If I get spam to my local e-mail client via POP, it got into my
webmail's Inbox folder since that's the only one that my POP client can
see. So, like you, I then use the webmail client to log into my account
and report that spam to Comcast to get their anti-spam filter updated.
Been so long since I got a spam into my webmail Inbox to get into my
local POP e-mail client that I can't remember when I had to do this last
time. I use to report spam to Spamcop, too, but now I only get about 1
a month, if that. I like not getting spam but it was actually fun to
report the crap.
However, I have added a few server-side rules (i.e., rules in the
webmail client), like:
- "discard - blacklist", action = discard (disappears from all folders)
These are repeat offenders from whom I don't want their e-mails.
Often they are from businesses I deal with that send me their spam,
like Tracfone who sends me offers, @
news.efax.com who sends me offers
because I have a free eFax account, and @
comcast.delivery.net which
are offers from Comcast (but the boobs don't even source it from their
domain and instead contract with
delivery.net to send this crap). In
total, I have just 4 blacklisted senders which has stay consistent for
a couple years.
- "keep - passcode in Subject", action = e-mail stays in Inbox
If a special string (passcode) is in the Subject header, keep that
e-mail. This is a unique string that is very unlikely to ever appear
in any e-mail that I will receive. It is something I tell very
trusted senders to add into the Subject header to ensure they get past
my spam filters (server- and client-side). It's like giving those
senders a skeleton key. If anyone abuses that trust, I change this
rule to use a different password. The abuser can't use the old one
anymore, has to run the gauntlet of spam filters (assuming they get
past the blacklist rule since it's likely their abuse means e-mail
communications get cutoff from them).
- "keep - safe sender", action = leave in Inbox
Unlike many local e-mail clients that let you maintain a list of known
good (safe) senders, Comcast's webmail doesn't have that. So I have
to use a rule to list my safe senders.
- "trash - no @ in From", action = delete the e-mail
If the sender doesn't provide an e-mail address in the From header, I
don't want their e-mail. While RFC standards require an e-mail
address, rare few SMTP server bother to check syntax.
- "trash - me in From", action = delete e-mail
My account's e-mail address is in the From header. I don't send
myself any e-mails. Those are from spammers pretending to me sending
e-mails to myself.
I used to have the following server-side filter:
- "Subject is blank", action = move to Spam folder
Test for Subject header NOT to have any characters A-Z, 0-9. That was
an old spam trick to make you read the body of an e-mail since you
might have the Preview pane disabled. Too often the Subject divulged
it was spam even when they tried to use one that wasn't spammy.
However, it's been so long that spammers used that trick that I removed
this filter. More often I have good senders that wrote the body of
their e-mail and forgot to add a subject.
I don't want to add too many server-side filters as this becomes too
aggressive. I'd end up having to revisit the webmail client too often
checking for false positives. I have more rules in my local e-mail
client but those move suspect e-mails into the Junk or Deleted Items
folders. Basically I take a subset of the rules in my local e-mail
client to use server-side in the webmail service. By letting the server
do the work, I can disable those rules (instead of delete them since I
may want them later) so my local e-mail client runs faster because there
is a reduced number of rules to exercise against every received e-mail.
When possible, I let the e-mail server do as much filtering as possible
but not get so aggressive that I end up have to use the webmail client
to check for false positives. If I have to check for those, I need to
reduce or modify my server-side rules.
Of course, the best means to keep spam from hitting your Inbox is to NOT
dole out your e-mail address to every joker that asks for it. Use
aliases instead (e.g., Spamgourmet). I don't even have to log into
Spamgourmet to dole out an alias. I can make them up on-the-fly using
no software, just some basic syntax rules on how to compose an alias
that you give to an untrusted, unknown, or temporary sender. Use a
*unique* alias with each sender. Then if that alias gets abused with
spam, you know exactly who betrayed you. My aunt doles her Hotmail
address out to everyone that asks for an e-mail address. She gets lots
of spam. I rarely give out my true e-mail address (maybe a dozen people
have it) and almost always give out an alias. If the alias hasn't been
abused in 6-8 months then I'll consider updating my account or contact
info with the sender to give them my true e-mail address; however, I
have aliases that have been defined for years with the same senders
although they are known but not fully trustworthy. If an alias gets
abused or was only temporarily, I just login to Spamgourmet to delete it
(well, you cannot delete aliases there but you can zero out how many
more e-mails get through an alias - if the receive count is zero, any
further e-mails to that alias go to the bit bucket).