I don't
> > >know the option, but it is in the docs or the config options), it
> > >seems to me - and this is a 'hostmaster' and 'postmaster' speaking -
> > >that if you cannot reverse look up one of your own hosts, you *might*
> > >have a DNS config issue...  That should be solved...
> >
> > Think "travelling user." You'll never have control over everything.
>
>Daniel:
>
>Even a traveling user *should* reverse lookup - even if they come from
>some unrelated ISP.  Or am I not understanding the"Travelling user"
>issue?

Yes, everyone SHOULD have INADDR. However, not everyone does.


>If some hotel room doesn't reverse lookup then THEY have a DNS issue
>that should be corrected.  I've made it a practice to tell these people
>what problems I, and my users, are experiencing.  Sometimes, they fix
>the issue, sometimes not.  Major chains seem to react well though.

While that's all fine, the question is whether there's any merit in filling 
logs on the mail server with information about 3rd parties who lack INADDR. 
My argument is that it's not worthwhile.

What's worse, is that many people run qpopper from inetd or xinetd, often 
with TCP Wrappers. Wrappers does a DNS lookup for its checks (or at least 
can be set up that way).

I have an Internet Draft which I've been working on, intended to be a BCP, 
that says everyone SHOULD implement INADDR, but at the same time it says 
use of INADDR as a part of any "security" check should be strongly discouraged.
-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Date: Wed, 2 Jan 2002 09:48:05 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: Canonical name error...

On Wed, Jan 02, 2002 at 10:09:44AM -0800, Gregory Hicks wrote:
> Even a traveling user *should* reverse lookup - even if they come from
> some unrelated ISP.  Or am I not understanding the"Travelling user"
> issue?
> 
> If some hotel room doesn't reverse lookup then THEY have a DNS issue
> that should be corrected.  I've made it a practice to tell these people
> what problems I, and my users, are experiencing.  Sometimes, they fix
> the issue, sometimes not.  Major chains seem to react well though.

  Nonetheless, this puts you in the position of "fixing the Internet". 
It's a rather large job.  While there's certainly nothing wrong with
letting people know what they need to fix, there's a lot of merit in
making your server deal well with the Internet as it exists now. 
(Which includes a whole lot of hosts with bad reverse-DNS.)

  -- Clifton

-- 
 Clifton Royston  --  LavaNet Systems Architect --  cliftonr at lava dot net
   WWJD?   "JWRTFM!" - Scott Dorsey (kludge)   "JWG" - Eddie Aikau

Date: Wed, 2 Jan 2002 09:51:04 -1000
From: Clifton Royston <cliftonr at lava dot net>
Subject: Re: Canonical name error...

On Wed, Jan 02, 2002 at 02:17:57PM -0500, Daniel Senie wrote:
> I have an Internet Draft which I've been working on, intended to be a BCP, 
> that says everyone SHOULD implement INADDR, but at the same time it says 
> use of INADDR as a part of any "security" check should be strongly discouraged.

I agree.  

One of the dangers of relying on INADDR, besides it becoming a nearly
useless check with the Internet in its present state, is that if it
*replaces* the raw IP address in logging, it allows an actual malicious
attacker to trivially conceal their point of attack.  I'm surprised how
many applications not only expect INADDR to be there but blindly rely
on it if it is.
  -- Clifton

-- 
 Clifton Royston  --  LavaNet Systems Architect --  cliftonr at lava dot net
   WWJD?   "JWRTFM!" - Scott Dorsey (kludge)   "JWG" - Eddie Aikau

Date: Wed, 2 Jan 2002 11:55:38 -0800
From: Randall Gellens <randy at qualcomm dot com>
Subject: Re: Emails are retrieved twice (sometimes 3 times) from

--============_-1202121079==_============
Content-Type: text/plain; charset="us-ascii" ; format="flowed"

At 9:09 PM -0800 12/30/01, Habib Abassi wrote:

>  Hi All,
>  Emails are retrieved twice (sometimes 3 times) from Qpopper vesion 3.1.2.
>  I am using a netscape library to connect and retrieve emails. After 
> retrieving email I mark it as deleted, it should be deleted after 
> quit, but when I connect again the same email still in mailbox and 
> I can retrieved again .... If any body kn ow why this is happen and 
> if there's a way to know that the email is marked as deleted before 
> retrieving it.
>  Thanks in advance

Try getting a trace of the session and see what is going on.

--============_-1202121079==_============
Content-Id: p05100300b85915700f22 at [192.168.1.5].0 dot 0
Content-Type: multipart/appledouble; boundary="============_-1202121079==_D============"

--============_-1202121079==_D===========
Content-Transfer-Encoding: base64
Content-Type: application/applefile; name="%qpopper_tracing"
Content-Disposition: attachment; filename="%qpopper_tracing"
 ; modification-date="Fri, 5 Oct 2001 10:27:42 -0800"

AAUWBwACAAAAAAAAAAAAAAAAAAAAAAAAAAMAAAADAAAAPgAAAA8AAAAJAAAATQAAACAA
AAAIAAAAbQAAABBxcG9wcGVyIHRyYWNpbmdFdVNuQ1NPbQEAAbkAAQAAAAAAAAAAAAAA
AAAAAAAAAAC9nLkDUEeeA7wy/wPFsgo
--============_-1202121079==_D===========
Content-Type: application/octet-stream; name="qpopper_tracing"
 ; x-mac-type="4575536E"
 ; x-mac-creator="43534F6D"
Content-Disposition: attachment; filename="qpopper_tracing"
Content-Transfer-Encoding: base64
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--============_-1202121079==_D============--
--============_-1202121079==_============--

Date: Wed, 02 Jan 2002 15:07:14 -0500
From: Daniel Senie <dts at senie dot com>
Subject: Re: Canonical name error...

At 02:51 PM 1/2/02, Clifton Royston wrote:
>On Wed, Jan 02, 2002 at 02:17:57PM -0500, Daniel Senie wrote:
> > I have an Internet Draft which I've been working on, intended to be a BCP,
> > that says everyone SHOULD implement INADDR, but at the same time it says
> > use of INADDR as a part of any "security" check should be strongly 
> discouraged.
>
>I agree.
>
>One of the dangers of relying on INADDR, besides it becoming a nearly
>useless check with the Internet in its present state, is that if it
>*replaces* the raw IP address in logging, it allows an actual malicious
>attacker to trivially conceal their point of attack.  I'm surprised how
>many applications not only expect INADDR to be there but blindly rely
>on it if it is.

Since qpopper ALWAYS logs both the IP address AND the INADDR (if present), 
I contend the logging message complaining about the INADDR failure is 
superfluous. It serves only to annoy the administrator and waste log space. 
In cases where the INADDR failed, the log actually shows that without the 
extra warning message. The IP address is logged twice on the line. So, 
anyone who wants to check on which hosts are not doing INADDR can easily 
find this information, without the extra logging line.
-----------------------------------------------------------------
Daniel Senie                                        dts at senie dot com
Amaranth Networks Inc.                    http://www.amaranth.com


Last updated on 2 Jan 2002 by Pensive Mailing List Admin