I am using Spam Assassin on my ISP as well as Spam Sieve on my laptop.
Spam Assassin is not very aggressive (it flags spam correctly nearly
100% of the time, but lets about 50% of the spam thru). The advantage of
using Spam Assassin is that it reduces the amount of spam I have to
download when traveling by 50% (I download and review it by hand only
when I have a high-speed connection).
Spam Sieve is really good, but it has consistently failed to recognized
"joe jobs"--where a spammer is using a return address at my domain, and
there are automatic replies to that address from the intended addressee.
I have special filters which seem to do a good job rejecting that source
of unwanted email.
To deal with .gif email I have a rule in PowerMail:
If Spam Rating is high
Move message into folder "Spam"
Don't notify new message
Don't show in Recent Mail
=> Move attachments to trash <=
Don't index message content
Don't apply subsequent filters to this message
The only problem with this is that PowerMail has a bug/feature lack that
once you have said "Don't index message content", there is no way to
later index the message if you identify the email as legitimate. (Other
than resetting the index, which I do every few months).
Please don't ever "bounce" spam--it only adds to the frustrations of us
"Joes" who have been "Jobbed".
A
On Mon, 29 Jan 2007 13:01:14 -0800 Richard Hart said:
>(1) Does any hosting service or ISP offer you the capability to bounce
>spam? I know many offer excellent SPAM sequestering tools, even deleting
>offending messages so you never have to see them.
>
>But what I'd like to do is have my mail server bounce messages from
>specific IP addresses. For this service, I would switch mail service
>providers in a heartbeat.
>
>(2) What is the best way to deal with the relatively recent phenomenon
>of GIF spam. You know the ones. The Subject and body are random text.
>The attachment is a graphic containing the spam pitch. I haven't been
>able to derive a common identifier for them in the headers, and they are
>not easy to define in a spam rule.
>
>Richard Hart
>
>
>
>
