Looks like the author has somewhat missed the point of the original article. Which is interesting because the author repeated the point to make it clear: > Jus

Looks like the author has somewhat missed the point of the original article. Which is interesting because the author repeated the point to make it clear: > Just because two pointers point to the same address, does not mean they are equal and can be used interchangeably. That is based on the C++ standard and its interpretation among mainstream compilers. The rest of the article goes on, at some length, to discuss ideas about how to define a abstract pointer model for a programming language in which pointers can be used more easily and safely and, by doing so, allow compilers to make optimizations which don't confusingly behave differently to unoptimised code. Which, seems to me at least, kind of the opposite of the characterisation as, sort of, "a bunch of wank in which compiler designers show us how clever they are with their useless micro-optimisations at the expense of making everything more difficult for us mere mortal programmers"