On Short Circuit Evaluation

I found the following quote on the Wikipedia page on Short-circuit evaluation¹.

The quote is from Edger W. Dijkstra²

The conditional connectives — "cand" and "cor" for short — are ... less innocent than they might seem at first sight. For instance, cor does not distribute over cand: compare

(A cand B) cor C with (A cor C) cand (B cor C);

in the case ¬A ∧ C , the second expression requires B to be defined, the first one does not. Because the conditional connectives thus complicate the formal reasoning about programs, they are better avoided.

— Edsger W. Dijkstra[2]

“There be a lot of long words in there; and I'm naught but a humble pirate³.”

Let's break it down.

(A && B) || C

(A || C) && (B || C)

Possibilities are:

A	B	C	(A && B) || C	(A || C) && (B || C)
false	false	false	false	false
true	false	false	false	false
false	true	false	false	false
true	true	false	true	true
false	false	true	true	true
true	false	true	true	true
false	true	true	true	true
true	true	true	true	true

So, for A = false and C = true,

(false && B) || true -> does not need to evaluate B

(false || true) && (B || true) -> does need to evaluate B.

And this while (A && B) || C == (A || C) && (B || C) holds true.

So that's interesting.

But I fear Edger W. Dijkstra may have been needlessly dogmatic here.

References

[1] Wikipedia - Short-circuit evaluation

https://en.wikipedia.org/wiki/Short-circuit_evaluation

[2] Texas Univerity - Edger W. Dijkstra

http://www.cs.utexas.edu/users/EWD/ewd10xx/EWD1009.PDF

[3] Wikiquote - Pirates of the Caribbean: The Curse of the Black Pearl

https://en.wikiquote.org/wiki/Pirates_of_the_Caribbean:_The_Curse_of_the_Black_Pearl