Thomas,
> nice idea - but : what, if there is a large number of URI's
> in the mail -
> depending on your setting of 'URIBLmaxdomains' the
> 'X-Assp-Detected-URI:'
> heaeder lines could easy oversize the header in a way that
> the MTA behind will reject the mail?
Yes, that is a negative for my method. See Grayhat's suggestion, which I
like.
> If you use the ASSP-OCR Plugin, the
> URIBLok sub will
> be called twice (again for the extracted OCR data).
Yes, that would be bad if one uses the plugin. I do not use the plugin now,
but might want to in the future.
> The added lines are depending on the setting of 'URIBLLog' !??
That is what I thought the "if" part after the mlog was checking.
> $this->{myheader}.=sprintf("X-Assp-Detected-URI:
> %s\r\n",$uri) if (! exists $domains{ lc $uri } && $URIBLmaxdomains &&
> scalar keys(%domains) <= $URIBLmaxdomains ) ;
>
> I think this will do it better - but it is not a solution for the
> 'oversize header' problem if 'URIBLmaxdomains' is set to a
> large value.
I think I understand the code above.
> You've added the code at a place where URI's are registered
> for checking -
> so even good and bad URI's are added to 'myheader'. Would it
> be better to
> only add header lines for bad URI's?
Yes, for my case, I want to see both good and bad. Any bad URI listed at a
trusted URI would have blocked the message. So, any bad URI were marked bad
by a URIBL that was not fully trusted. My concern is only for messages that
passed ASSP, but on manual inspection are identified as spam. Now I want to
take those URI and add them to a URIBL. See my suggestion to Grayhat, where
if this was implemented, I suggested three checkboxes and two headers, which
I will repeat for you later in this message.
Also, the location where I added the code can repeat the same URI many
times, which is not desirable. But for me, was simple code to add, as I can
barely write a line of perl code. I looked at two other locations to add the
code, but was not sure I could add the headers from those locations:
1. In sub lookup, at the "sending DNS(TXT)-query" could avoid multiple of
the same URI and outside the DNS server loop could avoid an entry for each
DNS server. But I did not see how I could add to headers from there.
2. In sub lookup, at the "Commencing $type checks" could avoid multiple of
the same URI. But I did not see how I could add to headers from there.
In particluar, in #1 and #2, I did not see $this that was valid and
available. So, I added it where I was capable.
As a side note, mlog uses "mlog(0," instead of "mlog($fh," so logged lines
do not include the message identity. Something I intended to check when I
had upgraded to the latest version. Currently still running X-Assp-Version:
2.0.2(1.2.26).
As I am splitting my attention between this message and the Patriots vs
Dolphins game, I hope this message makes sense :).
Michael Thomas
Mathbox
978-687-3300
Toll Free: 1-877-MATHBOX (1-877-628-4269)
------------------------------------------------------------------------------
Learn how Oracle Real Application Clusters (RAC) One Node allows customers
to consolidate database storage, standardize their database environment, and,
should the need arise, upgrade to a full multi-node Oracle RAC database
without downtime or disruption
http://p.sf.net/sfu/oracle-sfdevnl
_______________________________________________
Assp-test mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-test
