From kzl1313 at citiz.net  Sat Nov 26 09:52:42 2005
From: kzl1313 at citiz.net (William)
Date: Sat, 26 Nov 2005 16:52:42 +0800
Subject: [spambayes-dev] Marketing plan
Message-ID: <20051126085241.81C451E4009@bag.python.org>


Dear spambayes-dev at python.org:

We specialized in Email Marketing.

1. Targeted e-mail list

We may provide targeted e-mail list. We can customize the list
as you need.

2. Send out Targeted e-mail Marketing for you

We may send your e-mail message to your target clients! We will
customize your e-mail list and send out your message for you.

* We offer BP Hosting & mail server.


Regards!

William
Service Team
kezunli321 at yeah.net

From skip at pobox.com  Sun Nov 27 04:31:05 2005
From: skip at pobox.com (skip@pobox.com)
Date: Sat, 26 Nov 2005 21:31:05 -0600
Subject: [spambayes-dev] [Spambayes-checkins] spambayes WHAT_IS_NEW.txt,
	1.39, 1.40
In-Reply-To: <20051127021542.2FBEB1E4007@bag.python.org>
References: <20051127021542.2FBEB1E4007@bag.python.org>
Message-ID: <17289.10361.718229.350758@montanaro.dyndns.org>


    Tony> SpamBayes has changed to use ZODB as the default database backend,
    Tony> rather than dbm (usually bsddb).

Should be "... rather than anydbm (usually bsddb)" I think.

I've never used/installed ZODB.  Will I still get a bsddb database by
default?

Skip

From tameyer at ihug.co.nz  Sun Nov 27 05:39:34 2005
From: tameyer at ihug.co.nz (Tony Meyer)
Date: Sun, 27 Nov 2005 17:39:34 +1300
Subject: [spambayes-dev] [Spambayes-checkins] spambayes WHAT_IS_NEW.txt,
	1.39, 1.40
In-Reply-To: <17289.10361.718229.350758@montanaro.dyndns.org>
References: <20051127021542.2FBEB1E4007@bag.python.org>
	<17289.10361.718229.350758@montanaro.dyndns.org>
Message-ID: <40325008-8FDA-4A28-8DBE-BA86DEBDDF7E@ihug.co.nz>

>     Tony> SpamBayes has changed to use ZODB as the default database  
> backend,
>     Tony> rather than dbm (usually bsddb).
>
> Should be "... rather than anydbm (usually bsddb)" I think.

Ok, I'll change it to that; thanks :)

> I've never used/installed ZODB.  Will I still get a bsddb database by
> default?

I was wondering about that when I checked it in :)  Would having it  
default to ZODB if importable, otherwise anydbm be too magical?

=Tony.Meyer


From skip at pobox.com  Sun Nov 27 16:23:45 2005
From: skip at pobox.com (skip@pobox.com)
Date: Sun, 27 Nov 2005 09:23:45 -0600
Subject: [spambayes-dev] [Spambayes-checkins] spambayes WHAT_IS_NEW.txt,
 1.39, 1.40
In-Reply-To: <40325008-8FDA-4A28-8DBE-BA86DEBDDF7E@ihug.co.nz>
References: <20051127021542.2FBEB1E4007@bag.python.org>
	<17289.10361.718229.350758@montanaro.dyndns.org>
	<40325008-8FDA-4A28-8DBE-BA86DEBDDF7E@ihug.co.nz>
Message-ID: <17289.53121.767656.79720@montanaro.dyndns.org>


    >> I've never used/installed ZODB.  Will I still get a bsddb database by
    >> default?

    Tony> I was wondering about that when I checked it in :) Would having it
    Tony> default to ZODB if importable, otherwise anydbm be too magical?

I would hope not.

Skip


From nas at arctrix.com  Sun Nov 27 21:07:20 2005
From: nas at arctrix.com (Neil Schemenauer)
Date: Sun, 27 Nov 2005 13:07:20 -0700
Subject: [spambayes-dev] Idea for multi-user spambayes
Message-ID: <20051127200720.GA19780@mems-exchange.org>

I have an idea for a spambayes variation that should be more suited
to multi-user systems.  The goal is to make the DB somewhat
conditionalized based on recipient address.  In addition to storing
<token>, spambayes could also save (<recipient>, <token>).  When
scoring a message, the probability for (<recipient>, <token>) would
be added to the evidence as well as for <token>.

I'm looking at chi2_spamprob() and wondering if this is valid,
statistics-wise.  Is there some better way to include the
(<recipient>, <token>) evidence?

BTW, if this idea actually works, using (<sender>, <token>) may also
be helpful.

  Neil

From tim.peters at gmail.com  Sun Nov 27 21:23:18 2005
From: tim.peters at gmail.com (Tim Peters)
Date: Sun, 27 Nov 2005 15:23:18 -0500
Subject: [spambayes-dev] Idea for multi-user spambayes
In-Reply-To: <20051127200720.GA19780@mems-exchange.org>
References: <20051127200720.GA19780@mems-exchange.org>
Message-ID: <1f7befae0511271223o1d90e14dna41c4c6313f8765c@mail.gmail.com>

[Neil Schemenauer]
> I have an idea for a spambayes variation that should be more suited
> to multi-user systems.  The goal is to make the DB somewhat
> conditionalized based on recipient address.  In addition to storing
> <token>, spambayes could also save (<recipient>, <token>).  When
> scoring a message, the probability for (<recipient>, <token>) would
> be added to the evidence as well as for <token>.

Offhand I think it would make more sense to ignore <token> when a
(<recipient>, <token>) pair (for the same <token> and the given
<recipient>) is known.  For example, if a urologist trains on "penis"
as ham, it's not doing him a favor to fold in that it's spam to almost
everyone else.

> I'm looking at chi2_spamprob() and wondering if this is valid,
> statistics-wise.

There's really no sense in which chi2_spamprob() computes "a
probability" -- it works or it doesn't.  Heh.

> Is there some better way to include the (<recipient>, <token>) evidence?

Test some ;-)

> BTW, if this idea actually works, using (<sender>, <token>) may also
> be helpful.

Spam sender addresses typically change rapidly, while ham sender
addresses typically don't.  So I expect this would add major boosts to
the tokens sent by ham senders, and typically create a ton of hapaxes
from spam senders (due to the spam <sender> addresses constantly
changing).

From tameyer at ihug.co.nz  Mon Nov 28 01:58:55 2005
From: tameyer at ihug.co.nz (Tony Meyer)
Date: Mon, 28 Nov 2005 13:58:55 +1300
Subject: [spambayes-dev] [Spambayes-checkins] spambayes WHAT_IS_NEW.txt,
	1.39, 1.40
In-Reply-To: <17289.53121.767656.79720@montanaro.dyndns.org>
References: <20051127021542.2FBEB1E4007@bag.python.org>
	<17289.10361.718229.350758@montanaro.dyndns.org>
	<40325008-8FDA-4A28-8DBE-BA86DEBDDF7E@ihug.co.nz>
	<17289.53121.767656.79720@montanaro.dyndns.org>
Message-ID: <B3452A9D-145D-41EB-96E3-D5CA0F0B072A@ihug.co.nz>

>>> I've never used/installed ZODB.  Will I still get a bsddb  
>>> database by
>>> default?
>
>     Tony> I was wondering about that when I checked it in :) Would  
> having it
>     Tony> default to ZODB if importable, otherwise anydbm be too  
> magical?
>
> I would hope not.

I've checked in this change.  I'll also soon checkin a fix for the  
pop3proxy/imapfilter users so that when they change their type in the  
web interface, it automatically converts the database for them,  
rather than crashing (which has come up a number of times on  
spambayes at python.org).

=Tony.Meyer