Bitcoin Forum

Bitcoin => Bitcoin Discussion => Topic started by: Are-you-a-wizard? on August 06, 2011, 02:48:10 PM



Title: Additional use of processing power?
Post by: Are-you-a-wizard? on August 06, 2011, 02:48:10 PM
Would it be viable to use the processing power of the network to compress and decompress the blockchain data stored on our hard drives?

The network compresses all the data down to a really small size before we recieve it, and then the client requests for decompression as and when it's needed.


Title: Re: Additional use of processing power?
Post by: 2112 on August 06, 2011, 03:05:45 PM
The key protocol data is incompressible: correctly implemented cryptography is indistinguishable from noise. I have feeling that you completely don't know what you talking about.


Title: Re: Additional use of processing power?
Post by: hugolp on August 06, 2011, 03:17:52 PM
The key protocol data is incompressible: correctly implemented cryptography is indistinguishable from noise. I have feeling that you completely don't know what you talking about.

He probably has no idea about cryptography, but that is no reason to answer this way.


Title: Re: Additional use of processing power?
Post by: westkybitcoins on August 06, 2011, 03:38:22 PM
"noise" can still be compressed to reduce the size...

Sometimes it can.

If it's pretty random noise (which encryption strives for), attempting to compress it will likely result in an INCREASE of the size of the data.


Title: Re: Additional use of processing power?
Post by: qbg on August 06, 2011, 03:59:50 PM
"noise" can still be compressed to reduce the size...

Good encryption should be indistinguishable from random strings.  What is the chance that it could be compressed by at least one byte? There are 256^n strings with a length of n bytes. Since we want to be able to recover the original string, each compressed string must decompress to one decompressed string. There are 256^(n-1) strings of length n-1 bytes. This means that only (256^(n-1))/(256^n) = 1/256 strings of length n can be compressed by one byte. For more highly compressed strings, the chances get worse. This is why trying to compress the data would be useless.