Soeren Gerlach writes:
> > On Fri, Sep 05, 2003 at 08:55:26PM +0200, Soeren Gerlach wrote:
> > > Hi,
> > >
> > > > Changes since 2.5x:
> > > > [...]
> > > > - new Bayes tweaks -- tokenization of partial address and URI elements
> > >
> > > Are details available about this feature?
> >
> > Basicall
I've recently integrated a third-party program into SpamAssassin. I found
the best way to do this (in my case) was to create a program to run the 3rd
party app, take the results and add them into a X-Header and append to the
message. Using SA it's simple to create a rule based on X-headers. This
> On Fri, Sep 05, 2003 at 08:55:26PM +0200, Soeren Gerlach wrote:
> > Hi,
> >
> > > Changes since 2.5x:
> > > [...]
> > > - new Bayes tweaks -- tokenization of partial address and URI elements
> >
> > Are details available about this feature?
>
> Basically URIs are broken into
> - hostname
> - "wo
On Fri, Sep 05, 2003 at 08:55:26PM +0200, Soeren Gerlach wrote:
> Hi,
>
> > Changes since 2.5x:
> > [...]
> > - new Bayes tweaks -- tokenization of partial address and URI elements
>
> Are details available about this feature?
Basically URIs are broken into
- hostname
- "word" to
Hi,
> Changes since 2.5x:
> [...]
> - new Bayes tweaks -- tokenization of partial address and URI elements
Are details available about this feature? I'm asking because I'm currently
thinking about something new like DCC, but especially and only for URLs in
spam. Eventually this would become obsol