From: "Igor Chudov" <[EMAIL PROTECTED]>
On Sun, Jul 23, 2006 at 03:33:03PM -0500, Igor Chudov wrote:
I started receiving a lot of spam in my mailbox. That spam regards one
of the most frequently spammed mede cations, with its name somewhat
misspelled in the Subject:. I am afraid that perhaps some of my rules
stopped working (like network identification of open spam relays).
It is strange. Anyone else experienced something similar?
Further investigation revealed the following. I run SA on a sitewide
basis from root's procmailrc.
Here's the tests from the root's run:
X-Spam-Checker-Version: SpamAssassin 3.1.3 (2006-06-01) on
manifold.algebra.com
X-Spam-Level:
X-Spam-Status: No, score=0.1 required=3.0 tests=BAYES_50,FORGED_RCVD_HELO, HTML_MESSAGE
autolearn=ham version=3.1.3
X-Spam-Relay-Country: US ES
If I rerun SA manually from my own account, by doing
spamassassin < /tmp/badspam 2>&1 | less
I get the following headers and proper identification of that spam as spam:
X-Spam-Flag: YES
X-Spam-Checker-Version: SpamAssassin 3.1.3 (2006-06-01) on
manifold.algebra.com
X-Spam-Level: *****
X-Spam-Status: Yes, score=5.6 required=3.0 tests=HTML_MESSAGE,URIBL_BLACK,
URIBL_SBL,URIBL_WS_SURBL autolearn=no version=3.1.3
So, when I run it manually, it works great. Why does it not work from
/etc/procmailrc?
I believe both procmail and spamassassin object to being run as root.
If there is a hole in either that opens the machine to malicious attacks.
It's believed there are no such holes. But our good fellows doing the
development are darned good but they're not Godlike enough to write
code that is absolutely guaranteed to be bug free. Besides, even if
their code is bug free are you willing to bet perl itself is?
{^_^}
