What about "only" checking corpus files for DKIMWLAddresses or
DKIMNPAddresses matches only if AddDKIMHeader is enabled?  It would just
need to match the regex against the X-ASSP-DKIMidentity.

notspam would be checked against DKIMNPAddresses and spam would be checked
against DKIM NP and WL.  If a server's processor can handle it at rebuild
time, wouldn't this be a good thing to have as an option?

I don't know why we'd need to check denySMTPConnectionsFromAlways or
denySMTPConnectionsFrom during the rebuild process.   blackListedDomains,
redRe would be nice, but again, we need to balance performance here....

On Tue, Oct 26, 2021 at 3:36 AM Thomas Eckardt <thomas.ecka...@thockar.com>
wrote:

> >*without rebuild taking too much of a performance hit?*
>
> fastest case: all in the file model
>
> slowest case:
> - no file model
> - checking whiteRe, whitelist, npRe, DKIMWLAddresses, DKIMNPAddresses,
> redRe - for assp/spam
> - checking denySMTPConnectionsFromAlways, denySMTPConnectionsFrom,
> blackListedDomains for assp/notspam
>
> The slowest  case is 12 times slower than the fastest.
>
> Thomas
>
>
>
>
>
>
>
>
> Von:        "K Post" <nntp.p...@gmail.com>
> An:        "ASSP development mailing list" <
> assp-test@lists.sourceforge.net>
> Datum:        24.10.2021 02:40
> Betreff:        [Assp-test] Concept question: At rebuild, look at
> DKIMWLAddresses?
> ------------------------------
>
>
>
>
> *Would it be possible to have ASSP consider DKIMWLAddress matches during
> rebuild, removing matching messages from spam, without rebuild taking too
> much of a performance hit? *
>
> During rebuild, ASSP runs rb_whitelisted against each message in the spam
> corpus, and if a match is found against the whitelist, that message is
> removed from spam.  (right?) It's a terrific way to help keep the corpus
> clean after a whitelist addition.
>
> I rely heavily on DKIMWLAddresses - it's super helpful to consider a
> message whitelisted *only* when the DKIM signature matches.
>
> *If it's realistically possible and not ill conceived, removal of messages
> from spam where there's a DKIMWLAddress match would further clean up spam,
> and lead to more accurate HMM/Bayesian detections. *
>
> What do you think?
>
> Along the same lines, what about considering messages that match no
> processing rules: the regexes and DKIMNPAddresses for messages in both spam
> and notspam?
> _______________________________________________
> Assp-test mailing list
> Assp-test@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/assp-test
>
>
>
>
> DISCLAIMER:
> *******************************************************
> This email and any files transmitted with it may be confidential, legally
> privileged and protected in law and are intended solely for the use of the
> individual to whom it is addressed.
> This email was multiple times scanned for viruses. There should be no
> known virus in this email!
> *******************************************************
>
> _______________________________________________
> Assp-test mailing list
> Assp-test@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/assp-test
>
_______________________________________________
Assp-test mailing list
Assp-test@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/assp-test

Reply via email to