# The Deadlock Keeping Farcaster Alive

Source: https://nickysap.substack.com/p/the-deadlock-keeping-farcaster-alive

## Summary

The author argues that Farcaster is stuck in a stalemate between Neynar, the new steward of the protocol, and the Quorum/Hypersnap/Quilibrium camp, which acts as a forced-neutrality check on Neynar. Neynar's CEO Rish reworked a Farcaster API signature scheme after the author's Juke client asked for access to the audio Spaces primitive, giving the client feature parity within one week. The author says Hypersnap, a Snapchain fork, offers feature parity with Farcaster while the SNAP token has fallen more than 90% from its claim-day price, and that many builders and users are still on the fence.

## Article

I write a lot about Farcaster, the network currently declared dead and continuing to lose users and activity daily. I build a lot on the network as well, for reasons most would probably question. Part of it is stubbornness, part of it is belief in the underlying notion that decentralized social graphs should play a role in how we communicate, and another part is the feeling that there’s still something to be discovered.

As I see it, there are now three entities, two of which are intrinsically misaligned, ironically stabilizing the network. Neynar, the new stewards of the Farcaster ecosystem. Quilibrium/Hypersnap/Quorum, the would-be exit valve imposing forced neutrality. The builders (like me). Stuck in stasis. Stabilized by the push of each opposing force for the time being.

Merkle made numerous mistakes, but no single error caused their failure. A pattern of feigned stewardship under the guise of sufficient decentralization had compounded over several years. An unadulterated path would have focused on providing developers with first-class primitives on which new concepts could be built. Instead, they attempted to build a monolith that commodified any good external feature, subsumed all top-tier talent in the ecosystem, and gate-kept critical features that would have enabled true cross-platform unification. The Quorum camp’s ideological aversion to this pattern shouldn’t come as a surprise. However, their insistence that the new stewards of the protocol will meet the same fate simply because of their VC backing should be assessed with clear eyes.

Neynar has inherited all the good and bad decisions made by the Merkle team, and they’ve only just begun their descent into the madness of operating a multi-faceted consumer product underpinned by a decentralized social protocol. As the new stewards, they have the opportunity not only to right the ship but also to demonstrate that they are fresh-minded and bring a new approach. They’ve already been outspoken about focusing the network on builders. While that largely centers around developers, other types of creators can be grouped under this umbrella. Whether intentional or not, they exist in a counterbalance to the Hypersnap fork, and they’ve already demonstrated their willingness to bend towards neutrality.

I recently published an alt-client on the Farcaster network called Juke, built around creating native audio-centric experiences. Within a month of shipping my app, Farcaster deployed its own version of audio Spaces. To be clear, this was not the same pattern of commodification Merkle demonstrated. On the contrary, Rish (Neynar’s CEO and the new de facto head of Farcaster) reached out to me directly, asking how I felt about Farcaster App deploying their own version. Because of Farcaster’s gravity, I was inclined to agree, for the good of the network. After all, I built Juke because Farcaster lacked this primitive. This outcome was the exact intention of my build.

That said, Farcaster’s version did not immediately expose the primitive natively to the protocol. It was gated behind an authenticated endpoint accessible only with a signature generated by the Farcaster custody address. An address, mind you, that can only be accessed from within the Farcaster monolith. In the past, that would have meant either requiring developers to force users to import sensitive keys into their apps or omitting the feature from their clients altogether. Finding this to be the case, I reached out to Rish, who had the team rework the API’s signature scheme to enable external signers to submit valid requests to the Spaces endpoints. Within one week of my request, my client had feature parity, and I was able to build on top of the primitive by adding recordings and some other Juke-specific features.

Quorum (see also: Hypersnap, Quilibrium) is the escape hatch, the release valve, and the exit plan. Hypersnap is a fork of the Snapchain protocol, deployed on a custom L1. It enables 1:1 feature parity with the Farcaster Snapchain while offering enhanced flexibility across the stack. Cassie’s approach ensures that primitives that should be protocol-native will be (or already are). They are shipping at a breakneck pace, and their companion client, Quorum, is a robust, albeit slightly UX-deficient, alternative to the Farcaster App. Its very existence ensures Neynar’s neutrality. If Rish’s camp rallies against Quorum, they expose their alignment with past ideals. If they ignore it, they appear indifferent, which is perhaps worse.

I’ve dissented to token-driven economics across a distributed social layer ad nauseam. I’m not here to relitigate. But it is one of the core reasons I am not fully team Hypersnap. From an objective standpoint, the SNAP token was designed in the purest possible way: a fair launch, no VCs, no team allocation, and, ostensibly, a deterministic reward for participation rather than an airdrop. Yet on claim day, just about everyone (including me) dumped their allocation, as if it were an airdrop. Because, frankly, it was. Moving forward, there’s no incentive to buy or hold the token. Quorum’s team has stated in no uncertain terms that there will be no effort to promote the token, even at the most basic level (i.e., paying to claim the DexScreener page). SNAP is down more than 90% from its claim-day price with thin liquidity and minimal volume. No incentive to buy or hold means no incentive to participate for many. Of course, that’s the point. The token is designed to work as an anti-spam measure that rewards participation, not as a speculative asset. When assessed through that frame, sure, it’s working.

While some have rallied around the Quorum fork, it would seem they are a vocal minority. Most remain firmly on the fence, with feet in either camp. I, and I expect many others, are stuck in this stasis because we have no reason to believe history will repeat itself. So far, Rish and the Neynar team have appeared collaborative, malleable, and open to feedback. But because there are still primitives locked behind Merkle’s historical decision tree, apprehension is at least well-warranted.

Farcaster’s cast requesting feedback on channels is telling. Perhaps there is consideration being made about finally opening that primitive to the broader ecosystem. Will DMs make their way to the protocol? System-wide blocking? Perhaps a more frictionless way to register new users (read: signers)? As recently as last night, Cassie was proposing a new identification scheme and already has an FIP open on Hypersnap for channels.

One can’t expect Neynar to address every stale condition in a single day. But the clock is running. Many of the top builders have already left the ecosystem, and user retention continues to trend downward. Without a growing distribution layer, even building mini apps carries a negative expected value. The only independent Snapchain host, Uno, has moved on to building their own protocol. Yet, there are those of us showing up every day. Even Cassie’s adversarial approach demonstrates her deep desire to see Farcaster succeed.

The frustration I have with all of this is that I know every variable in this equation and still can’t solve for the outcome. As a client dev, I’m currently being stabilized by the forced neutrality of two opposing forces. But if one force yields, where do I end up?

Share