Re: data-destroiny malfunction: is that a "security" issue

["Steven M. Christey" <coley () linus mitre org>]

I don't think this is a security issue either, at least not in terms of
CVE coverage or how I define "vulnerability."  For there to be a
vulnerability, there needs to be some "attacker" role.  The attacker can
either actively force the error (as in a classic buffer overflow attack
against a server), passively benefit (as in an accidental information
leak), or perform some social engineering attack that has a high
likelihood of succeeding for a user during typical operations (e.g.
accepting a private message in an IRC client).

Some people might have very generic notions of vulnerability that include
inadvertent things that trusted people do to themselves, that don't
benefit any external attacker.  However, it seems that the people who
think this way are operating in environments with extremely high demands
for confidentiality, integrity, and availability.

The example you give demonstrates how sometimes, a plain ol' bug can be
much more serious than a security issue, depending on the priorities of
the victim.

- Steve