Hello,

is NobleCoin suffering an "attack" right now? I have a feeling it's an attack in the sense that a user with greedy intentions is succeeding in obtaining more blocks than would be fair.

Here's what I think is happening, he has around 0.5 - 1GHs of available hash power, and he is using it to hop between coins.

I know that the 1 block difficulty recalculation should make this attempt null, but I actually think it's working. The difficulty is around 20, then it falls to around 13, and slowly climbs back to 20. The problem is that during the slow climb it gives a large opportunity for the user to find blocks with easier difficulty, and when it stabilizes he can simply move the hash power to a different coin while it waits for the difficulty to fall (around 40 min AFAICT).

I don't think there's a quick fix, but the user is actually hurting the coin in some aspects. Firstly, it reduces the coin security (ie. in the climbing intervals there's also a chance that someone will have 51% more hash power than the hash rate of the network without the original attacker). Secondly, it is hurting the transaction time because during the fall-back period the next block can take about 40 min to be found.

Unfortunately, if this really is happening, the only solution I see is to add more hash power. Consider this a call-to-arms for the community! =D

Rofo, if there's no quick fix, don't bother. POS will fix it, and this is just another bump on the road. It's not as serious as the previous attacks, but if true we should be aware of a higher chance of 51% attack.

If such thing isn't happening, then feel free to ignore this post. I just thought it might be a good idea to point out something weird I'm seeing, but I'm still a noob here.

Thanks,

Janito

Hello,

I just thought of something to add to the brainstorm of additional layers of security. Might be a bad idea, but maybe it can inspire something else?

What if each wallet could "vote" on a range of PoS return rates? Each minted/staked/forged block will result in an interest rate, but what if that rate was dynamic, set by the user on the wallet app? Then, the other nodes of the network can accept or reject the new block, given an also user configured "allowed interest range". This way, even though a user can select its interest rate, it will only be accepted if it's in the range of at least 51% of (stake weighted?) nodes.

With time, the participating users of the network can lower their ranges in order to bring down inflation.

Of course, the problem isn't actually solved, just morphed from a technical problem into a social problem. How would you convince most users to keep acceptable levels of interest?

Anything useful from the idea?

Thanks,

Janito

Hello,

I'm a little excited by the idea of queued-mining, but I'm still blinded by the flaws I see. Help me shine a light on it and clear the way =)

The quoted justification still isn't valid to me. A malicious user could still use multiple VMs to get multiple blocks per round. If he had more development resources he could do even worse things, like using as little hash-rate as possible to have the maximum number of wallets. Say for example he has 1MH/s, what would prevent him from splitting them in one thousand 1KH/s wallets? Even if it was IP address limited, all that he needed to do was create a website that when accessed would use JavaScript to use the viewer's IP address to queue blocks, making botnets even easier.

I think there probably exists a clear explanation of this type of attack prevention, but I don't know where. Is there a whitepaper or something?

Thanks,

Janito

Hello,

I would like to add a little to this point, but I'd like to remind that I'm a noob here and this is based on my limited understanding.

The idea quoted is indeed great, the execution that's difficult to be done.

The simplest execution today would be to hard-fork every algorithm change, if I understood correctly. Rofo has already listed some arguments against hard-forks, and why he only wants to do the current planned hard-fork as an emergency measure. As far as I understood hard-forks require synchronization of all wallet versions, so all users participating in the coin's community have to have the wallet properly updated. This is not a problem for most enthusiast users or those that follow the community discussions, but not for the public that the coin is trying to achieve: non-enthusiast, non-technical, normal people. If they don't update the wallet for a hard-fork, or even if one of the "services" they use don't update the wallet (merchant, exchange, etc.) there is a real possibility of money loss.

That said, if there is another way to execute it, I think it's worthy of discussion. As a simple attempt: I would suggest that the algo be scriptable, and all wallets have a peer-to-peer way of determining the active script and therefore algo. Also, any person could submit new scripts implementing new algos, and the community would allow the switch to the new script without needing to change the wallet.

How to do this execution is beyond me. Consider for example it can't be a simple "vote", because otherwise SHA-1 farms would pressure the coin to keep the SHA-1 algo.

Anyway, I hope that this might help a little in the advance of the discussion regarding the point quoted in the beginning.

Thank you,

Janito

Hello,

just going to add another idea for brainstorming. Now that PoS is being planned, it might be a good idea to discuss possible alternatives to how the remaining coins will be distributed. As far as I can guess, they either won't (ie. cap at the available coins at the moment of the hardfork) or there will be a hybrid PoS/PoW for some time.

I would also suggest for them to be all "premined" for one specific usage: Noble Steps. The coins would then be gradually distributed among the active Noble Steps projects throughout a time period (one year, maybe more). How to do this distribution might be something to give more thought, but I initially thought of either giving a percentage of funds raised for each project or having a group of people choosing Project of the Month or something.

Anyway, just a wild idea to do things a little differently and possibly give more momentum to Steps. What do you guys think?

Thanks,

Janito

Hello,

just a quick idea. Don't know if it's a good one, but maybe it helps with future brainstorming.

What if we could set up a decentralized checkpointing system? The (non-average) user would specify a list of fixed checkpointing nodes they trust. Then, the user's node will promptly reject blocks that come from a different chain than the ones most of his trusted nodes have checkpointed.

In theory (or should I say, in my limited understanding), there wouldn't need too many nodes configured by non-average users to trust other nodes. Also, there might not be a need for many trusted checkpoint nodes. Each more active user of the forum can optionally set up one, and those who trust them can add them to their lists.

Obviously this would change the crypto-network from being totally distributed into something decentralized, but maybe it might strengthen against attacks.

Food for thought =)

Janito

Hello,

thanks for the answer. In the bigger cryptocurrency vision, it now makes sense to me how GPUs/ASICs/etc. have an important value, as a "centralization risk buffer" (couldn't think of a better expression).

Thanks for the enlightenment! I always look forward to read your posts (and the "The State of the Alt Cryptocurrency" document)!

I'll keep lurking here and also supporting as I can, by mining, buying and using the Marketplace =)

Janito

Hello,

noob here, but I would like to add my quick .02 cents.

If it ever reaches the decision point of what's safer, PoS or more hash, I think it's important to emphasize the following point:

- on one side, by buying a PoS coin, they (evil dudes) actually raise the value and they would probably not want to lose value when they decide to dump (although I know there may be very malicious people willing to suffer loses in order to do evil).
- on the other side, by buying GPUs/ASICs/etc. they buy something that is totally unrelated to the coin, they can do evil to as many coins as they want.

If this line of thought is incorrect, please enlighten me. I'm having a lot of fun learning about CryptoCurrency =)

Just wanted to bring this forward while discussion is early.

Thank you,

Janito