Farcaster

Another interesting Farcaster idea: move signers onto hubs

- Zero cost to add a new app
- Zero friction to add a new app 

Not feasible with CRDTs could accidentally cause a new signer to kick out your first one which would be really bad. 

But with snapchain this is doable. h.t to @stephancill for the idea.

This would be a great reversion

Right now, deciding to be a client (ie. an app that can write to hubs) requires not only betting on FC, but making users pay before they can use your app

By getting rid of the latter, it reduces risk to just platform risk and likely will enable FC read-only apps to become clients more easily

cypherpunk wannabe, engineer working on @frames /open-frames /opencast
https://stephancill.co.za create stories at https://stories.steer.fun

Oh interesting — I thought apps passed this cost onto users 

I might be mistaken, thanks for flagging!

Either way, not having to have a hot wallet sending meta transactions seems like a win

the onchain cost of adding a signer is basically negligible today so i don't think it will make a difference in that sense tbh (most apps are already subsidising this cost)

the cost to onboard net new users is a much higher barrier for farcaster apps imo

Every new app that wants to write data to a Farcaster hub needs a signer key that lets them sign messages on behalf of a user (otherwise you’d have to sign something every time you cast to prove it came from you)

Previously, signers were registered in hubs, at zero cost to apps (or the respective user). Downside was some issues with the CRDT model — read more here: https://github.com/farcasterxyz/protocol/discussions/103

As a result, signer registration had to be moved onchain, which has a gas cost to issue the transaction. Apps pass this cost down to users, which is not ideal

Now, because hubs are moving away from CRDTs to the Snapchain design, the downsides around registering signers in hubs goes away (presumably due to ordering and merkelized state)

This means we could move signers offchain again, making it free for users to use clients for the first time!

Product & /ux guy interested in new consumer tech | slowly building /firstdraft | markfishman.xyz

Can you eli5 for the less technical folks

Building /farcastle. Tryna help humans coordinate. Just /yeet it. TANSTAAFL! https://bootleggers.wtf/

Understandable -- the network is still tiny

Nobody wants to take more risk on top of execution risk

Every time I share what we are working on the first objection is platform risk