OK.

Time to move... 

Thanks.

I have a huge RULES.MER file.

And it will increase in size because I use it to eliminate SPAM and SPAM is increasing and increasing and increasing and ...

I now that this may not be the better way to manage SPAM but tom it works. I use SpamHalter in conjunction with the rules and I am happy with the solution.

But from some time ago I started getting the message "Infinite rule processing loop detected" and the rules processor stops processing the email message.

After doing a bunch of testing with my rules, I am pretty sure that the rules processor stops processing after a fixed number of rules and give up showing the infinite loop message.

I'm sure that an infinite loop wasn't achieved. 

My question is : is there a way to increase the number of the processed rules before an infinite loop is assumed ?

[[]] Maurício Ventura Faria 

> I now that this may not be the better way to manage SPAM but tom it works. I use SpamHalter in conjunction with the rules and I am happy
> with the solution.

I know you do not want to hear this but all I can tell you is that trying to do this with filters is a never ending task and you'll probably run out of disk space trying to use rules to solve the problem.  I would strongly recommend that you use Spamhalter, blacklists and Greywall to get the spam to a reasonable level and then use filtering to block the leakers.  

I personally use POPFileD with POPFile, Spamcop and Spamhaus blacklists and Greywall and 99.87% of the spam received by my system is caught.  I move these to one of my Mercury accounts so I can review the spam for false positives. This is pretty much a hands off operation and the users very seldom see any spam.  

I do IP address blocking once in awhile to stop the DoS type attacks but Mercury does do a pretty good job of handling this any way.

[quote user="Thomas R. Stephenson"] I would strongly recommend that you use Spamhalter, blacklists and Greywall to get the spam to a reasonable level and then use filtering to block the leakers.  [/quote] I do this. And Spamhalter does thje job. The problem are the false positive.
[quote user="Thomas R. Stephenson"]I move these to one of my Mercury accounts so I can review the spam for false positives.[/quote] I do the same as you. And I use rules while reviewing for false positives. I add rules to eliminate messages that I am sure are spam in order to facilitate reviewing false positives. And yes, this is an endless job, but with 10 minutes a day I can review the catched spam to avoid false positives. 

I made an small tool that adds rules to RULES.MER in an easy way ( to me ). I cut a piece of the spam email an the tool pastes it as a rule in RULES.MER.

My problem is that the rules processor is stopping processing all rules thinking it reached an endless lop when it does not.

> > I move these to one of my Mercury accounts so I can review the spam for false positives.
>
> I do the same as you. And I use rules while reviewing for false
> positives. I add rules to eliminate messages that I am sure are spam
> in order to facilitate reviewing false positives. And yes, this is an
> endless job, but with 10 minutes a day I can review the catched spam
> to avoid false positives. My problem is that the rules processor is
> stopping processing all rules thinking it reached an endless lop when
> it does not.

I guess I do not understand the problem, I have just 2 rules when reading the spam account and it has at least a couple of hundred messages a day put into the account.  I also only see a false positive about once a month.  Are you getting thousands a spams a day with a large percentage of false positives?

[quote user="Thomas R. Stephenson"]I guess I do not understand the problem[/quote]Possibly...[quote]I have just 2 rules when reading the spam account and it has at least a couple of hundred messages a day put into the account.  I also only see a false positive about once a month.  Are you getting thousands a spams a day with a large percentage of false positives?[/quote]Seems we have a similar environment, I also receive about 200 to 300 messages a day and have false positives 2 or 3 times a month.

But I revise all spam messages except those I called myself "superspam", what I do using rules.   In this way I have to revise about 10 messages a day in a universe of 200.

I spent less than 5 minutes a day doing this and I am able to identify and correct almost all false positives before users note... 

> Seems we have a similar environment, I also receive about 200 to 300
> messages a day and have false positives 2 or 3 times a month.
>
> But I revise all spam messages except those I called myself
> "superspam", what I do using rules.   In this way I have to revise
> about 10 messages a day in a universe of 200.
>
> I spent less than 5 minutes a day doing this and I am able to identify
> and correct almost all false positives before users note...

I guess I really do not understand. You must spend a whole lot more time editing and creating rules that may or may not be deleting messages that could have been false positives.  If you have a huge amount of  rules then there are many many chances to delete a good message with the bad.  Are you not defeating the actual purpose of moving these to an e-mail account instead of just deleting on detection?

FWIW, it takes me about 5 minutes to go through 200 spam messages looking for false positives.  Since I have this account added to my mailbox this is quite easy to do when I have a few minutes available.

Ok, thanks. 

Is source code available ? 

> Is source code available ?

Nope