Omilos is a client and indexer built on top of the sufficiently decentralized social network (DSN) Farcaster. This means that we are not in the business of on-boarding users, but in providing existing users a better experience on the network, without them having to create a new account and rebuild their social graphs. As long as those users are active on Farcaster, they can use our client seamlessly, they just have to login.

This article explains our reasoning behind our choice as there is a lot of options in term of decentralized social media.

Farcaster is "sufficiently decentralized", which means that it decentralizes only what is strictly necessary. The component that should be decentralized in a social network is a user's identity, which contains a pointer towards that identity's data.

As long as you have a guarantee over who is who, you can trust the pointer of that identity, which will point towards an authenticated collection of social media content. In the context of Farcaster, each account is registered on the blockchain, and contains a URL pointing towards the hub where that user's data resides.

The Hub is centralized, but they cannot manipulate the user's data as it is authenticated with the user's identity, using asymmetric cryptography. The only thing the Hub can do, is refuse to relay a user's data.

This technology choice is interesting as it allows for a certain scalability. If all the social media data was put on the blockchain, the network would get saturated very fast. During the world cup, 24 000 tweets were posted per second on Twitter, a lot higher than the mere ~10 transactions per second possible on Ethereum

The only bottleneck is how fast Hubs can communicate with each others, to ensure they can stay up to date with each user's data.

Because there is no constraints on how the data is stored in the protocol, each Hub can decide their retention policy and storage layer. At Omilos, we will try to archive as much of user's data as possible to allow for historical retrieval. Of course, this also leave the opportunity for Hubs to implement stronger data availability guarantees, like posting each content on a blockchain, resulting in a completely decentralized social network. We believe a good middle ground would be using special purpose data storage network such as Filecoin or Arweave.

What makes a social network, is its community. The prime example of that was the incredible explosion of the app "Clubhouse", which got very popular very quickly due to how exclusive it felt being on there, where you could end up in a coversation with Elon Musk or Jack Dorsey.

This is why we believe in Farcaster, because of its highly qualitative community, which is needed to bootstrap a social network. We want to make the on-boarding as simple as possible, that is why we are building a tool to bootstrap your social graph using your existing twitter account.

Lastly, the political neutrality of a protocol is of uttermost importance. Because the Farcaster is politically neutral at the base layer (there is, to our best knowledge, only decisions that involve technological choices), we feel confident to build on it. Other "decentralized" social networks have implemented choices on the protocol layer that can be influenced by political orientations. An example is a DSN making the decision to allow or not an NFT as profile picture. We argue that this is an opinion and has nothing to do in the protocol layer, therefore the choice should be abstracted away at the client level instead.

Other social networks like Mastodon have clear political bias, with the founders being very vocal about their political orientations.

So what's our vision, why are we building a DSN ? We are planning to build high quality clients, which abstract away all the blockchain logic in the backend. We want to onboard as many user as possible and make the UX simple enough to reach mainstream level.

We believe social networks can become so much better with a proper incentive structure between the network and its users, and the data portability Farcaster offers brings that. We will work on a subscription basis, as we don't want advertisement to pollute a public space. We plan to curate high quality contents and incentivize constructive behaviors, so you get the most out of your time.

All in all, we believe Farcaster is our best bet for a DSN to emerge and reach mainstream adoption. That's why we have decided to build on it. In future posts, we will dig deeper into the technological choices of Farcaster and explain why they make sense to use. In the meantime, please go to omilos.xyz and explore the alpha version of our client, and follow us on twitter for updates on exciting news.

Omilos is a client and indexer built on top of the sufficiently decentralized social network (DSN) Farcaster. This means that we are not in the business of on-boarding users, but in providing existing users a better experience on the network, without them having to create a new account and rebuild their social graphs. As long as those users are active on Farcaster, they can use our client seamlessly, they just have to login.

This article explains our reasoning behind our choice as there is a lot of options in term of decentralized social media.

Farcaster is "sufficiently decentralized", which means that it decentralizes only what is strictly necessary. The component that should be decentralized in a social network is a user's identity, which contains a pointer towards that identity's data.

As long as you have a guarantee over who is who, you can trust the pointer of that identity, which will point towards an authenticated collection of social media content. In the context of Farcaster, each account is registered on the blockchain, and contains a URL pointing towards the hub where that user's data resides.

The Hub is centralized, but they cannot manipulate the user's data as it is authenticated with the user's identity, using asymmetric cryptography. The only thing the Hub can do, is refuse to relay a user's data.

This technology choice is interesting as it allows for a certain scalability. If all the social media data was put on the blockchain, the network would get saturated very fast. During the world cup, 24 000 tweets were posted per second on Twitter, a lot higher than the mere ~10 transactions per second possible on Ethereum

The only bottleneck is how fast Hubs can communicate with each others, to ensure they can stay up to date with each user's data.

Because there is no constraints on how the data is stored in the protocol, each Hub can decide their retention policy and storage layer. At Omilos, we will try to archive as much of user's data as possible to allow for historical retrieval. Of course, this also leave the opportunity for Hubs to implement stronger data availability guarantees, like posting each content on a blockchain, resulting in a completely decentralized social network. We believe a good middle ground would be using special purpose data storage network such as Filecoin or Arweave.

What makes a social network, is its community. The prime example of that was the incredible explosion of the app "Clubhouse", which got very popular very quickly due to how exclusive it felt being on there, where you could end up in a coversation with Elon Musk or Jack Dorsey.

This is why we believe in Farcaster, because of its highly qualitative community, which is needed to bootstrap a social network. We want to make the on-boarding as simple as possible, that is why we are building a tool to bootstrap your social graph using your existing twitter account.

Lastly, the political neutrality of a protocol is of uttermost importance. Because the Farcaster is politically neutral at the base layer (there is, to our best knowledge, only decisions that involve technological choices), we feel confident to build on it. Other "decentralized" social networks have implemented choices on the protocol layer that can be influenced by political orientations. An example is a DSN making the decision to allow or not an NFT as profile picture. We argue that this is an opinion and has nothing to do in the protocol layer, therefore the choice should be abstracted away at the client level instead.

Other social networks like Mastodon have clear political bias, with the founders being very vocal about their political orientations.

So what's our vision, why are we building a DSN ? We are planning to build high quality clients, which abstract away all the blockchain logic in the backend. We want to onboard as many user as possible and make the UX simple enough to reach mainstream level.

We believe social networks can become so much better with a proper incentive structure between the network and its users, and the data portability Farcaster offers brings that. We will work on a subscription basis, as we don't want advertisement to pollute a public space. We plan to curate high quality contents and incentivize constructive behaviors, so you get the most out of your time.

All in all, we believe Farcaster is our best bet for a DSN to emerge and reach mainstream adoption. That's why we have decided to build on it. In future posts, we will dig deeper into the technological choices of Farcaster and explain why they make sense to use. In the meantime, please go to omilos.xyz and explore the alpha version of our client, and follow us on twitter for updates on exciting news.