Topic: delete (Read 1716 times)
My guess would be that somebody wanting to amass a million would not place a single buy order, they would spread it out and grab thousands at a time with automated orders. Looking at the SC markets and the range of price points with support you could easily pull this off. 240 BTC would give you 2% of your first million SC on a single exchange as I type this. Run that rate out with 24,000 new SC's coming online every day, with lots of happy miners looking to dump for a quick buck? Yeah, you could become an Arbiter Node pretty cheap.
A million SC2 is a whopping $4700. That won't even buy a halfway serious server, let alone protect one.
Do you really think this? Price will raise after some thousands are bought, you will have to spent much more to get 1 million SC2...
Well it's difficulty for normal people to understand I guess. A reorg can happen currently where two trusted nodes both sign a block and send it out at the same time, so for 2 blocks the network is split. A 2 block split is OK provided anyone dealing with SolidCoin looks for 6 confirmations (which the two major exchanges do, and is what most default to). Considering our supernodes also will likely have the exactly same transactions in them (as they aren't malicious) it's really quite moot besides the efficiency factor.
You mean you took a fully redundant peer to peer network and introduced a single point of failure. One node fails and the whole network just stops. One node that can be found with a network sniffer and a dozen installed clients at different parts of the network.
And you can't fix this without introducing a blockchain split on every second block..
The degree by with you fucked this up is simply staggering.
A million SC2 is a whopping $4700. That won't even buy a halfway serious server, let alone protect one.
Obviously not a scam considering being an Arbiter node is voluntary and you pay to protect your network. If you don't want to pay the CPF you can choose not to become an Arbiter node, all it takes is to not amass a million SC's into a single wallet on your mining client. Then done deal, your not an arbiter node and you don't pay the CPF fee.
Details on this? I thought that the rules simply were "have a million solidcoins, become a trusted node".
That is the rules, but those rules do not prohibit anyone from gaining that status, in theory everyone could get that status, the catch being that once you have it, you are in essence opting in to pay the CPF fee since only the trusted nodes pay it. If you don't like the CPF fees, then you fragment your wallet such that all of your coins are not hoarded into a single place which in and of itself is dangerous since now your wallet gets big googly eyed attackers interested in it. To be an Arbiter node will take on a lot of optional risk but with that kind of $$ even at low exchange rates you had dang well be able to afford top notch protection for your server(s) and the CPF fee would be peanuts to you.
So yes the Arbiter nodes will be decentralized just in a limited fashion as we get more Solidcoin millionaires+ who want to take on the risk and CPF fees at the benefit of acting as additional network security.
What I don't get is why would anyone who has amassed a million solidcoins really want to be an Artbiter node? Why would you want to put all your coins in one basket and ask for attention PLUS have to pay a fee for each trusted block you find? The reason to help protect the network so you help preserve your investment is pretty weak. If I had a million solidcoins, I'd keep them in a thousand separate address and keep a low profile. The only reason to become an arbiter node is if you get something out of it.
For example, I can imagine that if I become an arbiter node, I will only release a trusted block confirming my own mined blocks and release them as a pair. And create some kind of mining cartel and steal blocks from the poor. That's the only thing I see happening. So the future of solidcoin is either a mining cartel OR keep with the appointed arbiter node with unspendable coins where CH/RS gets to pick his trusted nodes. Neither of those futures are very appealing to me.
Have you considered having the trusted nodes assigned to their own difficulty target? Instead of always using diff 1, do regular difficulty retargets on the trusted chain based on how quickly each trusted block was solved after their previous normal block, over x amount of trusted blocks. This is even reverse-compatible with the current chain. Just start having clients enforce it in a later version.
No need to always reinvent the wheel...
No need to always reinvent the wheel...
I think this would make a trusted node attack easier because an attacker who has a lot of hashing power could make it impossible for any other trusted node to generate blocks.
In that case, the other solution is to do away with having the trusted nodes actually make blocks altogether (it's a stupid idea since it forces the network to be idle for a good amount of time) and instead move toward having trusted nodes sign the forks that they consider correct. See:
https://bitcointalksearch.org/topic/proof-of-stake-instead-of-proof-of-work-27787
https://bitcointalksearch.org/topic/proof-of-stake-brainstorming-37194
Have you considered having the trusted nodes assigned to their own difficulty target? Instead of always using diff 1, do regular difficulty retargets on the trusted chain based on how quickly each trusted block was solved after their previous normal block, over x amount of trusted blocks. This is even reverse-compatible with the current chain. Just start having clients enforce it in a later version.
No need to always reinvent the wheel...
No need to always reinvent the wheel...
I always wondered why it was diff 1 also, I also suspect that Coinhunter has the ability to GPU mine this chain also. A single 5830 or 5850 could provide excellent protection against a 51% attack but would be useless against a trusted node ddos.
Why the diff 1 Coinhunter?
It would make WAY more sense if the blocks were worth nothing or 0.00000001
... but that would remove the scam reason for them existing.
i.e. they are a scam since the OBVIOUS way for them to NOT be a scam was not used.
Seriously, if they were not worth something then there would be no reason for anyone to specifically want them, thus they would truly be what CH pretends they are: a '51% attack safeguard' based on a closed controlled block-chain.
However, the fact that the blocks themselves are under the control of only a few and they are controlled by only one - well you might as well go use a frigging bank - you can trust a bank or the laws they must exist under WAY more than someone who created SC1.0 (yeah people don't change they just try to find better ways to do what they did before, SC1.0 was a scam plastered with lies and deceit - as is SC2.0 created by the same person for the same reason, he hasn't suddenly changed ...)
Have you considered having the trusted nodes assigned to their own difficulty target? Instead of always using diff 1, do regular difficulty retargets on the trusted chain based on how quickly each trusted block was solved after their previous normal block, over x amount of trusted blocks. This is even reverse-compatible with the current chain. Just start having clients enforce it in a later version.
No need to always reinvent the wheel...
No need to always reinvent the wheel...
Over at Solidcointalk Coinhunter seems to be preparing his followers that an attack on the Trusted Nodes will stop the network. Tells them they will not lose any coins. He admits that the Trusted Nodes are an issue that can be attacked. Interestingly enough he also talks about when more than a single trusted node comes online, the second ones starts a small re-org. I do believe he has stated that Solidcoin was "Hack Proof", "Attack Proof" and designed without flaws. Now he says different. The trusted node issue sounds like a gaping design flaw to me.
http://solidcointalk.org/topic/305-realsolid-is-there-any-reason-for-concern/page__view__findpost__p__3105
http://solidcointalk.org/topic/305-realsolid-is-there-any-reason-for-concern/page__view__findpost__p__3105
Well it's difficulty for normal people to understand I guess. A reorg can happen currently where two trusted nodes both sign a block and send it out at the same time, so for 2 blocks the network is split. A 2 block split is OK provided anyone dealing with SolidCoin looks for 6 confirmations (which the two major exchanges do, and is what most default to). Considering our supernodes also will likely have the exactly same transactions in them (as they aren't malicious) it's really quite moot besides the efficiency factor.
We'll look to eliminate these reorgs to a bare minimum because it's an efficiency thing, you want the network working together on the same trusted block instead of having multiple and then finding the real winner one block later. Currently due to our layering of supernodes re-orgs only happen rarely, so network efficiency is rather high. That said we are implementing new methods to further reduce them.
It's really quite irrelevant if people want to attack SC2. With SC1 I couldn't sleep some nights thinking about how insecure it was, it's different with SolidCoin 2.0 , getting caught up on a lot of sleep now.
If you want to think it's insecure then that's fine, up to you to decide what is secure or not given the information out there. 
delete
Jump to: