BitShares Forum
Main => General Discussion => Topic started by: bytemaster on January 07, 2015, 09:21:17 am
-
http://bytemaster.bitshares.org/article/2015/01/07/The-Most-Decentralized-Proof-of-Stake-System/
-
You should take a look at this research into NXT forging, where it shows the number of actual forgers is mostly above 200: https://nxtforum.org/general/network-analysis/ which is roughly twice as decentralised as the 101 delegates.
-
You should take a look at this research into NXT forging, where it shows the number of actual forgers is mostly above 200: https://nxtforum.org/general/network-analysis/ which is roughly twice as decentralised as the 101 delegates.
Yes... but what is the distribution... did 1 guy produce 99% and the other 199 produce 1%?
-
Some guy ask me to ask this question :
Can BTS add UDP protocol to enhance the stability of the network ?
Because the current TCP only protocol can not perform well when the network situation fluctuates even for a small period of time .
-
You should take a look at this research into NXT forging, where it shows the number of actual forgers is mostly above 200: https://nxtforum.org/general/network-analysis/ which is roughly twice as decentralised as the 101 delegates.
Yes... but what is the distribution... did 1 guy produce 99% and the other 199 produce 1%?
Given the (not so) recent events of buyouts of PTS/DNS/VOTE? decided solely by developers it should be accounted for points in favor of centralisation for Bitshares.
I'm in no way judging or accusing but until there is a vote inside the blockchain for such important decisions and it is demonstrated to work - bitshares remain a little bit more centralised in my mind.
I'm not saying this cannot happen to other project though but the fact that it happened to bitshares and currently there is no voting mechanism inplace is a little troubling.
-
Some guy ask me to ask this question :
Can BTS add UDP protocol to enhance the stability of the network ?
Because the current TCP only protocol can not perform well when the network situation fluctuates even for a small period of time .
Sure.. it is on the TODO list. This seems off topic.
-
You should take a look at this research into NXT forging, where it shows the number of actual forgers is mostly above 200: https://nxtforum.org/general/network-analysis/ which is roughly twice as decentralised as the 101 delegates.
Yes... but what is the distribution... did 1 guy produce 99% and the other 199 produce 1%?
Given the (not so) recent events of buyouts of PTS/DNS/VOTE? decided solely by developers it should be accounted for points in favor of centralisation for Bitshares.
I'm in no way judging or accusing but until there is a vote inside the blockchain for such important decisions and it is demonstrated to work - bitshares remain a little bit more centralised in my mind.
I'm not saying this cannot happen to other project though but the fact that it happened to bitshares and currently there is no voting mechanism inplace is a little troubling.
The core developers of any young alt-system have the ability to direct the project however they wish. Including Nxt and Peercoin. If the project is not mature and needs love and care of a group of dedicated developers then it will always be centralized.
What you described is not a property of DPOS, it is a property of social relationships and economics.
-
You should take a look at this research into NXT forging, where it shows the number of actual forgers is mostly above 200: https://nxtforum.org/general/network-analysis/ which is roughly twice as decentralised as the 101 delegates.
Yes... but what is the distribution... did 1 guy produce 99% and the other 199 produce 1%?
I think your article would carry more weight if it contained these stats, with referenced research.
-
Great post! You are really hitting it home with these articles, and I think it will have a huge impact in the following months as the insight you provide trickles down into the crypto community.
The weak points of the article is 1) lack of real data regarding PPC and NXT. If anyone could dig this up, it would really give the article a boost, and 2) the switch to talking about "in theory" 100% approval for Delegates after mainly doing practical analysis. These are really minor points.
The flow of the article is really good as it takes something complex but essential for everyone in this space and makes it easily comprehensible. In the PPC/NXT section however I feel this powerful clarity stumbling a bit in the following sentences,
"This means Peercoin and Nxt would need to generate about $36,000 in transaction fees per year or about 1 transaction per second at a minimum."
"Because transaction fees are distributed proportional to balance, we can assume that anyone with less than 1% of the stake will not be able to produce profitably and thereby be excluded from the consensus process."
These two sentences went a little too fast for me. Maybe unpacking them will ease the flow?
"steeling* profits" typo
-
What a great read. Really enjoying absorbing the logic in these blogs. :)
Gets my brain up and running in the mornings
-
"This means Peercoin and Nxt would need to generate about $36,000 in transaction fees per year or about 1 transaction per second at a minimum."
Isn't an assumption about the height of the tx fee necessary here?
Nxt, Peercoin, and likely Ethereum would all depend upon a wealthy elite with enough stake to profitably validate the network.
From what the article has provided as information up to this point and with no information about what security mechanism Ethereum will use it isn't true for ethereum (just creates unnecessary hostility).
-
Network costs in Peercoin are not paid by minters, but by everybody who holds stake without taking part in the minting, because they miss out on the 1% reward. Unlike in Nxt, Peercoin does not reward minters with the transactions fees, but with new coins. Transaction fees get destroyed.
Also, I can run a Peercoin and an NXT client on the same machine, cutting costs in half.
Peercoin can also be run on a Raspberrry Pi, which consumes way less than $30 a month in electricity costs. I can also use a machine to mint that is running anyway for other purposes.
-
At scale you cannot run it on a pi. Bandwidth costs eat majority of costs. Subsidizing costs via inflation just changed who pays, it does not get rid of costs.
-
Can you run a full btc node on a pi?
-
Can you run a full btc node on a pi?
Can you do that for bitshares ?
-
Can you run a full btc node on a pi?
Can you do that for bitshares ?
If it had enough RAM, we could do it at todays scale. We are not attempting to run full nodes on small devices because OUR goal is 1000 TPS trading and large scale.
-
At scale you cannot run it on a pi. Bandwidth costs eat majority of costs. Subsidizing costs via inflation just changed who pays, it does not get rid of costs.
For the foreseeable future, yes, because Peercoin is meant as a backbone currency, not for everyday transactions like Bitcoin. That's why Peershares and Nubits do not run on the Peercoin blockchain.
Hi bytemaster,
I may sound a bit harsh in this post, but please read it to the end. I very much respect the hard work you do for Bitshares, although I am not quite knowledgeable about it yet.
But the whole point you are making in your blog post is pointless in my opinion, and it seems you are building up a strawman argument to draw attention away from other respected cryptoprojects to benefit your own.
Security of any of the major Proof-of-stake systems is much higher than in PoW systems and network costs are minimal in any Proof-of-stake system. Why are there several hundred nodes online in both Nxt and Peercoin if it is too expensive to run a node as you say?
So that's why I regard the whole argument of your blog post about the network costs of Nxt, Peercoin and Bitshares as a strawman-argument.
Such a strategy may seem favourable to you in the short term, but for Crypto as a whole this is not wise.
I may remind you that we have got a hard time in crypto, because our real enemy are central banks, but the Bitcoin-weapon has some flaws.
We are already in a situation where in the short and medium term no Crypto 2.0 project will be able to surpass Bitcoin because there are simply too many. So people stick with Bitcoin regardless of its flaws.
And with a blog post like your's you aren't doing crypto any good.
We shouldn't become like the Judean people's front and the people's front of Judea:
https://www.youtube.com/watch?v=gb_qHP7VaZE
-
My goal is not to harm them, but to defend ourselves.