Potential Bug in the Dependency Analysis Chapter

[jsalzbergedu]

Hello,

I think I may see a potential issue/bug in the dependency analysis chapter.

On page 606, there are six derivations in the calculational design of the potential dependency semantics of an assignment statement. In the first derivation, you write "definition of inclusion". On the second definition, you write "definition (47.16) of the future seqval [[y]]". On both of those derivations, I follow your reasoning.

Following that, you justify your next step with 47.18 so that diff (a.b, c.d) iff (1) a =/= c or (2) a = c and b =/=d.

While this is true of the definition of diff, I cannot see that it is the case in this rule. Both of the parameters to the diff have a sequence that is only one value each, so I cannot see how you could split into a.b, c.d.

Any clarification or addition to the corrigendum would be greatly appreciated.

Thank you!

• Jacob Salzberg