Re: [mod-security-users] behviour models

[Alexx A. <zm...@ya...>]

Kiran, are you developing application that use
positive security model? 

What appproach are you oging to use?

--- kiran k <kir...@ya...> wrote:

> Ivan:
>    
>   I am afraid,  dont understand design document,
> very little detail.  first approach says developer
> to publish the constraints, if known they would
> check them in the application itself and no need for
> WAF right ? 
>    
>   Again in the email thread below you will look int
> the audit logs. no need to go through the audit logs
> for each request, only once after training, you
> build rules out of historic data. How to correlate
> rules out of these logs is challenge. More info on
> the design please.
> 
>   
> Alexx Alexx <zm...@ya...> wrote:
>   Let me join your exciting discussion! 
> It's a topic I'm very intrested in!
> 
> --- Ivan Ristic wrote:
> 
> > I am afraid I don't understand your question. You
> > can read about my
> > ideas here:
> >
>
http://www.modsecurity.org/blog/archives/2005/11/positive_securi.html
> > (but that's not implemented). Also read this
> >
>
http://www.cs.ucsb.edu/~vigna/publications/2005_kruegel_vigna_robertson_CN05.pdf
> 

----------------
Best regards, 
Alexander

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com