Varun Srinivasan on Warpcast

Content pfp

0 reply

0 recast

2 reactions

Varun Srinivasan pfp

Varun Srinivasan

I wrote a bit about Farcasters consensus model, and the specific tradeoffs we had to navigate. Planning on sharing more broadly tomorrow, any feedback is appreciated! https://www.varunsrinivasan.com/2024/04/28/the-goldilocks-consensus-problem

25 replies

65 recasts

370 reactions

Gabriel Ayuso pfp

@gabrielayuso.eth

Great write up! 🙏 I'd love to learn more about how the deltagraph approach is different from other CRDT-based eventual consistency systems such as some NoSQL DBs. I'd also like to learn about some of the main challenges with keeping nodes up-to-date on the global state as the amount of deltas increases.

1 reply

0 recast

2 reactions

MJC pfp

great piece, learned a lot. how does the deltagraph receive and validate the blockchain data?

1 reply

0 recast

2 reactions

Royal pfp

Extremely well written summary. There are a few places where it seems some of the technical underpinnings are just known, CDRTs for example, that might benefit from a one sentence explainer when they are introduced but even that feels a bit nit-picky. Definitely helped me understand what is going on a lot better.

0 reply

0 recast

1 reaction

Vladyslav Dalechyn pfp

Vladyslav Dalechyn

Awesome illustrations! Gotta research the eclipse attack now.

0 reply

0 recast

1 reaction

jj 🛟 pfp

4. Reliance on sync mechanism to find missing deltas at some point will hit a ceiling as the number of nodes continues to increase. Will you ever limit the number of writable nodes?

1 reply

0 recast

0 reaction

jj 🛟 pfp

3. Charging rent for data storage will be a barrier for wide scale adoption especially trying to get into countries where the exchange rates are much higher. Is this the long term plan?

2 replies

0 recast

0 reaction

jj 🛟 pfp

2. Security concerns and reliance on local consensus. How would you address the security trade offs like Sybil or network partitioning attacks? Where either reliability is affected because of numerous manipulated data flows, identities or nodes. Or potentially manipulating node networking

1 reply

0 recast

0 reaction

jj 🛟 pfp

Finally had a chance to read it. 1. About ordering. You’re saying order doesn’t matter in casts but in replies order does matter in maintaining a sequence or posts. Would this be affected?

1 reply

0 recast

0 reaction

behkod.base.eth pfp

behkod.base.eth

I'd appreciate it if you could elaborate more on this "edit"ablity of casts (either providing links to discussions, docs, whatever). An article on "Thinking MEV on FC/Deltagraphs" would also be really interesting to start a series of thoughts [If someday this design pattern is used for games, MPs as you said)

1 reply

0 recast

1 reaction

𝚣𝚘𝚣 🎩 pfp

𝚣𝚘𝚣 🎩

Very informative thread. Haven’t been following technical design recently and this is 🔥

0 reply

0 recast

2 reactions

Pepo pfp

Very cool, thanks. One question, "who" and "how" collects the rent paid?

1 reply

0 recast

1 reaction

borged.eth pfp

Thanks, very informative! I am now much curious about the roadmap for future scaling.

0 reply

0 recast

0 reaction

Pierre H. — q/dau pfp

Pierre H. — q/dau

Thanks @v for the great article! spot a little typo: “A social network network doesn’t need perfect ordering.” Or was it an intended pun 😆

0 reply

0 recast

0 reaction

Simone Pompei pfp

@simmonsstummer

Very interesting

0 reply

0 recast

0 reaction

y0b pfp

Great to have this resource. 50 $degen

0 reply

0 recast

0 reaction

Taylor pfp

Great article, @v. Even for a non-technical reader, your explanation of the architecture makes a lot of sense.

0 reply

0 recast

0 reaction

flea pfp

Fantastic read! /justbuild

0 reply

0 recast

0 reaction

stellabelle 🔪 pfp

stellabelle 🔪

interesting, didn’t know what CRDT means

0 reply

0 recast

0 reaction

CARDELUCCI🎩 pfp

Thank you for this 100 $degen

0 reply

0 recast

0 reaction