UNOCS - Update Posted in Sticky
-
[quote name=“Tuck Fheman” post=“23377” timestamp=“1374958356”]
My problem is, pretending that this isn’t about profit, when all signals indicate otherwise. Again, that’s not an issue for me personally. Profit all you want, hence why I said I will too. I’m not in this solely for profit, but I’m sure not going to turn profit down when I see an opportunity (like this one).…
I routinely ditch most all of a cryptocoin to purchase more at a lower price. Doesn’t everyone?!
I assumed we all picked our spots, sold off and re-bought more lower (profit).
Maybe I’m doing something wrong, but [b]I’m trying to obtain the most cryptocurrency I can[/b] (profiting) [b]while subverting my masters into non-existence[/b].
You’re more than welcome to sit on your existing stash or buy more with fiat. If my way is childish, so be it. I’m a child prodigy then! =)
[/quote]Yeah, I thought the same thing too. It’s actually part of a healthy ecosystem: The price dips, the price corrects, the price soars, the price corrects. All of it driven by speculation, and here I’m being accused of speculating in a way that people don’t like so they tell me I shouldn’t be here? You can keep your coins but you can’t silence speculators. Meanwhile I’ll be buying a small island and inviting Tuck over so we can drink mojitos and feel sorry for the people who held onto their coins through the dip instead of buying low and selling high.
[move][b]It really is all about the profit. Get used to it.[/b][/move]
We’re not selling good vibes here. We’re developing a currency for the next generation that will free our children from government oppression and centralization of authority. Wait, what’s that you say? We’re centralizing the authority with the new features we’re rolling out? And you don’t want people to talk about their trading habits for fear of [u][i]you losing your own value[/i][/u]?
*facepalm*
-
[quote name=“wesphily” post=“23413” timestamp=“1374986405”]
The trick with check pointing is that you don’t have to subscribe if you don’t want to. Also, the number of people agreeing on the checkpoint will keep it decentralized.Finally, PoS does not stop 51% attack. All it does is slow it down.
[/quote]
[s]
Im quite excited regarding how the checkpoint will work, regarding the pure technical part of it.
Will it be evenly distributed toward third parties/distributed among many persons and such to keep it decentralized?[/s]Edit:
Actually this question has nothing to do in the unocs thread -
[quote name=“wesphily” post=“23413” timestamp=“1374986405”]
Finally, PoS does not stop 51% attack. All it does is slow it down.
[/quote]Vague statement. No instrument can stop 51% attacks completely if attack cost is negligible totally. If someone can afford to spend a milliard euros on mining hardware, deployment facilities, some personnel and electricity, he can destroy anything Bitcoin included. PoS alone doesn’t stop attacks either. PoW + PoS do it better than PoW alone. PoW + PoS + advanced checkpointing do it even better than PoW + PoS. And so on. Our security objective is to make 51% attacks as expensive as possible with valid miners affected as little as possible.
-
[quote name=“wesphily” post=“23413” timestamp=“1374986405”]
The trick with check pointing is that you don’t have to subscribe if you don’t want to. Also, the number of people agreeing on the checkpoint will keep it decentralized.Finally, PoS does not stop 51% attack. All it does is slow it down.
[/quote]Then all it takes is one person not subscribing to disagree, creating forks in the chain. What then? Sucks to be that disenfranchised guy? What if a majority of miners form and turn off checkpointing, and then someone on high rolls back some transactions with a checkpoint? You lose > 51% of your hashpower, making your new forked shorter chain vulnerable to a 51% attack!
You’ve not solved any problems with this solution, you’ve just created more problems and a forked chain that only you and your ‘centralized list’ of mining pools deem valid. The problem IS that more than 51% of the work is controlled already, so all you’ve done is centralize the power to decide the fate of the chain AWAY from miners. If someone is deciding who those people are and maintaining that list, then the power is centralized to the owner of the list. Someone has to say, “Hey, we should checkpoint -this- chain in order to eliminate -that- fork.” and that’s centralization. Someone has to maintain the list of voters, and that’s centralization. It’s bad news all the way to the bank. The design is 1 CPU cycle==1 vote, and you’ve made that equation meaningless, because now 99% of the network can disagree with you, the central authority, and still ‘lose’.
If I were an attacker, I’d take massive advantage of this. I’d start by forking the chain with a longer chain. This would cause transactions to be rolled back, while new ones get added. Then you checkpoint a point in the old chain AFTER the fork in order to get the network back on the chain, causing all the transactions between now and the time of the first fork to get rolled back. At this point, you have 2 active forks: Those who didn’t subscribe are on the forked, longer chain, and those who did are on the original, shorter chain, dividing your network power. Then I’d just do it again, and again, and again, and again, each time causing transactions to be rolled back, and at a reduced cost since you’ve divided your PoW resources with your checkpoint. I’d probably attack the original chain again, since Cryptsy would undoubtedly be on it having re-enabled FTC again after the attack, and pull it off again with less effort because some portion of your hashes are still mining the longer forked chain since they didn’t subscribe.
This is basically your worst case scenario, and you’ve made it EASIER for the attacker to pull off by introducing ACS. Every time it happens, confirmed transactions are unconfirmed, people lose money, and confidence plummets. This is without question the worst possible move to make.
Finally, PoS does not ‘slow down 51% attacks’, it makes them monetarily unfeasible and highly unproductive since the attacker would have to own 51% of the PoW AND 51% of the currency itself. It’s a GREAT solution, and if inflation is a concern, PoB is a great solution for that.
-
[quote name=“Kevlar” post=“23422” timestamp=“1375006405”]
[quote author=wesphily link=topic=2934.msg23413#msg23413 date=1374986405]
The trick with check pointing is that you don’t have to subscribe if you don’t want to. Also, the number of people agreeing on the checkpoint will keep it decentralized.Finally, PoS does not stop 51% attack. All it does is slow it down.
[/quote]Then all it takes is one person not subscribing to disagree, creating forks in the chain. What then? Sucks to be that disenfranchised guy? What if a majority of miners form and turn off checkpointing, and then someone on high rolls back some transactions with a checkpoint? You lose > 51% of your hashpower, making your new forked shorter chain vulnerable to a 51% attack!
Wait, didn’t the attacker already have 51% to begin with? Well now we’re screwed…
You’ve not solved any problems with this solution, you’ve just created more problems and a forked chain that only you and your ‘centralized list’ of mining pools deem valid. The problem IS that more than 51% of the work is controlled already, so all you’ve done is centralize the power to decide the fate of the chain AWAY from miners. If someone is deciding who those people are and maintaining that list, then the power is centralized to the owner of the list. Someone has to say, “Hey, we should checkpoint -this- chain in order to eliminate -that- fork.” and that’s centralization. Someone has to maintain the list of voters, and that’s centralization. It’s bad news all the way to the bank. The design is 1 CPU cycle==1 vote, and you’ve made that equation meaningless, because now 99% of the network can disagree with you, the central authority, and still ‘lose’.
If I were an attacker, I’d take massive advantage of this. I’d start by forking the chain with a longer chain. This would cause transactions to be rolled back, while new ones get added. Then you checkpoint a point in the old chain AFTER the fork in order to get the network back on the chain, causing all the transactions between now and the time of the first fork to get rolled back. At this point, you have 2 active forks: Those who didn’t subscribe are on the forked, longer chain, and those who did are on the original, shorter chain, dividing your network power. Then I’d just do it again, and again, and again, and again, each time causing transactions to be rolled back, and at a reduced cost since you’ve divided your PoW resources with your checkpoint. I’d probably attack the original chain again, since Cryptsy would undoubtedly be on it having re-enabled FTC again after the attack, and pull it off again with less effort because some portion of your hashes are still mining the longer forked chain since they didn’t subscribe.
This is basically your worst case scenario, and you’ve made it EASIER for the attacker to pull off by introducing ACS. Every time it happens, confirmed transactions are unconfirmed, people lose money, and confidence plummets. This is without question the worst possible move to make.
Finally, PoS does not ‘slow down 51% attacks’, it makes them monetarily unfeasible and highly unproductive since the attacker would have to own 51% of the PoW AND 51% of the currency itself. It’s a GREAT solution, and if inflation is a concern, PoB is a great solution for that.
[/quote] -
[quote name=“Kevlar” post=“23422” timestamp=“1375006405”]
[quote author=wesphily link=topic=2934.msg23413#msg23413 date=1374986405]
The trick with check pointing is that you don’t have to subscribe if you don’t want to. Also, the number of people agreeing on the checkpoint will keep it decentralized.Finally, PoS does not stop 51% attack. All it does is slow it down.
[/quote]Then all it takes is one person not subscribing to disagree, creating forks in the chain. What then? Sucks to be that disenfranchised guy? What if a majority of miners form and turn off checkpointing, and then someone on high rolls back some transactions with a checkpoint? You lose > 51% of your hashpower, making your new forked shorter chain vulnerable to a 51% attack!
You’ve not solved any problems with this solution, you’ve just created more problems and a forked chain that only you and your ‘centralized list’ of mining pools deem valid. The problem IS that more than 51% of the work is controlled already, so all you’ve done is centralize the power to decide the fate of the chain AWAY from miners. If someone is deciding who those people are and maintaining that list, then the power is centralized to the owner of the list. Someone has to say, “Hey, we should checkpoint -this- chain in order to eliminate -that- fork.” and that’s centralization. Someone has to maintain the list of voters, and that’s centralization. It’s bad news all the way to the bank. The design is 1 CPU cycle==1 vote, and you’ve made that equation meaningless, because now 99% of the network can disagree with you, the central authority, and still ‘lose’.
If I were an attacker, I’d take massive advantage of this. I’d start by forking the chain with a longer chain. This would cause transactions to be rolled back, while new ones get added. Then you checkpoint a point in the old chain AFTER the fork in order to get the network back on the chain, causing all the transactions between now and the time of the first fork to get rolled back. At this point, you have 2 active forks: Those who didn’t subscribe are on the forked, longer chain, and those who did are on the original, shorter chain, dividing your network power. Then I’d just do it again, and again, and again, and again, each time causing transactions to be rolled back, and at a reduced cost since you’ve divided your PoW resources with your checkpoint. I’d probably attack the original chain again, since Cryptsy would undoubtedly be on it having re-enabled FTC again after the attack, and pull it off again with less effort because some portion of your hashes are still mining the longer forked chain since they didn’t subscribe.
This is basically your worst case scenario, and you’ve made it EASIER for the attacker to pull off by introducing ACS. Every time it happens, confirmed transactions are unconfirmed, people lose money, and confidence plummets. This is without question the worst possible move to make.
Finally, PoS does not ‘slow down 51% attacks’, it makes them monetarily unfeasible and highly unproductive since the attacker would have to own 51% of the PoW AND 51% of the currency itself. It’s a GREAT solution, and if inflation is a concern, PoB is a great solution for that.
[/quote]The attackers chain will die when they stop hashing. Their fork will only exist as long as they are running the attack, once they stop and the pools regain the majority any clients that moved to the attackers chain will revert back. With the major exchanges and pools subscribed to the same chain it will be no benefit to subscribe to a different one. If people really want to run their own chain off of Feathercoin they could, but they would not be able to play with the subscribed network. Good luck to them but they will not get in the front door of BTC-e :)
PoS breaks our inflation model. I will look into PoB.
Everyone cheer up. So many sad faces at the minute. It cannot be sunshine and lollipops all the time. I will be here working away on Feathercoin through think and thin. We have good things coming after we get Advanced Checkpointing implemented. There seems little point progressing some things until we have some protection in place. Having stayed quiet for a couple of weeks has seen the attackers move to other coins. Let’s keep our heads down for a little bit longer.
-
Thank you bushstar for explaning it! :D
You should definitly prioritize to be more active on the forum, its nice to get some straight facts from time to time for people to keep focus,
and cheere the coin forward. -
[quote name=“Bushstar” post=“23448” timestamp=“1375035875”]
Having stayed quiet for a couple of weeks has seen the attackers move to other coins.
[/quote]yep. I can’t imagine who would be targeted next … [url=http://digitalcoin.co/forums/index.php/topic,4.0.html]http://digitalcoin.co/forums/index.php/topic,4.0.html[/url]
-
[quote name=“Bushstar” post=“23448” timestamp=“1375035875”]
[quote author=Kevlar link=topic=2934.msg23422#msg23422 date=1375006405]
[quote author=wesphily link=topic=2934.msg23413#msg23413 date=1374986405]
The trick with check pointing is that you don’t have to subscribe if you don’t want to. Also, the number of people agreeing on the checkpoint will keep it decentralized.Finally, PoS does not stop 51% attack. All it does is slow it down.
[/quote]Then all it takes is one person not subscribing to disagree, creating forks in the chain. What then? Sucks to be that disenfranchised guy? What if a majority of miners form and turn off checkpointing, and then someone on high rolls back some transactions with a checkpoint? You lose > 51% of your hashpower, making your new forked shorter chain vulnerable to a 51% attack!
You’ve not solved any problems with this solution, you’ve just created more problems and a forked chain that only you and your ‘centralized list’ of mining pools deem valid. The problem IS that more than 51% of the work is controlled already, so all you’ve done is centralize the power to decide the fate of the chain AWAY from miners. If someone is deciding who those people are and maintaining that list, then the power is centralized to the owner of the list. Someone has to say, “Hey, we should checkpoint -this- chain in order to eliminate -that- fork.” and that’s centralization. Someone has to maintain the list of voters, and that’s centralization. It’s bad news all the way to the bank. The design is 1 CPU cycle==1 vote, and you’ve made that equation meaningless, because now 99% of the network can disagree with you, the central authority, and still ‘lose’.
If I were an attacker, I’d take massive advantage of this. I’d start by forking the chain with a longer chain. This would cause transactions to be rolled back, while new ones get added. Then you checkpoint a point in the old chain AFTER the fork in order to get the network back on the chain, causing all the transactions between now and the time of the first fork to get rolled back. At this point, you have 2 active forks: Those who didn’t subscribe are on the forked, longer chain, and those who did are on the original, shorter chain, dividing your network power. Then I’d just do it again, and again, and again, and again, each time causing transactions to be rolled back, and at a reduced cost since you’ve divided your PoW resources with your checkpoint. I’d probably attack the original chain again, since Cryptsy would undoubtedly be on it having re-enabled FTC again after the attack, and pull it off again with less effort because some portion of your hashes are still mining the longer forked chain since they didn’t subscribe.
This is basically your worst case scenario, and you’ve made it EASIER for the attacker to pull off by introducing ACS. Every time it happens, confirmed transactions are unconfirmed, people lose money, and confidence plummets. This is without question the worst possible move to make.
Finally, PoS does not ‘slow down 51% attacks’, it makes them monetarily unfeasible and highly unproductive since the attacker would have to own 51% of the PoW AND 51% of the currency itself. It’s a GREAT solution, and if inflation is a concern, PoB is a great solution for that.
[/quote]The attackers chain will die when they stop hashing. Their fork will only exist as long as they are running the attack, once they stop and the pools regain the majority any clients that moved to the attackers chain will revert back. With the major exchanges and pools subscribed to the same chain it will be no benefit to subscribe to a different one. If people really want to run their own chain off of Feathercoin they could, but they would not be able to play with the subscribed network. Good luck to them but they will not get in the front door of BTC-e :)
PoS breaks our inflation model. I will look into PoB.
Everyone cheer up. So many sad faces at the minute. It cannot be sunshine and lollipops all the time. I will be here working away on Feathercoin through think and thin. We have good things coming after we get Advanced Checkpointing implemented. There seems little point progressing some things until we have some protection in place. Having stayed quiet for a couple of weeks has seen the attackers move to other coins. Let’s keep our heads down for a little bit longer.
[/quote]That’s not true though. The attackers chain will be longer, so it will be held on to by all the non-subscribing clients, who will continue to post transactions to it. At some time in the future, LONG AFTER the attack has stopped, the CENTRALIZED list of pools will eventually overcome the longer chain with their shorter chain, assuming the pools all subscribe. At that time, all the transactions on what was the longer chain will be unconfirmed. All it takes is one rogue pool to make the situation take longer to play out, causing more damage in the form of more transactions unconfirmed. This all assumes no one is asleep at the checkpointing wheel, and that the guy doing the checkpointing can be trusted and isn’t himself compromised. You didn’t address how during this entire process people are literally losing their coins on these switching chains. Did you think for one minute that when a million dollars is on the line the attacks won’t become more sophisticated?
This is EXACTLY what the attacker wants, and you’re giving it to him on a silver platter. Congratulations, you have just eliminated the very thing which makes decentralized cryptocurrencies potent: decentralization, one cpu cycle=1 vote, no chargebacks, and replaced it with centralization, 1 person = all the votes, and transactions that can become unconfirmed hours after confirmation. It’s a hackers paradise, and a thief’s wet dream: Guaranteed rollbacks. All I gotta do is spend the coins on the attackers chain before the checkpoint is put into place, and 15 minutes later when the checkpoint goes out I get them all back. Free money!
Why doesn’t anyone believe me when I tell you this is the worst thing you could possibly do?
-
Kevlar, that’s very fire and brimstone :) No offense, but please do not have a full blown panic attack on the public forums.
So perhaps we would be safer if the checkpointing was fully centralised and hard coded into the client? No one could then create their own feeds to create their own chains. As for centralised control of the checkpoints, this is a trade off that we will be making to guarantee the integrity of our blockchain. It is not as if we are fully decentralised, there are several DNS Seed nodes hard coded into the client which I maintain. This is a trade off that gives the network as a whole stability and saves people manually using the addnode option.
Considering the scale of the attacks and the harm this does to public confidence a trade off of this nature is a small price to pay for blockchain security.
-
[quote name=“Kevlar” post=“23295” timestamp=“1374920392”]
[quote author=Markus11 link=topic=2934.msg23289#msg23289 date=1374918348]
Selling your coins changes what exactly?
If youre unsatisfied performane the moves which satisfy you but please dont share it here … all you do is creating bad mood - capitalistic pov … thays what i call support
[/quote]No, that’s actually exactly what this forum is for: The sharing of opinions related to the currency, and not just the ones you happen to like either. I’ve done nothing but support the members of this community with my efforts, and you better believe that a capitalistic PoV is appropriate when we’re [u][i][b]discussing the performance of a currency[/b][/i][/u].
[/quote]Some people associate capitalism with manipulating, bullying, complaining, isolating. Wonder why that is…
Kevlar has been a big support. If things aren’t moving forward it’s likely because people are sitting on their butts. Truth is we’re all going blind here. We are glued to the old ways of doing business and it’s going to take some time to figure out what each of us wants.
-
Advanced checkpointing is something that can be removed once the network is big enough (secure), is that correct, or not, or not that simple?
