Topic: How good is Decred? - page 5. (Read 8708 times)

Well $18+ million dollars is a lot of money to attack a $50 million dollar network... come on.

You are 100% correct though, Decred, like bitcoin, relies on a majority of honest hash power and stakers.

No we can't agree, because if you took the time to actually analyze deeply what happens with selfish mining strategies and stake combined, the PoW is going to become rapidly centralized but you will have no way to actually know it.

I am not really concerned about double-spending. More so I am concerned about attacks when it comes time to make protocol change decisions. Also the controlling stake+PoW will extract the maximum fees and they will monopolize everything.

It is really about milking the thing to extract all the value out of it. The system is designed to establish cartels which are rent seekers. Which goes against everything we hope to achieve with blockchains.

I hate shit like this, so it is difficult for me to agree. Sorry.

They are a form of centralization... that was never denied.  We would like better decentralization of pools / better distribution of pool users.

I'm not sure why you are so angry man.  I'm not arguing with your math; you're obviously not a fool. 

However, it is prohibitively expensive to do what you describe.  Can we agree on that?

Tell that to the girls by the xerox machine.

Meanwhile those who understand math and game theory know that the adjective (not name) prescribed to you is accurate.

You guys are faking some level of competency that you don't have.

There is no free lunch in consensus designs. You don't go merging PoS and PoW without introducing crazy game theory weaknesses. You guys are just lazy to actually find the weaknesses. And you even deny the math that was presented in the PoA paper. Pretending that staking pools aren't a form of centralization (as if the person who can't be online can delegate the responsibility without causing centralization).

You always know when someone has lost when they resort to name calling.  Have a nice day brother.

You're an idiot. You deny the PoA math. And who ever wrote your white paper has not properly analyzed the vulnerabilities when combined with selfish mining.

LOL, come on man, just crawl back in your cave.

A little FUD is nice; we don't get much at Decred, but show up with better material next time

Apparently you don't know what the word centralization means.


That was written by an idiot. Who wrote that?

That is not a correct analysis of the math and game theory. The attacking stake has a percentage of the stake miners and they can use this along with the selfishing mining strategy to amplify it.

This is simply not true!  If users are not online when their tickets are called to vote, they lose the subsidy.  It's not like traditional PoS where you only receive subsidy when you are staking.  There is a heavy monetary incentive to be online 100% of the time.  This is why Decred has stake pools to allow for offline staking.  

From the whitepaper (that you still did not read):

"It should also be noted that many well known mining
attacks, such as selfish mining [26] and stubborn mining
[27], will no longer function advantageously in a system
where there is effective decentralization of stake mining
and no PoW-PoS miner collusion. This is simply
because it is impossible to generate secret extensions
to blockchains without the assistance of stake miners."


I'm not here to argue with you about vulnerabilities.  Every system has them.  Hell, there are several ways to attack bitcoin.

But people just don't want to spend 18 million usd to destroy decred (and lose their investment).  Over time, it will only become more expensive.

Being good at copying things doesn't mean we should trust them (you) with knives, i.e. consensus algorithm design.

They (you) should stay near the girls at the xerox machine.


Obviously you didn't read what I wrote.

There is no way you will keep any where near 50% of the users online and participating in staking.

So run the numbers with a more realistic participation percentage and you can see that this design can't possibly scale up to mass adoption. In the power-law distribution only about 33% of the wealth will be serious owners and of them many are not going to bother staking since that will end up pretty much an unprofitable activity due to competition. So in reality the math is going to look more like:

((1/0.33 - 1) × 0.20)³ = 0.06

Where the attacking whales only need 6% of the hashrate to double-spend. And I didn't even get into the math of selfish-mining attacks which are much much more damning. The security is absolute dogshit. The selfish mining is going to be much worse and quickly centralized the PoW control of the coin (even though nobody can detect this).

The only way this design makes any sense is for a very centralized stake where we trust that overlord. Which is precisely what you guys are counting on and you damn well know it. Else you are really really bozos if you can't do some simple math.

Whales can profitably attack their own coin by shorting it before they do. They can turn the price into a yoyo, going long and short as they know the timing. Centralized shit is a clusterfuck.

I hope you are proud to work on this shit. 

Not technically "merge mining", as Decred does not have aux-pow and eth uses another pow algo... however, I agree.  The dual miner does contribute significantly to our network's security.

Do not step on the gecko!

The only reason I have Decred is because it can be merge mined with Ethereum...that itself doesn't contribute to the discussion here but I think that it's probably important to note that a lot of the network strength in Decred comes from ETH merged-mining.

Yes, those "bozos" are the developers... they happen to be well respected in the bitcoin community for their work on btcsuite.  

BTC Suite Github:
https://github.com/btcsuite

Decred Github:
https://github.com/decred

That vulnerability is discussed in our whitepaper, which you can read here (you obviously didn't):
https://decred.org/dtb001.pdf

Conducting such an attack would require one to hold over 50% of active staking tickets (which would cost upwards of 18 million usd at this point in time assuming all tickets were purchase at the 30 day vwap [ once again... extremely difficult to do without significantly increasing ticket prices / the price of the attack])
It would also require a large percentage of the PoW hashrate to be controlled by the attacker...  again, prohibitively expensive.

Decred is based on PoA which is a research paper written by 4 respected computer scientists including Meni Rosenfeld. However, the security has a major flaw which they admit in their equation.

Notice in their example if a whale had 20% of the stake, then assuming 50% of the stake is online and participating then that whale would need more than 8 times the honest hashrate in order to double-spend, i.e. 88.8% of the hashrate (because 8 x 11.2 > 88.8%), where N = 3 is the number of stakeholders that sign each block.

((1/0.2 - 1) × 0.5)³ = 2³ = 8

But notice as a whales' (or colluding whales) percent of the stake climbs towards 50%, the design goes pear shaped and the security collapses:

((1/0.5 - 1) × 0.5)³ = 0.125

Which means you'd only need 12.5% of the hashrate to double-spend. The reason this is so, is because with the control over stake, the attacker can deny to respond to blocks found by the honest miners.

The problem is that to increase the leverage on the side of higher security at low levels of dishonest stake, by increasing N to 3 or greater, this causes the leverage to bite you back at higher levels of dishonest stake as shown above. And it is also damn bad at even lower levels of dishonest stake:

((1/0.33 - 1) × 0.5)³ = 1

Okay that is standard 51% attack but that assumes 50% of the stakeholders are participating in mining! Which isn't going to be likely for any mass adopted coin. Certain all the tokens on the exchanges aren't mining. At  33% participation it goes pear shaped again:

((1/0.33 - 1) × 0.33)³ = 0.29

Actually I am very disappointed that Meni would sign onto work like this. Because when you combine this with selfish-mining strategies on withholding PoW blocks and releasing them late, the security is just horrendous.

This shitcoin is currently seeing a huge rise in price because n00bs don't know how to read a white paper. And thus they don't realize they are buying insecure shit that has no chance whatsoever of being the next big thing.

Its sad actually. Because this shit might continuing going up in price even though it stinks real bad.



Are these the bozos developers?

---

Note Charlie Lee admitted at the above Redditard that "PoA is an experiment and may not work".

They have hyped promo: The new Decred Promo Video... Never seen anything like it.

But he shills a lot:

I was busy before. I will give a real answer now

Decred stands for Decentralized Credits

If any coin truly understands decentralization this is it. The mixed PoW and PoS allows room for ASICs to later develop and strengthen the blockchain against attacks while keeping a check on a known BTC issue, when a centralized group of mining whales call all the shots. The PoS system and its governance over the PoW mining is FAR more inclusive than Dash masternodes which only allows wealthy whales to have a say over the little guys. BTC has an awful consensus issue with 95% of miners needing to come to one decision for BTC to upgrade. In DCR just 30 credits is enough to enter PoS and earn DCR by having a vote in the future of the network. It's a currency by the people for the people more than any other coin can claim to be.

This stuff is not something for a cryptocurrency newbie to understand. I turned an eye to this when I was new just mining and holding what seemed to be the best moves. I did not understand the obsession with the overused word decentralized around here. I would read PoS and wonder wait do they mean Piece Of Shit? With some time reading up on coins, history lessons, seeing where some coins prevail or fail over governance issues and consensus, or centralization of various forms, it all began to make sense to me. It's clear to me that DCR has among the highest long term potentials in the entire lot

What makes Decred "good" is that it's not another clone.

It's inspired by Bitcoin but added with new features that the devs believe is beneficial for the long term. The best part of Decred is their open governance and hybrid proof-of-work and proof-of-stake mining system that ensures a small group can't dominate the flow of transactions or make changes to Decred without the input of the community. You're able to vote and have a voice, lots of potential for this project and out of 99% of the alts out there, this one will probably continue to rise and in my opinion is a worthy hold.

Had limited experience with it, but as I recall the transaction fees are insanely high. It seemed to start pumping around the time Coblee started shilling for it on twitter.  Has some unique governance system or some such thing.

haha, its def a good token, its its uniqueness that keeps it around and (now) raising in value tho
the PoS ticket system is just cool, GREAT way to tie up some value with GREAT APR%
Mining is unique, own algo, thats a selling point (usually)
and don't forget how they started, with the most profitable airdrop ever! (mine was worth 1BTC when i sold it)
DCR is sweet, tho pretty tecky, but hey: Crypto(graphy), right?