If using MercuryE for SMTP delivery Mercury will rely on the routing settings in Windows for reaching external destinations. Any firewall or proxy server will need to be set to allow traffic through port 25. Additionally, the Internet provider must not be blocking port 25. If port 25 for some reason can't be used the alternative is to use MercuryC and relay outgoing messages through the Internet provider's SMTP server.

/Rolf 

Thanks Paul and Rolf and sorry for the delay to reply, too much work these days.

[quote user="PaulW"]

But how can you avoid creating backscatter - sending a reply to a 'spoofed' address?

[/quote]

Sure, but the alternative is too have wrong addresses mails not sent back to the senders, then the sender may think that his/her mail was received.

I tried to link a rule to the 'misc' mailbox but it is not seen as a mailbox linked to a user by Mercury, it 's just seen as a garbage folder.

I'm trying to write a program  that would process the .cnm files recorded in this folder and would send back a bounce message to most of them (but not to the bounce messages senders, ie mailer-daemon) and then would delete those files, the program would run each hour as a scheduled task.

Regards

[quote user="Konrad Hammerer"]

At the moment, I'm just thinking about a secured connection, not a secured content.[/quote]

I've always wondered what the purpose of that is for server to server connections?

I'm not using that part of list handling either but two things to check could be that the unsubscribe message was sent as plain text (not HTML) and that the sender was in fact registered as a member of that list.

/Rolf 

Well, something in the message didn't agree with Outlook, Mercury, Exchange or some antivirus software that may have been someplace in between, but it's unfortunately rather impossible to say more than that without examining the message itself.

/Rolf 

The idea behind such messages is to trick a reputable mail server to re-send a spam message (as part of the non-delivery notification) to the real recipient, placed in the reply-to header of the message. This rather common phenomena is known as backscatter.

/Rolf 

I guess that this has the same effect as the debug switch in the clamd config file and I already tried that. Did not result in any more useful information...

Konrad

Failed to mention that my first rule is one that stops filter processing if the message originates from our local domain.  I don't want to disable it in the production environment so need to figure out another option.  I haven't taken time to look at it since my initial test.  It may be a mute point since I haven't received one of these crasher messages in several days.  I'm hoping so.

I'am agree with Phil..you can use Pegasus Mail to search very easy [*-)]

Many thanks!

This must be set within gmail.
Please read this:
"Sending mail from a different address": http://support.google.com/mail/bin/answer.py?hl=en&answer=22370

It seems to me that the Envelope-to header is expected to contain the real recipient of the message as defined in the SMTP envelope from the original sender. If the host that first receives the messages puts some other address there you will have to talk to whoever manages that host to have them either put the proper address in that header or add one more header with it.

/Rolf 

No apology required, I was, after all, being rude. IMO, not directly to the OP, but it could be construed as such.

It is a pet peeve of mine when people use meaningless jargon to describe things they clearly don't understand, presumably to give the appearance that they do, to others that don't.

This phenomenon is exacerbated by the proliferation of "user-friendly" systems that remove the user from the actual process so much that there is no way to know how it really works and things just happen by magic, effectively reducing the OP's question to "Can Mercury cast this spell on my mail, like google & exchange can?"

Oops, i think I've done it again.

Maybe I'm just getting old and cranky ... :D

Thanks a lot, folks 

I will leave everything as it is.

Connection being refused due to incorrect credentials?

Check the log file for MercuryI to see if it shows any entries that will provide a clue (e.g: password failure entries).  Might also want to enable session logging for a more detailed log.

[quote user="Greenman"]Regarding the spam issue - that is exactly why I do not want all the mail to be downloaded. These people will ultimately need to manage the Mercury installation. It will be far easier for them to review the spam folder on the web site and forward false positives than it would be to manage all the aspects necessary to effectively filter out spam, and then still have to review the spam account for false positives.[/quote]

IMHO, leaving it on the server is only an option if detection of false positives is not a concern.  Bringing it down, filtering it to a SPAM user and adding the SPAM user mailbox as an added mailbox to someones folder list avoids the 'out of sight, out of mind' syndrome.  If you do bring it down, add filters to auto remove all the vulgar ones, nobody needs to see that crap!  I would be happy to share my filters with you if you decide to go that route.

Hi Rolf,

[:S] A shame but thanks for the information.

I'll have to find a way round it since Mercury is good for what I want to do.

Maybe a longer username is something for the future !!

I am running the 64 bit version of Win7.  I have an XP system that I can use for a test platform.  I'll try that.

I don't see any options for encryption key length when generating self-signed certs within Mercury32.  The service I see running on Win7 is called "Cryptographic Services" and I can't tell if that's using the crypt dlls or not.

 I'll post back after I test on XP.  I appreciate all the help.

We had a problem with delivery issues where emails were being refused because we did not have reverse DNS set up when we first began managing our email. I manage the DNS for several domains and have never been charged for setting up RDNS records. I'd look for a different ISP.

Found the problem.  My test message was copied to self for use on subsequent tests.  I looked fine on resends but when I opened it I could see that there was a tab after "password" in the password line.  That tab looks like a space in the new message window.