Bill Cole <[email protected]> writes:

> On 2026-01-21 at 10:50:51 UTC-0500 (Wed, 21 Jan 2026 10:50:51 -0500)
> Greg Troxel <[email protected]>
> is rumored to have said:
>
>> I just got an email from (last name replaced by doe in case this is
>> legit:
>>
>>   From: Netcom <[email protected]>
>
> It's not, and the supposed name is 'cooper' in my copy.

Matches mine and thanks for taking the time to write back.

> It was apparently sent via MailChimp (apparently now an Intuit
> company?) to everyone on this mailing list. e.g. it was sent to the
> unique address I use for this list.

Ah, interesting.  You would think spamming the SA community would appear
to be a bad idea.

>> Has anyone else been seeing spam via mailchimp, that is wrongly
>> getting lots of negative points?
>
> Not a lot, which is likely because I have custom local rules and score
> adjustments which result in rejection of most MC mail, including ALL
> of their garbage not sent to a tagged address.

Makes sense -- but perhaps if the default ruleset gives points to ISIPP
rules which certify mailchip, things should change so that people with
defaults have better outcomes?

>> I'll write to isipp.com support and see how that goes.
>
> Seems like the right choice.

I have written.

> You may wish to weaken the IADB "OPTIN" and "VOUCHED" rules, as they
> really are not very accurate.

Thanks; I intend to weaken those and the OPTIN_GT50.  An emitter having
more than 50% optin to me is an assertion that they are a spammer :-)

Are these scores determined by masscheck?  It seems they are too large.
My personal somewhat snarky take for now:

  ## adjust down "certified" scores (that certify spammers)
  # if 50% is optin, then the other 50% is spam!
  score RCVD_IN_IADB_OPTIN_GT50         1.0     # was -1.2
  # I'll allow 100% optin to be neutral (assuming GT50).
  score RCVD_IN_IADB_OPTIN              -1.0    # was -2.1
  # If you're paying to be vouched for, that's a clue.
  score RCVD_IN_IADB_VOUCHED            0.1     # was -2.2



It would be nice if there were a

  dnsbl_skip_domain_negative    *.mcsv.net

which would say, for that domain, ignore any negative score results.

Reply via email to