auth_cvm_unix_local: AUTH module for Bruce Guenther's CVM

Comments/suggestions on the attached AUTH module? [...] =head1 NAME auth_cvm_unix_local - SMTP AUTH LOGIN module using Bruce Guenther's Credential Validation Module (CVM) http://untroubled.org/cvm/ [...] Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Go

Re: Great news from YAPC

John Peacock wrote: Matt Sergeant wrote: A birdy tells me that qpsmtpd will be the default smtp server in the next release of the e-smith SME server. That birdy is correct, or you could have asked me instead :-) Apparently that's 10K new users (not sure whether that was end-users or actua

Re: Great news from YAPC

John Peacock wrote: Gordon Rowell wrote: 7.0alpha is using 0.29, and rolling 0.30 would be great. That's two, or is it three, votes for rolling 0.30 now. How many people are actually running the current trunk? I'm happy to roll trunk into the next 7.0alpha, but I'd only

0.30rc1 (was Re: Great news from YAPC)

Ask Bjørn Hansen wrote: [...] I made a tag that I'll copy to 0.30 then unless something comes up: http://svn.perl.org/qpsmtpd/tags/0.30rc1 Hmm - the MANIFEST doesn't list my new plugins (e.g. plugins/auth/cvm_unix_local) - I don't know if anything else was missed in that tag. We also

Re: 0.30rc1 (was Re: Great news from YAPC)

Ask Bjørn Hansen wrote: [...] Oops. http://svn.perl.org/qpsmtpd/tags/0.30rc2/ And initial RPMs can be found here: http://www.gormand.com/smeserver/WIP/ Thanks, Gordon

PATCH: Don't reveal version in SMTP greeting

Although useful for debugging, I don't think we should disclose the version number. I'd also be happy (or is that happier?) with: 220 $me ESMTP ready Gordon --- lib/Qpsmtpd/SMTP.pm.orig2005-07-02 17:00:52.0 +1000 +++ lib/Qpsmtpd/SMTP.pm 2005-07-02 17:03:30.0 +1000 @@ -104,

PATCH: Sanitise reply from check_spamhelo

Not totally appropriate for a commercial mail server, IMO. Gordon --- plugins/check_spamhelo.orig 2005-07-02 17:07:43.0 +1000 +++ plugins/check_spamhelo 2005-07-02 17:08:56.0 +1000 @@ -29,7 +29,7 @@ for my $bad ($self->qp->config('badhelo')) { if ($host eq lc $bad) {

PATCH: Check relayclient in check_badrcptto

--- plugins/check_badrcptto.orig2005-07-02 17:13:11.0 +1000 +++ plugins/check_badrcptto 2005-07-02 17:13:18.0 +1000 @@ -7,6 +7,9 @@ sub check_for_badrcptto { my ($self, $transaction, $recipient) = @_; + + return (DECLINED) if $self->qp->connection->relay_client();

PATCH: Check relayclient in rhsbl

--- plugins/rhsbl.orig 2005-07-02 17:15:11.0 +1000 +++ plugins/rhsbl 2005-07-02 17:15:29.0 +1000 @@ -9,6 +9,8 @@ sub mail_handler { my ($self, $transaction, $sender) = @_; + return (DECLINED) if $self->qp->connection->relay_client(); + my $res = new Net::DNS::Resolv

PATCH: Check relayclient in dnsbl

--- plugins/dnsbl.orig 2005-07-02 12:17:32.0 +1000 +++ plugins/dnsbl 2005-07-02 17:17:00.0 +1000 @@ -17,6 +17,8 @@ sub connect_handler { my ($self, $transaction) = @_; + return (DECLINED) if $self->qp->connection->relay_client(); + my $remote_ip = $self->qp->connect

PATCH: config.sample for badrcptto_patterns

--- config.sample/badrcptto_patterns.orig 1970-01-01 10:00:00.0 +1000 +++ config.sample/badrcptto_patterns2005-07-02 17:24:07.0 +1000 @@ -0,0 +1,5 @@ +# Format is pattern\s+Response +# Don't forget to anchor the pattern if required +! Sorry, bang paths not accepted

Re: PATCH: Don't reveal version in SMTP greeting

Ask Bjørn Hansen wrote: [...] Why? If it's for security, will it really make a difference? Does it give any information out that an attacker can use? If there ever is a security problem in qpsmtpd (unlikely, but I suppose possible), wouldn't the attacker just hit SMTP servers at random fo

Re: PATCH: Don't reveal version in SMTP greeting

Andrew Pam wrote: On Sat, Jul 02, 2005 at 08:25:08PM +1000, Gordon Rowell wrote: Sure. I don't think it's a big issue. I'm not a believer in security through obscurity, but I also don't believe in giving such things away for free. It can really help debugging in the fi

Re: PATCH: Check relayclient in ...

John Peacock wrote: [...] These three I'm not going to commit (without discussion), because I don't like the code duplication involved. Rather than patching every plugin to respect the check_relay() setting (and yes, I know, I have done this recently), and remember to add that code to any *ne

Re: PATCH: Don't reveal version in SMTP greeting

Robert Spier wrote: [...] How about a compromise? Use the return $msg from the connect hook (if any)? How about the contents of config('smtpgreeting'), as per qmail-smtpd? [...] smtpgreeting SMTP greeting message. Default: me, if that is supplied; otherwise qmail-smt

Re: PATCH: config and plugin directories

Peter J. Holzer wrote: [...] This one makes the config directory and the plugin directory configurable. The config dir is taken from an env. variable QPSMTPD_CONFIG, if it exists. [...] Yes, please! For the SMEServer, I install qpsmtpd in /usr/lib/qpsmtpd/ and build symlink trees from /var/s

Re: PATCH: pid file and aux groups

Peter J. Holzer wrote: This patch adds a pid for forkserver (this was previously posted by someone else to the mailinglist) > [...] PID files are evil, unreliable things which are often stale and introduce race conditions. You don't need one under supervise, and IMO, we're better off without

Re: PATCH: pid file and aux groups

n the state of a process. FTR - in SMEServer land, we've moved from supervise to the runit workalike: http://smarden.org/runit/ as the latter doesn't have djb's license restrictions, and has some nice extra features, such as a finish script. Well worth a look, IMO. Thanks,

Re: PATCH: Don't reveal version in SMTP greeting

Robert Spier wrote: How about the contents of config('smtpgreeting'), as per qmail-smtpd? [...] smtpgreeting SMTP greeting message. Default: me, if that is supplied; otherwise qmail-smtpd will refuse to run. The first word of smtpgreeting should be the curr

Re: IRC Channel

Matt Sergeant wrote: I've created an IRC channel for talking about qpsmtpd - #qpsmtpd on irc.perl.org But please keep the questions on the list...some of us sleep when the good stuff is happening on IRC channels. Thanks, Gordon

Re: Great news from YAPC

Charlie Brady wrote: On Fri, 1 Jul 2005, Gordon Rowell wrote: John Peacock wrote: Matt Sergeant wrote: A birdy tells me that qpsmtpd will be the default smtp server in the next release of the e-smith SME server. That birdy is correct, or you could have asked me instead :-) And why

PATCH: Hide version, if config('smtpgreeting') exists

If config('smtpgreeting') exists: - Display it in the connect response - Don't display the version in the 'help' response Thanks, Gordon --- SMTP.pm.orig 2005-07-07 14:25:53.0 +1000 +++ SMTP.pm 2005-07-07 15:03:57.0 +1000 @@ -104,8 +104,9 @@ return $rc; } elsif (

PATCH: Direct bugs to the list, not to ask

Or should we have a bug reporting address/tracker? Thanks, Gordon --- SMTP.pm.orig 2005-07-07 15:06:28.0 +1000 +++ SMTP.pm 2005-07-07 15:06:59.0 +1000 @@ -336,7 +336,7 @@ $self->respond(214, "This is qpsmtpd " . $self->version, "See http://smtpd.develoop

Re: RC scripts

$0 {start|stop|restart|status|sigalrm|sigcont|sighup|sigint|sigkill|sigstop|sigterm|sigusr1|sigusr2|svdisable}" ;; esac #!/bin/sh #-- # copyright (C) 1999-2005 Mitel Networks Corporation # Copyright (C) 2005 Gordon Rowell &l

Re: RC scripts

Matt Sergeant wrote: On 7 Jul 2005, at 18:03, Gordon Rowell wrote: We (SMEServer) have single RC script called "daemontools" and we symlink all of our daemontools/runit processes to that. All daemon specific startup is in the run scripts. Anyway, attached. Hmm, I'm th

Re: About smtpgreeting

John Peacock wrote: [...] That's easy enough to add. However, I've attached a slight rewrite to Guillaume's patch so that now it handles multiline config files properly and has all of the default values in a hash (which makes it very easy to eventually abstract out completely as a language fi

Re: 0.31 release candidate 2

Peter J. Holzer wrote: [...] [0] At the usual place: resp. . I can rebuild them for other Redhat/Fedora releases if somebody is interested. And if anyone is using my qpsmtpd RPMs (which

[OT] supervise/runit (was Re: Out of memory! ?)

supervise (and more), without djb's licensing issues. IMO supervise/runit is clean and simple and it works. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 St Pauls NSW 2031 Australia "The test of our progress is not whether we

Re: [OT] supervise/runit (was Re: Out of memory! ?)

Matthew Harrell wrote: [...] I don't have too much of a personal bias against it - it's mostly that a lot of these systems we were dealing with were controlled and managed by end users who had a minimal familiarity with Linux systems. init.d scripts like that are commonplace, relatively obvious,

0.31rc2 : Mails smaller than 10k are not scanned

This one took a little digging. I upgraded to 0.31 (from 0.29) in SME Server 7.0beta4, and we've had bugs raised about ClamAV and SpamAssassin not working. They are, just not for small mails... If messages are smaller than 10k, the scanners don't get to see them: https://sourceforge.net/tracke

Re: 0.31rc2 : Mails smaller than 10k are not scanned

thanks for the complete fixes. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 St Pauls NSW 2031 Australia "The test of our progress is not whether we add more to the abundance of those who have much; it is whether we provide enough

Re: Newest version, "any day now"

John Peacock wrote: > [...] FWIW, I've been running branches/0.31 for months now, with hardly any lost mail [that wasn't my own damn fault]. ;-) > [...] Umm - if (hardly any != no), I'm concerned. Putting aside things which are of your own doing, do you know of failure cases in the 0.31 bran

PATCH: check_badrcptto should not apply to relay clients

As done in check_goodrcptto Gordon [EMAIL PROTECTED] tmp]$ diff -u check_badrcptto.orig check_badrcptto --- check_badrcptto.orig2005-11-04 11:31:43.350962930 +1100 +++ check_badrcptto 2005-11-04 11:32:19.161737046 +1100 @@ -2,6 +2,7 @@ sub hook_rcpt { my ($self, $transaction, $

Re: PATCH: check_badrcptto should not apply to relay clients

Ask Bjørn Hansen wrote: On Nov 3, 2005, at 16:34 , Gordon Rowell wrote: As done in check_goodrcptto I'm not sure I understand the use case? We (SME Server) generate a badrcptto config file which blocks mail to auto-generated generic addresses like [EMAIL PROTECTED] from external us

Re: check_delivery plugin

tes. They can move to a proper location when approved by our code police :-) Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 St Pauls NSW 2031 Australia "The test of our progress is not whether we add more to the abundance of th

Re: check_delivery plugin

John Peacock wrote: [...] 3) Keeping all files related to a single plugin in a dedicated directory would make it easier to have a layout like this: unsupported/ check_delivery/ README - describe limitations and TODO's for this plugin plugins/ check_delivery - actual p

contrib/unsupported/unstable tree (was Re: check_delivery plugin)

Ask Bjørn Hansen wrote: [...] 1) It does seem a little "heavy" to me too. 2) If it's a good idea, why isn't it a good idea for the regular plugins? (or: if it's not a good idea for the regular plugins, why is it for the unsupported plugins?) John's proposal gives small sandboxes to each d

Re: Thank You for qpsmtpd ... and a deployment note....

Andrew W. Donoho wrote: > [...] On an FC4 system, there is always tension between the one, true RPM path and getting things done. In my case, because I had already made the commitment to djb's dnscache, utilizing your standard execution model via daemon tools was a key parameter in choos

Re: check_delivery and dot-qmail-default

plugins provide most, though admittedly not all, of the features of check_delivery. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 St Pauls NSW 2031 Australia "The test of our progress is not whether we add more to the abundance of

Re: check_delivery and dot-qmail-default

Peter J. Holzer wrote: > [...] I think validating recipients at the MX is something like a minimum requirement for anyone running a mail server. Accepting an email and then sending a bounce is very rude. Agreed. I meant, but didn't say, "There are other ways to reduce/remove backscatter than c

Re: Is this too hardball?

Bob Dodds wrote: [...] By "address literal" is OK, they mean IP, right? > [...] See section 4.1.3 and 4.1.2 - address literals must be enclosed in (square) brackets: [...Section 4.1.2...] address-literal = "[" IPv4-address-literal / IPv6-address-literal

Re: Is this too hardball?

John Peacock wrote: Bob Dodds wrote: http://perlq.org/ yellow band shows that require_resolvable_fromhost is a productive filter. Except that is a different test (based on the same section of the RFC, however). I've seen badly configured servers (*cough*Notes and Exchange*cough*) which used

Re: Where do plugins look for config files?

Robin Bowes wrote: Hi, I'm writing a check_validrcpt plugin that uses a cdb file so I need to specify where that cdb file is. Have a look at check_goodrcptto which I mentioned on the list a few days back. http://www.nntp.perl.org/group/perl.qpsmtpd/4114 It should be fairly simple to allow

Re: maildb plugin

Ulrich Stärk wrote: Hi list, I just wrote a plugin to fulfil a customer's wish to archive every incoming and outgoing mail. The archive consists of two parts: - local file system store where the mail gets stored as is, with full headers and attachments - database that stores metadata for each m

Re: maildb plugin

Peter J. Holzer wrote: [...] Message header field names are not case-sensitive. Actually, from my reading of it, Mail::Header::_tag_case() should ensure that the header is handled as Message-ID regardless of the actual case(s) used in the mail. Gordon

Re: New plugin: denybounce

Johan Almqvist wrote: [...] I'll think about it once more, I think I should probably turn the logic around both in check_badrcptto_patterns and in my plugin, so that check_badrcptto_patterns becomes check_goodrcptto_patterns (a list of patterns that describes all legit addresses) I built

Re: New plugin: denybounce

Johan Almqvist wrote: On Jan 23, 2006, at 10:48, Gordon Rowell wrote: check_goodrcptto extn - I don't have this, it wasn't in the distro. I'll submit it on Gavin Carr's behalf unless he's listening. Gavin? I also provided a link to the plugin set we use in

Submitting plugins (was Re: New plugin: denybounce)

7;s what I put in my plugins: =head1 AUTHOR Copyright 2005 Gordon Rowell <[EMAIL PROTECTED]> This software is free software and may be distributed under the same terms as Perl itself. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 S

Re: Submitting plugins (was Re: New plugin: denybounce)

Ask Bjørn Hansen wrote: On Jan 24, 2006, at 1:08 PM, Gordon Rowell wrote: - License statement - either as per qpsmtpd or as per Perl or similar open license No, it really should be MIT licensed ("as per qpsmtpd") to go in the distribution. There are a few exceptions (only yo

[OT] Licensing, license texts (was Re: Submitting plugins (was Re: New plugin: denybounce))

Ask Bjørn Hansen wrote: [...] Even worse if one piece of software uses multiple completely different licenses. ;-) Sure. As a distro maintainer you should appreciate software licensed with the MIT/X11 license... Ah, but which version of that beastie? :-( Life would be made quite a bit

Re: tmp dir and logs

Sydney Bogaert wrote: [...] Is this behaviour portable ? It certainly is across all *nix variants - it's a fairly standard way to ensure cleanup of temporary files. The downside is that it makes it harder to leave the temporary files around if you need to (e.g. for later diagnosis of errors

Re: More than enough already (Re: Mail submisstion - second instance of qpsmtpd vs stunnel)

Les Mikesell wrote: [...] Sorry, most lists aren't so hostile to the mention of things invented elsewhere. > [...] Most lists are hostile to having developer time wasted: http://lists.contribs.org/mailman/public/devinfo/msg08129.html This list is no exception, and neither is the SME Serv

Contribs tree/contributed plugins (was Re: [announce] qpsmtpd 0.32)

ooks great), but we end up with plugins all over the net which never make it into the main tree. And those links will go stale as sites go away. Has the idea of a contribs tree been formally abandoned, or has it just not happened? Thanks, Gordon -- Gordon Rowell [EMAIL PROTEC

Re: qpsmtpd contribs

ugins directory. The goal here is to ensure that we can collect plugins in SVN with minimal effort on the part of those who have SVN write access. The contribs tree could be maintained by someone other than the current trunk maintainers. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http:

Re: [perl #38747] RCPT TO addresses with spaces are not correctly parsed

erse-path = Path Forward-path = Path Path = "<" [ A-d-l ":" ] Mailbox ">" And RFC821: ::= ::= ::= "<" [ ":" ] ">" We (SME Server) have been running with Peter Holtzer's RPMs which

Re: [perl #38747] RCPT TO addresses with spaces are not correctly parsed

sends with angles. I have yet to see a legitimate mail rejected. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 St Pauls NSW 2031 Australia "The test of our progress is not whether we add more to the abundance of those who have muc

Re: [perl #38747] RCPT TO addresses with spaces are not correctly parsed

unambiguous. Thanks, Gordon -- Gordon Rowell [EMAIL PROTECTED] http://www.gormand.com.au Gormand Pty Ltd PO Box 239 St Pauls NSW 2031 Australia "The test of our progress is not whether we add more to the abundance of those who have much; it is whether we provide enough for those who

Re: Wiki update

x27;t call itself qmail. The fact that only he can improve qmail is a significant issue. One of the (many) things which interests me about qpsmtpd is the Postfix backend and one of my longer-term goals for the SME Server is to migrate from qmail to Postfix. We have already moved from daemontools

Re: SysV-style init.d/rc-Scripts

John Wang wrote: [...] I think it would be a useful addition since I imagine most people want it to be autostarted and init.d tends to be more popular than daemontools. Sure, but it's not an either/or choice. You can have SysV startup and process supervision - that's how we do it in the SME Se

[OT] Comments in RT and RT for bug tracking (was Re: [perl #38806] [RESOLVED] Inadequate validation of authenticationdata)

John Peacock wrote: I would have thought that you would have gotten the previous message from RT (which I thought was the correct "Requester gets this" and not the, to my mind, mostly useless "just add some random note to RT that no one will ever see without hunting it down"). Sorry... It may

Re: google calendar email rejected

Charlie Brady wrote: On Tue, 18 Apr 2006, Michael Holzt wrote: Now one can discuss if a date header is required, but at least there is very likely no error in the filtering code. What is there to discuss? Nothing, in my mind. RFC2822 and RFC822 are both quite clear: http://www.contribs.o

Re: Case significance

Andrew Pam wrote: Apologies if this has been discussed before. I have a number of clients with mailboxes on my server who are complaining that they are losing important incoming emails. On further investigation, this turns out to be caused by the senders incorrectly capitalising the username. I

Re: Case significance

Andrew Pam wrote: > [...] qpsmtpd is bouncing it due to the lack of a local mailbox. I have a plugin that checks for the existence of local mailboxes, but even if I made this case-insignificant and accepted the mail, it would promptly be rejected by qmail-local. I think there's some other issu

Re: Case significance

Andrew Pam wrote: On Tue, Apr 25, 2006 at 10:37:37AM +1000, Gordon Rowell wrote: [...] Most MDAs are case-insensitive, but gateways to other mail systems cannot assume that downstream mail systems are. I know that, but in this case the mail is for local delivery. Sorry - I chopped a

Re: Case significance

Andrew Pam wrote: > [...] qpsmtpd is bouncing it due to the lack of a local mailbox. I have a plugin that checks for the existence of local mailboxes BTW: I use Gavin Carr's check_goodrcptto for this task, though you do need to generate the list). When a decision is made about the plugin re

Re: Case significance

Andrew Pam wrote: [...] Are you sure you understand my requirement? I believe I do, but maybe I am missing something. I want all email, regardless of case, to be delivered to a (lowercase) local mailbox. Yes, and that happens for me. A bunch of plugins validate the mail and mail which pas

Re: qpmtpd plugins how can I alter the sender address ?

Pascal SCHIRRMANN wrote: [...] One automated system on my computer send mail in the form [EMAIL PROTECTED] name}.{domain name} Unfortunately, my system name isn't known in public DNS (and this is not a good solution for me anyway). Hi Pascal, You could probably do it in qpsmtpd, but in my o

[OT] avoid fetchmail/multidrop (was Re: qpmtpd plugins how can I alter the sender address ?)

Pascal Schirrmann wrote: I totally agree with you two, and I just have a try on the fetchmail-devel list. But, even if this change is accepted, it could be a long run before a new version of fetchmail is included in SME 7. And in between, I'm searching a solution, because failure message of

Re: [PATCH] High-performance qpsmtpd daemon

Lars Roland wrote: Removing the daemonize stuff completely is going a bit to far (at least from my point of view), it should never be required (or desirable due to its crappy license) to have supervise installed FYI - we've been here before on this list - you don't need to use supervise: http:

Re: qpsmtpd-forkserver, runit, Mac OS X

. This is what we use in the SME Server config (which should use chpst instead of softlimit): #!/bin/sh #-- # copyright (C) 1999-2005 Mitel Networks Corporation # Copyright (C) 2005 Gordon Rowell <[EMAIL PROTECTED]> # #

Re: Account testing in Outlook fails with check_basicheaders

Michael Holzt wrote: > [...] > Redmond called Outlook does not set a Date: Header in the mail which is > generated for testing. Therefore check_basicheaders denied the mail which > in turn led to the error message described. > Yep - been there: http://www.contribs.org/bugzilla/show_bug.cgi?id

qpsmtpd based "distro" (was Re: Per-user config; spamassassin; performance)

Skaag Argonius wrote: > [...] I'm going to start working on this. If anyone cares to join, you'r welcome ;-) Have you looked at the SME Server - www.smeserver.org / www.contribs.org It doesn't have everything on your list, but it has quite a few and is an excellent base for the rest. Thanks

ANNOUNCE: plugin check_smtp_forward

I've written a new plugin which (mostly) replaces the smtp-forward plugin. See this bug for background: http://www.contribs.org/bugzilla/show_bug.cgi?id=1850 I looked at enhancing the smtp-forward plugin to handle multiple internal mail servers, but it got ugly very quickly. Rather than writ

Re: domain literals (Re: require_resolvable_fromhost and numeric MXes in 0.32)

Charlie Brady wrote: [...] If you wish to accept mail for domain literals you will need to configure them in /var/qmail/control/rcpthosts and/or locals. I'd hope that qpsmtpd will do the right thing if you do. And in SME Server land, we only accept [EMAIL PROTECTED] (courtesy of Gavin Carr'

Re: Temporary comms breakdown between qpsmtpd and spamd

Charlie Brady wrote: [...] It does mean that qpsmtpd's plugin is adding useless X-Spam-Status headers rather than detecting and reporting that spamd hadn't reported back correctly. Worse than useless, IMO, as it's an explicit 'No' Gordon -- Gordon Rowell

Re: Temporary comms breakdown between qpsmtpd and spamd

. Thanks, Gordon -- Gordon Rowell Gormand Pty Ltdhttp://www.gormand.com.au/ SME Server development and support http://www.smeserver.com.au/ Perl development, systems and network consulting

Re: check_basicheaders question

Matthew Harrell wrote: I use this plugin because it does seem to stop a nominal amount of suspicious mail. But I see it also catches mail from some companies trying to send out autogenerated messages. When I look through the plugin I don't see anything about whilisted hosts so I was thinking a