search

polkadot-fellows / RFCs

Proposals for change to standards administered by the Fellowship.
https://polkadot-fellows.github.io/RFCs/
Creative Commons Zero v1.0 Universal
115 stars 55 forks source link

Conviction Voting Delegation Modifications #35

Open Leemo94 opened 12 months ago

Leemo94 commented 12 months ago

This RFC proposes changes to OpenGov delegations.

antonkhvorov commented 12 months ago

👍 🚀 💯

noah-foltz commented 12 months ago

Keep in mind with # 4, I'm not sure costs would decrease much because all of the logic has to be done for every track anyway, with exception of a couple things that don't change like the nonsense check, the lock, and the balance check. I'd definitely advocate for it from a user experience perspective though.

And # 1 is definitely needed as well. It definitely alleviates worries delegators have about disagreeing with a specific vote even though they agree with a delegate generally.

shawntabrizi commented 12 months ago

Keep in mind with # 4, I'm not sure costs would decrease much because all of the logic has to be done for every track anyway, with exception of a couple things that don't change like the nonsense check, the lock, and the balance check. I'd definitely advocate for it from a user experience perspective though.

And # 1 is definitely needed as well. It definitely alleviates worries delegators have about disagreeing with a specific vote even though they agree with a delegate generally.

indeed, the decrease in costs may not be huge, but there should be some number of redundant storage item reads / writes which can be reduced by up to 15x, since we can assume all of those redundant reads will be in the overlay.

It seems okay to at least make a more accurate estimate as to the weight / fee costs improvements before moving forward with implementing 4.

Otherwise, perhaps the real solution here would be to have more advance benchmarking / weight systems which can detect redundant storage reads by default and refund weights.

xlc commented 12 months ago

can you make this .md file, otherwise Github can't preview it and I don't want to read raw markdown file.

gavofyork commented 12 months ago

I'll underline @xlc 's remarks. As it stands, this is just a wishlist. For it to be taken seriously it must include implementation specifics, including how to achieve bounded-complexity compute and state-changes on all operations.