Why don't we use the computing power for something useful?

[Matt Corallo]

Being serious here...

Maybe these other community projects should help the bitcoin network. I download the 'protine' project, but might also run hashes 10 minutes out of every hour (user setting). In this way you get paid to help out other projects.

You can already do that, just write a script which runs a bitcoin miner and a distributed computing project at the same time and adjusts the priority of each process every x minutes.  You should be able to cut their cpu usage down to almost nothing if there is a big enough difference in niceness.

[ByteCoin]

... where "break" means it is vulnerable to an attack that takes longer than brute force.

You meant the exact opposite of what you said.

Anything worth throwing computing power at is probably not easily verifiable.

Correct. A suitable problem would have to have at least the properties outlined in this post
http://forum.bitcoin.org/index.php?topic=203.msg3669#msg3669

ByteCoin

[gigitrix]

tl;dr: We would, if we could, but we can't.

[jo.bo.co]

i posted this in a different thread, but it fits better here.

Since the bitcoin network is considered secure by the userbase right now (and was considered secure a year ago, too), is there a security need to constantly increase the difficulty? Is the increasing difficulty superfluous from a SECURITY standpoint?

Because if it is, could the bitcoin client not be forced to work out a certain number of scientific (or otherwise useful) calculations for every block generated? I'm not a computer scientist, but here's my rough idea: Basically, the server would only accept blocks if your client sends them with a certain numbers of valid keys. Those keys would be distributed to your client by a server upon the completion of a set number of completed tasks from a "useful" project like seti or protein folding by your client.

This way, the difficulty of actually creating bitcoins could be constant, and the architecture for that would not have to be seriously modified. As time went on, and more people set up mining rigs, the number of required keys would adjust, raising the ratio of GPU time spent on scientific calculations to bitcoin calculations and effectively raising the difficulty.

I feel this would be great from a scientific standpoint- but also from a PR standpoint. Bitcoin would have few detractors if it were everything it is AS WELL AS the largest source of computer power available for scientific calculations.

The transaction fees that will exist in the future could even be avoided if laboratories were charged a small fee (much smaller than what they would otherwise pay) for use of the computing time.

[BitterTea]

The difficulty increases naturally with the advance of technology. That technology can be used to secure the network or to attack it.

[FreeMoney]

You can't know how much difficulty is needed. It isn't like potential attackers are publishing how much power they could put together.

[martinderm]

I actually like the idea of throwing in some folding. But the main point of Bitcoin is that there is no central authority, so who should decide what project is added. On the other hand, someone had to decide that there are 50 coins reward for a completed block. But the other problems of "useful work" described here still remain. I guess some energy waste is the price to pay for a decentraliced secure digital currency.

Edit: True, its not a waste! see next post.

[BitterTea]

It's not waste, it's securing the network. Is computing for SETI a waste? What about any other project that doesn't render something useful?

[benjamindees]

We are using our computing power for something useful.

We're automating an industry that wastes over 1.3 trillion dollars a year.

[kjj]

Ok.  When someone looks at a block, they can tell exactly how much work was done, and they can verify which work was done.  No other scheme has that, and no other scheme will work.

[martinderm]

...they can tell exactly how much work was done...

Thats not entirely true, since the work required to find a valid hash is random, the amount of work done can only be guesed.
I agree with you on the rest though. After thinking about the system for a while whan realizes that there is almost no alternative...

[kjj]

...they can tell exactly how much work was done...

Thats not entirely true, since the work required to find a valid hash is random, the amount of work done can only be guesed.
I agree with you on the rest though. After thinking about the system for a while whan realizes that there is almost no alternative...

Meh.  They can tell how much work it would take to duplicate the result with any given probability.  The difference between that and the actual amount of work done, over time, becomes a philosophical issue.  In the real world, they are essentially the same.