BitShares Forum
Main => General Discussion => Topic started by: clayop on January 21, 2015, 01:17:59 am
-
With 11% of stake, I can obtain about 65 delegate positions. If I attempt to use these 65 delegates maliciously to make profit, what happens? I'm asking because many people keep asking me about this and I cannot fully understand this process.
Thanks in advance
-
With 11% of stake, I can obtain about 65 delegate positions. If I attempt to use these 65 delegates maliciously to make profit, what happens? I'm asking because many people keep asking me about this and I cannot fully understand this process.
Thanks in advance
we should consider how to encourage people to vote more.
-
It's essentially a 51%-attack of the "total voting-stake" ..
countermeasure, as BTSdac stated, would be to increase the "total voting-stake" by encouraging people to vote
-
With 11% of stake, I can obtain about 65 delegate positions. If I attempt to use these 65 delegates maliciously to make profit, what happens? I'm asking because many people keep asking me about this and I cannot fully understand this process.
Thanks in advance
You mean like, what if someone put a gun to Bytemaster's head who controls 15% of the supply? :)
-
With 11% of stake, I can obtain about 65 delegate positions. If I attempt to use these 65 delegates maliciously to make profit, what happens? I'm asking because many people keep asking me about this and I cannot fully understand this process.
Thanks in advance
You mean like, what if someone put a gun to Bytemaster's head who controls 15% of the supply? :)
He doesn't control that much though, especially after the severance packages he paid out to the core developers. More like 8% imo..
Edit: From recent voting history I can say BM's current active voting power is only about 5.7%. He may of course control more funds, especially vested funds, but that's what he's been using recently to vote at least.
-
With 11% of stake, I can obtain about 65 delegate positions. If I attempt to use these 65 delegates maliciously to make profit, what happens? I'm asking because many people keep asking me about this and I cannot fully understand this process.
Thanks in advance
You mean like, what if someone put a gun to Bytemaster's head who controls 15% of the supply? :)
He doesn't control that much though, especially after the severance packages he paid out to the core developers. More like 8% imo..
Edit: From recent voting history I can say BM's current active voting power is only about 5.7%. He may of course control more funds, especially vested funds, but that's what he's been using recently to vote at least.
It means that someone will need a few more guns to control the other devs' heads.
-
I think a lot of people are currently not voting out of convenience. Once there is a serious attack, that puts their investment at risk they would probably get their sh** together.
Despite that I would expect drastically increased voting participation once there is a more usable wallet and offline transaction signing.
-
我想，你先有11%的股份再说吧。你拥有11%的股份，还希望垮掉，我觉得你的心态有问题、
-
The main thing is making cold storage voting safe and easy I think.
Other possible ideas to increase voting..
1. To claim your yield you need to vote.
2. A BTS lottery that requires voting to enter. A lottery can at times get very large which will incentivise people to vote when it does.
-
in fact, techically speaking at this moment you don't need 51% of stake to attack BTS, 16% are enough. if you have that much stake you can vote all of your 101 delegates in, which means you take full control of the whole dpos network since right now the delegate which gains most votes is only supported by less than 16% stake. when you do so, you can have all of your delegates disabled block production.
or with a 10% stake you can control over the half of delegates and sign on a alternative "main" chain (?)
yes, I think so. with 10% stake you can control 57 delegates at this moment, then reject all the blocks signed by the rest 44 delegates and also exclude the transactions which don't vote for you. but we are only talking about this technically.
-
我想，你先有11%的股份再说吧。你拥有11%的股份，还希望垮掉，我觉得你的心态有问题、
That's true but then it's peanuts to those existing third parties who might be threatened by BitShares. I don't expect in reality any are so corrupt or feel so threatened that they would launch an attack before it became too expensive for even them but it's worth limiting risks wherever they are.
-
An immediate and effective solution to this issue is actually very simple. Malicious delegates will likely be delegates that don't produce blocks or produce blocks and steal 100% of the inflation. Both scenarios are by themselves quite unlikely and not actually that hurtful to BitShares, but most importantly they are easy to detect and it's possible to see what accounts control the stake that voted for them. To ensure the malicious stake doesn't continue ruining things a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
-
Yup, couple of days downtime max, attacker would've spent 3 million dollars for a temporary disruption, and we would have made a strong point to any literal-minded person that the social zeitgeist will prevail.
-
a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
That's a viable solution now, but not when bitshares is at the scale of bitcoin. Can you imagine asking all the businesses running bitcoin clients to all update simultaneously whenever something like this occurs?
-
By the time we are the size of bitcoin the cost of the attack will be 300 m usd. I don't see it happening. Certainly more expensive than buying all of the btc hash power for a year.
-
An immediate and effective solution to this issue is actually very simple. Malicious delegates will likely be delegates that don't produce blocks or produce blocks and steal 100% of the inflation. Both scenarios are by themselves quite unlikely and not actually that hurtful to BitShares, but most importantly they are easy to detect and it's possible to see what accounts control the stake that voted for them. To ensure the malicious stake doesn't continue ruining things a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
Destroying all the stake is less than ideal. I don't think there is a way how to discriminate between votes really intended for the malicious delegate and votes selected with "vote random subset" and "vote as delegates recommend".
I know it would be probably just a small portion of votes, but in principle it wouldn't be right.
-
An immediate and effective solution to this issue is actually very simple. Malicious delegates will likely be delegates that don't produce blocks or produce blocks and steal 100% of the inflation. Both scenarios are by themselves quite unlikely and not actually that hurtful to BitShares, but most importantly they are easy to detect and it's possible to see what accounts control the stake that voted for them. To ensure the malicious stake doesn't continue ruining things a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
Destroying all the stake is less than ideal. I don't think there is a way how to discriminate between votes really intended for the malicious delegate and votes selected with "vote random subset" and "vote as delegates recommend".
I know it would be probably just a small portion of votes, but in principle it wouldn't be right.
"vote random subset" means "vote for a random few of the delegates that I've approved in my wallet." So even when you have that option checked (as you generally should for the sake of your privacy), you're never voting for delegates that you haven't already specifically approved.
I'm not sure how "vote as delegates recommend" works.
-
An immediate and effective solution to this issue is actually very simple. Malicious delegates will likely be delegates that don't produce blocks or produce blocks and steal 100% of the inflation. Both scenarios are by themselves quite unlikely and not actually that hurtful to BitShares, but most importantly they are easy to detect and it's possible to see what accounts control the stake that voted for them. To ensure the malicious stake doesn't continue ruining things a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
Thanks. This is clear answer for me.
-
I'm not sure how "vote as delegates recommend" works.
From:
http://wiki.bitshares.org/index.php/DPOS/ApprovalVoting
Vote as Delegates Recommended
Some users publish a set of delegates (a slate) they recommend. Theses delegates fulfill certain criteria that are defined by the particular users. As an example: A user only recommends delegates whose real-world identities are known and verified. Another user recommends delegates that are trusted members of the bitsharestalk.org forum. And so on. If that user is a delegate and you vote for him with a wallet_transfer, you can also vote for all of his recommended delegates by choosing
Vote as Delegates Recommended
-
An immediate and effective solution to this issue is actually very simple. Malicious delegates will likely be delegates that don't produce blocks or produce blocks and steal 100% of the inflation. Both scenarios are by themselves quite unlikely and not actually that hurtful to BitShares, but most importantly they are easy to detect and it's possible to see what accounts control the stake that voted for them. To ensure the malicious stake doesn't continue ruining things a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
Stealing is not objective. What about a dev that doesn't really work (effectively)?
-
An immediate and effective solution to this issue is actually very simple. Malicious delegates will likely be delegates that don't produce blocks or produce blocks and steal 100% of the inflation. Both scenarios are by themselves quite unlikely and not actually that hurtful to BitShares, but most importantly they are easy to detect and it's possible to see what accounts control the stake that voted for them. To ensure the malicious stake doesn't continue ruining things a hard fork can be manually circulated among the community that permanently destroys all the stake that voted for the malicious delegates. BitShares would be back up and running normally quite quickly, and the only person who would really be hurt would be the attacker.
Stealing is not objective. What about a dev that doesn't really work (effectively)?
Or, what about an account squatter who receives funds from others when they are careless?
-
First of all an attacker attempting to steal 100% of the dilution and produce nothing would have to first pay 2 weeks worth of pay * 101 slots. That would cost about $100,000 or more. As long as the network could vote them out in less than 2 weeks that isn't a problem.
Only an attacker that takes over and then stops including transactions would require a "hard fork". I suspect that it would be trivial for any of the previously elected delegates to black list transactions that vote for the attacker. They could then easily hard fork out the transactions that voted in the attacking delegates and block any future transactions that would vote for them.