​

Farcaster Topics

​

Decentralized Channels — @dwr

• Updates on moderation
  • Big changes last night, testing a trending algorithm as an offering, on a test basis, ability to toggle recent feed
  • Settings are easy to decentralize and communicate to other clients if they want to implement them
  • Setting up moderation has been hard and tricky to do, Warpcast might implement some changes to help make it smoother
  • New updates from Automod and Airstack will also help with moderation UX flow

​

Long Casts — @vrypan, @df

• FIP: https://github.com/farcasterxyz/protocol/discussions/171
  • High level thoughts from V
    • Amazing to see how much response has come through
    • Why introduce this? Bring more interesting new stuff to Farcaster
    • Metric used: Whatever drives more users to use it, how easy for clients to create
    • Possible use cases
      • Option 1: People either share long form articles on things like paragraph, ideal is feeding post inline to Farcaster. FIP kinda supports that
      • Option 2: Making the content in app like in Warpcast. Difference is composing inside the app vs another app like Paragraph
  • FIP 171 supports both use cases
  • Warpcast perspective
    • Model of a pointer to a json object
    • From a client pov, how do you compose a long cast within Warpcast?
    • Team wants to build some client mocks before implementing
  • Has the potential to reach into other forms of content
  • Over the next two weeks Farcaser team will dive further into questions and stages to implement

​

Hubble — @Sanjay

• Link Defragmentation changes from Warpcast
• Shuttle updates
  • Replicator deprecation is on its way
  • Adding some features to help fix missing data in DBs

​

Warpcast Updates

​

Feed Improvements — @akshaan, @cmlad.eth

• ML-based feed rolled out to 100%
  • Takes what you haven’t read and show you the most interesting stuff on the first page
  • So far pretty promising results and have made it 100%
  • If you encounter any weird bugs or posts that don’t make sense, take a screenshot and report to @v or @dwr
• Testing replies in-feed
  • 50% user base test right now to show replies in the feed.
  • Author replies or replies from people you follow goes into priority
  • Want to see how the feed changes numbers on people following other people etc
• New tests: more channel content, more signals
  • Injecting more channel content vs people you follow in that channel. Seeing if there’s opportunity to inject interesting content into your feed
  • Training ML on which users and channels you care about the most and pushing the rest further down the feed

​

Frames — @horsefacts

• High level exploration — new compact entry point for frames
  • Looking at ways to make smaller entry points for frames
  • Frames take up a lot of space on the feed, and if its not interesting then people may not use the app
  • Trying a method to collapse as a call to action which expands into a larger frame
  • Sets up for more interesting concepts if you can have more of an application by opening a frame. Goes beyond the 4 buttons and an image
  • Starting with a design experiment with existing frames, get feedback
  • Proposal for new meta tags will be coming soon