"Lee Rhodes" <lee@xxxxxxxxx> writes: > Alas, the reinterpret-cast produces an "invalid reinterpret_cast" compile > error (is this possibly a bug?). That is sort of weird -- my impression was that reinterpret_cast was _always_ valid as it didn't change type attributes (e.g. const/volatile), and maintained certain necessary invariants like both types being scalars or whatever... No? -Miles -- Quidquid latine dictum sit, altum viditur.
