On Fri, Sep 21, 2007 at 10:24:40AM -0400, Steven Adair wrote:
Not in my book. I guess the people on this list are working off too many
different definitions of 0day. 0day to me is something for which there is
no patch/update at the time of the exploit being coded/used. So if I code
an exploit for IE right now and they don't patch it until April September
2008, it's a 0day exploit for a year. It's not necessarily new and it
doesn't have to be used maliciously.
If I code an exploit (for which there is no patch) and use it on my own
servers, does that mean it's not 0day? I don't think so. If my WordPress
blog gets owned by pwnpress, that's not 0day.. there's patches/updates for
everything on there. It just makes me an idiot for not upgrading. Now if
I get hit with some WP exploit that's not patched, then that's another
[0-day] story.
The reason malicious use before there's a patch is significant is that it
indicates a greater risk profile for users of the software in question.
If it's being actively used to compromise systems, you can't just sit
around waiting for a patch and expect to call your systems "secure" in
any sense of the term: you have to find a work-around, or remove the
vulnerable systems from the environment in which they're vulnerable
(normally, this means "the Internet"). That's why the term zero day is
important, and why it should not be misused to refer to something
demonstrated in a lab somewhere but not publicly disclosed in any detail.
That's why it's important to differentiate from exploits "in the wild"
and discovered vulnerabilities or proofs of concept.
--
CCD CopyWrite Chad Perrin [ http://ccd.apotheon.org ]
Baltasar Gracian: "A wise man gets more from his enemies than a fool from
his friends."
