|
Title: compression not a waste of money Post by: yogi on September 15, 2011, 11:55:02 PM I don't know if this is a good idea or not, someone may even have already suggested it, but...
As bitcoin grows and more transactions take place, anyone running the bitcoin client will start having there resources monopolized. They will then stop using the client and the number of nodes will diminish leaving the network vulnerably to bot attacks. So my suggestion is, create a version of bitcoin that rewards people with coins for solving compression problems which would make the network more scalable. Compressed data would consume less bandwidth and less HDD space. Also the electric used to mine would be spent on something useful. What do you think? Title: Re: compression not a waste of money Post by: Kansattica on September 16, 2011, 06:54:53 PM Mining is already s something useful: Maintaining the block chain. If you make the miners work toward something else, then Bitcoin is kinda boned.
Title: Re: compression not a waste of money Post by: MadHacker on September 16, 2011, 07:55:36 PM If you want to save on harddrive space you could always compress the directory that holds the block chain data.
myself I'm not worried about bandwidth... If i nuke all my data except for my wallet.dat file then it only takes about 1/2 hour to re enumerate all the confirmations (had to do that a few weeks back) Title: Re: compression not a waste of money Post by: kjj on September 16, 2011, 08:00:46 PM I don't know if this is a good idea or not, someone may even have already suggested it, but... As bitcoin grows and more transactions take place, anyone running the bitcoin client will start having there resources monopolized. They will then stop using the client and the number of nodes will diminish leaving the network vulnerably to bot attacks. So my suggestion is, create a version of bitcoin that rewards people with coins for solving compression problems which would make the network more scalable. Compressed data would consume less bandwidth and less HDD space. Also the electric used to mine would be spent on something useful. What do you think? Much of the block chain is random numbers and outputs from cryptographic functions. Both of these are very high entropy data and don't compress very well. I just ran bzip -9 on my blk0001.dat and I got 21.36% savings (606,522,351 bytes -> 476,975,806 bytes). For comparison, a recent test showed that pruning could reduce the block database by over 70%. The bigger problem is that no reversible compression systems that would meet our needs as a security function. Hashing was picked because it has a whole bunch of desirable properties, like the avalanche effect (http://en.wikipedia.org/wiki/Avalanche_effect), and that the outputs look very much like random numbers, allowing us to scale the difficulty in a predictable way. Title: Re: compression not a waste of money Post by: Alex Fenner on September 17, 2011, 07:45:53 PM It would work.
Title: Re: compression not a waste of money Post by: yogi on September 18, 2011, 12:07:08 PM Thank you for you answer 'kjj', while I believe it would be possible to create a compression mechanism with all the required properties I think the added complexities would not be worth the gain in data size.
I have seen an idea called HubCoin which may provide a solution to the scaling problem. It would be interesting to know how much resources the users of the network are prepared to donate in order to predict its tolerance. My resources are limited, and so I will switch the client off when it exceeds:- 2gb bandwidth per month (my ISP limits me to 10gb a month) 0.5ghz cpu (i only have 2.5ghz) 50gb hdd Maybe this would be worth a poll on a separate thread with the aim a generating a network tolerance graph? |