Re: Categorically exhaustive reasoning applied to the decision to abort

On 3/26/2024 3:57 AM, Mikko wrote:

On 2024-03-25 22:50:08 +0000, olcott said:
 

On 3/24/2024 11:42 AM, Mikko wrote:

On 2024-03-24 03:39:12 +0000, olcott said:
 

On 3/23/2024 9:54 PM, immibis wrote:

On 24/03/24 03:40, olcott wrote:

On 3/23/2024 9:34 PM, immibis wrote:

On 24/03/24 03:15, olcott wrote:

On 3/23/2024 8:40 PM, immibis wrote:

On 24/03/24 00:29, olcott wrote:

On 3/23/2024 5:58 PM, immibis wrote:

On 23/03/24 16:02, olcott wrote:

(b) H(D,D) that DOES abort its simulation is correct
     (ABOUT THIS ABORT DECISION)
     because it would halt and all deciders must always halt.

 To be a decider it has to give an answer.
 To be a halt decider it has to give an answer that is the same as whether the direct execution of its input would halt.

 That would entail that

 Tough shit. That is the requirement.

 I proved otherwise in the parts you erased.

 You proved that the requirement is not actually the requirement?

 I proved that it cannot be a coherent requirement, it can still
be an incoherent requirement. Try and think it through for yourself.

 Every program/input pair either halts some time, or never halts.
Determining this is a coherent requirement.

 That part is coherent.

 The part that this determination must be done by a Turing machine
using descriptions of the program and input is coherent, too.
 

 Every decider is required by definition to only report on what
this input specifies.

 Not true.

 int sum(int x, int y){ return x + y; }
The show how sum(3,4) reports the sum of 5 + 6.