https://old.reddit.com/r/logic/comments/1q1sokp/ on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming
something that doesn't exist.
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming
something that doesn't exist.
god that's such a lazy as fuck reply
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming
something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that you
can do what has been shown can't be done.
Thus, your whole argument becomes unsound, as it seem is also an apt description of yourself.
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming
something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that you
can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with paradoxical input, in order to refute the proof that we can't do total halting
analysis ...
because it has been shown that total halting analysis can't be done, due
to paradoxical input???
Thus, your whole argument becomes unsound, as it seem is also an apt
description of yourself.
idk how to make the fact ur *still* just begging the question richard,
any more fucking clear
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming
something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that you
can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do total
halting analysis ...
Because it CAN'T. If you could, you might be able to, but you can't just assume you can.
because it has been shown that total halting analysis can't be done,
due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it will
get wrong, it is impossible to make a decider that will get every input right, since we have shown that there is at least one it will get wrong.
Why is that so hard to understand?
Thus, your whole argument becomes unsound, as it seem is also an apt
description of yourself.
idk how to make the fact ur *still* just begging the question richard,
any more fucking clear
Because I am not. Your problem is your logic is just broken as you need
to assume a lie.
The problem is that the proof exsits that you can't make the machine you want to.
To try to claim you can. you need to show it, not just ask someone to
give you a case you can get right, as that isn't how the proof is based.
The problem is YOU are "Begging the question", as to begin with, your criteria of your API is incorrect.
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming >>>>>> something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that
you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do
total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you can't
just assume you can.
because it has been shown that total halting analysis can't be done,
due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it will
get wrong, it is impossible to make a decider that will get every
input right, since we have shown that there is at least one it will
get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that you *CANNOT* construct a paradox in regards to???
Thus, your whole argument becomes unsound, as it seem is also an apt
description of yourself.
idk how to make the fact ur *still* just begging the question
richard, any more fucking clear
Because I am not. Your problem is your logic is just broken as you
need to assume a lie.
The problem is that the proof exsits that you can't make the machine
you want to.
To try to claim you can. you need to show it, not just ask someone to
give you a case you can get right, as that isn't how the proof is based.
The problem is YOU are "Begging the question", as to begin with, your
criteria of your API is incorrect.
i'm just showing how we can construct interfaces which *CANNOT* be paradoxed.
the fact this happens to refute ur *only* proof of halting
undecidability is truly not my fucking problem bro
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in claiming >>>>>>> something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that
you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do
total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you can't
just assume you can.
because it has been shown that total halting analysis can't be done,
due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it
will get wrong, it is impossible to make a decider that will get
every input right, since we have shown that there is at least one it
will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that you
*CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
And, when you admit that "contradictory" isn't a valid answer, as the--
copy the input is built on is a different machine then the one doing the actual decision, you it the same problem.
All you are doing is showing you don't actually understand the problem
Thus, your whole argument becomes unsound, as it seem is also an
apt description of yourself.
idk how to make the fact ur *still* just begging the question
richard, any more fucking clear
Because I am not. Your problem is your logic is just broken as you
need to assume a lie.
The problem is that the proof exsits that you can't make the machine
you want to.
To try to claim you can. you need to show it, not just ask someone to
give you a case you can get right, as that isn't how the proof is based. >>>
The problem is YOU are "Begging the question", as to begin with, your
criteria of your API is incorrect.
i'm just showing how we can construct interfaces which *CANNOT* be
paradoxed.
But that isn't the problem.
the fact this happens to refute ur *only* proof of halting
undecidability is truly not my fucking problem bro
Nope, since the proof isn't based on an interface.
Just shows you don't understand what you are talking about.
And, your interface tries to justify two answers that can never be
right, one of which can't even be computed.
After all, the input program WILL either Halt or Not.
It is NOT contradictory to the decider it is given to, since that could
be different and give a right answer, it only makes the one that it was built on wrong.
And, if the decider CAN determine that this input is one of the
undecidable machines, it has actually determined that it doesn't halt.
(as ALL Halting input are deciable)
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in
claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that >>>>>> you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do
total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you can't
just assume you can.
because it has been shown that total halting analysis can't be
done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it
will get wrong, it is impossible to make a decider that will get
every input right, since we have shown that there is at least one it
will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that you
*CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think we
can't, then the only actual thing that has been paradoxed is the
*presumed interface* to the total halting decider, dick, not an *actual* machine that exists.
not even gunna read the rest cause we're clearly reaching the point of
you not being able to argue anything close to coherency, which is just fucking useless to me. a modicum of coherency with mistakes can be *extremely* useful to me...
but if you're so caught up trying to play contrarian to me, that you're
now claiming you've constructively referenced something in a proof that
that disproves an ability to construct that very thing,
well then we're pretty fucking close to done, ehh???
And, when you admit that "contradictory" isn't a valid answer, as the
copy the input is built on is a different machine then the one doing
the actual decision, you it the same problem.
All you are doing is showing you don't actually understand the problem
Thus, your whole argument becomes unsound, as it seem is also an
apt description of yourself.
idk how to make the fact ur *still* just begging the question
richard, any more fucking clear
Because I am not. Your problem is your logic is just broken as you
need to assume a lie.
The problem is that the proof exsits that you can't make the machine
you want to.
To try to claim you can. you need to show it, not just ask someone
to give you a case you can get right, as that isn't how the proof is
based.
The problem is YOU are "Begging the question", as to begin with,
your criteria of your API is incorrect.
i'm just showing how we can construct interfaces which *CANNOT* be
paradoxed.
But that isn't the problem.
the fact this happens to refute ur *only* proof of halting
undecidability is truly not my fucking problem bro
Nope, since the proof isn't based on an interface.
Just shows you don't understand what you are talking about.
And, your interface tries to justify two answers that can never be
right, one of which can't even be computed.
After all, the input program WILL either Halt or Not.
It is NOT contradictory to the decider it is given to, since that
could be different and give a right answer, it only makes the one that
it was built on wrong.
And, if the decider CAN determine that this input is one of the
undecidable machines, it has actually determined that it doesn't halt.
(as ALL Halting input are deciable)
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in
claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption that >>>>>>> you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do
total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you can't >>>>> just assume you can.
because it has been shown that total halting analysis can't be
done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it
will get wrong, it is impossible to make a decider that will get
every input right, since we have shown that there is at least one
it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that you
*CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think we
can't, then the only actual thing that has been paradoxed is the
*presumed interface* to the total halting decider, dick, not an
*actual* machine that exists.
No, we paradoxed every pretender to the throne.
Note, you can't even BUILD a machine based on an "interface", as your
result is not an acual program, as a program (in this system) actually includes all of the algorith/code it uses.
not even gunna read the rest cause we're clearly reaching the point of
you not being able to argue anything close to coherency, which is just
fucking useless to me. a modicum of coherency with mistakes can be
*extremely* useful to me...
In other words, you admit being too stupid to understand.
but if you're so caught up trying to play contrarian to me, that
you're now claiming you've constructively referenced something in a
proof that that disproves an ability to construct that very thing,
well then we're pretty fucking close to done, ehh???
We started done, since you haven't even started on a rebutal for what
you are trying to rebut.
And, when you admit that "contradictory" isn't a valid answer, as the
copy the input is built on is a different machine then the one doing
the actual decision, you it the same problem.
All you are doing is showing you don't actually understand the problem
Thus, your whole argument becomes unsound, as it seem is also an >>>>>>> apt description of yourself.
idk how to make the fact ur *still* just begging the question
richard, any more fucking clear
Because I am not. Your problem is your logic is just broken as you
need to assume a lie.
The problem is that the proof exsits that you can't make the
machine you want to.
To try to claim you can. you need to show it, not just ask someone
to give you a case you can get right, as that isn't how the proof
is based.
The problem is YOU are "Begging the question", as to begin with,
your criteria of your API is incorrect.
i'm just showing how we can construct interfaces which *CANNOT* be
paradoxed.
But that isn't the problem.
the fact this happens to refute ur *only* proof of halting
undecidability is truly not my fucking problem bro
Nope, since the proof isn't based on an interface.
Just shows you don't understand what you are talking about.
And, your interface tries to justify two answers that can never be
right, one of which can't even be computed.
After all, the input program WILL either Halt or Not.
It is NOT contradictory to the decider it is given to, since that
could be different and give a right answer, it only makes the one
that it was built on wrong.
And, if the decider CAN determine that this input is one of the
undecidable machines, it has actually determined that it doesn't
halt. (as ALL Halting input are deciable)
On 1/3/26 2:56 PM, Richard Damon wrote:
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in
claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption
that you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do >>>>>>> total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you
can't just assume you can.
because it has been shown that total halting analysis can't be
done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it >>>>>> will get wrong, it is impossible to make a decider that will get
every input right, since we have shown that there is at least one >>>>>> it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that
you *CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think we
can't, then the only actual thing that has been paradoxed is the
*presumed interface* to the total halting decider, dick, not an
*actual* machine that exists.
No, we paradoxed every pretender to the throne.
yes, the *presumed interface*, dick
Note, you can't even BUILD a machine based on an "interface", as your
result is not an acual program, as a program (in this system) actually
includes all of the algorith/code it uses.
i'm not even gunna bother trying to unpack this cause you clearly enjoy having ur head stuck up ur own asshole
not even gunna read the rest cause we're clearly reaching the point
of you not being able to argue anything close to coherency, which is
just fucking useless to me. a modicum of coherency with mistakes can
be *extremely* useful to me...
In other words, you admit being too stupid to understand.
but if you're so caught up trying to play contrarian to me, that
you're now claiming you've constructively referenced something in a
proof that that disproves an ability to construct that very thing,
well then we're pretty fucking close to done, ehh???
We started done, since you haven't even started on a rebutal for what
you are trying to rebut.
And, when you admit that "contradictory" isn't a valid answer, as
the copy the input is built on is a different machine then the one
doing the actual decision, you it the same problem.
All you are doing is showing you don't actually understand the problem >>>>
Thus, your whole argument becomes unsound, as it seem is also an >>>>>>>> apt description of yourself.
idk how to make the fact ur *still* just begging the question
richard, any more fucking clear
Because I am not. Your problem is your logic is just broken as you >>>>>> need to assume a lie.
The problem is that the proof exsits that you can't make the
machine you want to.
To try to claim you can. you need to show it, not just ask someone >>>>>> to give you a case you can get right, as that isn't how the proof >>>>>> is based.
The problem is YOU are "Begging the question", as to begin with,
your criteria of your API is incorrect.
i'm just showing how we can construct interfaces which *CANNOT* be
paradoxed.
But that isn't the problem.
the fact this happens to refute ur *only* proof of halting
undecidability is truly not my fucking problem bro
Nope, since the proof isn't based on an interface.
Just shows you don't understand what you are talking about.
And, your interface tries to justify two answers that can never be
right, one of which can't even be computed.
After all, the input program WILL either Halt or Not.
It is NOT contradictory to the decider it is given to, since that
could be different and give a right answer, it only makes the one
that it was built on wrong.
And, if the decider CAN determine that this input is one of the
undecidable machines, it has actually determined that it doesn't
halt. (as ALL Halting input are deciable)
On 1/3/26 6:02 PM, dart200 wrote:
On 1/3/26 2:56 PM, Richard Damon wrote:
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in >>>>>>>>>>> claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption >>>>>>>>> that you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with
paradoxical input, in order to refute the proof that we can't do >>>>>>>> total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you
can't just assume you can.
because it has been shown that total halting analysis can't be >>>>>>>> done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that it >>>>>>> will get wrong, it is impossible to make a decider that will get >>>>>>> every input right, since we have shown that there is at least one >>>>>>> it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that
you *CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think we
can't, then the only actual thing that has been paradoxed is the
*presumed interface* to the total halting decider, dick, not an
*actual* machine that exists.
No, we paradoxed every pretender to the throne.
yes, the *presumed interface*, dick
Nope, the particular machine.
On 1/3/26 3:33 PM, Richard Damon wrote:
On 1/3/26 6:02 PM, dart200 wrote:
On 1/3/26 2:56 PM, Richard Damon wrote:
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in >>>>>>>>>>>> claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption >>>>>>>>>> that you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with >>>>>>>>> paradoxical input, in order to refute the proof that we can't >>>>>>>>> do total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you >>>>>>>> can't just assume you can.
because it has been shown that total halting analysis can't be >>>>>>>>> done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that >>>>>>>> it will get wrong, it is impossible to make a decider that will >>>>>>>> get every input right, since we have shown that there is at
least one it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that >>>>>>> you *CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think we >>>>> can't, then the only actual thing that has been paradoxed is the
*presumed interface* to the total halting decider, dick, not an
*actual* machine that exists.
No, we paradoxed every pretender to the throne.
yes, the *presumed interface*, dick
Nope, the particular machine.
you haven't actually built the machine, so you can't directly paradox
it, cause u don't have a number that actually represents the machine, as
u don't think that number exists...
On 1/3/26 6:37 PM, dart200 wrote:
On 1/3/26 3:33 PM, Richard Damon wrote:
On 1/3/26 6:02 PM, dart200 wrote:
On 1/3/26 2:56 PM, Richard Damon wrote:
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/
on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in >>>>>>>>>>>>> claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption >>>>>>>>>>> that you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with >>>>>>>>>> paradoxical input, in order to refute the proof that we can't >>>>>>>>>> do total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you >>>>>>>>> can't just assume you can.
because it has been shown that total halting analysis can't be >>>>>>>>>> done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that >>>>>>>>> it will get wrong, it is impossible to make a decider that will >>>>>>>>> get every input right, since we have shown that there is at >>>>>>>>> least one it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface that >>>>>>>> you *CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think
we can't, then the only actual thing that has been paradoxed is
the *presumed interface* to the total halting decider, dick, not
an *actual* machine that exists.
No, we paradoxed every pretender to the throne.
yes, the *presumed interface*, dick
Nope, the particular machine.
you haven't actually built the machine, so you can't directly paradox
it, cause u don't have a number that actually represents the machine,
as u don't think that number exists...
Why should I build the machine?
You are the one that says you can make a machine to match your definition.
Since the "paradox" machine depends on what machine it is making wrong, until you provide such a machine, we can't build the counter to it yet.
It seems you just don't understand what the proof is saying.
On 1/3/26 5:38 PM, Richard Damon wrote:
On 1/3/26 6:37 PM, dart200 wrote:
On 1/3/26 3:33 PM, Richard Damon wrote:
On 1/3/26 6:02 PM, dart200 wrote:
On 1/3/26 2:56 PM, Richard Damon wrote:
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/ >>>>>>>>>>>>>>> on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in >>>>>>>>>>>>>> claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the assumption >>>>>>>>>>>> that you can do what has been shown can't be done.
so i can't show how total halting analysis can coexist with >>>>>>>>>>> paradoxical input, in order to refute the proof that we can't >>>>>>>>>>> do total halting analysis ...
Because it CAN'T. If you could, you might be able to, but you >>>>>>>>>> can't just assume you can.
because it has been shown that total halting analysis can't >>>>>>>>>>> be done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider that >>>>>>>>>> it will get wrong, it is impossible to make a decider that >>>>>>>>>> will get every input right, since we have shown that there is >>>>>>>>>> at least one it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface >>>>>>>>> that you *CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think >>>>>>> we can't, then the only actual thing that has been paradoxed is >>>>>>> the *presumed interface* to the total halting decider, dick, not >>>>>>> an *actual* machine that exists.
No, we paradoxed every pretender to the throne.
yes, the *presumed interface*, dick
Nope, the particular machine.
you haven't actually built the machine, so you can't directly paradox
it, cause u don't have a number that actually represents the machine,
as u don't think that number exists...
Why should I build the machine?
You are the one that says you can make a machine to match your
definition.
Since the "paradox" machine depends on what machine it is making
wrong, until you provide such a machine, we can't build the counter to
it yet.
It seems you just don't understand what the proof is saying.
lost track of the convo there, boomer?
On 1/3/26 8:44 PM, dart200 wrote:
On 1/3/26 5:38 PM, Richard Damon wrote:
On 1/3/26 6:37 PM, dart200 wrote:
On 1/3/26 3:33 PM, Richard Damon wrote:
On 1/3/26 6:02 PM, dart200 wrote:
On 1/3/26 2:56 PM, Richard Damon wrote:
On 1/3/26 5:46 PM, dart200 wrote:
On 1/3/26 2:34 PM, Richard Damon wrote:
On 1/3/26 3:32 PM, dart200 wrote:
On 1/3/26 12:05 PM, Richard Damon wrote:
On 1/3/26 2:54 PM, dart200 wrote:
On 1/3/26 11:36 AM, Richard Damon wrote:
On 1/3/26 1:12 PM, dart200 wrote:so i can't show how total halting analysis can coexist with >>>>>>>>>>>> paradoxical input, in order to refute the proof that we >>>>>>>>>>>> can't do total halting analysis ...
On 1/3/26 6:02 AM, Richard Damon wrote:
On 1/3/26 6:00 AM, dart200 wrote:
https://old.reddit.com/r/logic/comments/1q1sokp/ >>>>>>>>>>>>>>>> on_the_halting_ghost_detector_revised/nxev5ln
better yet, actually respond on reddit Efyc
Why, someone else has already pointed out your error, in >>>>>>>>>>>>>>> claiming something that doesn't exist.
god that's such a lazy as fuck reply
But true.
Yes, your arguement is based on starting with the
assumption that you can do what has been shown can't be done. >>>>>>>>>>>>
Because it CAN'T. If you could, you might be able to, but you >>>>>>>>>>> can't just assume you can.
because it has been shown that total halting analysis can't >>>>>>>>>>>> be done, due to paradoxical input???
Right, BECAUSE we can make an input for any given decider >>>>>>>>>>> that it will get wrong, it is impossible to make a decider >>>>>>>>>>> that will get every input right, since we have shown that >>>>>>>>>>> there is at least one it will get wrong.
Why is that so hard to understand?
idk, y is so hard to understand i'm suggesting an interface >>>>>>>>>> that you *CANNOT* construct a paradox in regards to???
But the paradox isn't vs an "interface" but vs a program.
given we haven't built a total halting decider, because we think >>>>>>>> we can't, then the only actual thing that has been paradoxed is >>>>>>>> the *presumed interface* to the total halting decider, dick, not >>>>>>>> an *actual* machine that exists.
No, we paradoxed every pretender to the throne.
yes, the *presumed interface*, dick
Nope, the particular machine.
you haven't actually built the machine, so you can't directly
paradox it, cause u don't have a number that actually represents the
machine, as u don't think that number exists...
Why should I build the machine?
You are the one that says you can make a machine to match your
definition.
Since the "paradox" machine depends on what machine it is making
wrong, until you provide such a machine, we can't build the counter
to it yet.
It seems you just don't understand what the proof is saying.
lost track of the convo there, boomer?
Nope, but I think you never understood the problem.
| Sysop: | Amessyroom |
|---|---|
| Location: | Fayetteville, NC |
| Users: | 54 |
| Nodes: | 6 (0 / 6) |
| Uptime: | 15:44:29 |
| Calls: | 742 |
| Files: | 1,218 |
| D/L today: |
3 files (2,681K bytes) |
| Messages: | 184,203 |
| Posted today: | 1 |