[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [WEB SECURITY] Cross Site Scripting
- From: Francois Larouche <francois.larouche-ml@xxxxxxxxxxxxxxxxxxxx>
- Subject: Re: [WEB SECURITY] Cross Site Scripting
- Date: Thu, 20 Sep 2007 12:35:29 -0700
I don't entirely agree with you
Always encode output in case that input validation failed OR the data
comes from another source that you have no control (sharing a DB for
instance).
Besides as many mentioned black list is a limited system. To add to
this, nowadays it's trivial with a human launching and monitoring a set
of signatures against a parameter to find the loop hole in the black
list. I'm sure that many out there has already written such a script or
tool. If in other hand someone is trying by hand all the signatures it
will then entirely depend on its expertise and patience.
I do believe that black lists are just raising the bar... Nothing else
nothing more. First step to a safer world I guess.
As for being alerted that you are under attacks under a few days... If
you use black list it doesn't mean you have a IDS or whatever system
that detects attack, or even that someone will monitor it 24/7 or that
once he sees that something is going on it's not already too late. I
wouldn't bet my business on that, would you?
As a final note, I entirely agree that the white listing is one of the
best solution _when_ you can apply it on simple data that always follows
the same pattern. (ZIP code, Integer, tel number, postal code, and so
on) But it can be a nightmare for the developers to keep that white list
valid when it comes to more complex inputs. The worst problem might
occur when the white list is now blocking valid data. For example, a
white list refuse to accept purchases of more than 9999$ (4 digits) but
your business has grown and now you have purchases ranging over this. I
don't think the CEO will be happy to learn that the company has lost
those sells just because of a over zealous developer... But I guess you
can imagine the problem.
My two cents...
Francois
Agree...
1. input validation
2. Output encoding if we have to accept malicious characters (that's a
real problem of web site social network - too large acceptable range
of characters)
3. Blacklist...
By experience, the advantage of blacklist is not protection against
XSS (we can bypass this kind of filter easily) BUT don't minimise the
advantage of blacklisting...
Blacklisting is a very powerfull detection system... Often hackers try
patterns before finding an issue...
We can expect that the blacklist can detect some issue.
You will be alerted that you will be under attacks under a few days...
----------------------------------------------------------------------------
Join us on IRC: irc.freenode.net #webappsec
Have a question? Search The Web Security Mailing List Archives:
http://www.webappsec.org/lists/websecurity/
Subscribe via RSS:
http://www.webappsec.org/rss/websecurity.rss [RSS Feed]
Brought to you by http://www.webappsec.org
Search this site
|