Re: [Asrg] Spam button scenarios

Martijn Grooten <> Mon, 08 February 2010 20:49 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 07EAB3A7255 for <>; Mon, 8 Feb 2010 12:49:24 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -5.136
X-Spam-Status: No, score=-5.136 tagged_above=-999 required=5 tests=[AWL=1.307, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, SUBJECT_FUZZY_TION=0.156]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id KYvoA27Tx2hq for <>; Mon, 8 Feb 2010 12:49:23 -0800 (PST)
Received: from ( []) by (Postfix) with ESMTP id 2090D3A70AF for <>; Mon, 8 Feb 2010 12:49:23 -0800 (PST)
Received: from (localhost.localdomain []) by localhost (Postfix) with SMTP id 4A960700D3 for <>; Mon, 8 Feb 2010 20:50:26 +0000 (GMT)
Received: from ( []) by (Postfix) with ESMTP id ED7C9700A4 for <>; Mon, 8 Feb 2010 20:50:25 +0000 (GMT)
Received: from ([fe80:0000:0000:0000:e1bd:d3c1:]) by ([]) with mapi; Mon, 8 Feb 2010 20:50:25 +0000
From: Martijn Grooten <>
To: Anti-Spam Research Group - IRTF <>
Date: Mon, 8 Feb 2010 20:50:24 +0000
Thread-Topic: [Asrg] Spam button scenarios
Thread-Index: Acqo/LA9EgouJYFWRZOSQreWZvdEBwAAJAnQ
Message-ID: <>
References: <alpine.BSF.2.00.1002080111310.16135@simone.lan> <> <>
In-Reply-To: <>
Accept-Language: en-US, en-GB
Content-Language: en-US
acceptlanguage: en-US, en-GB
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: Re: [Asrg] Spam button scenarios
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: Anti-Spam Research Group - IRTF <>
List-Id: Anti-Spam Research Group - IRTF <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Mon, 08 Feb 2010 20:49:24 -0000

> > One way to look at this is: Which choice benefits the user most?
> >
> > Since the Yahoo mail is being forwarded to Gmail, I would think it
> > would be more valuable to have Gmail doing the spam filtering.
> Why?  If _either_ of them blocked it, the user wouldn't see it.
> Training Yahoo helps the user if he later switches to having someone
> else pick up his mail from Yahoo.  Training google doesn't.
> Training google to see mail forwarded from Yahoo as spam can have
> adverse effects for him.

Firstly, I think the scenario didn't talk about forwarding, but rather about fetching email using POP3 (or possibly IMAP). Secondly, if Gmail did receive forwarded mail and then considered a click on the TiS button (which has always been there in Gmail) to automatically mean "this IP address is a spammer", then it's doing something wrong. (And given then Gmail itself can be set easily to forward email, they're unlikely to act this way.)

> So it's better to train Yahoo.  (Now, if you believe that google
> trains much better than Yahoo, that's an important difference.)

But the only thing your MUA can be sure about is where it received the email from (Gmail's IMAP). It may have a good reason to think that Gmail fetched the email from Yahoo, but it has to trust Gmail for that. Which might be okay if Gmail is really Gmail and Yahoo is really Yahoo, but not if Gmail is a random IMAP server $x and Yahoo is a random POP3 server $y and you as an end user haven't explicitly told the MUA how it receives email.

So I think the only right thing to do is to send the report to Gmail.

Gmail (or $x) then knows that (and how) it received the mail from Yahoo (or $y) and hopefully knows it is best to forward the report.


Virus Bulletin Ltd, The Pentagon, Abingdon, OX14 3YP, England.
Company Reg No: 2388295. VAT Reg No: GB 532 5598 33.