Warpcast

Genuine question, no troll:
Why is Warpcast not using XMTP protocol for DC ?

Technowatermelon. Elder Millenial. Building Farcaster. 

nf.td/varun

If you are considering 1, why not just use XMTP?

Instead of a Warpcast-only DC functionality, I'd rather see those 6-12 months spent adding an open source DC capability to Hubs, so all FC clients can use it.

lots of discussion around this a year ago

building /anser https://anser.social | prev. Ledger

lots of discussion around this a year ago 

https://warpcast.com/boscolo.eth/0x1d1986a3

With the benefit of hindsight, this was the right decision.

DC/GC in Farcaster should:
1️⃣ Be based on FIDs & FC KeyStore. (XMTP has its own ID layer.)
2️⃣ Use Hubs for message delivery. (XMTP servers require additional payment.)
3️⃣ Use

Building /bright-moments - minting onchain art IRL | /purple #15 | FID 129 #magikarp

Rome, Metropolitan City of Rome Capital, Italy

With the benefit of hindsight, this was the right decision.

DC/GC in Farcaster should:
1️⃣ Be based on FIDs & FC KeyStore. (XMTP has its own ID layer.)
2️⃣ Use Hubs for message delivery. (XMTP servers require additional payment.)
3️⃣ Use https://github.com/farcasterxyz/protocol/discussions/99

Answering the question "where's the beef?"

1. XMTP supports FID today. XMTP is not a walled garden - it supports all EVM identities, including FIDs.

2. XMTP has never required payment. Re: hubs - I'm curious why you believe a real-time messaging use-case must use the same backend as a social networking app.

3. I'm curious why MLS, an open IETF standard with Signal-like security properties, is insufficient.

1. Registration/storage price charged by Farcaster. That cost is waived if the user registers via Warpcast, but that’s not a cost that XMTP can afford to pass onto other apps, especially those without sybil resistance.

2. Currently we subsidize all infrastructure costs, but you’re right that there is no free lunch. In the long-term, it is important to us that apps can maintain a free tier as well as paid options - Warpcast has the right idea here, we just need to generalize it.

3. Great question. A small percentage of MLS payloads (i.e. commits) require ordering consensus. This is currently achieved by a centralized service, but we are working on sequencing these using an L2 blockchain.

1. Why didn't XMTP just use the FC ID contracts instead of duplicating that layer?

2. There was a proposal at one time to charge. If that's no longer the case, how the economics of the network function?

3. 💯 (tho I am curious how XMTP solved the GKA without a centralized service?)

These are very valid points.
I still haven't made my mind on this choice.
100 $DEGEN