Good afternoon, Stephen,
On Wed, 10 Dec 2003, Stephen Westrip wrote:
> I am trying to set up sa-learn in a site-wide configuration. I have a Red
> Hat 9 server, SpamAssassin 2.61 and MIMEDefanf 2.39. I have got SA to work
> fine and our spam has dropped considerably, but I would also like to use
]
[mailto:[EMAIL PROTECTED] On Behalf Of Matt
Kettler
Sent: 10 December 2003 18:15
To: Stephen Westrip; [EMAIL PROTECTED]
Subject: Re: [SAtalk] Using sa-learn in a site-wide configuration
At 12:04 PM 12/10/2003, Stephen Westrip wrote:
>What exactly do I need to do to make this work? I have read l
At 12:04 PM 12/10/2003, Stephen Westrip wrote:
What exactly do I need to do to make this work? I have read lots about
adding 'bayes_auto_learn 1' and other bits and pieces to put in the cf file
but whatever I try the Bayes DB never gets added to.
did you install DB_File? if not, bayes won't go.
Dear All,
I am trying to set up sa-learn in a site-wide configuration. I have a Red
Hat 9 server, SpamAssassin 2.61 and MIMEDefanf 2.39. I have got SA to work
fine and our spam has dropped considerably, but I would also like to use
sa-learn.
What exactly do I need to do to make this work? I have
On Friday 10 October 2003 09:55, Terry Shows wrote:
> I am relatively new to the SAtalk list, so forgive this "stupid"
> question:
>
> Just what is HAM?
Ham is "real meat"; the stuff you want. As opposed to spam, which is
"phony meat"; the stuff you don't want.
Disclaimer: None of this has anyt
e: [SAtalk] using sa-learn...
Jeff Lasman wrote:
>
> I'm still not sure if I should run through "ham" as well as "spam" or
> not. Is there a specific answer for this?
>
> Thanks.
>
> Jeff
Hi Jeff,
Bayes won't kick in until it has >20
Jeff Lasman wrote:
>
> I'm still not sure if I should run through "ham" as well as "spam" or
> not. Is there a specific answer for this?
>
> Thanks.
>
> Jeff
Hi Jeff,
Bayes won't kick in until it has >200 ham/spam in the database.
You should sa-learn --ham against as wide a range of your normal
> I'm still not sure if I should run through "ham" as well as "spam" or
> not. Is there a specific answer for this?
Yes you should run ham as well, because the Bayes algorithm needs to
compare againt both "spammy" and "hammy" words to determine how to
classify new mail. You cannot train it with
I did install SA-2.60 and I can't locate sa-learn, except for a file
sa-learn.raw in the sources directory?
what did I do wrong ?
thanks,
petre
On Friday 10 October 2003 17:56 Anno Domini, Jeff Lasman wrote using one of
his keyboards:
> On Friday 10 October 2003 07:32, Martin Schroeder wrote:
On Friday 10 October 2003 07:32, Martin Schroeder wrote:
> On 2003-10-10 07:14:56 -0700, Jeff Lasman wrote:
> > And one more thing... though I don't see it mentioned, I'm
> > presuming from the sa-learn manpage that I don't have to remove the
> > already existing SpamAssassin headers from either t
Jeff Lasman wrote:
If I should, then I'll probably uses messages sent to this list, as I
know there's no spam hidden in them; I can't guarantee that about any
of my other folders. Is that a reasonable assumption?
Should I try to put through the same number of messages? Or doesn't it
matter?
On 2003-10-10 07:14:56 -0700, Jeff Lasman wrote:
> And one more thing... though I don't see it mentioned, I'm presuming
> from the sa-learn manpage that I don't have to remove the already
> existing SpamAssassin headers from either the spam or the ham; that
> sa-learn will ignore them. Is that
I'm very new to SpamAssassin. I've read 'man sa-learn' but I have a few
questions about it...
I've been running it for a few days now and I have a folder I call
'spamuncaught', every email put into it by hand, that contains 210
messages.
If I run these through sa-learn as spam, won't I be mar
On Thu, Aug 21, 2003 at 01:34:00AM +0300, Jefferson Cowart wrote:
> Reading the documentation for sa-learn I don't see how to tell
> it to use the site-wide databases in stead of the per-user db.
> How do I use site-wide bayes db?
If you run SpamAssassin as spamd user on the system, su to spamd u
Reading the documentation for sa-learn I don't see how to tell it to use
the site-wide databases in stead of the per-user db. How do I use
site-wide bayes db?
Thanks
Jefferson Cowart
[EMAIL PROTECTED]
---
This SF.net email is
William Turrell wrote on Tue, 10 Jun 2003 11:18:35 +0100:
> Is there a way I can make sa-learn read .dbx files, or a batch program I can
> use to convert a .dbx file into plain text messages?
>
There's something called dbxtract, should be available from here:
http://www.oehelp.com/
I don't know
http://mbx2mbox.sourceforge.net/
>
> -Original Message-
> From: William Turrell [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, June 10, 2003 5:19 AM
> To: [EMAIL PROTECTED]
>
[]
>
> Is there a way I can make sa-learn read .dbx files, or a
> batch program I can use to convert a .dbx file
I am running Outlook Express 6 on Windows XP, with SAProxy 1.2.0 (which uses
Spamassassin 2.55).
I have have two spam folders:
- "Spam" - everything identified by Spam Assassin
- "Spam to learn" - false negatives I want to pass through sa-learn
The problem is Outlook Express stores each folder as
On Thu, May 29, 2003 at 12:34:50PM -0700, Dale Harris wrote:
> So I have spamassassin 2.55 installed. However, it doesn't look like
> spamc is passing any information on to sa-learn. I do temporary store
spamc doesn't do learning. spamd can autolearn, but that's it.
> spamassassin -d | sa-lear
So I have spamassassin 2.55 installed. However, it doesn't look like
spamc is passing any information on to sa-learn. I do temporary store
the spam in it's own folder... so it is it valid to pipe messages to:
spamassassin -d | sa-learn --spam --single
Is what I want to do to populate my Bayesi
20 matches
Mail list logo
