*Is this possible? PiCoin! Calculates digits of pi with your computing power!*

[smeagol]

PiCoin
PiCoin... A new altcoin that will use your mining power to calculate digits of pi, rather than wasting it on random calculations!  High difficulty at start for higher value, and no maximum to coins, though it will get exponentially higher as time goes on until it is pretty much impossible to make any more!

Development
Anyone know if this is feasible and would be willing to develop this?  PM me and post your thoughts below!

[yacoin]

Wow! fucking nice man!

[smeagol]

you know if it is possible?

[BitJohn]

There is a guy working on this coin.........you grab this off of Verns forums?

[Eb0la]

I see my birthday.. home address.. and the answer to the universe all in that one picture!

[rbaggio]

what an awesome idea

[yacoin]

PiCoin
PiCoin... A new altcoin that will use your mining power to calculate digits of pi, rather than wasting it on random calculations!  High difficulty at start for higher value, and no maximum to coins, though it will get exponentially higher as time goes on until it is pretty much impossible to make any more!

Development
Anyone know if this is feasible and would be willing to develop this?  PM me and post your thoughts below!

http://gallery.bridgesmathart.org/sites/default/files/styles/large/public/bridges2012/dreimann/pi.jpg

The image is TRIPPY as hell. Even if you're not on LSD. Just stare at the center

[achillez]

soooo cool

[megablue]

i was wondering the same thing as well... it is definitely possible but is it feasible?

Challenges:
- it would require us to start from the next unknown digit
- what if somebody already have a longer list that doesn't get published in public and decided to use it as advantage?
- perhaps, it can and should be pair with hashing algo so that even someone with the next digit doesn't have full advantage


Since it is floating point heavy calculation, nvidia gpu can finally be useful in mining.

[spoid]

problems will probably overlap with those of sciencecoin/curecoin etc.
seti@homeCoin and more! spacecoin. aliencoin.

[smeagol]

i was wondering the same thing as well... it is definitely possible but is it feasible?

Challenges:
- it would require us to start from the next unknown digit
- what if somebody already have a longer list that doesn't get published in public and decided to use it as advantage?
- perhaps, it can and should be pair with hashing algo so that even someone with the next digit doesn't have full advantage


Since it is floating point heavy calculation, nvidia gpu can finally be useful in mining.

Start from a random digit (1 to 1 million) and pair it with a hashing algorithm.  Use random number generator to decide which digit to start from

If you are willing to help develop this PM me!

[ntkrnl]

pretty cool idea

[smeagol]

problems will probably overlap with those of sciencecoin/curecoin etc.
seti@homeCoin and more!

sciencecoin and cure coin both give only a portion of profits to the miners.  PiCoin will give all profits to miners

[smeagol]

also, PiCoin will have an actual purpose.  SciCoin and CureCoin still use pointless calculations.

[Rannasha]

Calculating pi is not really a useful purpose. It has been calculated to 10 trillion (10^13) digits already. Practical use of increased accuracy in even the most precise applications is completely gone after 30-40 digits.

edit: Additionally, the whole idea behind crypto-currency is to have a task that is hard to perform, but easy to check. Finding a block that has certain properties when hashed is very difficult, but once found, it's simple to verify that the result is correct (just a single hash-operation on the found block, which is very fast). There is no way to verify whether the 28389283751th decimal of pi is really 4 or whether i'm just cheating (unless you have a table with all digits, in which case a miner can just use that).

There's a reason why hash-functions are used for all crypto-currencies and simply throwing out the "useless hash-computations" and replacing it with something else isn't going to work.

[cre8r]

The basic principle of Proof of Work scheme is - hard to do the work and easy to check that it is done right. There is no easy way to check that newly discovered digits of PI are not just taken from the air, so it will not work.

[turtle83]

The basic principle of Proof of Work scheme is - hard to do the work and easy to check that it is done right. There is no easy way to check that newly discovered digits of PI are not just taken from the air, so it will not work.

Exactly. I thought about picoin months ago.... and while bitcointalk was loading in my browser i realized it would be just as hard to validate the PoW as to do the work itself.. so pointless exercise.

[spndr7]

. There is no way to verify whether the 28389283751th decimal of pi is really 4 or whether i'm just cheating (unless you have a table with all digits, in which case a miner can just use that).

Exactly. I thought about picoin months ago.... and while bitcointalk was loading in my browser i realized it would be just as hard to validate the PoW as to do the work itself.. so pointless exercise.

Verification of n th digit can be directly found using  Bailey–Borwein–Plouffe formula. The formula can directly calculate the value of any given digit of pi without the need to calculate the preceding digits.

Pi is an irrational number and can be written as generalised continued fraction as --
(from thread https://bitcointalk.org/index.php?topic=271568.0 )




Recursive function implementing 1st continued fraction for n iterations

float rec(float x,int n){
        float ret;
        if(x>=n)
        ret=1;
        else
        ret=(1+2*x)+((x+1)*(x+1)/rec(x+1));
        
        
        return(ret);
        }

int main(){
    float x;
        
    x=4/(1+(1/rec(1,8)));     //here n=8
    printf("\npi=%f",x);
  
  
     return(0);
     }

Here 8 iterations produce first 3 correct digits of pi.


It can be seen that as the number of iterations are gradually increased,computational complexity increases exponentially.Block reward could be adjusted accordingly.

Posted in https://bitcointalk.org/index.php?topic=271568.0

Other current problems from Experimental Maths could also be used for p2p alt coins.

[miffman]

PiCoin sounds great!

[SALHERO]

Ok, somebody check this before? 

http://en.wikipedia.org/wiki/Super_PI

http://www.techpowerup.com/downloads/366/super-pi-mod-v1-5/