we can fork and makes it X11 (I am pretty sure it would be possible...)
Of course. Just tell me. Why?
X11 is a POW algo. Why should it be implemented in a POS system?
Pure POS has a set of serious flaws that need adressing. Migrating to X11 is utterly meaningless.
I also dont understand why we would need POW algo like X11
Which are the flaws you are referring?
Generally the cost to do an attack on a fully POS system is minimal.
A wallet can vote for several different generated blocks without any cost of doing so. It is a similar attack vector like 51% attacks in POW, just without the calculation cost.
Mind, you still need a sizeable amount of coins or at least nodes to do such an attack.
Nevertheless, it is without calculation price.
This is new to me! I thought that you could only attack if you would have more than 51% of the coins!
I am not completly sure about every detail. I am sure though, that you don't need 51% of all coins.
Maybe 51% of all minting coins, but never more.
And the minting ones aren't that many after all.
I think the metric more relevant is the amount of nodes you control. A Botnet would have a compareably easy approach on attacking a POS system.
So, POS is far from invincible
wiki say that you need 51% of all coin to do it, but doing so will damage yourself, which is stupid
Wiki is wrong, or rather unclear if it is stated like that in there.
You will at the very most need 51% of all minting coins. And that only if you want to ensure that your attack will succeed 100%. If you don't believe that, think for a moment. Create a coin with 100 Million units and sink 50 Million of these. Now the coin is safe as nobody can own 51 Million of the coin? I don't think so.
In any way. The problem here is, that with a large amount of coins you can basically flood the network with blocks and vote for all your own blocks, hoping that other nodes will vote for your blocks as well.
An attack of this kind is very cheap, as you don't need serious calculation power to generate a block. And you don't need a large amount of coins.
This is a problem that needs adressing if pure POS should work at any time.
if 1/2 are sinked(sinked you mean that they are out of the supply count right), then the new 51% is 1/4 of the to old total or 1/2 of the new one
Not really. What you interpret as supply count are only the coins staking/minting.
And it is really that. The staking coins.
Imagine an extreme example.
There is only one node with one coin. Then this one node with the one coin creates all blocks and thus determines what is going on.
other % supply attack are surely possible , but he must own a large number of stake, he can't do an attack with 1% of the cap
Of course. If only 1% of the cap is minting, then he can do the attack.
Inactive or cold stored coins don't count in this calculation.
If the attacker owns all the nodes in the network it is of no concern how many coins of the total cap he owns. He is the network. He decides which blocks are accepted and which are rejected. He generates all blocks.
Moving on from this simply conclusion you can surely see why only the coins actively participating in the minting process are of concern to the question how many coins are needed to effectively attack the network.
how can he own all the node? you mean no one beside him is minting
approx 640k comm coins(0.08%) are minted every day, he can do an attack if he own 320k+ at least, but to do this he need 400M+ comm coin
the hashpower in a pos coin is your stake not your mint, so compared to the traditional pow you need a certain % to do an attack, but i don't know if other %attack like finney
https://en.bitcoin.it/wiki/Proof_of_StakeI'm not talking about the minted, but the mint
ing coins.
If he owns most of the mint
ing coins, he owns the network.
I go back to the beginning. Your logic dictates, that the network would be forever safe, if the dev would have lost 500 Million instead of 200 Million. Obviously this is ridiculous.
And yet again. The wiki is no good source for in detail questions.
It's good enough for people who take a quick look and leave again. As soon as you ask "and how/why exactly is it like it is written here?" the wiki becomes pretty lacking.
but those lost coin are not in the network anymore, it's like they don't exist anymore nodes can't relies on them, so an attack it's still possibile even in that scenario
are you sure about the fact that someone can own the network if he own a large % of the minting coin? because that's seems to easy to me, % of minting coin for a certain time is a very very low number
anyway to control a great % of the minting coin, the diff must be really low(also competitors should have a low stake), and he should have a right % of the total supply
There is a reason why peercoin switched to an almost completly centralized model as the amount of POW compared to POS dimnishes. They have the same problem over there. It is also the reason why pure POS hasn't been done all that often in the past. You can do it now, as there are enough people who jump a train like this without knowing what exactly it is.
Mind, it is not trivial to own the network. It is just a lot cheaper calculation and investment wise. By now nobody has done it I think. Pure POS is niche enough to be safe for the time being.
And you also need to code a new wallet which can use the flaws of the POS system. That is another point. The attack vector I'm talking about is academical for the time being.
Oh, and the coins are in the network. The clients just reject any transaction from the adress they are stored in.
The network is a consensus machine. If enough nodes have consensus about something, it will be done.