> A few more tests later, I have found the SA results of 1 to be spam,
> and my global.cfg line scores a hit at 3, this 3 gets added to the
> result of SA's 1 and therefore I get 4 points.
There should be no _addition_ of Declude-defined weights to external
test result codes! I don't even know how your test definition could be
doing that. Certainly, that's not the intended use of SPAMC32.
What's confusing you seems to be the way of pairing Declude's external
test types ('external nonzero' or 'external weight') with the related
SPAMC32 options (default or '-e').
A Declude 'external nonzero' test is designed for tests whose return
codes you want Declude to convert to binary pass/fail-- it's either
'pass' (return code 0) or 'fail' (return code 1+). [A test that only
returns a 0 or 1 (never 1+) can alternately be defined as an 'external
1' test, but I think most people just use the catchall 'nonzero'
type.] After you define a a test as nonzero, from the manual:
> You can add weights, by adding two numbers at the end of the line:
> the first one is the standard weight (the weight if the test fails),
> the second is the "negative weight" (the weight if the test does not
> fail, usually 0)
On the other hand, a Declude 'external weight' test is designed for
tests whose return code you want to reuse _as_ the Declude weight for
the test. This is designed for tests that return a range of values
that represent increasing spamminess, such as SPAMC32 with the -e (and
optionally the -et) option. In these cases, you're passing the SPAMD
weight, up to the -et threshold, directly back to Declude, which is
what you seem to want to do.
> Hummm... I was hoping to get the real score of SA, max of 10, and
> score that value against the email message for the weighting system.
> So, for now, it's ham or spam :)
Only because you haven't defined the test correctly in Declude. :)
The -et switch on its own is useless, because that just sets an upper
threshold if -e is in use; -et requires -e, but -e does not require
-et (I'm not going to change the behavior at this point, but I
understand why you thought that -et implied -e.) You need both.
And if you want to get the SPAMD score in raw form, don't think about
-lt or -ht. These switches set client-side thresholds that allow you
to override the server-wide spam/ham threshold set on a central SPAMD
server. But you're already overriding SPAMD's spam/ham decision by
passing the raw SPAMD result without any simplification. The -lt and
-ht are useful if multiple SPAMC32s are running against the same SPAMD
rulesets and need different sensitivities (for example, if you're
running SPAMC32 from different client scanning boxes at a hosting
provider). You don't need them.
--Sandy
------------------------------------
Sanford Whiteman, Chief Technologist
Broadleaf Systems, a division of
Cypress Integrated Systems, Inc.
e-mail: [EMAIL PROTECTED]
SpamAssassin plugs into Declude!
http://www.imprimia.com/products/software/freeutils/SPAMC32/download/release/
Defuse Dictionary Attacks: Turn Exchange or IMail mailboxes into IMail Aliases!
http://www.imprimia.com/products/software/freeutils/exchange2aliases/download/release/
http://www.imprimia.com/products/software/freeutils/ldap2aliases/download/release/
---
This E-mail came from the Declude.JunkMail mailing list. To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type "unsubscribe Declude.JunkMail". The archives can be found
at http://www.mail-archive.com.
