On Fri, Dec 13, 2002 at 06:30:08PM +0100, Tony L. Svanstrom wrote:
> Exactly, which is why I, on my own lists, always set a reply-to to the list if
> there's no reply-to set already; meaning that unless the sender says otherwise
> the reply goes to the list, just as he wants it to.
Hint: Broken a
[EMAIL PROTECTED] ([EMAIL PROTECTED]) had this to say on 12/13/02 at 21:43:
> Maybe I need to rephrase that: It informs the *sender*. What I like about
> exit code 100 is that it doesn't queue the message and clog my queue.
> Anyway, I've tested this. It definitely passes the echo back. Here i
On Fri, 13 Dec 2002, Mike Leone wrote:
> [EMAIL PROTECTED] ([EMAIL PROTECTED]) had this to say on 12/13/02 at 18:20:
> >
> > This varies depending on the MDA. With maildrop (used with Maildir), you
> > would have something like:
> >
> > if ((/^X-Spam-Status:.*Yes/))
> > {
> > echo "Your Email
On Fri, 13 Dec 2002 16:20:56 -0800,
[EMAIL PROTECTED] wrote
with a subject of "Spamassassin-talk -- confirmation of subscription --
request 624005":
> Spamassassin-talk -- confirmation of subscription -- request 624005
>
> We have received a request from 24.76.225.37 for subscription of your
> ema
[EMAIL PROTECTED] ([EMAIL PROTECTED]) had this to say on 12/13/02 at 18:20:
>
> This varies depending on the MDA. With maildrop (used with Maildir), you
> would have something like:
>
> if ((/^X-Spam-Status:.*Yes/))
> {
> echo "Your Email was Rejected by our SPAM filters. Sorry."
> EXIT
Jason Haar <[EMAIL PROTECTED]> writes:
> As such, only the *USER* can differentiate between
> the two. If the *USER* receives SCE, then they must filter it off *before*
> applying SA rulechecks - otherwise they will loose mail.
Which is where Bayes (as in CVS) shows its advantage in that the user
Matt Kettler ([EMAIL PROTECTED]) had this to say on 12/13/02 at 15:36:
> I'd say that's perfectly reasonable.. Unfortunately some of mailing lists
> flat-out munge any existing reply-to's.. which is the bad part everyone
> objects to.
>
> ie: if I explicitly set a reply-to of myself because I w
This varies depending on the MDA. With maildrop (used with Maildir), you
would have something like:
if ((/^X-Spam-Status:.*Yes/))
{
echo "Your Email was Rejected by our SPAM filters. Sorry."
EXITCODE=100
exit
}
to "./Maildir/."
On Fri, 13 Dec 2002, James D. Stallings wrote:
>
> Do
Yo!
I'm very surprised SpamAssassin doesn't do better on this (attached)
style of messages. As you can see in the headers, score is very low with
4.6, of which 3.5 are because I run bogofilter together with sa (other
scores are only adjusted a few tenths).
I receive these buggers quite often and
Could you explain the last part of allowing spamd reading accounts
user_prefs to modify existing rules. (Is new whitelist_to a new rule?) What
is the other site-wide setting allows user_prefs?
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, December 1
On Fri, Dec 13, 2002 at 10:11:28AM -0500, bunger wrote:
> 1. There appears to be multiple versions of spamd running around, each doing
>something different. One appears to be an smtp relay and one appears to be just a
>middle-ware piece without the ability to pass messages to another host.
The
If you want to bounce mail that you are sure is spam,
you'll want to integrate SA with a milter like
mimedefang. Then you can just do:
if ($hits >= $req) {
return action_bounce("Message looks like SPAM, rejected");
}
You can also keep a copy of the message in quaranteen for a
few days in case
On Fri, Dec 13, 2002 at 03:06:36AM -0800, John Rudd wrote:
>
> On Thursday, Dec 12, 2002, at 14:02 US/Pacific, Theo Van Dinter wrote:
> > (so someone could, sign up for an account
> >with someone else's email address and then sign up for newsletters,
> >but ...)
>
>
> Correction. Not "someone c
At 10:00 AM 12/13/2002 -0500, Vivek Khera wrote:
> "MK" == Matt Kettler <[EMAIL PROTECTED]> writes:
Interesting tactic. I assume it goes in my razor config somewhere, or
is it an SA parameter?
Here's what I'm using in ~/.razor/razor-agent.conf:
min_cf = ac + 10
note that
On Fri, 13 Dec 2002 the voices made James D. Stallings write:
> Does anyone know if there is a way that I can have SA send a
> message that is detected as spam back to the spammer stating
> you are not welcome and this message is blocked...??
1. SA doesn't send nor delete e-mails, it just tags '
Perhaps useful to know for others: Perl 5.005_03 needs "Errno.pm" in spamd
2.43. This does not show up on needed dependencies, nor in the "make test"
run. The test just fails to start spamd (and you do not get to see why, cuz
it redirects STDERR to /dev/null).
- Mark
System Administrator
I found this piece of non-spam in my spam folder today.
I only found it after he sent a reply(with passcode) to
my autoreply. I guess I should watch my spam folder
closer. Anyways, here it is, and unfortunately, I can't
really think of a solution to avoid this type of problem
other than bouncing
Well, that's not really something SA itself can do at all.. SA is just a
filter, it can't deliver, delete, or do anything else but change the
content of an email. However the MUA/MTA that you are calling SA from could
be capable of such things. A good procmail recipe should be able to do it..
U
At Fri Dec 13 17:04:17 2002, Harold Hallikainen wrote:
>
> Last week I posted a section of California law that permits ISPs
> to sue spammers if the spammer has been notified that the ISP does not
> accept spam. The law makes mention of automatic notification by the
> receivning email server
At 10:55 AM 12/13/2002, you wrote:
Does anyone know if there is a way that I can have SA send a
message that is detected as spam back to the spammer stating
you are not welcome and this message is blocked...??
Or is that asking for trouble.
I think it's been brought up before - SA has no sendi
I'd say that's perfectly reasonable.. Unfortunately some of mailing lists
flat-out munge any existing reply-to's.. which is the bad part everyone
objects to.
ie: if I explicitly set a reply-to of myself because I wanted private
replies, and the list over-wrote that, I'd be a bit bent outa shape
At Fri Dec 13 18:46:52 2002, Steve Thomas wrote:
>
> | email systems. One responder on this list said he/she included something
> | in the HELO response on the SMTP server. Does anyone know of any standards
> | in this area as to what a legally enforcible "no spam" SMTP response would
>
>
> I Am
That is, it.. more-or-less..
Razor2 supports nilsimsa "fuzzy" hashes, but right now those are disabled
server-side because of bugs in the hash algorithm resulting in a high
collision rate.
Thus ephemeral sigs are what razor2 uses, for now. It is defeatable, but it
does require the spammer to c
On Fri, Dec 13, 2002 at 11:08:07AM +0100, Ralf Hildebrandt wrote:
> * Albert Croft <[EMAIL PROTECTED]>:
>
> > Question 1:
> >What has the performance of spamassassin (particularly the
> > spamc/spamd combination) been like for mailservers handling
> > medium-to-large quantities of in-bound m
Does anyone know if there is a way that I can have SA send a
message that is detected as spam back to the spammer stating
you are not welcome and this message is blocked...??
Or is that asking for trouble.
Thanks
Jim
Get your own "800" number
V
| email systems. One responder on this list said he/she included something
| in the HELO response on the SMTP server. Does anyone know of any standards
| in this area as to what a legally enforcible "no spam" SMTP response would
I Am Not A Lawyer
I don't really know about the enforceability of
???
Just filter it normally - procmail doesn't care if you're running mail
through SA or not. In fact, the "standard" way to run SA is *from* procmail.
| -Original Message-
| Is there an easy way to have my mail filtered by procmail after it is
| processed by SpamAssassin. I am calling S
On Fri, 13 Dec 2002, Administrator wrote:
> Is there an easy way to have my mail filtered by procmail after it is
> processed by SpamAssassin. I am calling Spamassassin on a sitewide basis,
> but would like the ability to filter mail after it is processed by
> SpamAssassin.
Sure...it's in the sa
The last I heard on this was that the recommendation was to add "; NO UCE"
to the end of your SMTP banner. At one point at least one spamware honored
that.
--
Kent Hamilton [EMAIL PROTECTED]
Manager - Systems & Networking
Hunter Engineering Company
> -Original Message-
> From: Harold
On Fri, 13 Dec 2002 the voices made Mike Leone write:
> However, 90+% of the time, you want the reply to be public; that's why you're
> on a public mailing list, no? So why set the defaults to cater to a special
> case, in stead of the majority case?
Exactly, which is why I, on my own lists, alw
Last week I posted a section of California law that permits ISPs
to sue spammers if the spammer has been notified that the ISP does not
accept spam. The law makes mention of automatic notification by the
receivning email server, indiciating that notification would be
sufficient. However, th
Hi,
I have a set of 5 emails in my inbox, all from a [EMAIL PROTECTED]
address, and all bypassing my filtering. The largest is 2.6k, so it
didn't get excluded by my procmail recipe or by spamd. spamd/spamc
version is 2.43 (but I have seen it happen with 2.31 also). I am at a
loss to begin where
Is there an easy way to have my mail filtered by procmail after it is
processed by SpamAssassin. I am calling Spamassassin on a sitewide basis,
but would like the ability to filter mail after it is processed by
SpamAssassin.
Any help would be greatly appreciated.
Michael
Tom Allison ([EMAIL PROTECTED]) had this to say on 12/13/02 at 07:36:
> [EMAIL PROTECTED] wrote:
> >--On Thursday, December 12, 2002 3:07 PM -0500 Tom Allison
> ><[EMAIL PROTECTED]> wrote:
> >
> >| My question is if anyone knows anything about running
> >| spamassassin/spamc/spamd under sieve.
>
Duncan Findlay ([EMAIL PROTECTED]) had this to say on 12/13/02 at 00:40:
> Fair enough, I guess. On the Debian lists, Mail-Followup-To is the
> header everyone lives by. (Probably because we all use mutt)
Not everyone. :-) When I read my email from work, I do it via Squirrelmail,
a web client. A
Bob Proulx ([EMAIL PROTECTED]) had this to say on 12/13/02 at 02:00:
> Duncan Findlay <[EMAIL PROTECTED]> [2002-12-12 23:57:48 -0500]:
> > On Thu, Dec 12, 2002 at 09:56:23PM -0500, Matt Kettler wrote:
> > > If you really want people to reply on-list your should add a Reply-To
> > > header that co
I have been struggled for the past few days trying to find a simple smtp relay plugin
for SpamAssassin that will do the following:
1. Listen on a pre-specified IP and port
2. Scan any incoming message for Spam
3. Pass *good* messages onto another pre-specified IP and port
4. Pass *bad* messag
> "MK" == Matt Kettler <[EMAIL PROTECTED]> writes:
MK> Razor is notorious for having some quantity of invalid submissions.. Not
MK> too many, but some... (note: I'm excluding the null-mime-block issue from
MK> this, since that's now fixed). Usually bad submitters get bad CF ratings,
MK> but
> Razor2 uses ephemeral sigs by default, which only hashes some subset of
> the body of the email. ie: bytes 50-200, then later it might be
> 120-900.
Well, I hope that's not it -- that's pretty trivial to work around too ;)
--j.
---
This sf
> "DF" == Duncan Findlay <[EMAIL PROTECTED]> writes:
DF> Funny, I don't want people to CC me on replies (or even address them
DF> To me). That's why I'm not in my Reply-To or Mail-Followup-To: header.
DF> Yet, strangely enough, I do get CC'd most of the time.
When I hit "reply" to this last m
I solved my own problems:
On Fri, 13 Dec 2002, Ian G Batten wrote:
> 1. Even though I have terse reporting turned on, which under 2.43 and
For whatever reason, under 2.50 I need report_header to be explicitly
set to zero. Under 2.43 and before I didn't.
> 2. Although the Bayesian filtering, w
> "BA" == Bob Apthorpe <[EMAIL PROTECTED]> writes:
BA> mailing lists, so it's in your organization's best interest to move to
BA> 'confirmed opt-in' (or in greasy DMA-speak, 'double opt-in') as soon as
BA> reasonably achievable.
There's nothing greasy DMA-speak about "double opt-in". Describ
> "TVD" == Theo Van Dinter <[EMAIL PROTECTED]> writes:
TVD> Another thing to make that subset smaller is double opt-in (ie: verify the
TVD> initial opt-in). We're actually not doing that yet, although I hear it's
TVD> on the board for next year. (so someone could, sign up for an account
TVD>
> I think the time taken to scan a message vastly outweighs the time
> taken to open up a new connection.
fwiw, the 3-message TCP connection overhead is not a big deal on
a LAN. Reusing a connection a la HTTP keepalive is important
over the net or a WAN, but for spamd it's overkill.
--j.
--
--On Friday, December 13, 2002 6:47 AM -0500 Tom Allison
<[EMAIL PROTECTED]> wrote:
| What version of amavisd-new are you using? I have heard there are a lot
| of patches to resolve, or has this been addressed?
Latest + patches, as listed at referenced site.
Amos
--
Albert Croft said the following on 13/12/02 09:26:
Two questions, which I hope someone may be able to point me in the
direction of resources for.
Question 1:
What has the performance of spamassassin (particularly the
spamc/spamd combination) been like for mailservers handling
medium-to-lar
Juan Quesada wrote:
Hello,
I have two questions:
1) If a legitimate email is marked as spam, I want
Spam Assassin to return an email to the sender stating
that the email was marked as spam. We want to do this
inititally so we can whitelist anyone who complains.
2) Also, I can I copy the 60_whit
[EMAIL PROTECTED] wrote:
--On Thursday, December 12, 2002 3:07 PM -0500 Tom Allison
<[EMAIL PROTECTED]> wrote:
| My question is if anyone knows anything about running
| spamassassin/spamc/spamd under sieve.
Sure, sort of. Though, I'm using amavisd-new[1] with postfix[2]. It's
not that it run
On 2002-12-10 09:32:54 -0500, rODbegbie wrote:
> Adam Henry wrote:
> >> From: [EMAIL PROTECTED]
>
> The "From:" header doesn't end with @ebay.com, therefore the eBay checks
> weren't performed.
I have the same problem with ebay.de .
How does one write a rule GENUINE_EBAY_DE_RCVD?
header GENUINE
I've upgraded a personal system from 2.43 to 2.50 in order to play with
the Bayesian stuff. The configuration is Redhat 7.2, Sendmail 8.12.6,
spamassassin_milter with a lot of my own hacks in. 2.43 and
predecessors have been running correctly for the past year.
I've encountered two issues:
1.
On Thursday, Dec 12, 2002, at 08:37 US/Pacific, Theo Van Dinter wrote:
As an example, things I was thinking about were:
- Choose good 'from' addresses
- Choose descriptive and non-spammy-looking subjects
- Include clear information about why the person is receiving the
newsletter and how t
On Thursday, Dec 12, 2002, at 14:02 US/Pacific, Theo Van Dinter wrote:
(so someone could, sign up for an account
with someone else's email address and then sign up for newsletters,
but ...)
Correction. Not "someone could", "people do". It happens. A lot
(though, maybe not through your ser
* Albert Croft <[EMAIL PROTECTED]>:
> Question 1:
>What has the performance of spamassassin (particularly the
> spamc/spamd combination) been like for mailservers handling
> medium-to-large quantities of in-bound mail (i.e., averaging on the
> order of tens of thousands or more emails per d
Two questions, which I hope someone may be able to point me in the
direction of resources for.
Question 1:
What has the performance of spamassassin (particularly the
spamc/spamd combination) been like for mailservers handling
medium-to-large quantities of in-bound mail (i.e., averaging on th
> I'm guessing that someone has a mis-configured spamtrap that they
> wound up bouncing one of these messages to. There seems to be a lot of
> misconfigured spamtraps out there...
Just to bring this up...
In my opinion any spamtrap whose address appears on a Web page is
misconfigured, unless you'
> The behavior you see happens when someone replies, removes the SATalk
> subject tag, and replies to you directly (usually including a CC
> or To: to
> the list as well).
That can happen, but not in this case:
Thread-Topic: sql support
Thread-Index: AcKiFSjH4NQNOiAMTDahdq4U92XGBA==
From: "Dallas
On Thu, 12 Dec 2002 the voices made Duncan Findlay write:
> Funny, I don't want people to CC me on replies (or even address them
> To me). That's why I'm not in my Reply-To or Mail-Followup-To: header.
> Yet, strangely enough, I do get CC'd most of the time.
It's standard behavior when you reply
57 matches
Mail list logo
