waku-org / research

Waku Protocol Research
MIT License
3 stars 0 forks source link

It's all resource pricing? #27

Closed s-tikhomirov closed 7 months ago

s-tikhomirov commented 9 months ago

There is a talk from EthCC that made a lasting impression on me and I think is quite relevant:

https://www.youtube.com/watch?v=YoWMLoeQGeI

TLDR: incentivization in blockchains is about fairly rewarding actors for the resources they spend. The question is: what resources exactly and how to quantify them?

We can approach our incentivization question (#21) from a similar angle. Currently, clients occupy the server's resources but don't pay for that, such as:

TODO: think about how to quantify these resources, are there others, etc.

alrevuelta commented 9 months ago

Agree on "fairly rewarding actors for the resources they spend" but there is another set that is "resources that you can prove you spent". And imho we should look at the intersection of both sets.

For example, proving that you are spending "x bandwidth" is not trivial. Same for CPU. So perhaps we should look into things that are easy to prove and that indirectly represent your "efforts".

An example I like a lot is the Ethereum one. No one incentivizes you for gossiping blocks, attestations, storing blocks, or even serving them to new nodes syncing. But people do it because its a requirement for other higher level duties that are incentivized.

s-tikhomirov commented 9 months ago

things that are easy to prove and that indirectly represent your "efforts"

I agree, one caveat though is that the gap between what I actually spend and what I use as a proxy to prove what I've spent can be exploited by attackers.

s-tikhomirov commented 7 months ago

Closing in favor of #35.