Sujet : Re: HHH(DDD) computes the mapping from its input to HHH emulating itself emulating DDD --- anyone that says otherwise is a liar
De : richard (at) *nospam* damon-family.org (Richard Damon)
Groupes : comp.theory comp.ai.philosophyDate : 18. Nov 2024, 04:19:45
Autres entêtes
Organisation : i2pn2 (i2pn.org)
Message-ID : <254d3e7be0462ba8225ec0eb4804941ea635770d@i2pn2.org>
References : 1 2 3 4 5 6 7
User-Agent : Mozilla Thunderbird
On 11/17/24 9:47 PM, olcott wrote:
On 11/17/2024 8:26 PM, Richard Damon wrote:
On 11/17/24 8:46 PM, olcott wrote:
On 11/17/2024 4:04 PM, Richard Damon wrote:
On 11/17/24 4:30 PM, olcott wrote:
On 11/17/2024 2:51 PM, Richard Damon wrote:
On 11/17/24 1:36 PM, olcott wrote:
void DDD()
{
HHH(DDD);
return;
}
>
_DDD()
[00002172] 55 push ebp ; housekeeping
[00002173] 8bec mov ebp,esp ; housekeeping
[00002175] 6872210000 push 00002172 ; push DDD
[0000217a] e853f4ffff call 000015d2 ; call HHH(DDD)
[0000217f] 83c404 add esp,+04
[00002182] 5d pop ebp
[00002183] c3 ret
Size in bytes:(0018) [00002183]
>
DDD emulated by any encoding of HHH that emulates N
to infinity number of steps of DDD cannot possibly
reach its "return" instruction final halt state.
>
This applies to every DDD emulated by any HHH no
matter the recursive depth of emulation. Thus it is
a verified fact that the input to HHH never halts.
>
>
>
I will also add, that since you have dropped your requirements on HHH (or are seeming to try to divorse yourself from previous assumptions) there are MANY HHH that can complete the emulation, they just fail to be "pure functions".
>
>
The damned liar despicably dishonest attempt to get away
with changing the subject away from DDD reaching its final
halt state.
>
>
Which is just what YOU are doing, as "Halting" and what a "Program" is are DEFINED, and you can't change it.
>
>
YET ANOTHER STUPID LIE.
A SMART LIAR WOULD NEVER SAY THAT I MEANT
PROGRAM WHEN I ALWAYS SPECIFIED A C FUNCTION.
>
>
But then you can talk about "emulation" or x86 semantics, as both of those are operations done on PROGRAMS.
>
No stupid I provided a published paper that includes the
termination analysis of C functions.
Look again at what they process. C functions that include all the functions they call.
LLVM is equivalent to x86 code. PDF page 27
And show where the emulate a non-leaf function that it doesn't have the code for the functions it calls.
This seems to be the thing that you are fundamentally missing, "Program" doesn't mean has a "main" function, program means it is the complete code with a defined input/output.
Automated Termination Analysis of C Programs
https://publications.rwth-aachen.de/record/972440/files/972440.pdf
Yep, just shows you don't understand what the paper is talking about.
And note, the "processing" they are doing is NOT "emulation" of actually trying to see what each "instruction" does, but a transformation of each block of code to see how program state transforms as the program progresses, and under what conditions.
This is because for termination analysis, you don't have "values" that you are processing so you can do "emulation", but only a list of pre-conditions that you transform to post-conditions to try to prove that the program WILL terminate for all inputs (or for what class of inputs).
You just continue to demonstrate that you don't understand what you are talking about, but just parrot by rote things not learned, so you don't understand in what context they apply.
Sorry, you are just sinking your reputation.