You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by NGSS <ng...@gmail.com> on 2008/07/02 07:55:35 UTC
bad rules that likely to result in more false positives
Thanks for the response.
Yah I think it is just too aggressive, I included a handful of rules
Is there any forum or website that discuss about (lists of ) rules that is
likely to result in more false positives ?
-----Original Message-----
From: Matus UHLAR - fantomas [mailto:uhlar@fantomas.sk]
Sent: Tuesday, July 01, 2008 3:35 PM
To: users@spamassassin.apache.org
Subject: Re: ways to check reason/error of rejected/bounced emails with
calling customers
Please teach your mailer to wrap lines in a sane way...
On 01.07.08 11:46, NGSS wrote:
> Many of our clients started to have problem sending emails to us after I
> inserted more strict
> SA rules . Previously our system was flooded with spams. So I decided to
> inserted them to the
> Existing emails. After this the spams had reduced significantly. But I
know
> more worry about false
> Positive and rejected (or sometimes disappeared emails) .
> I can't call all of them to get them to send me the bounced/error
messages.
> So I wonder if there is a
> Way to check for the rejected emails and why they are being rejected? So
at
> least I know what reason
> For the rejects and will be able to fine-tine it further.
Seems you set up your MTA too agressively - probably reject mail with too
low score. However you did not provide enough informations for us to help
you.
What's "existing emails"? Did you train global BAYES filter on received
spam?
Did you feed it enough of hams to avoid FPs? Did you play with scores? What
did you set required_score to?
Did you fiddle with other settings like trusted_networks and
internal_networks to set up proper trust path? Did you turn on all possible
network rules?
--
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Micro$oft random number generator: 0, 0, 0, 4.33e+67, 0, 0, 0...
Courtesy copy (was: bad rules that likely to result in more
false positives)
Posted by SM <sm...@resistor.net>.
At 10:34 04-07-2008, Robert - elists wrote:
>I don't get why the list software doesn't take care of this so we never have
>this issue.
It's a question of choice. If the list software takes care of it,
then the MUA cannot set a preference. There is a tendency to "fix"
list software for the failings of the MUA or because of the inability
of users to make use of email features.
>Isn't one of the people that administrates this list bright enough to make
>that happen?
No. :-)
>One, it is the right thing to do, and two, then we wouldn't have to listed
>to all this baloney on the list about it anymore.
It's not the right thing; it's a matter of opinion. This type of
discussion happens on most mailing lists every now and then.
The header mentioned in the discussion is a non-standard
header. There is already a header added by this mailing list which
can be used to reply to the list only. Some MUAs support it.
Some people, usually those asking a question, ask for a courtesy copy
of the reply. Some prefer not to receive it as they follow the mailing list.
Regards,
-sm
RE: bad rules that likely to result in more false positives
Posted by Bernd Petrovitsch <be...@firmix.at>.
On Fre, 2008-07-04 at 10:34 -0700, Robert - elists wrote:
[....]
> hmmmm
>
> I don't get why the list software doesn't take care of this so we never have
> this issue.
Because it's a decision of the mail sender (read: you and me) if he
wants a private copy (additionally to the public one over the
mailinglist).
> Isn't one of the people that administrates this list bright enough to make
> that happen?
Not possible - see above.
The only thing the ML manager can do is: Look at the recipients (in the
To: and Cc: line) and suppress the copy to you over the ML if the email
address is already found there.
The only drawback is: You then get ML emails partly directly and partly
over the ML. Since ML tend to add headers (e.g. list-help,
list-unsubscribe, List-Post and List-Id), this may brake mail filter
rules at the receivers end.
Mailman has a user-specific option for this, I don't know about ezmlm.
Bermd
--
Firmix Software GmbH http://www.firmix.at/
mobil: +43 664 4416156 fax: +43 1 7890849-55
Embedded Linux Development and Services
Re: bad rules that likely to result in more false positives
Posted by mouss <mo...@netoyen.net>.
Robert - elists wrote:
> [snip]
> I don't get why the list software doesn't take care of this so we never have
> this issue.
>
This has been debated to death here and elsewhere. please search the
archives and/or google before bringing this religious debate again.
> Isn't one of the people that administrates this list bright enough to make
> that happen?
>
The following lists work the same way as SA:
- postfix
- amavisd-new
- bogofilter
- dovecot
- courier
- debian
- FreeBSD
- NetBSD
- IETF lists (ietf-smtp, ...)
- spam-l
- securityfocus lists
....
do you think you are bright and all these people are stupid? don't you
think this is very unlikely (exceptionally when you see that the IETF
lists are configured this way!).
> One, it is the right thing to do,
No, the right thing is my religion. everything else is bad ;-p
come on...
> and two, then we wouldn't have to listed
> to all this baloney on the list about it anymore.
>
If everybody uses Mutt, we wouldn't have this "baloney" either ;-p
Note that "even" Thunderbid correctly handle the Mail-Followup-To
header. so if you really have a problem with console mailers, you can
still use TB :)
> The list software and or a list admin also occasionally signs up addresses
> that are not opt'ed in...
>
>
I don't understand this part.
Please let's kill this thread.
- Matus and few others will keep shouting when people send them private
copies.
- The rest of us don't care
- Earth will keep turing...
RE: bad rules that likely to result in more false positives
Posted by Robert - elists <li...@abbacomm.net>.
>
> I'm sorry, but what MUA recognizes those? Why don' t you set Reply-
> To: which will be honored by all MUAs?
>
snip
>
>
> He is acted as is common and expected. Others who, like you, don't
> want private copies set Reply-To.
>
> --
> Jo Rhett
> Net Consonance : consonant endings by net philanthropy, open source
> and other randomness
>
hmmmm
I don't get why the list software doesn't take care of this so we never have
this issue.
Isn't one of the people that administrates this list bright enough to make
that happen?
One, it is the right thing to do, and two, then we wouldn't have to listed
to all this baloney on the list about it anymore.
The list software and or a list admin also occasionally signs up addresses
that are not opt'ed in...
- rh
Re: bad rules that likely to result in more false positives
Posted by Matus UHLAR - fantomas <uh...@fantomas.sk>.
> On Jul 3, 2008, at 12:14 AM, Matus UHLAR - fantomas wrote:
> >>>Please, don't send private replies, I did not ask for
> >>>them.
> >
> >On 02.07.08 21:32, Jari Fredriksson wrote:
> >>Its impossible to know who wants them, and who does not.
> >
> >my mail headers contain Mail-Followup-To: header that is only sent
> >to the
> >list. That means that replies should be sent to the list.
On 03.07.08 14:30, Jo Rhett wrote:
> I'm sorry, but what MUA recognizes those?
Mutt.
> Why don' t you set Reply-
> To: which will be honored by all MUAs?
Because Reply-To: has different meaning and side-effects.
> >If anyone wants private copies, (s)he should ask for them. This is a
> >mailing
> >lists and all members receive all mail posted to it. Even non-
> >members can
> >read it all in archives.
> He is acted as is common and expected.
No, people are expected to keep discussions on the list. Sending private
copies may result in discussing off-the-list.
> Others who, like you, don't
> want private copies set Reply-To.
Reply-To: HAS DIFFERENT MEANING AND SIDE-EFFECTS
--
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
My mind is like a steel trap - rusty and illegal in 37 states.
Re: bad rules that likely to result in more false positives
Posted by Sahil Tandon <sa...@tandon.net>.
Matus UHLAR - fantomas <uh...@fantomas.sk> wrote:
[...]
> > > He is acted as is common and expected.
>
> No, people are expected to keep discussions on the list. Sending private
> copies may result in discussing off-the-list.
True. So reply-all leads to mail on-list (notwithstanding the mail sent
off-list).
> > > Others who, like you, don't want
> > > private copies set Reply-To.
>
> Reply-To: HAS DIFFERENT MEANING AND SIDE-EFFECTS
I don't understand what you mean -- what side effects? But in this case,
please reply off-list to me if you'd like; this is off-topic.
> > Bingo! :) Maybe Matus and Benny will get it now.
>
> Maybe you and Jo will finally get it now.
Agree to disagree. :-) Let's close this thread.
--
Sahil Tandon <sa...@tandon.net>
Re: bad rules that likely to result in more false positives
Posted by Matus UHLAR - fantomas <uh...@fantomas.sk>.
On 03.07.08 17:58, Sahil Tandon wrote:
> Jo Rhett <jr...@netconsonance.com> wrote:
>
> > On Jul 3, 2008, at 12:14 AM, Matus UHLAR - fantomas wrote:
> >>>> Please, don't send private replies, I did not ask for
> >>>> them.
> >>
> >> On 02.07.08 21:32, Jari Fredriksson wrote:
> >>> Its impossible to know who wants them, and who does not.
> >>
> >> my mail headers contain Mail-Followup-To: header that is only sent to the
> >> list. That means that replies should be sent to the list.
> >
> > I'm sorry, but what MUA recognizes those?
Mutt.
> > Why don' t you set Reply-To: which will be honored by all MUAs?
Because Reply-To: has different meaning and side-effects.
> >> If anyone wants private copies, (s)he should ask for them. This is a
> >> mailing
> >> lists and all members receive all mail posted to it. Even non-members can
> >> read it all in archives.
> >
> > He is acted as is common and expected.
No, people are expected to keep discussions on the list. Sending private
copies may result in discussing off-the-list.
> > Others who, like you, don't want
> > private copies set Reply-To.
Reply-To: HAS DIFFERENT MEANING AND SIDE-EFFECTS
> Bingo! :) Maybe Matus and Benny will get it now.
Maybe you and Jo will finally get it now.
--
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
The only substitute for good manners is fast reflexes.
Re: bad rules that likely to result in more false positives
Posted by Sahil Tandon <sa...@tandon.net>.
Jo Rhett <jr...@netconsonance.com> wrote:
> On Jul 3, 2008, at 12:14 AM, Matus UHLAR - fantomas wrote:
>>>> Please, don't send private replies, I did not ask for
>>>> them.
>>
>> On 02.07.08 21:32, Jari Fredriksson wrote:
>>> Its impossible to know who wants them, and who does not.
>>
>> my mail headers contain Mail-Followup-To: header that is only sent to the
>> list. That means that replies should be sent to the list.
>
> I'm sorry, but what MUA recognizes those? Why don' t you set Reply-To:
> which will be honored by all MUAs?
>
>> If anyone wants private copies, (s)he should ask for them. This is a
>> mailing
>> lists and all members receive all mail posted to it. Even non-members can
>> read it all in archives.
>
> He is acted as is common and expected. Others who, like you, don't want
> private copies set Reply-To.
Bingo! :) Maybe Matus and Benny will get it now.
--
Sahil Tandon <sa...@tandon.net>
Re: bad rules that likely to result in more false positives
Posted by Jo Rhett <jr...@netconsonance.com>.
On Jul 3, 2008, at 12:14 AM, Matus UHLAR - fantomas wrote:
>>> Please, don't send private replies, I did not ask for
>>> them.
>
> On 02.07.08 21:32, Jari Fredriksson wrote:
>> Its impossible to know who wants them, and who does not.
>
> my mail headers contain Mail-Followup-To: header that is only sent
> to the
> list. That means that replies should be sent to the list.
I'm sorry, but what MUA recognizes those? Why don' t you set Reply-
To: which will be honored by all MUAs?
> If anyone wants private copies, (s)he should ask for them. This is a
> mailing
> lists and all members receive all mail posted to it. Even non-
> members can
> read it all in archives.
He is acted as is common and expected. Others who, like you, don't
want private copies set Reply-To.
--
Jo Rhett
Net Consonance : consonant endings by net philanthropy, open source
and other randomness
Re: bad rules that likely to result in more false positives
Posted by Matus UHLAR - fantomas <uh...@fantomas.sk>.
> > On 02.07.08 13:55, NGSS wrote:
> >> To: 'Matus UHLAR - fantomas' <uh...@fantomas.sk>,
> >> users@spamassassin.apache.org
> >
> > Please, don't send private replies, I did not ask for
> > them.
On 02.07.08 21:32, Jari Fredriksson wrote:
> Its impossible to know who wants them, and who does not.
my mail headers contain Mail-Followup-To: header that is only sent to the
list. That means that replies should be sent to the list.
> Someone who does
> not sit here and read all messages thru may be very greatful of a reply to
> his email address.
If anyone wants private copies, (s)he should ask for them. This is a mailing
lists and all members receive all mail posted to it. Even non-members can
read it all in archives.
--
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
99 percent of lawyers give the rest a bad name.
Re: bad rules that likely to result in more false positives
Posted by Jari Fredriksson <ja...@iki.fi>.
> On 02.07.08 13:55, NGSS wrote:
>> To: 'Matus UHLAR - fantomas' <uh...@fantomas.sk>,
>> users@spamassassin.apache.org
>
> Please, don't send private replies, I did not ask for
> them.
>
Its impossible to know who wants them, and who does not. Someone who does not sit here and read all messages thru may be very greatful of a reply to his email address.
Re: bad rules that likely to result in more false positives
Posted by Matus UHLAR - fantomas <uh...@fantomas.sk>.
On 02.07.08 13:55, NGSS wrote:
> To: 'Matus UHLAR - fantomas' <uh...@fantomas.sk>,
> users@spamassassin.apache.org
Please, don't send private replies, I did not ask for them.
> Yah I think it is just too aggressive, I included a handful of rules
> Is there any forum or website that discuss about (lists of ) rules that is
> likely to result in more false positives ?
ANY rules could lead to false positives. That's why it's better to have more
rules with lower scores. See:
http://wiki.apache.org/spamassassin/ContributingNewRules
http://wiki.apache.org/spamassassin/WritingRules
--
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Due to unexpected conditions Windows 2000 will be released
in first quarter of year 1901