Without realising it Olcott has actually confirmed rather than refuted the Halting Problem:

In x86utm, H simulates D(D), detects the nested recursion as non-halting, aborts, and returns 0 (non-halting). But when D(D) runs for real:

* It calls H(D,D).
* H simulates, aborts the simulation (not the real execution), and returns
0 (non-halting).
* D, receiving 0 (non-halting), halts.

Thus, the actual machine D(D) halts, but H reported "does not halt". H is
wrong about the machine's behavior which aligns with the diagonalization paradox at the heart of extant Halting Problem proofs.

/Flibble

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)