Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 7 Feb 1997 11:09:19 +0900 (JST)
From:      Michael Hancock <michaelh@cet.co.jp>
To:        Warner Losh <imp@village.org>
Cc:        current@FreeBSD.ORG
Subject:   Re: How paranoid is Theo? (was Re: Karl fulminates, film at 11. == , thanks) 
Message-ID:  <Pine.SV4.3.95.970207105604.5960A-100000@parkplace.cet.co.jp>
In-Reply-To: <E0vsf8S-0007Q2-00@rover.village.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 6 Feb 1997, Warner Losh wrote:

> : Cool.  You can establish a contract of trust between called functions and
> : calling functions and not waste time fixing things that aren't
> : exploitable.
> 
> That's true to a point, but it takes a lot of time, sometimes, to
> establish that path of trust, especially in a large program.  And the

I see your point.  On the other hand, establishing that path of trust
would lead to more understanding.  (Yeah, I know, time is a luxury) 

> time isn't wasted (unless you are in firedrill mode) fixing those
> buffer overflows.  Your programs are just more robust :-).

Here are my rankings for three dimensions of software quality:

1) Correctness
2) Performance
3) Robustness

Fixing all buffer flows is:

1) Robustness
2) Correctness
3) Performance

In some cases, robustness is contrary to correctness.  But in this case I
guess it isn't so I won't say anymore and we can just agree to disagree.

Regards,


Mike Hancock




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.SV4.3.95.970207105604.5960A-100000>