----- Original Message ----- 
From: Matt

> I think that I've pointed out the caveats many times
> over on blocking with SBL.  SBL is though more
> accurate than my system as a whole, and I have
> never seen a true false positive with it.

> I've asked this several times; has anyone ever seen
> a false positive with SBL?  I've not ever received a
> single reply to that question, though this is the 3rd
> time I've asked it now.

Because people didn't respond doesn't mean anything.  All RBLs produce
false-positives.  How could they not, they are run by humans.

> I think your advice is well founded, however it is a
> generalization and exceptions may apply.

There are no exceptions when it comes to anything run by humans, there WILL
be errors.

Just from yesterdays logs, legitimate mailing list messages blocked by SBL:
20 Subject: RE: [MS SMS] What are YOU doing to remove spyware? 2004 Edition
19 Subject: RE: [MS SMS] OT: Football
12 Subject: RE: [myOT] Alias
10 Subject: RE: [MS SMS] SMS 2003: WMI
9 Subject: RE: [MS SMS] Installing a DP over the wire.....
9 Subject: RE: [myOT] MMS 2004
7 Subject: RE: [MS SMS] VBS Question
6 Subject: RE: [myOT] Stargate season opener tonight...
5 Subject: RE: [MS SMS] Central Site
5 Subject: RE: [MS SMS] Error in scan tool
4 Subject: RE: [MS SMS] ROI for 2003
4 Subject: RE: [MS SMS] SMS 2: Clients failed to connect to APM server
4 Subject: RE: [MS SMS] XP clients
3 Subject: RE: [MS SMS] SMS and Tablet PC
3 Subject: RE: [MS SMS] SMS on VMWare
3 Subject: RE: [MS SMS] SMS2003 - How to re-trigger advertisement on client
2 Subject: RE: [MS SMS] MakeColl.exe for SMS 2003?
2 Subject: RE: [MS SMS] OT: Anyone from the UK going to the MMS?
2 Subject: RE: [MS SMS] OT: Read Receipts on List messages
2 Subject: RE: [MS SMS] SMS SUSFP Updates
2 Subject: [MS SMS] XP clients
1 Subject: RE: [myOT] New Bill and Monica pics...
1 Subject: RE: [MS SMS] advanced client prestaging
1 Subject: RE: [MS SMS] FW: Clarification on recent email from Shavlik
Technologies
1 Subject: RE: [MS SMS] FW: Clarification on recent email from Shavlik
Technologies
1 Subject: RE: [MS SMS] FW: Clarification on recent email from Shavlik
Technologies
1 Subject: RE: [MS SMS] Security scan tool upgrade
1 Subject: RE: [MS SMS] SMS 2003 MP problem
1 Subject: RE: [MS SMS] SMS Office Updates
1 Subject: RE: [MS SMS] SUS Distribute software updates wizard doesn't show
up
1 Subject: NWCYCLING:  FW: 2004 Mt. Hood Cycling Classic
1 Subject: NWCYCLING:  USCF Rulebook 2004 Changes - online
1 Subject: NWCYCLING:  WSBA Junior Informational Meeting Tonight
1 Subject: MEDITECH Issue PP #3714704 - Open
1 Subject: [partb-l] HCPCS codes
1 Subject: [NPinfo] Interesting article on the physician shortage.
1 Subject: [myOT] Test
1 Subject: [myOT] Alias
1 Subject: [MS SMS] VBS Question
1 Subject: [MS SMS] SUS Distribute software updates wizard doesn't show up
1 Subject: [MS SMS] SMS2003 - How to re-trigger advertisement on client
1 Subject: [MS SMS] SMS Office Updates
1 Subject: [MS SMS] SMS and Tablet PC
1 Subject: [MS SMS] SMS 2003: WMI
1 Subject: [MS SMS] SMS 2003 Bug
1 Subject: [MS SMS] ROI for 2003
1 Subject: [MS SMS] Query help needed
1 Subject: [MS SMS] OT: Read Receipts on List messages
1 Subject: [MS SMS] OT: Guest Account
1 Subject: [MS SMS] OT: Anyone from the UK going to the MMS?
1 Subject: [MS SMS] MakeColl.exe for SMS 2003?
1 Subject: [MS SMS] Installing a DP over the wire.....
1 Subject: [MS SMS] Holy Replicating Servers Batman!
1 Subject: [MS SMS] Couple backup questions
1 Subject: [MS SMS] Central Site
1 Subject: [MS SMS] Adobe Acrobat 6 Deployment

I also found at least a dozen personal messages that were flagged by SBL,
but were delivered anyway because of the way we weight our tests.  Again,
this is just from yesterday.

Instead of applying a huge weight to a single test, why not apply a small
weight to may tests?  That way you at least get corroboration from multiple
tests, thus negating the human factor.

Bill

---
[This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]

---
This E-mail came from the Declude.JunkMail mailing list.  To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type "unsubscribe Declude.JunkMail".  The archives can be found
at http://www.mail-archive.com.

Reply via email to