🚀 go-pugleaf

On 1/10/2010 8:38 PM, |-|ercules wrote:
> "Sylvia Else" <sylvia@not.here.invalid> wrote ...
>> On 1/10/2010 6:48 PM, Sylvia Else wrote:
>>> On 1/10/2010 5:43 PM, |-|ercules wrote:
>>>> "Sylvia Else" <sylvia@not.here.invalid> wrote
>>>>>
>>>>> What is the "FIRST QUESTION FIRST" protocol?
>>>>>
>>>>> Sylvia.
>>>>
>>>>
>>>> The web forms at MR.com are a compromise solution. They separate the
>>>> first
>>>> question, page number and line number FROM the third page where
>>>> 3 questions are simultaneously submitted, BUT the 3 questions are
>>>> submitted
>>>> almost immediately afterwards.
>>>>
>>>> Failing that, my next protocol would be more tedious and I would send
>>>> the answer to the first (unknown by me) question before the 3 options
>>>> are worked out.
>>>>
>>>> Then there is no way God could mix up the questions and answer the
>>>> wrong
>>>> one!
>>>
>>> Glad you think so, but it seems to me that you're arguing that God can
>>> get confused about the order of questions if they're all posted at once,
>>> but there's no danger that God would get confused if it happens that the
>>> questioner has *thought* of all the questions in advance. Given that
>>> this is allegedly a mind-reading power, whether on your part or on
>>> God's, that seems a rather rash position to take. However, I have a
>>> solution. See below.
>>>
>>>>
>>>> You can either try MR.com and see if the protocol works, knowing I have
>>>> a copout
>>>> if it doesn't. Or you can do a single final test where I answer you
>>>> before you
>>>> send the 3 options, over usenet.
>>>>
>>>> There is a problem with people knowing my answer and then selecting the
>>>> dummy
>>>> questions to BETTER FIT than their original question, but since your 5
>>>> cosmology
>>>> Qs are preset it won't be an issue. If you use a random number to
>>>> select
>>>> the
>>>> two dummy questions, (after hearing my channeled answer) then that
>>>> would
>>>> be a valid test assuming you don't cheat.
>>>
>>> Let me try to get this clear, which a couple of modifications of my own.
>>>
>>> a) Before each trial, you will post a 10 decimal digit string. It is in
>>> your interest that it be random, but that is not a requirement.
>>>
>>> b) Either before or after (a) I will choose one of the 5 cosmology
>>> questions. After (a) I will post an MD5 of a 30 decimal digit string, of
>>> which the first digit is the question number, the next 19 digits are of
>>> my choosing, and the last 10 digits are the ones you posted. The reason
>>> for including the last 10 is explained below.
>>>
>>> c) You will channel an answer for the numbered question, and tell me
>>> that you have done so. You will not tell me what the answer is (you
>>> implied otherwise, but I can't see a need for you to tell me, and it
>>> eliminates one form of cheating on my part).
>>>
>>> d) At some time before I need them, including possibly before I choose
>>> the question for which I post the MD5, I'll choose 2 other cosmology
>>> questions out of the 5. I will now present the numbers of all three
>>> shuffled. Since I won't know what your channelled answer is, I cannot
>>> cheat by choosing better matching questions. There's also no need for
>>> them to be chosen randomly, though I would be inclined to anyway.
>>>
>>> e) You'll tell me which of the 5 I posted the MD5 for.
>>>
>>> Since I've previously asked the 5 questions, and you've channelled
>>> answers to them, you're obviously not objecting to their reuse.
>>> Presumably we can therefore reuse them multiple times to perform a
>>> sequence of trials.
>>>
>>> IMPORTANTLY, it is assumed that God can read from my mind the specific
>>> question for which I've chosen to post the MD5. That posting is a
>>> discrete act on my part. There is one question, and I'm posting the MD5
>>> for it. There can be no question of God's getting confused.
>>>
>>> Consequently the answer at the end is either right or wrong, and it's
>>> impossible for me to cheat by any kind of mental deception.
>>>
>>> The only way I could cheat would be by being able to choose random keys
>>> such as to create collisions between MD5s despite the restriction on the
>>> source text. It is accepted by you that this is not something I am
>>> likely to be able to do, and certainly not within the available time.
>>> Note in particular that your 10 decimal digit string prevents me from
>>> launching a birthday-party attack on the MD5 algorithm.
>>>
>>> A failure on your part to achieve more than chance over a suitably long
>>> sequence of trials would be EVIDENCE that neither you, nor God, can read
>>> my mind in the required way. There would be NO NULL RESULT.
>>>
>>> Sylvia.
>>>
>>>
>> Sadly, I've subsequently realised that there is a way for me to cheat.
>>
>> Let's see whether Herc can figure it out.
>>
>> This would be a whole lot easier to test of God weren't so totally
>> feeble.
>>
>> Sylvia.
>
>
> God's been saying for 10 years "HEY I CAN TALK" and 10,000 skeptics scoffed
> and never bothered to listen. Who's feeble?

It's just that He seems remarkably easy to hoodwink, or just plain 
confuse, in these tests - at least, according to you.

>
> Your MD5 idea is no good. If you limit a private key to 15 numbers you
> can't
> birthday attack a collision.

I'm happy with any number of digits that are long enough to prevent you 
from doing an exaustive search, and which you are happy cannot be 
birthday party attacked.

  Just look at page 2 on mindreading.com,
> you're reinventing the wheel with balsa wood. What would an initial
> number 5 imply

Since we're using just 5 possible questions, and they can be listed in 
some order in advance, each can be identified by a number, between 1 and 
5. I cannot for the life of me see why it would be necessary to copy and 
paste the text of the question for the purpose of this test, unless God 
can't count to 5, or has a really bad memory for lists. An initial 
number of 5 would indicate that the question chosen is the fifth in the 
list. If you want the text copied and pasted, then OK. It just seems 
like a waste of effort.

I take it you haven't figured out how I could cheat using the above 
protocol.

OK, I'll leave that for you to think about some more, but consider that 
for many years you claim to have been channelling answers to questions 
that you are interested in getting the answer to. You consider that the 
answers are definite, even if no one else does.

Consider also that you clearly really really want to be able to get one 
of these protocols working, so as to prove your point.

So if *you* asked the question "Is Sylvia cheating in the current round 
of the test", you could certainly be considered to be interested in the 
answer. If you then channelled an answer, you should be able to 
determine whether or not I am cheating, either absolutely, with one 
channel, or with sufficient probability with a number of channels. In 
any trial, before you gave your answer, you could decide to reject that 
trial on the grounds that I was cheating in it. It would then not count, 
and we'd move on to the next trial.

So in all the accepted trials, we'd have definite hit-miss result, and 
you'd have a channelled answer indicating that I hadn't cheated.

So if you failed the test overall, it would either be because you didn't 
have a better than average chance of determining which question I chose, 
or because your channelling to determine whether I was cheating didn't 
work. The latter would obviously bring the whole chanelling concept into 
question.

To avoid the situation where you claim I'm cheating consistently when 
you see that things aren't going your way, but before the result is a 
statistically significant fail, you wouldn't get any feedback on your 
hit-miss rate until the end of the test.

Sylvia.