Observational purity

[bearophile]

Walter Bright:

> "observational purity" seems like another word for "logical const". This has 
> been debated here many times.

They show how to enforce "observational purity". I don't remember people discussing this here.

They use observational purity just for methods called by contracts, to find a trade-off between flexibility and analyzability. I don't remember this discussed in this newsgroup.

So, it's not a so boring paper as you say.

Bye,
bearophile