[Discussion] Consensus Mechanism during tail emission

The poll seems to be heading towards a tail emission in 4-5 years.

It is foreseen that this would require a change in consensus model to ensure the blockchain security is maintained when the block subsidy ends as it is unclear whether a masternode model is sustainable with a 1 FIRO/block emission.

While this is very preliminary, I thought it would be good to open up discussions on potential consensus models such as

a) Merged mining
b) Komodo notarization route (this is what Piratechain uses)
c) Pure PoW with a max reorg length
d) Proof of Stake

Currently I’m leaning towards a merged mining model to leverage chains with larger networks while remaining lean on infrastructure. It’s also a relatively proven model with Dogecoin being merge mined with Litecoin.

To me for a consensus model to be considered a viable candidate it should be:
a) Secure even with tail emissions
b) Censorship resistant (shouldn’t be dependent on only a handful of nodes)
c) Easy to implement (we don’t want to spend years coming up with a new consensus mechanism) (for those who remember the years we spent on MTP)
d) Maintains fast finality

Hope to hear some good suggestions!

I’ll have to look into option b) before making up my mind.
Right now, option a) seems like the obvious choice. It also may help expose Firo to wider communities.

1 Like

I am a big fan of Pirate Chain, and I think dPOW is a very strong system. It greatly reduces the possibility of a 51% attack making it almost impossible.

Pirate Chain has also floated the idea of tail emissions to be in a future upgrade.

1 Like

I’m not a big fan of DPoW as its security claims aren’t what they seem as the weak point is the notary nodes which is what you’re basically trusting the system to. While it is okay for some uses cases it isn’t exactly censorship resistant.

@sgp writeup on this is pretty comprehensive.

This is a good summary of the pros and cons @freQniK @Fiendish

2 Likes

For mass adoption, I think the max supply should be around 100 Millions or more. Tail emission 1 per block is too slow. I suggest keeping the status quo until forever, or tail emission up to 4 firo/block.

About concensus, I suggest Byzantine Fault Tolerance for speed and 1 block finality.

Do you have any data or examples of this?

1 FIRO/block is less than 1% inflation per year which is generally considered to be the sweet spot for a stable monetary supply.

Altering the supply from 21.4 to 100 million for mass adoption without solid data would be detrimental since I’m not sure how a higher supply contributes to mass adoption.

100m is way too much. This whole thing about mass adoption bothers me. No one coin needs to be mass adopted nor scale to 6 billion users. The likely outcome is when the time comes that you can pay for your coffee with crypto, you’ll have a choice of what crypto to pay with just like a lot of online payment gateways currently offer. As long as the entire industry scales, which it already does, there is no need for one coin to be the de facto means of payment.

I’ve read before that 1% is a sweet spot, you could get a way with a little higher even.

Also, im not well versed on dPoW as i didnt know there was a set of nodes holding the whole chain up. Knowing this, but still being a Pirate Chain supporter, i would try to avoid this architecture at all costs. Beyond censorship reasons, it also makes them a target for attack which would cripple the chain if there is no failsafe mechanism.

1 Like

BFT because of the need for each validator to communicate with every other validator means that these consensus mechanisms only can support up to a couple of hundred validators at best. This is an issue basically termed as ‘communication complexity’. There are several variants of BFT such as PBFT (100-200 validators), Tendermint (300-500), Hotstuff (high hundreds), dBFT (~100) but all do not scale well into the thousands.

For a privacy coin, this isn’t a really good model despite its performance benefits. Also FYI, we already have single block finality through chainlocks.

Solana has a hybrid consensus that uses Tower BFT with Proof of History can scale to thousands of validators but they do require very high requirements to run. Also because it’s written in a different language, it would be a significant lift to port this over for Firo which I think would be better spent elsewhere.

I’ve been thinking about this for a while - and figured I should weigh in with some thoughts.

We obviously have 4-5 years to implement something - but certainly need to start thinking about it now - since it may not be a quick solution.

The problem we are going to have - is at current market prices for Firo - the tail emission of 1 Firo/block may not be enough to sustain any kind of secure infrastructure. With some (good) luck we may see a large increase in Firo’s price over the next year or two, and this may allow for more options to be achievable.

Comments on Reuben’s post:

a) Merged mining

I’ve been a miner for almost 7 years - and looked at mining merged mined coins a few times - but never did - because the main coins usually needed a lot of processing power (usually an ASIC or large GPUs), and the secondary coin (which is usually the one I was really interested in) was thought of by the miners as an ‘also-ran’ - basically a throw-away coin that was not important - to be dumped at whatever price they could get for on an exchange. This would not be a very desirable position to place Firo.
I don’t think Dogecoin is a good example of merged mining working - since the only reason why Dogecoin is doing so well is because of Elon Musk’s endorsement, and Tesla accepting Dogecoin as their only crypto payment option. (We really need to get ‘Uncle Elon’ to adopt Firo as well … hehehe).

b) Komodo notarization route (this is what Piratechain uses)

Really not a fan of PirateChain - but that being said - I took a look at Komodo’s notarization process - and if the Firo tail emission block reward is only 1 Firo I don’t see how anyone would want to run a notary node, even if the number of nodes are limited and elected like Komodo’s are. Why would this be all that different than the nodes we have now - and it would restrict who could receive the block rewards to only a few nodes - and would also increase the risk of censorship to the coin. (targeting 50 notarization nodes is a lot easier to do than trying to attack/shutdown 1000, 2000, 3000 world wide distributed nodes).
Finding a way of increasing work and potential income sources for the existing MasterNodes would make more sense. MasterNodes would continue to receive a share of the block reward (even after tail emissions kick in) but could also receive additional rewards for providing other services (yet to be determined what they will be - but something like distributed communications or some kind of DEX order-book hosting, or some other form of distributed computing services - we need to get creative here somehow).

c) Pure PoW with a max reorg length

We have such a secure network with PoW + MNs - and why would we want to go back and risk 51% attacks again. Pure PoW really is NOT secure - even with limits on reorg lengths - when the attack happens - the damage is done.

d) Proof of Stake

Pure Proof of Stake still runs the risk of 51% attacks if an individual or group of people control 51% of the total Staked coins. Not all coins in circulation are staked. Anyone wanting to cause harm to the blockchain at times may not need all that much staked Firo to get control.

Additional thoughts:

The miners will currently settle into the new 5% of the current block reward = 0.3125 Firo/block. When tail emissions kick in, if mining rewards are kept at 0.3125, this will leave 0.6875 Firo for whatever other infrastructure is being used to secure the Firo network (let’s say MasterNodes). Prorating the current number of nodes (~4000) and current rewards (3.125 before increase) and assuming a similar take-up profile after tail emissions kick-in (0.6875 Firo/block) in theory will only leave about 880 nodes instead of the current ~4000 nodes running. Reuben has indicated that this number of nodes on the network would likely not be sufficient to secure the network safely (and I would agree). We have no idea what the future will have in store for the price of Firo; however, if the price of Firo were to say double (which is certainly possible, or even better) then - in theory - the number of nodes that people will want to host could be in the range to sufficiently secure the Firo network even with such a small tail emission.

Instead of merged mining maybe as an alternative we could see if ‘merged noding’ would work (yes - I just made up that term - I don’t like it - but you get the point). Have Firo master nodes also work on securing other networks and receive rewards for that network as well - just a thought (Firo would be the primary coin - NOT the ‘also-ran’).

My gut feel is - that we need to find a way to continue to use the current technology (PoW mining + Master Nodes) to secure the Firo network after tail emissions begin.
This would continue to provide Censorship resistance, be Easy to implement, and will Maintain fast finality. The only issue will be how to ensure security of the Firo network with only tail emissions (I hate to say it - but making tail emissions 2 Firo/block may be enough to provide incentive to run enough nodes - especially if the price of Firo increases).

We still have some time to come up with something to make this all work. I still like the idea of evolving Firo’s eco-system into something that allows additional services to be performed by the Master Nodes and having those nodes earn additional service fees in addition to the node’s share of the tail emissions. It would have to be enough to give incentive to run enough nodes to properly secure the Firo network. We need to get creative - but still retain Firo’s vision of providing the best privacy preserving cryptocurrency and ecosystem available.

4 Likes