So TPTB at work here would like some insight
into how text is extracted from a submission
for a signature. Specifically, the Email.Trojan-xx
signatures. Of the handful of viruses I submitted,
two are causing false positives. Email.Trojan-48 is
the text of a legal disclaimer belonging to our
customer who reported it to us as spam. It is not a
part of the original virus email, which was attached
and contained a Trojan.Downloader.Agent-1297 virus.

Likewise, Email.Trojan-36 is a quote from a customer's
.sig, not a part of the virus email. Email.Trojan-37
is the text of the viral email, along with Trojan.Autorun-287
to cover the executable part of the email.

Two questions then:

1) Is there an automated process for generating text
signatures or does a human look at it (if you'd rather
not reveal that, I can understand, I just need to tell
them that)? It would seem fairly evident that a .sig or
disclaimer that occurs before the viral email that's an
attachment should not be treated as virus material.

2) Can you remove the Email.Trojan-36 and Email-Trojan-48
signatures?

Thanks. I'll strip out customer stuff from now on before
submitting, just so there's no misunderstanding.

-- 
Brian Bebeau
Trustwave
http://www.trustwave.com

_______________________________________________
Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net
http://www.clamav.net/support/ml

Reply via email to