From quinlan@pathname.com Fri Feb  4 03:58:56 2005
From: Daniel Quinlan <quinlan@pathname.com>
To: discuss@lists.surbl.org
Subject: [SURBL-Discuss] rule for mixed case URI scheme
Date: Thu, 03 Feb 2005 18:58:47 -0800
Message-ID: <16898.58599.424614.185555@proton.pathname.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============5988604207789442526=="

--===============5988604207789442526==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit

Something close to this will be in 3.1, so you'll want to remove the
rule then, maybe name it something else too.

  uri URI_SCHEME_MIXED_CASE	/^(?![a-z]{3,6}:|[A-Z]{3,6})[A-Za-z]{3,6}:\//
  describe URI_SCHEME_MIXED_CASE	URI scheme has mixed uppercase and lowercase

The mass-check results are good, hits about 1% of spam with basically no
false positives in 92,000 hams from 6 people.

Daniel

-- 
Daniel Quinlan
http://www.pathname.com/~quinlan/

--===============5988604207789442526==--


From Robert@menschel.net Fri Feb  4 15:56:50 2005
From: Robert Menschel <Robert@menschel.net>
To: discuss@lists.surbl.org
Subject: [SURBL-Discuss] Re: rule for mixed case URI scheme
Date: Fri, 04 Feb 2005 06:56:12 -0800
Message-ID: <251294977.20050204065612@Menschel.net>
In-Reply-To: <16898.58599.424614.185555@proton.pathname.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============0337492178548450781=="

--===============0337492178548450781==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

Hello Daniel,

Thursday, February 3, 2005, 6:58:47 PM, you wrote:

DQ> Something close to this will be in 3.1, so you'll want to remove
DQ> the rule then, maybe name it something else too.=20

DQ>   uri URI_SCHEME_MIXED_CASE	/^(?![a-z]{3,6}:|[A-Z]{3,6})[A-Za-z]{3,6}:\//
DQ>   describe URI_SCHEME_MIXED_CASE	URI scheme has mixed uppercase and lower=
case

Thanks for the heads up.  WI'll drop this into the SARE uri.cf update
expected in around a week from now, and we'll migrate it to an x31
(don't use with SA 3.1) file when appropriate.=20

Bob Menschel




--===============0337492178548450781==--


From Robert@menschel.net Wed Feb  9 07:24:54 2005
From: Robert Menschel <Robert@menschel.net>
To: discuss@lists.surbl.org
Subject: [SURBL-Discuss] Re: rule for mixed case URI scheme
Date: Tue, 08 Feb 2005 22:24:42 -0800
Message-ID: <14410653805.20050208222442@Menschel.net>
In-Reply-To: <16898.58599.424614.185555@proton.pathname.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============3309354709170194756=="

--===============3309354709170194756==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

Hello Daniel,

Thursday, February 3, 2005, 6:58:47 PM, you wrote:

DQ> Something close to this will be in 3.1, so you'll want to remove
DQ> the rule then, maybe name it something else too.=20

DQ>   uri URI_SCHEME_MIXED_CASE	/^(?![a-z]{3,6}:|[A-Z]{3,6})[A-Za-z]{3,6}:\//
DQ>   describe URI_SCHEME_MIXED_CASE	URI scheme has mixed uppercase and lower=
case

DQ> The mass-check results are good, hits about 1% of spam with
DQ> basically no false positives in 92,000 hams from 6 people.

Found a ham hit in a corpus I'm analyzing:

<a href=3DNotes:///8525679900664A0D/38D46BF5E8F08834852564B500129B2C/991A65E0=
B32F1E3385256F94004AC059>Link</a>

I'm guessing that's an internal link used by Lotus Notes.  A lot of
private information in the email, so it's not one I'm able to submit,
and I'm still not ready to rejoin the nightly mass-check.

Bob Menschel




--===============3309354709170194756==--