xDai Stable Chain xDai Stable Chain BlockScout BlockScout TokenBridge TokenBridge

xDai Telegram community Q&A

From Legend (@rumblerumble)

Questions 1:

Q: If usage would pick up and blocks would get full fees would simply rise, right?"
A: Not really. Transactions will wait in mempool. Some transactions will get more priority.

My question: Wouldn’t that mean transaction time may exceed 5 seconds?

Question 2:

In a perfect world where xDai gains massive adoption and becomes the dominant P2P payment protocol, I imagine the “blocks with empty capacity” will no longer be the case as most if not every block will be full. How will xDai maintain it’s 5-second transaction time, extremely low gas fees (validators raising fees)?

Question 3:

Are you concerned at all that the zk protocol may be a potential issue with regulators if xDai becomes the norm?

Question 4:

Can the default MAX_VALIDATORS = 19 be changed to a higher number or is it permanent?

Question 5:

What happens to the existing exclusive group of 10 validators in the newly implemented POSDAO model? Will they maintain their exclusivity as validators and not subject to staking requirements?

Question 6:

Why the decision to introduce a 7-day time for each epoch instead of hours/lesser number of days? Any game theory to justify 7 days?

Question 7:

What will happen to xDai when Ethereum fully implements Casper?

Question 8:

“As additional stake-based chains and networks are configured and launched, the DPOS token will be transferable via the TokenBridge, providing users the opportunity to place DPOS token stakes across multiple networks.”

I’m trying to understand what you mean by “additional stake-based chains and networks”. Could you provide some examples? Do you mean xDai pegged to Euro (which is dependent upon Maker implementing Multi Collateral)? xDai pegged to BTC (which is not part of Dai)?

Question 9:

I presume part of the reason why the fees are so low is due to the willingness of the validators to subsidize, but until what point will that continue? When they stop subsidizing, how will that affect transaction fees?

Question 10:

How will the general public sale participants (1M DPOS unlocked) be able to compete with private investors (8.5M DPOS w/ 25% unlocked) when it comes to staking since staking pool size plays a major role despite a randomness factor? Let’s assume that he/she wants to be a validator and no intention to delegate

Question 11:

Why did Helena decide to develop/run their dApp on xDai chain instead of POA if xDai is intended for P2P payment?

Question 12

Additional incentives (including transaction fees for validators and bridge entrance/exit
fees for validators and delegators) will also ensure permanent compensation in exchange for protecting the xDai DPOS chain.

Is the bridge fee going to be implemented upon staking launch?

Question 13

Bearing a black swan situation in the crypto space, can the emission schedule be postponed / slowed down to minimize the bleeding?

Question 14

Can the 19 validators of a particular POSDAO sidechain (eg. xDai) be validators at a different POSDAO sidechain? If yes, what if a validator turns out to be a bad actor, how will that reputation be transferred to the other sidechain? Will all sidechains implementing POSDAO consensus have the same max 19 validators?

Question 15

All sidechains using the POSDAO model will use DPOS correct? And theoretically speaking, we could have xDai1 xDai2 xDai3 all running concurrently to serve an evergrowing and popular xDai sidechains?

Question 16

Then how will that affect a developer who use xDai1 to build it’s dApp in relation to transactions occuring in xDai2 xDai3?

Question 17

Who is/are xDai’s closest competitor(s)?

Question 18

Which one is more cost/time efficient? Vertical vs horizontal? In a congested network, which gets priority?

Question 19

“Through the open market, the candidates will buy the necessary number of tokens and become new validators. In this way, a feedback loop will be formed and the network can resist censorship.”

If the G20 countries agree (and I think they will) to target privacy coins, won’t that negatively affect the candidate pool since the source of these candidates will be restricted to non G20 countries?

Question 20
Who invented POSDAO? The POA team? Open source license?

Question 21
If a project develops on xDai sidechain, can it take advantage from an xUSDT sidechain? If yes, how will that relationship look like?

Questions 1:
My question: Wouldn’t that mean transaction time may exceed 5 seconds?

Block time won’t exceed 5 seconds, but transaction time can. If there are a lot of transactions in mempool, some of them will have to wait for their turn.

Question 2:
How will xDai maintain it’s 5-second transaction time, extremely low gas fees (validators raising fees)?

We can try to increase block gas limit (by default it is 10 million). We haven’t yet done stress tests for, e.g., 100 million limit, but we think this can be possible in general (if validators nodes would work on powerful enough cloud instances).

Question 4:
Can the default MAX_VALIDATORS = 19 be changed to a higher number or is it permanent?

It is supposed to be a constant, but yes, this could be increased if needed.

Question 5:
Will they maintain their exclusivity as validators and not subject to staking requirements?

POSDAO doesn’t have exclusive rules for anyone, so these validators are subject for the common staking requirements.

2 Likes

Expanding on Question #5 a bit

During the first staking epoch, the initial validators (assuming they want to continue and aren’t banned :slight_smile: ) will need to place stake into their own pools. This stake must be at least CANDIDATE_MIN_STAKE of $DPOS (amount TBD).

During this first week, additional candidates may also enter the protocol by launching a node and staking at least the CANDIDATE_MIN_STAKE. When staking epoch #1 ends:

  • If the number of viable candidates is 19 or fewer, then all initial validators who placed stake will be retained for the next staking epoch, along with any new candidates who meet the requirements.

  • If there are more than 19 candidates (including new candidates and the initial validators), a new validator set will be selected from all eligible participants based on a random value and weighted by total pool size (candidate stake + delegator stake). Depending on the selection outcome, some initial validators may be excluded from this - and subsequent - staking epochs.

Questions 1:

Q: If usage would pick up and blocks would get full fees would simply rise, right?"
A: Not really. Transactions will wait in mempool. Some transactions will get more priority.

My question: Wouldn’t that mean transaction time may exceed 5 seconds?

Transaction time may exceed 5 seconds if there is no free capacity in the scarce resource of blocks. The question is how validators and the protocol mitigate this type of denial of service.

The network can be scaled in by optimizing the protocol and parameters of the validators and consensus. The network can be scaled out by launching parallel networks like xDai. The security model of xDai is scalable horizontally. The incentive layer for new sidechains is provided by “multichain staking token” which can be bridged between networks.

Question 2:

In a perfect world where xDai gains massive adoption and becomes the dominant P2P payment protocol, I imagine the “blocks with empty capacity” will no longer be the case as most if not every block will be full. How will xDai maintain it’s 5-second transaction time, extremely low gas fees (validators raising fees)?

As I’ve answered in Q1, the network can be scaled both horizontally and vertically. Validators and delegators of the network have an incentive mechanic to move their stakes to horizontal chains.

If demand for xDai will increase and vertical scalability will not be enough then xDai2 can be launched with the same type of staking and native tokens.
The same scaling model in a centralized way is widely used in MMORPG such as World of Warcraft.

Here is an example of the realms status of World of Warcraft.
It’s the same game on each realm but with the different state (players, items). Players can move from realm to realm with their state if they’d like to play in different timezone or population.

Question 3:

Are you concerned at all that the zk protocol may be a potential issue with regulators if xDai becomes the norm?

In a consensus, such as the POA on which the xDai network operates, validators select new validators based on the criteria inherent in the idea of the network. This set of validators is decentralized because everyone in the group has the same rights and decisions as a quorum. The downside of this model is that the group is exclusive and can be censored at the group level.

The new POSDAO consensus is more inclusive and protected from censorship at the group level. xDai plans to switch to this algorithm three months after the staking token is initialized.

Let’s imagine that validators of xDai with POSDAO will have pressure from regulators in specific jurisdictions in a form of “cease and desist” order because of the ZK transactions and illegal activities on the chain.

What will the validators do in this situation? Most likely, they will obey orders and stop doing what they do.
A new opportunity will open up in the market for candidates to become validators in the network. Through the open market, the candidates will buy the necessary number of tokens and become new validators. In this way, a feedback loop will be formed and the network can resist censorship.

Question 4:

Can the default MAX_VALIDATORS = 19 be changed to a higher number or is it permanent?

This parameter is derived empirically from the analysis of other networks and protocols. The number of nineteen is selected so that the network can withstand the six consensus faults according to the 3*f+1 formula where f is the number of possible faulty validators, including those that do not work or are in the byzantine state.

The parameter can be changed through the protocol upgrade on both the xDai network and subsequent DPOS-based networks.

Question 6:

Why the decision to introduce a 7-day time for each epoch instead of hours/lesser number of days? Any game theory to justify 7 days?

This parameter is selected empirically based on system limitations and analysis of other protocols. Theoretically, the parameter can be changed in the future through a protocol upgrade. Such a change will change the emission curve and the probability of such a change is minimal.

During the epoch, stakes of validators and delegators are frozen. The longer the epoch, the longer the tokens will be withdrawn from circulation within the epoch for a longer time. The tokens withdrawn from circulation do not create pressure on sales in the open market.

Question 7:

What will happen to xDai when Ethereum fully implements Casper?

The xDai protocol works on top of the Ethereum 1.0 protocol. As far as we know, the Ethereum Foundation has no plans to implement Casper on Ethereum 1.0

We plan to upgrade POSDAO to HBBFT consensus https://github.com/poanetwork/hbbft which is censorship-resistant, asynchronous, leaderless protocol, highly scalable, with onchain randomness. We implemented and audited HBBFT consensus. and working on integration into the parity-ethereum client.

Question 8:

“As additional stake-based chains and networks are configured and launched, the DPOS token will be transferable via the TokenBridge, providing users the opportunity to place DPOS token stakes across multiple networks.”

I’m trying to understand what you mean by “additional stake-based chains and networks”. Could you provide some examples? Do you mean xDai pegged to Euro (which is dependent upon Maker implementing Multi Collateral)? xDai pegged to BTC (which is not part of Dai)?

There are several types of stable tokens, for example, algorithmic, like Dai or with reserves, like USDT. We see interest in launching an xDai-type network for the USDT token and probably this network will be called xUSDT and have an USDT for the native token and a DPOS token for the stacking token.

Question 9:

I presume part of the reason why the fees are so low is due to the willingness of the validators to subsidize, but until what point will that continue? When they stop subsidizing, how will that affect transaction fees?

Each validator can set the transaction fee at its own discretion. Despite the cost of gas in 1 Gwei, users sometimes set a higher cost of gas, such as 100 Gwei, which created an award for the 0.197173 xDai validator for the https://blockscout.com/poa/dai/blocks/5211420/transactions block.

Changes in transaction fee costs can be a tool to mitigate an attack such as denial of service

Question 10:

How will the general public sale participants (1M DPOS unlocked) be able to compete with private investors (8.5M DPOS w/ 25% unlocked) when it comes to staking since staking pool size plays a major role despite a randomness factor? Let’s assume that he/she wants to be a validator and no intention to delegate

According to the white paper, delegates receive an award evenly in the pool depending on the weight of their bets.
It is obvious that those with fewer tokens will receive a proportionately smaller award. The market of staking token is open for everyone after the token is added to the open market. Participants from the public round will be able to buy tokens from participants of the private round, if they see interest in it.

Question 11:

Why did Helena decide to develop/run their dApp on xDai chain instead of POA if xDai is intended for P2P payment?

Here is their official announcement https://media.consensys.net/prediction-markets-platform-becomes-first-dapp-on-xdai-sidechain-f26e6d691799

Quote:

[…] significantly improving the user experience by making transactions faster and cheaper. The average block time on the xDai sidechain is 5 seconds, and gas price is fixed at 1 Gwei and paid with the xDai stablecoin.

Question 12

Additional incentives (including transaction fees for validators and bridge entrance/exit
fees for validators and delegators) will also ensure permanent compensation in exchange for protecting the xDai DPOS chain.

Is the bridge fee going to be implemented upon staking launch?

Bridge fees mechanic is implemented already in the TokenBridge.
The code is audited by SmartDec and published in 2.3.3 version https://github.com/poanetwork/poa-bridge-contracts/releases/tag/2.3.3

Enabling of bridge fees is planned upon staking launch.

Question 13

Bearing a black swan situation in the crypto space, can the emission schedule be postponed / slowed down to minimize the bleeding?

POA/xDai team will follow the designed emission schedule. Future governance of public networks could modify the schedule by a protocol upgrade.

It is theoretically possible that 19 xDai network validators will also be 19 parallel network validators. The probability of such a state is low due to the chaotic nature of states with a large number of unknown and loosely connected actors.

Reputation will not be transferred between parallel networks. A good actor in one network can be a bad actor in a parallel network.

Parallel networks can have more than 19 validators.

Question 15

All sidechains using the POSDAO model will use DPOS correct? And theoretically speaking, we could have xDai1 xDai2 xDai3 all running concurrently to serve an evergrowing and popular xDai sidechains?

Yes, all sidechain using the POSDAO model launched by xDai Foundation will use DPOS.
Although, due to the nature of open-source software and permissive license other projects could utilize consensus and model without using the token. E.g. Artis https://artis.eco/en/ networks plans to use POSDAO for their own chain and token without using $DPOS token for staking https://twitter.com/lab10collective/status/1159358958982877184

Another example is Lukso, https://www.lukso.network/ which is also exploring migration to POSDAO consensus with their own staking and native tokens https://twitter.com/feindura/status/1113033290028650496

Question 16

Then how will that affect a developer who use xDai1 to build it’s dApp in relation to transactions occuring in xDai2 xDai3?

This pattern for multichain dapps is not fully discovered.
DApps can query and submit transactions to more than one network. This approach is often used in wallets, where a single user has assets and transactions on different networks. Likely API providers will provide multichain functionality for DApp developers.

Question 17
Who is/are xDai’s closest competitor(s)?

In DeFi based (with locked funds) Payments, our competitors are

As L2 Ethereum sidechains for smart contracts:

Question 18

Which one is more cost/time efficient? Vertical vs horizontal? In a congested network, which gets priority?

Vertical scalability is cost and time more efficient.
The possibilities of this approach are limited.

In a congested network likely vertical scalability will work first with increasing block size/ decreasing block time.

Questions19

“Through the open market, the candidates will buy the necessary number of tokens and become new validators. In this way, a feedback loop will be formed and the network can resist censorship.”

If the G20 countries agree (and I think they will) to target privacy coins, won’t that negatively affect the candidate pool since the source of these candidates will be restricted to non G20 countries?

Staking token will be available for purchase on decentralized exchanges such as Uniswap https://uniswap.exchange. Candidates will be able to get staking tokens in a censorship-resistant manner.

Question 20
Who invented POSDAO? The POA team? Open source license?

POSDAO is a research project by POA Team. The reference implementation https://github.com/poanetwork/posdao-contracts will be opensourced under a double MIT/Apache2 license https://github.com/poanetwork/posdao-contracts after security audits.

Question 21
If a project develops on xDai sidechain, can it take advantage from an xUSDT sidechain? If yes, how will that relationship look like?

I don’t see much difference in what token is used as a stable token. both networks will be equally attractive for smart contract developers. There may be a difference when using payments, as USDT is more liquid than Dai.

1 Like

Comments on concerns:

“The xDai Chain project seems for the moment focused on Ethereum 1.0, which is great on short term but becomes an unattractive long term investment.”

We are focusing on Ethereum 1.0 scalability. It’s the most advanced platform for smart contracts with

“Most of the work seems to either come from the POA project or to almost be completed with only seed funding, so why this new token and additional raise?”

The xDai project was created by the POA team.

Fundraising helps the project to obtain resources for adoption, to create a community, to create a market valuation of the DPOS token, to create a basis for cryptoeconomics of stacking.

“POSDAO is not a leaderless consensus mechanism which opens the network to some attack vectors, i.e. the validator nodes are selected for 1 week, making them attractive targets for malicious actors during that time”

In the current POSDAO consensus algorithm there is no need for validators to have open ports. The probability of an IP-level attack on a validator infrastructure is similar to an attack on a cloud provider infrastructure.

Validators in POSDAO are anonymous and a phishing attack is not easily feasible in a short time, for example, by email.

“Staking DPOS brings uneven revenue to holders, based on random candidate selection; the DPOS weekly snapshot for rewards distribution encourages periodical price spikes”

Random selection of validators takes into account the size of stake of candidates and delegated coins. In most cases, candidates with a higher reputation will be selected by consensus for the role of validators of the network.