Re: xkcd: CrowdStrike

Mark Jackson  <mjackson@alumni.caltech.edu> wrote:

>
Well, absent a countervailing force the capitalist imperative
discourages carrying the cost of robustness, and eventually eliminates
it entirely.  Do you have a suggestion other than regulation?

The government could regulate.  But on the other hand, the government
also could put money into development of reliable computing systems
and code verification techniques.  But more importantly, into transferring
that information into the hands of people willing to make actual products.

We've had actual verified kernels since the seventies, although for
very limited applications (and having to use interrupts makes everything
much much harder... Honeywell avoided a lot by avoiding interrupt-driven
I/O at the expense of performance).  Don't even get me started about the
iAPX 432 which was a bad system with some great ideas that were never
carried forward.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."