Re: "Bugs" (Was: Anyone could give a complete and yet small

[Ken Garlington <GarlingtonKE@lmtas.lmco.com>]

Larry Kilgallen wrote:
> 
> In article <32DA7CB4.456@lmtas.lmco.com>, Ken Garlington <GarlingtonKE@lmtas.lmco.com> writes:
> > Marin David Condic, 561.796.8997, M/S 731-93 wrote:
> >>
> >> Richard Riehle <rriehle@NUNIC.NU.EDU> writes:
> >> >  Following on the proposition stated by Mr. Maggio, we may ask the
> >> >  question, "Does all software contain mistakes?"
> >> >
> >>     Oh.. I don't know. It would seem to me that one could write a
> >>     sufficiently small program which contained no "mistakes" -
> >>     provided one was real careful about stating the requirements. Does
> >>     anyone doubt that a "perfect" program could be written to satisfy
> >>     the requirement: "Output the result of adding 2 and 2"?
> >
> > Output - in what base? to what device? :)
> 
> Those are minor details.
> 
> I can show any program to be bug-free if you let me do the specification
> after the fact, especially if I can continue to revise the specification.

This only says you can generate perfect _specifications_, not perfect
programs.
Surely you don't contend that a program is perfect just because it meets
its
specification?

> At worst, if the program is inconsistent from run-to-run, I can specify
> it as a pseudo-random number generator :-)

Better yet, classify it as a partial CPU test. That way, it doesn't even
have to
produce random results!

> 
> Larry Kilgallen

--
LMTAS - The Fighter Enterprise - "Our Brand Means Quality"
For job listings, other info: http://www.lmtas.com or
http://www.lmco.com