Philip Chee wrote:
On Thu, 13 Sep 2012 07:23:26 -0500, Richard Owlett wrote:
For more than a half century I've been a learn by experiment
learner. Before posting my initial message I had looked at
headers of messages that were tagged as possible scam and
those which were not. I spotted no difference. Afterwards I
did a little more experimentation to determine exactly what
SeaMonkey found objectionable.
SeaMonkey thinks the existence of a numeric URL in the body
of a message indicates a possible scam.
After reading your post, I sent myself two one line
messages. The body of one was "www.example.com". The body of
the other was "http://1.2.3.4";. [Obviously without quotation
marks]
[Note to new users following this thread - You can see the
complete headers either by using CNTRL-U while reading the
message or by a Right-click in message body and choosing
"Forward" in the menu.]
Currently our scam detection is a hard coded piece of code that looks
for a limited number of patterns including URLs using numeric IPs
e.g."http://1.2.3.4";. Thunderbird's code was originally similar but they
moved long ago to use the Firefox/Google "Safe Browsing" API. This
downloads regular updates from Google's safe browsing servers.
Phil, I find it interesting that Google has a "Safe Browsing" API, yet
on several of my Usenet groups, and even here on some of the moz groups,
a lot of the Spam seems to come via Google-Groups!!
--
Daniel
_______________________________________________
support-seamonkey mailing list
support-seamonkey@lists.mozilla.org
https://lists.mozilla.org/listinfo/support-seamonkey
