SANS Institute - CWE/SANS TOP 25 Most Dangerous ProgrammingErrors

[vanderaj at owasp.org (vanderaj vanderaj)]

Tom,

> From the business' point of view, they really don't care if widget X
has weaknesses, they want to know how to make money by buying and
using widget X. They assume X is safe by default, even though it's
not. They've been doing fast and crappy for so long, and made heaps of
money from it, that it's a hard sell in many places to do the safer
thing until the horse has bolted.

The only examples where folks buy widget X over widget Y is those
folks in operational risk who have to make a financial allowance for a
probable risk difference between X and Y. For example, if one
satellite launch system blows up one time every four launches, and
another blows up one time every eight launches, you'd go with the
second or you'd have to budget for the likelihood of having to replace
your satellite a bit more with the first one.

In our industry, we have still yet to make a compelling, measurable
and thus believable case that there's a TCO benefit from buying more
expensive, but safer software. Most folks believe all software is
safe, despite the fact that it is not. Until that time, CWE and
similar *weakness* patterns are a derivative of the actual cost of
ownership, and not the actual benefits.

That's why I've gone gung ho into "build it right the first time"
mode. I doubt we'll get the accurate metrics required for proof that
safer software is cheaper (over time), so it's best that we simply get
safer software - period. That's why I will be working with the
frameworks and code repositories rather than the 0day crowd. In my
view, there is zero value in vulnerability disclosure, discussion, or
discovery. It's like shooting fish in a barrel.

thanks,
Andrew

> Will business start to talk CWE as they already talk CVE?
>
> Discussion/Debate/Thoughts
>
> Tom Brennan