On Sat, Jan 04, 2003 at 06:02:24PM -0500, Theo Van Dinter wrote:
> In thinking about it now, it does look like we need another variable to
> set the paths correctly. My issue is that I need to install into another
> location but have the scripts update to point at the system-wise points.
> Your is
On Sat, 4 Jan 2003 the voices made Adrian 'Dagurashibanipal' von Bidder write:
ADvB> I find bogofilter does a very good job currently, no false positives at
ADvB> all; and most false negatives get caught by sa and are fed back to
ADvB> bogofilter so it can learn...
Take a closer look at those fa
On Sat, Jan 04, 2003 at 11:24:19PM +0100, Denis Ducamp wrote:
> - -DINSTALLSITELIB="$(SITELIBEXP)"
> + -DINSTALLSITELIB="$(INSTALLSITELIB)"
>
> But I have no idea if this is good for a system wide installation of
> spamassassin...
Since I specifically put that in
On Sat, Jan 04, 2003 at 08:41:22PM +0100, Denis Ducamp wrote:
> On Sat, Jan 04, 2003 at 01:49:37PM -0500, Theo Van Dinter wrote:
> > On Sat, Jan 04, 2003 at 07:31:40PM +0100, Denis Ducamp wrote:
> > > are loaded from /usr/lib/perl5/site_perl first. Looking at the spamassassin
> > > installed script
interesting effect with this spam... I'm set to report_safe 1. However,
the message cam to me with the headers really screwed up. The result was
that my server rules for filtering on X-Spam-Status didn't work.
Should SA make better attempt to not reproduce corrupt headers/etc.?
-- Nathan
--
The biggest problem, even with using the RBLs with SA, is trying to
teach my users what SA is... "a mail filter to identify spam". Too many
of our users think it's a magic bullet (or worse, they seem to think
that a human somehow makes an individual decision as to spam/nonspam),
and get damn un
On Fri, Jan 03, 2003 at 03:19:11PM -0500, Chris Santerre wrote:
> Well I'm reading more and more with the slow holiday time. I keep coming
> back to Theo and kludge.net as great info. Particularly the post Theo made
FYI: that's kluge.net ... I wonder how much mail the guys at kludge.net
get for m
On Fri, Jan 03, 2003 at 10:26:08PM +0100, Tony L. Svanstrom wrote:
> CS> -SMTP(access bd, RBLS, simple header checks, ect)
> CS> -Procmail
> CS> -SA
> CS> -MUA filter
> CS> -Delete key :-)
>
> [*] I'm really trying to cut down on the number of perl-processess each and
> every e-mail ends up starti
On Sat, 2003-01-04 at 18:13, Ben Jackson wrote:
> On Sat, Jan 04, 2003 at 12:40:38PM +0100, Adrian 'Dagurashibanipal' von Bidder wrote:
> > [the following false positive:]
> > Thank you for your message regarding
> > Systematic scanning from 209.241.48.162
>
> I have a personal SA rule that'
Yo!
I didn't follow this mailing list long enough - how does the bayes
algorithm of 2.5 compare to bogofilter?
I find bogofilter does a very good job currently, no false positives at
all; and most false negatives get caught by sa and are fed back to
bogofilter so it can learn...
cheers
-- vbi
-
On Sat, Jan 04, 2003 at 01:49:37PM -0500, Theo Van Dinter wrote:
> On Sat, Jan 04, 2003 at 07:31:40PM +0100, Denis Ducamp wrote:
> > are loaded from /usr/lib/perl5/site_perl first. Looking at the spamassassin
> > installed script show it uses that directory :
> > $ grep 'use lib' ~/saCVSusr/bin/spa
On Sat, Jan 04, 2003 at 07:31:40PM +0100, Denis Ducamp wrote:
> are loaded from /usr/lib/perl5/site_perl first. Looking at the spamassassin
> installed script show it uses that directory :
> $ grep 'use lib' ~/saCVSusr/bin/spamassassin
> use lib '/usr/lib/perl5/site_perl'; # substituted at 'make'
On Sat, Jan 04, 2003 at 11:51:02AM -0500, Theo Van Dinter wrote:
> On Sat, Jan 04, 2003 at 04:54:58PM +0100, Denis Ducamp wrote:
> > Then when I run "~/saCVSusr/bin/spamassassin < msg.txt 2>&1 | less" I get :
> >
> > configuration file "/home/ducamp/saCVSusr/share/spamassassin/20_body_tests.cf"
>
On Sat, Jan 04, 2003 at 09:13:02AM -0800, Ben Jackson wrote:
> I don't see any way for the default rlueset to have any of the rules
> that I find most effective for avoiding false positives:
>
> - mentions my IP address
> - uses my real name
> - includes part of my address
> - mentions keyword
On Sat, Jan 04, 2003 at 12:40:38PM +0100, Adrian 'Dagurashibanipal' von Bidder wrote:
> [the following false positive:]
> Thank you for your message regarding
> Systematic scanning from 209.241.48.162
I have a personal SA rule that's worth -5 for my cable modem IP address.
I don't see any wa
On Sat, Jan 04, 2003 at 04:54:58PM +0100, Denis Ducamp wrote:
> Then when I run "~/saCVSusr/bin/spamassassin < msg.txt 2>&1 | less" I get :
>
> configuration file "/home/ducamp/saCVSusr/share/spamassassin/20_body_tests.cf"
>requires version 2.50 of SpamAssassin, but this is code version 2.43. May
Hi,
I'd like to test the cvs version and install it with a line such as :
make distclean && cvs -z3 update -d -P -A && echo y | perl Makefile.PL \
~/PREFIX=~/saCVSusr SYSCONFDIR=~/saCVSetc && make && echo make test && rm \
-rf ~/saCVSusr ~/saCVSetc && make install
Then when I run "~/saCVSusr/bin
On 2003-01-04 12:40:38 +0100, Adrian 'Dagurashibanipal' von Bidder wrote:
> I know this is probably not very relevant with 2.5 release so soon.
> Anyway - see the attached message. Would have scored even higher (8.7)
> with 2.43 default scores.
Tell them to generate valid dates and use a senders n
At Sat Jan 4 12:52:09 2003, [EMAIL PROTECTED] wrote:
>
> Hi
>
> I'm new to SpamAssassing but so far I got it running. A very nice piece of
> software.
> I checked the online documentation but I couldn't find an answer.
> Can I automatically drop mail that has been identified as spam?
> I would n
Hi
I'm new to SpamAssassing but so far I got it running. A very nice piece of
software.
I checked the online documentation but I couldn't find an answer.
Can I automatically drop mail that has been identified as spam?
I would not like to define that per user but to be valid for everybody.
thanks,
Yo!
I know this is probably not very relevant with 2.5 release so soon.
Anyway - see the attached message. Would have scored even higher (8.7)
with 2.43 default scores.
--
featured link: http://fortytwo.ch/gpg/subkeys
--- Begin Message ---
Thank you for your message regarding
Systematic
On Fri, Jan 03, 2003 at 02:58:09PM -0600, Rich Puhek wrote:
> I found that I can't outright block .cn and .kr without some customers
> getting very irritated at me. Even if I could, maintaining a whitelist
> would be a PITA for entire countries (or, worse, for RBLS like
> unconfirmed.ordb.org.)
22 matches
Mail list logo
