[TUHS] Bugs in V6 'dcheck'

[scj at yaccman.com]

Doug is quite right that we were much more relaxed about security then. 
THe academic world was very much inspired at the time by the idea that one
could prove programs correct using mathematics.  I remember having some
very spirited arguments on this topic with some academics.  My argument
was roughly "Proving programs is stupid.  Look at Unix.  It's insanely
useful, but most programs can't be proved correct because they aren't!". 
Buffer overflow was one of the bugs I had in mind.  It was embarrassing in
later years to read about all the hackers exploiting these bugs.

In our defence, with only <= 64KB for programs and data and the slow
machines of the day, dynamic allocation and subscript checking were often
impractical...

Steve



> At 7:24 ?pm -0400 2014/05/31, Doug McIlroy wrote:
>>Does anybody
>>remember who published that deep new insight and/or where?
>
> Probably this:
> B.P. Miller, L. Fredriksen, So, B. "An Empirical Study of the Reliability
> of UNIX Utilities", Communications of the ACM 33, 12 (December 1990)
>
>
> --
> Kevin Schoedel <schoedel at kw.igs.net> VA3TCS
> _______________________________________________
> TUHS mailing list
> TUHS at minnie.tuhs.org
> https://minnie.tuhs.org/mailman/listinfo/tuhs
>