Everyone who writes safety critical software should read this

[Joseph Rushton Wakeling]

On Wednesday, 30 October 2013 at 00:28:28 UTC, Brad Roberts wrote:
> As long as you're relying on trust, you're in trouble.  Trust 
> and verify.  Of course, you have to trust the verification, but 
> that trust can in turn be validated (harder to falsify stress 
> to failure results than "yeah, it'll work" assertsions).  It's 
> part of why testing exists.

Of course -- in fact, verification serves to enhance and sustain
trust, the two are complementary.

But not relying on blind trust doesn't make it any less daft to
employ people you don't have trust in.