Pinned
We will exit the validator set at the end of this month.
We’ve been quietly securing the network and keeping its promise of decentralization for the last 8 months. This chapter comes to a close soon.
When the time is right, we will move our resources to our other current focuses.
Thanks to everyone for the opportunities over the years! We’ve learned a lot.
1
Good example of this was that we just turned off mini apps one day for all users and hid them in the feeds.
Over the next week, about 20 users~ all said we had a better feed than Merkle with zero change in the recsys.
One of the things I learned at Spotify is to detect churn and ruthlessly drive to lower it.
We were probably 60% of the way to parity to Merkle with our native build. If we wanted to, we would have gotten to 90%~ and been competitive.
We were more focused on the *data* of Farcaster than the client.
Getting good at data means you don’t have to ask what to build. You can just predict it and feature flag.
3
1
10
In 2024 we built Uno as a native Farcaster client after building the MVP in React Native.
The team was rate limited by the dev process and the infra to support RN.
Elite teams don’t want to wait for RN devs to fix their stuff. And they also don’t want to patch internal packages..
1
1
15
2
4
20
I will buy an iPhone Duo.
2
1
13
Muse -> Meta -> Link -> Stripe -> Tempo.
Lots of layers!
13
Been spending more time on Twitter. Follow me there.
x.com/christopherwxyz
2
1
8
Also nine months ago, pre-Merkle-to-Tempo 😳
1
10
Well, well, well.
4