Why not allow empty code blocks?

[Steven D'Aprano]

On Mon, 25 Jul 2016 01:20 pm, Rustom Mody wrote:

>> so ultimately, it all comes down to testing anyway.
> 
> All??

Ultimately, yes. It all comes down to testing. How else do you know that
your program to flernge the widget *actually* flernges the widget or not?


> There is a famous quote by Dijkstra:
> «Testing shows the presence, not the absence of bugs»

Correct. And as Knuth said:

"Beware of bugs in the above code; I have only proved it correct, not tried
it."

You cannot *prove* the absence of bugs in a large, complex program, because
how do you know your proof is correct? Your automatic prover is a program,
which will contain bugs. If you don't use an automatic prover, then how do
you know you didn't make a mistake in your manual proof?

Ultimately, any program beyond a certain level of complexity can only be
*suspected* to be correct.

> Or if you prefer things of a more ‘practical’ (so-called_ nature:
>
http://www.testingexcellence.com/reasons-automated-tests-fail-to-find-regression-bugs/





-- 
Steven
“Cheer up,” they said, “things could be worse.” So I cheered up, and sure
enough, things got worse.