How Bitcoin could become a super-sized Wayback Machine Researchers have proposed a system which could see Bitcoin users earning their trendy tender by replicating vital data sources rather than crunching pointless algorithms. The new system, dubbed Permacoin by a team of University of Maryland and Microsoft researchers, would substitute the current requirement to obtain Bitcoins …
When I first heard of "bitcoin mining" and had not looked it up, I was under the impression that they were "earned" by doing something useful like this and not simple solving a pointless puzzle designed to create logarithmically increasing scarcity.
I doubt the limiting factor in practice would be the end user's storage space though, network bandwidth is going to make the practicalities of accessing TB-sized data sets distributed on home user's PCs a challange.
..the limiting factor in practice would be the end user's storage space though, network bandwidth is going to make the practicalities of accessing TB-sized data sets distributed on home user's PCs a challange.
I think that the idea is that it would be a backup of last resort rather than primary storage. Multiple replication would mean that some sort of BitTorrent-style peer protocol could mitigate the problems caused by wet-piece-of-string ADSL connections somewhat.
With a lot of broadband accounts having ~1GB/day upload limits, you are looking at just under 3 years to upload a TB of data, even assuming 24/7 connections with no down-time.
Yes, multiple sources would spread that burden around, but even so it is still a major problem. How many users, let alone businesses, can wait for months to get data back?
Waiting is better than not at all, and one would hope that not all blocks were on the equivalent of a memory stick in rural Lincolnshire.
As you correctly point out, though, the 'value' of any crypto-currency (especially BitCoin) is entirely in the efficiency of the network. You might not mind waiting 15 minutes for confirmation if you're BCing some chewing gum or a pint, but if you were waiting that long for the price of your house then you'd sweat.
When you invent a functioning cryptocurrency that also folds proteins or searches for extra-terrestrial intelligence, then let us know. There's almost certainly a Nobel Prize waiting for you.
If, on the other hand, you just want other people to do the stuff that you want them to, rather than the stuff that's making them money and changing the future of commerce, then there's nothing to stop you from setting up a global SETI@coin/ProteinFolding@coin network, and buying the resulting coins from miners at a price that the market recognises as justifying the costs and risks of that enterprise. The question is: are you willing to put your money where your mouth is?
Of course, there's nothing stopping you using NTFS reparse points to effectively pretend to have unlimited numbers of copies of the same file, thus giving you unlimited money! (Especially if the 'machine' you connect to the bitcoin network is a virtual machine with an abstract hard drive... /sigh)
I suspect that the protocol could check for that using checksums to verify that data was actually being stored fully and completely. In fact it would need to do this anyway really, so I doubt that it would be easy to adopt the 'cheap high capacity chinese pendrive from ebay' strategy you're suggesting.
This post has been deleted by its author
If I sold my permacoin would I also be selling my chunk of stored data?
What kind of duty of care would I have for my data?
If my hard drive blows up do I also loose my coins?
I love the concept (especially as I have a super fast university internet connection at my disposal) and distributing valuable data is a great idea but I do not understand how you would release the value in the coins.
Well the network would obviously use a high level or redundancy, effectively your storage contribution would be one part of a very large virtual raid array, if you like. So you would have no duty of care.
As for earning coins, I guess it would have to be based on both time availability and storage, in which case coins would not be lost when you 'switched off' (or blew up) your storage - basically you would be paid for the contribution you make to the 'virtual drive', and you should be able to enter or leave as you wish - nothing else would be practical.
Instead of typing a server address in the address bar, you type a data address. The data is then found at the nearest server and displayed.
You would be the authority of your own data on your own NAS server or your ISP's server space and would be able to change it or delete it at will, the changes or deletions would then be replicated out to other servers in different regions.
Youtube, Facebook, and other sites which only exist to horde other people's data then disappear in a puff of smoke.
Where's my flying car by the way?
Backing up the 200TB library of congress? Too bad there aren't hard drives that can store a few terabytes on their own available for next to nothing, then any public library that cared to could replicate the library of congress. Or that there aren't cloud providers who measure their storage space in millions of terabytes, who could replicate it hundreds of times all over the world.
While using bitcoins as a backup mechanism is moronic, changing the algorithm used so useful work is done instead of meaningless computation would be a great improvement. If people are going to invest millions in custom hardware, at least have that custom hardware producing results that are useful to the scientific world.
As a side benefit, if there isn't a single algorithm but a range, either that custom hardware would have to be more general purpose, and thus useful to scientists solving other problems, or bitcoin mining wouldn't be taken out of the hands of regular people and relegated to specialized ASIC rigs only.
It would be funny if botnet operators hoping to make a buck mining bitcoins stumbled upon the solution to some major scientific problem. Think about all those bots harnessed for good, instead of (only) enriching the criminal element.
OK, so I read the whitepaper or whatchamacallit, and some of the technical aspects went over my head, but I couldn't help but come away with the impression that the whole thing is very speculative, whereas Bitcoin was surprisingly solid right from the original published idea. This plan is definitely intriguing, but they've left out some key parts.
For instance they start out talking about a "trusted dealer", which if I understand right is the entity who wants their data backed up. But this kind of central authority clearly won't work in something intended to replace Bitcoin. They address this by saying that role will be filled by the network itself, which must mean, as this article says, that users would be able to nominate data to be stored, but details about how this would work were sorely lacking.
Also, in my opinion, to be useful, there needs to be some built in method of actually retrieving the data. It's all very well that it can be proved your business records are safely backed up, but when your office burns down you need to be able to actually get them back! It sounds like they're saying you would have to manually arrange this on the side, which is hardly convenient.
I have read the paper and I am still a little**1 confused about how it works.
First things first, I really dislike the inherent uselessness in the BTC computations. It serves no beneficial purpose other than as an overhead in ensuring scarecity. But it does this by using frightful amounts of energy and wasting computing resources that could otherwise be used in useful pursuits (Folding / Climate Modeling / etc). At best you could call it an overhead; a necessary evil but finding something useful to do with those computations should be a high priority. Otherwise, the only people who will be able to profit from it**2 are those who run massive distributed botnets or cryptolocker malware who therefore don't need to pay input costs.
So I really WANT to like this, but there is one glaring problem that I can't see addressed. Let us assume that I agree to store X Bytes of data which is valued at a given amount Y. I can prove that I have stored X by signing some random challenges with my private key. Now I can believe a few things about all this:
* That due to the random challenge I can not predict which subset of X will be challenged.
* That due to my private key being needed to sign the challenge, I would not want to pool it in the cloud.
Now after some time, I want to sell my Y currency to buy a good/service. I am assuming that the buyer has the necessary mechanisms to prove that I am indeed the owner of the coin and the transaction goes ahead.
Now that I am no longer the owner of the coin, where is my motivation to not just wipe the data I am currently holding to realise the X Bytes of storage (or to use those same X bytes to store some other coin)? If I was to do this, there would be a coin that was not backed by the promised stored data which goes counter to the purpose of this.
How does the proposed solution prevent me from doing this? Surely I don't need to upload all of the data associated with the coin to the recipient in order to transact? Otherwise, where is the benefit in trading vs just mining fresh coins.
**1 by a little, I mean a lot.
**2 talking about mining here, not speculation; I will leave that for others to "solve".
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
