Farcaster Devs

Neynar

If you want to subscribe to our Kafka data pipeline but don't want to pay for all of the messages you're not interested in, such as reactions, now you don't have to. In addition to our aggregated offering, we have broken out farcaster events into separate Kafka topics for users, casts, reactions, follows, and signers to save you $ and bandwidth. Reach out to @rish if you're interested!

(niche feature, bit more iykyk)

For those streaming raw events straight from a hub gRPC, you know the events need to be massaged and joined with other data before they can be used in your app.

To solve for that, we now have a real time Kafka stream where developers can consume fully hydrated events and use them directly in their applications.

Screenshot below shows raw hub event on the left and Kafka stream event on the right, you'll notice even the embed metadata is hydrated in Kafka.

The Kafka pipeline stores the events for longer so if your app missed something, you can pick back up from where you left off. Overall, should make development a lot easier compared to wrangling raw hub streams.

Reach out if interested. There is a node example you can plug and play to get started: https://github.com/neynarxyz/farcaster-examples/tree/main/neynar-webhook-kafka-consumer

h/t @flashprofits.eth @shreyas-chorge for the work

🪐

https://docs.neynar.com/docs/from-kafka-stream

We've offered a managed Kafka streaming service for Farcaster events for a bit now - benefit being able to read fully hydrated objects real time instead of raw hub events

So far, the only option was to subscribe to the full stream or not subscribe at all - same hubs

Now, it's possible to subscribe by event type e.g. get cast.created or reaction.created events only (same as webhooks)

So you end up with a stream that has rich objects with all metadata and only has the objects you need so you don't need to filter (also lowers cost when consuming)

Reach out if interested 🪐

see more:
- https://farcaster.xyz/rish/0x7c2997ec
- https://docs.neynar.com/docs/from-kafka-stream

Cohost of /gmfarcaster, a live stream Farcaster news show.  Engineer. Mom x 3. I love to share the wonder and beauty of daily life.

building farcaster infra @ /neynar 🪐 | casting @ /rish

@adrienne if you could please be prepared to explain this tomorrow that’d be great lol

bringing mini apps to 1B users @base              dylansteck.com

building /gmfarcaster network |
blockchain ENT prof |  ⌐◧-◧ Nouncillor, Nouner, BuilderDAO community resident /builder |

https://www.gmfarcaster.com/

@dylsteck.eth if you could please bring back the hub and explain this to everyone that’d be great lol

you would think that would be helpful, it really should be lol

lol 

just in case helpful, the 9 mo old cast linked above has most of the context for kafka and then this one is about added on filters