The reason SHA256 was chosen is that it is a hash function. Hash functions are meant to be easy to calculate but difficult or near impossible to reverse and near impossible to optimize.
MD2 was designed by cryptographers and has tons of research regarding it's security. It is now broken.
Protein folding basically has no research regarding it's security as a hash function, because it isn't designed as a hash function.
Protein folding almost certainly has a large optimization problem, an attacker likely could find ways to increase his mining power by orders of magnitude. If the currency became valueable, it is inevitable that someone would make solving this problem constant time.
Now, is making protein folding more a efficient a good thing? Absolutely! But the only way it can be made more efficient is by making this currency useless.
Essentially, the only way this currency can be useful to DNA folding is if the currency becomes useless for securing money.
Edit: It appears that foldingcoin is on top of the bitcoin blockchain so miners aren't even there for consensus. They are simply creating a speculative asset that has an increasing monetary supply that equates to donations to protein folding. The miners are completely unnecessary for security.
> Protein folding almost certainly has large optimization problem, an attacker likely could find ways to increase his mining power by orders of magnitude.
As you mentioned, this is a good thing. And as far as the effects on the currency go, this doesn't seem any different from Bitcoin miners moving from CPUs to GPUs to ASICs.
> If the currency became valueable, it is inevitable that someone would make solving this problem constant time.
Protein folding / structure prediction is NP-complete, so seems unlikely.
>As you mentioned, this is a good thing. And as far as the effects on the currency go, this doesn't seem any different from Bitcoin miners moving from CPUs to GPUs to ASICs.
The optimization "problem" is a problem because it can wipe out the currency. The transition to better hardware was gradual, while a software optimization could literally make one persons CPU more efficient than the rest of the world combined.
>Protein folding / structure prediction is NP-complete, so seems unlikely.
While it may be difficult to actually fold the protein, folding the protein such that it breaks one of the assumptions I listed may be trivial.
If we could simply use NP-hard proven problems as hash functions, we would.
I do agree that many altcoins are "pump and dump", as i have been a miner for years now. But also a lot of these altcoins never became incorporated and masked their identities, IPs, ext... We are fully open with who we are, with pictures to our names and we are 100% compliant with US laws, we are a registered non profit with my name and information on their.
That being said, we are held accountable for our actions as many other altcoins are not, which is how they get away with scamming people.
Even if we do fail in the future, we have at least brought the idea of Folding to many miners who had never heard of it before. Look at our team, http://folding.extremeoverclocking.com/team_summary.php?s=&t... but since we allow other team members to earn FLDC, we just hit 21 million FAH credits yesterday making us the 6th fastest FAH team in the program. That is an accomplishment in itself :)
Your lack of an IPO doesn't imply pump and dump. Making an altcoin that is deceptive and speculative.
"Mine Medicine, Not Hashes" implies that this is somehow useful for medicine. It really is just a pump and dump where a portion of the money goes to protein folding.
Actually non of the money goes towards protein folding, its the power of the folders that does. All of the FLDC (money) goes directly to those that fold.
As for the IPO, we dont need one. Since a counterparty asset is very easy to create, our overhead is low. Thats why we are running a fundraiser, we plan on implementing additional features, but really no features are required to distribute FLDC. The features will help better the currency and the economy surrounding the currency, but unless people wish to give us money to do these things, then we will simply keep coding ourselves and distribute FLDC everyday.
Other blockchains require a lot of maintenance and patches, we do not. Almost no overhead
>Actually non of the money goes towards protein folding, its the power of the folders that does. All of the FLDC (money) goes directly to those that fold.
What else could you possible think I meant when I said the money goes towards protein folding?
>Other blockchains require a lot of maintenance and patches, we do not. Almost no overhead
You have the Bitcoin blockchain as overhead and you have the security of the Bitcoin blockchain, however you have created a speculative asset that is Bitcoin plus a mandatory tax for protein folding.
If the system updates difficulty in the way Bitcoin does, they'd get a week's worth of coins quickly, then the difficulty would climb to compensate. That's not a big problem.
But with FLDC, you can have both :) The ASIC miners for Bitcoin cannot fold proteins, we are trying to bring in the Altcoin miners still using CPU and GPUS
"Protein folding / structure prediction is NP-complete, so seems unlikely."
Are you sure about that? I am not sure that a protein structure can be checked in polynomial time, given there is no algorithm to check a protein structure besides manual labor.
You can compute the free energy but that's not cheap. Still that doesn't give you a definitive answer to folded or not folded. It tells you more folded or less folded.
Ah yes the ANTON :) We would love for this to become readily available to the public. However it is expensive at this point. But I love this source and will be calling this company very soon, thank you.
Here is how we invision the ANTON could become readily available for purchase:
There is no direct translation from one to another, but a common consensus is 1 hash equals 12,700 FLOPS when comparing the 2 side by side http://en.wikipedia.org/wiki/Talk%3AFLOPS#Bitcoin_.22FLOPS.2... . The FAH grid computing network has 46 PetaFLOPS and is known as the world's most powerful computing network outside of the Bitcoin mining network.
Now at the time before ASICs and FPGAs started hitting the market in December 2012, the Hash rate of the BTC network was at 26 TeraHASH’s in mostly GPU and CPU power. Based on a rough comparison 12.7 PetaFLOP = 1 TeraHASH the potential computational power that could be added to FAH is
26 TeraHASH 12.7 PetaFLOPS = 330 PetaFLOPS.
Imagine if that power was harnessed for molecular protein folding. Most of this power was redirected to altcoin mining after the SHA ASICs came out, since there was no profit motive for folding. FoldingCoin looks to bring a profit motive for people to fold proteins by distributing FLDC along with other Counterparty tokens.
Venture capitalists could see this as an opportunity to invest in the creation of economic sized Anton Supercomputer, which is an ASIC molecular protein simulating machine that can fold more efficiently than standard computing hardware. This very thing happened to Bitcoin mining when it became exponentially profitable: venture capitalists invested in the creation of ASIC miners to compute SHA256 at a more efficient rate than standard computing hardware.
Some formulations of protein folding are NP complete. That doesn't mean those models always correspond to reality. There's no evidence that nature is solving an NP hard problem when folding proteins (See the paper NP complete problems and physical reality by Scott Aaronson)
An attacker could alter the amount of points that they receive. It has happened in the past with Folding@home, but they have put measures in place to prevent this from happening again. Although it isnt a guarantee that someone will not find a work around, we would become instantly aware and a patch could be put in place to correct the problem. https://folding.stanford.edu/home/faq/faq-points/#ntoc3
Actually folding is sort-of similar to a one-way function. It's difficult to compute, but easy to verify. The verification is the calculation of the energy.
You give the same protein sequence to multiple miners, and the one with the minimum energy after time X wins the money. This will force the miners to find lowest energy solutions.
The hashing function is the Bitcoin network. FLDC is simply in 140 bytes of data inside a BTC transaction. Thats why we chose to go with Counterparty, all computational power that users donate to the FAH program goes 100% in the project, and the existing miners for BTC are the ones who hash FLDC
Mentioned this in my first post. Basically the currency is a speculative asset that doesn't do anything for DNA folding since the amount you lose from monetary supply increase is could just be donated through bitcoin to DNA folding.
Your right, the currency directly does not do anything, but in the crypto currency space people love the altcions. There are arguments for both sides on the value and the need for altcoins but the fact remains that people mine them.
The way our coin helps is by convincing the mass amount of computational power that is dumped into altcoins that will eventually fade away, or become scam coins, to be better used towards something that will benefit the world.
The BTC miners i do not believe waste energy as they are providing a fantastic transaction ledger.
Stanford. Indeed, this is centralization. But as others have mentioned, it's much more like "exchange your CPU time for equity in our security" than it is like "participate in our distributed, decentralized currency." No sustainable decentralized currency could use inputs like protein folding, SETI, etc.
Counterparty is really innovative, and this is a great way to leverage it. Unfortunately, it's really easy to confuse with typical altcoins which really use miners.
Well said. A lot of my time in forums is spent explaining the benefits of Counterparty. It is a fantastic platform that allows anyone with any skill set to create an altcoin, known as an asset or token in the CP world.
This is great because it allows our development team to work on features, rather than a blockchain and wallet system that can be very hard for a small team of 5 to maintain. Please read this page if you havent already about the pros and cons of Counterparty http://foldingcoin.net/fldc-vs-alts/
Real quick, this forum does not allow me to post a lot of responses, so I encourage everyone to email me rross@foldingcoin.net or join in the discussion here https://bitcointalk.org/index.php?topic=781352.0 but i will still answer as often as this forum allows me to per day :)
The miners are there for a reason. Looking at the Counterparty protocol you will see that since the CP assets live inside the individual BTC transactions, that unless the miners verify the transaction to be accepted, then the assets will work in the same way that BTC does and not go through.
You didn't really address why you think the miners are there for a reason, you just asserted it.
Realistically, what you have done here is made a centralized currency in which you issue new coins for yourself then we trust you donate the coins to protein folding.
We can already donate to protein folding with Bitcoin so I don't see much use in making an altcoin that does it in a less useful way.
The miners are there because Counterparty assets are built inside of the bitcoin blockchain. So without the miners hashing the blockchain, the transactions for CP assets (FLDC) will not go through.
As for why not just donate to standford? You absolutely could and i am sure they would love that :) But we are trying to harness the 330 PetaFLOPS of CPU and GPU computational power used in the other altcoins. So by giving an incentive to fold to the miners, the hopes are they will move over to folding, rather than hashing at blockchains that in many cases die off or get abandoned.
MD2 was designed by cryptographers and has tons of research regarding it's security. It is now broken.
Protein folding basically has no research regarding it's security as a hash function, because it isn't designed as a hash function.
Protein folding almost certainly has a large optimization problem, an attacker likely could find ways to increase his mining power by orders of magnitude. If the currency became valueable, it is inevitable that someone would make solving this problem constant time.
Now, is making protein folding more a efficient a good thing? Absolutely! But the only way it can be made more efficient is by making this currency useless.
Essentially, the only way this currency can be useful to DNA folding is if the currency becomes useless for securing money.
Edit: It appears that foldingcoin is on top of the bitcoin blockchain so miners aren't even there for consensus. They are simply creating a speculative asset that has an increasing monetary supply that equates to donations to protein folding. The miners are completely unnecessary for security.