product liability (was 'we should all be uncomfortable with the extent to which luck..')

Michael Airhart mairhart at cisco.com
Wed Jul 25 21:21:04 UTC 2001


Not in anything I would pilot or ride in!!!!

Think about the Class 5 switches and IBM mainframes.  It took YEARS to get 
anything done because of the critical nature of their functions.  If people 
are going to put their lives in the balance, then they should spend the 
time to research their choices.  People who just toss up a web server, run 
their business, and then complain when it crashes obviously didn't take it 
too seriously.  Same with networks.  It costs major $ to assure no single 
point of failure and quit frankly most people won't spend the $..  Or they 
can't recoup the costs by selling service..  It's not easy, I understand that.

Look at the traffic controller industry.  There is a significant amount of 
work to assure that you NEVER get a double green / crossing traffic 
situation.  The technology in there is 30+ years old.  If the 68000 chip 
croaks, the old tech relays prevent light violations.

Man we are WAY off any operational topic here.  My apologies to the group 
trying to get some work done here..

Michael (Speaking for myself ONLY)

At 02:03 PM 7/25/2001 -0700, you wrote:
>On Wed, 25 Jul 2001, Michael Airhart wrote:
> > The analogy is further flawed in that the comparison is between problems in
> > the real world that can cause real injury or death to real people versus a
> > piece of software that operates in a virtual world...
>
>m$ stuff is being used in flight control systems now, if I remember
>correctly. the navy also had a spin using nt in a "smart ship" trial, much
>to their detriment.
>
>-Dan
>
>--
>[-] Omae no subete no kichi wa ore no mono da. [-]




More information about the NANOG mailing list