Feedback Request for 'R' Protocol Proposal

So if 66% of the stake does not run DAL nodes, is the new rewards model not activated? As far as 10% for dal participation, seems reasonable, but I wonder if it’ll be enough to incent bakers to upgrade hardware (if necessary) or rent a vps to run the dal.

If it were so, they would signal it with the LB switch

2 Likes

Thank you for your thoughtful feedback, I appreciate your input and suggestion about taking the 33% of liquidity baking rewards and moving those to DAL. We’ll continue to refine and clarify the proposal so every voice matters to us. All suggestions, including yours, will be summarised and presented for further discussion during governance iterations.

2 Likes

Agora is our main platform for feedback because it keeps discussions focused and free from the noise of larger channels like Reddit. This helps us address relevant voices efficiently without spreading ourselves too thin. While we do monitor other channels, having a centralized space ensures no feedback gets lost and allows us to respond thoughtfully.

By keeping Agora as the main hub, we can work together more effectively to improve Tezos. Thank you for contributing here—it truly makes a difference!

The 21.1 version of Octez makes this setup compatible with the operation of the DAL. Indeed, improvements have been made to resolve issues related to latency that were encountered in the past.

Thanks for your feedback @BakingBangkok.
It covers several points, which I’ll try to respond to:

  • Regarding deploying it on a machine hosted at home, it’s up to the operator. We briefly explained here why we no longer advise against deploying DAL and node on the same machine. The same principle applies to hosting at home. We can come back to the community with more details if necessary.

  • As we’ve seen in other responses in this thread, the figure of 10% of rewards allocated to DAL is a value intended for discussion, and not a definitive value. If this figure seems excessive to the community and they converge on a lower figure, we will adopt the latter. We also don’t rule out making multiple proposals with a wide range of values to leave the choice to governance.

  • As for using the revenue from Liquidity Baking to compensate participation in DAL, as @beatalipska said earlier, we take it as a request from you, and this thread is the perfect place to formulate them.

1 Like

This is just a generic response, Beata. Take a moment to think about it. Answers to questions are posted one after another, making it difficult to follow the discussion threads. It’s unclear which points are being addressed and which are not. There’s no indication of priority either (for example, on Reddit, upvoted questions appear at the top). Agora works well for single-topic discussions, but feedback threads aren’t well-suited for it. We’re dealing with multiple topics here. To use Agora effectively, you’d need to break down the proposed changes into separate posts.

I completely agree with the points raised regarding the functional limitations—there’s no question about the validity of your concerns. However, introducing an additional platform for these discussions would require significant time and resources to implement effectively.

1 Like

Doesn’t it require significant time and resources for TT to implement everything effectively?

What does it really take to succeed on Reddit? It doesn’t sound too complicated—create a subreddit, post feedback prompts there, and engage with the discussions there instead of here. The only difference between Reddit and Agora is that whoever runs Agora won’t have the analytical data they have now.

Yes, it clearly is up to operators to decide. This is why we no longer advise against deploying DAL and RPC node on the same machine, but rather prefer to explain the ins and outs, as well as our analysis of the matter. This way, everyone can form an informed opinion.

1 Like

Cool. Just make sure to really explain the ins and outs. So far, I’ve seen plenty of ‘we don’t advise against…’ but zero ‘BUT’.

Thanks for that feedback @TezoSpanish, we take it as a change suggestion for this proposal. We are keeping track of all of them, and we value them a lot.

1 Like

Yes, the proposal actually consists of distributing the rewards between baking and DAL.
The figure of 10% is intended as a starting point for discussion, not a final figure, though. We would prefer it to be determined through this feedback round, so that it would capture constraints and needs such as the one you expressed.

Please find further details in this other answer:

Thanks for your feedback!
We take this as a change request for this proposal, which is very welcome.

Thank you for your support for Etherlink! We’re focused on making setup easy for home bakers and will provide clear guidance. Your point about 1-day cycles increasing accounting efforts is noted, and we’ll explore ways to simplify this. Extending the activation period to 6 days for emergencies is a great suggestion, and we’ll consider it in discussions. Your feedback is invaluable—thank you!

1 Like

Reducing withdrawal delays for Etherlink users is indeed critical for improving user experience, and this feedback will be taken into account.

2 Likes

I agree that losing a fraction of rewards if we don’t run a DAL node is not a great incentive. Are there any other posibilities to make it more attractive?
What it we don’t achieve the 66%? I just think about the initial plan to attract 50% stake for adaptive inflation but we keep hanging ~20%. What to do in a similar case for the DAL? Can rewarding DAL node operators with an adaptive incentive as well? One that goes down if we reach the 66%?
Or, I’m just thinking out loud here: Reward early adopters! Avoid raising inflation and reward them with shares of projects in the Tezos Ecosystem. Maybe with Uranium (even though I’m not a fan of it) or NFT’s with recognizable value.

5 Likes

yes, it is certainly a good idea not to fine but to reward, for example, as in other networks. In AVAX, for deploying an additional node, the validator receives a premium or bonus and thus they reach the required number of nodes. This is better than fining.

4 Likes

There has been some debate in community about DAL bandwidth usage

Looking at tezos x roadmap there are three milestones for DAL 1 mb/s, 10 mb/s and 100mb/s

Could you provide more clarity on required bandwidth for each stage

Are these values constant data usage or just burst speeds?

How much data in TB are you expecting each stage to consume? //monthly

1 Like

Hello @Primate,
Thanks for your message, which is particularly comprehensive and rich in feedback as well as proposed solutions.
If we haven’t responded to it yet, it’s precisely because of the broad scope it covers. We’re going to start responding in a grouped manner from now on to avoid the proliferation of individual messages, which some readers have found difficult to follow.
Rest assured that the topics you’ve listed here will be covered in this response.

This also applies to all of you who haven’t received a response yet in this thread, and whom we thank warmly.

3 Likes