Enhancement to ACP
-
Wouldn’t this create the risk of allowing people to pay to become the ACP node? What’s the risk of someone being the ACP node for a turn?
-
It’s not really paying, as they simply have to place an amount of coins in a special wallet to become eligible as an ACP node. If they withdraw the coins, they will loose the right to ACP sign blocks.
As there will be more than one ACP node in the network, being an ACP node grants the right to sign one block with ACP and then wait until it’s the turn of the node again. Depending on the number of ACP nodes, nobody has the power to force major changes in the block chain, as the maximum number of blocks that could be invalidated by signing a different branch of the BC is 5 if every 5th block is ACPed as it is the case at the moment. If there are 51 active ACP nodes active in the network, one node would be allowed to ACP sign a block every 255 minutes or every 4.25 h
A possible attacker could inject 5 own blocks or force the BC to a different branch every 4.25 hours, which is not a valuable target.
Also I have a voting system for active ACP nodes in mind, comparable to the mechanism DASH has implemented, I just don’t have an idea how to implement that.
With a voting system in place and more ACP aspirants than active ACP nodes, the power a single ACP node has over the network is diluted even more and the whole mechanism is distributed.
Over time the ACP nodes could also provide additional services besides signing blocks with ACP.
One I could think of is that the net of ACP nodes provide a distributed escrow service.
Prerequisite for additional services would be a reward system for beeing ACP node.
-
I would like to hear more about this. I’m still wondering what advantage it would have over DPoS.
-
You still have the issue of mining waste. Slim to zero profits for miners means the whole point of block incentive is broken for all but the most efficient farms.
-
-
an ACP node is a full node. Can get some reward when user run it ?
-
Fixed number of nodes ? keep top 101 ACP node in network list .
-
-
I would like to hear more about this. I’m still wondering what advantage it would have over DPoS.
Over time it can provide a very similar function to DPOS. I see it as a evolution to implement DPOS features or may be even DPOS without the need to create a new BC.
You still have the issue of mining waste. Slim to zero profits for miners means the whole point of block incentive is broken for all but the most efficient farms.
Correct, this doesn’t change the hashing algorithm, so the current situation remains.
On the other hand DPOS excludes most people from mining totally, as they are not in the top 101 delegate lists, and a high entrance deposit (~ 1000 USD for bitshares and DASH?) hast to be made to even register as a delegate.
Currently we still have several hundred or thousands of miners, with DPOS we have 101, even if these chance time by time. This also can be discouraging for people.
-
-
an ACP node is a full node. Can get some reward when user run it ?
-
Fixed number of nodes ? keep top 101 ACP node in network list .
- As I mentioned above, we could implement a mechanism, that every 10th block would be mined by an ACP node, for example with lowest difficulty, so not much power would be wasted. The ACP node mined block would be the Checkpoint for the BC at the same time.
- Yes, I was thinking about a fixed number of ACP nodes. I thought about a selection mechanism, may be based on forum reputation, or whatever measure. There will be better ones than forum reputation. We just need t agree on the measure.
- For using the 101 top nodes as ACP nodes, it could be the measure for the selection/election process I thought about.
- Question: what is the definition of top node?
- uptime? can’t be read from the BC
- number of blocks found/announced?
- wallet value?
-
-
Over time it can provide a very similar function to DPOS. I see it as a evolution to implement DPOS features or may be even DPOS without the need to create a new BC.
Correct, this doesn’t change the hashing algorithm, so the current situation remains.
On the other hand DPOS excludes most people from mining totally, as they are not in the top 101 delegate lists, and a high entrance deposit (~ 1000 USD for bitshares and DASH?) hast to be made to even register as a delegate.
Currently we still have several hundred or thousands of miners, with DPOS we have 101, even if these chance time by time. This also can be discouraging for people.
That’s not true about excluding people. What about the purpose of DPoS in empowering the community? People need both tokens (FTC) for spending and community shares (from BC2) for voting and having a voice. Turning FTC’s tokens into voting rights just won’t work. Shares should be distributed as far and wide as possible to help the community. We want the larger community to have a voice right? The distribution of tokens is already heavily centralised, you don’t want people to avoid spending FTC because doing so weakens their already weak voice in the community.
People are already excluded from mining, they are forced to mine at a loss. As it stands they have no incentive to participate unless they are super wealthy with a mining farm on free juice. That’s not inclusive! Inclusive would be saying, “hey, the community acknowledge how hard you are working, have our votes. You deserve to be a 101 delegate”. One encourages participation and the other encourages more centralisation upon the wealthy.
-
IMO, this thread should be discussed in public.
-
That’s not true about excluding people. What about the purpose of DPoS in empowering the community? People need both tokens (FTC) for spending and community shares (from BC2) for voting and having a voice. Turning FTC’s tokens into voting rights just won’t work. Shares should be distributed as far and wide as possible to help the community. We want the larger community to have a voice right? The distribution of tokens is already heavily centralised, you don’t want people to avoid spending FTC because doing so weakens their already weak voice in the community.
I don’t see a need to have voting rights, if we limit the max number of coins that are allowed to use for a given voting.
it could be just 1 coin for a voting.
People are already excluded from mining, they are forced to mine at a loss. As it stands they have no incentive to participate unless they are super wealthy with a mining farm on free juice. That’s not inclusive! Inclusive would be saying, “hey, the community acknowledge how hard you are working, have our votes. You deserve to be a 101 delegate”. One encourages participation and the other encourages more centralisation upon the wealthy.
What about the registration fee (probably not the rigth word, but I have no better), that one must put at stake to become a delegate?
This is a hurdle already to become a delegate and earn he right to mine coins.
For bitshares it is USD 1000 and if I remember right for DASH it is the equivalent to the number of coins mined in two weeks, which is at current rate also > USD 10002nd:That mining is not profitable, unless you get energy at nearly no cost is a different topic and related to the curreent exchange rate, therefore I’d like to keep this out of the discussion at the moment.
- One topic is to secure the block chain and decentralize the processes used to do that as wide as possible and 2) a second one is to make mining profitable and less power consuming.
For me the first step is to concentrate on 1) and work on 2) later.
One way to archive 2) could be:
during implemenation of 1) we decide to give an intensive for eACP (= enhanced ACP) nodes to be an eAcP node by getting every 10th in a round robin mode of the selected eACP nodes. block mined by an eACP node at minimum difficulty or even by different algorithm.
this would make it interesting to e an eACP node and also reducing power cost for eACP nodes, as these would need to only GPU mine.over time we can increase the rate of eACP mined blocks enhancing the power savings.
in the extreme we could end up in a schema, that is 90% comparable to DPOS. -
IMO, this thread should be discussed in public.
I fully agree. My intention is to discuss this internally, come to a conclusion or as close to a conclusion as possible and let the whole community discuss that conclusion.
I deem that way more efficient that a large public discussion from the start.
Also I hoped for a more technical discussion, where points like- Can we implement such a solution in the current code base?
- How much effort will it be to implement?
- How much time would be needed?
If the answer to the first question above is ‘no’, there is no need to go public with this thread, but we need to think about other ways to move forward.
Also it may be worth to have a hangout session just for this discussion and then go public
If the other team members favor a direct public discussion, we also can go that route.
-
We are not limited by how others choose to implement DPoS, we are only limited by how we choose to implement DPoS.
Anything ‘off’ blockchain should be avoided in regards to voting. Sure you could do voting on the blockchain using a single coin. You could just use a counter party layer and have the tokens programmed to return. Counter Party on BTC’s chain does that now. But how would you make sure that people were not using multiple IDs? At least with stake voting you need huge resource to influence only a small percentage of the vote for delegates.
Energy costs are the elephant in the room. The whole point of crypto currency is that the network security has an incentive. We have lost the bulk of our community miners because they can’t compete. Until you fix that, there is nothing in it for the ‘little guy’ and the network will become centralised over time. So whilst the ‘little guy’ can pick up a GPU and hash away with no barrier to entry, he still has no incentive to do so. How many of our Team are actively mining?
-
I think we all agree, that PoW in it’s current flavor is power intensive and we should find a way to mine coins with less energy.
We could do that by shortening the interval between eACP mined blocks and reduce overall power consumption, as these blocks can be mined either at low difficulty with a GPU or even CPU or we could use a totally different hashing algorithm for these blocks, which requires less power to be calculated.
-
Decided to move the topic to the public discussion now to get more thoughts, ideas and improvements discussed
-
I know how to get rid of the ACP entirely to stay decentralised and secure fairly. However I have put my development efforts in this direction on hold half a year ago and not going to resume unless we get funding. It’s not going to happen like with NeoScrypt which has been a large time killer and financial loss for me the last year. It could be much better if I just did my real life job and got paid for it. I hope you’re not surprised.
-
I know how to get rid of the ACP entirely to stay decentralised and secure fairly. However I have put my development efforts in this direction on hold half a year ago and not going to resume unless we get funding. It’s not going to happen like with NeoScrypt which has been a large time killer and financial loss for me the last year. It could be much better if I just did my real life job and got paid for it. I hope you’re not surprised.
Hi Ghost,
I suggest to talk with markmod ;)
Meanwhile, can you explain your idea in more details?
m.
-
Hi Ghost,
I suggest to talk with markmod ;)
Meanwhile, can you explain your idea in more details?m.
I know. I can tell the details to those involved in the actual development, but not in public right now.
-
Very good to know, that there is a solution to replace ACP :)
I’m very interested to hear about your solution.
-
for all members, can you answer for follow questions:
1. What is the problem?
2. What is the CAUSE of the problem?
3. What are all possible solutions to the problem?
4. What solution do you suggest? -
for all members, can you answer for follow questions:
1. What is the problem?
-
centralization
2. What is the CAUSE of the problem? -
one person holding key to the network
3. What are all possible solutions to the problem? -
so far I think abotut:
a, POW+0%POS
b, DPOS
in that architectures ACP is not necessary
4. What solution do you suggest?I like adding 0% POS and eventually switching to DPOS as we establish solid and stable position
-
