It has been more than two years since I (Joshua) first started RSS3. It has been a truly incredible and meaningful journey - we’ve kept our heads down, working tirelessly on shipping the RSS3 Mainnet. And today, we are finally ready to share some insights into the upcoming ultimate Open Information infrastructure.
For the newcomers to the RSS3 community recently, RSS3 is inspired by Aaron Swartz and the original RSS protocol. RSS3 started in 2021 and has been aiming to build the Open Information infrastructure, paving the way for a new generation of applications designed for the Open Web. At that time, none of the modules that RSS3 aimed for - social, search, or AI - was big. The only blockchain-based content publishing project was Mirror, publishing around 70 articles a day.
While evolving toward the RSS3 Mainnet, RSS3 has been the humble driving force behind lots of innovations in the fields with our Testnet:
- The manifesto toward Open Information
- The advocation toward platformless media
- The introduction of the RSS3 specification, which later turned into the RSS3 UMS
- The first browser extension syncing content to decentralized networks with Re: ID
- The first aggregated profile page with Web3 Pass
- The first experience bridging Web3 social to Web2 with Mask Network
- The first feed of tracking on-chain activities with Revery
- The first global-scope Web3 search engine with hoot.it
- The first AI training environment based on Web3 content
- The first Snap bridging Web3 social to MetaMask
- The first GPT bridging Web3 content to GPT-4
And as the RSS3 ecosystem continued to evolve and expand, the world of Web3 grew dramatically as well. Unlike 2 years ago, the two information-driven verticals, social and AI, have become the most promising fields within the new Open Web.
There is still a substantial amount of work remains to be done, for RSS3 to become the Open Information Layer that benefits billions of Internet users. However, with the upcoming launch of the RSS3 Mainnet, that goal is closer than ever.
Designing a network like RSS3 presents numerous challenges, especially when examining Web3 (or the broader concept of the Open Web) in its current state; what is one critical piece to make it the next-generation Internet? Historically, every Internet revolution began with novel methods of creating information, followed by innovative ways of utilizing and consuming it.
For Web1, it was “writing HTML webpages” and “RSS + Google.”
For Web2, it was “posting on apps” and “Twitter + Instagram + TikTok.”
Now, as Web3 develops, a new paradigm of creating information emerges: users across different protocols and networks are generating data with diverse standards. The process is open (similar to Web1) and facilitated by service providers for a better experience (similar to Web2).
If we want to build the information layer for Web3, one of the fundamentals is the ability to cover as much Open Information as possible and ensure its readiness for any applications. This is the bedrock for building social media, search engines, and LLMs with Web3, setting the stage for future innovative applications and projects.
Additionally, embracing the Web3 ideology, this information layer must be open and decentralized, allowing any participant to join and contribute without a single entity exerting complete control.
Finally, this network should foster an environment where ownership and value are verifiable, constructible, and materializable - this is vital for network participants to foster a diverse ecosystem.
The full technical specifics will be available in the forthcoming RSS3 Mainnet Whitepaper. However, we are eager to offer some early insights into the RSS3 Mainnet right here.
A simplified architecture of the RSS3 Mainnet looks like this:
On the top level, the RSS3 Mainnet includes the Data Sublayer (DSL) and the Value Sublayer (VSL). The DSL consists of RSS3 Nodes constantly indexing and structuring Open Information for various modules, while the VSL handles value and ownership with our own blockchain.
The RSS3 DSL is a decentralized network built to oversee Open Information’s entire life cycle on the RSS3 Mainnet.
Serving Nodes (SNs) are permissionless components responsible for indexing and structuring Open Information.
- Operation: Anyone can operate an SN to join the Network permissionlessly, reflecting the spirit of the Open Web.
- Indexing: Each Node decides its own information coverage and operates its own set of workers that continuously index data from what we refer to as Permissionless Data Sources. This ensures that quality information is always readily available.
- Serving: Structured Open Information is served to consumers via the Global Indexer (GI), combined with high availability.
Global Indexers (GIs) provide performance and quality assurance for the DSL. It dynamically finds the optimal path for incoming requests based on multiple factors, including the scope of the request and the quality of SNs.
- Record of Work and Proof Onchain: The history of work and slashes for all SNs are submitted to the VSL for settlement, ensuring a fair and transparent process.
- Network Transparency Metrics: The DSL upholds its commitment to openness and transparency by providing open access to metrics that reflect all aspects of the Network.
Consumers pay request fees in $RSS3 to the DSL. All fees are automatically handled by the VSL and distributed to SNs according to their record of work submitted by the GI.
The RSS3 VSL is a modularized Ethereum Layer 2, built to handle value and ownership of Open Information (social, search, AI, and beyond).
Blockchain is the technology naturally designed to handle value and ownership with consensus. Therefore, we’ve decided to have VSL in the form of a blockchain. There are quite a few approaches to building a blockchain, and we ended up having an Ethereum layer 2: its EVM is the most widely used environment for smart contracts, it offers great potential liquidity, and $RSS3 for the current Testnet is issued on Ethereum Mainnet - it will smoothly evolve as the utility token for the RSS3 Mainnet.
For Ethereum rollups, there are still ongoing debates on what scaling solution is the best. We’ve decided that, for now, the RSS3 VSL will be built with OP Stack for its efficiency and ecosystem. As a relatively mature solution powering multiple L2s, Optimistic rollup has been well-proven. And we’ve adjusted OP Stack to serve the purpose of VSL best.
Although L2s save developers an immense amount of gas fees per transaction in comparison to direct deployment on Ethereum Mainnet, the cost is still too high. The Value Sublayer will be the destination for the RSS3 network settlements and developers building search engines, explorers, and activity feeds–meaning constant, high-frequency microtransactions. We’ve made the strategic decision to build the RSS3 VSL with Celestia underneath. This means access to Open Information is truly democratized, and developers building on the Value Sublayer will enjoy ~100x less fees.
The RSS3 VSL will have $RSS3 as its gas token, making it the first native gas L2 with Optimistic rollup and Celestia DA. The sequencer will take $RSS3 and convert it into $ETH and $TIA. And if that results in sequencer revenue, it will go to public goods and ecosystem grants.
The RSS3 Mainnet is set to go online in the first quarter of 2024, and we are looking forward to sharing more information about the network and how it will be built with you.
The introduction of the Mainnet is set to unlock for all of the RSS3 community:
- Set up a node and contributed to the network
- Build your own search engine verticalized to its own need
- Train a language model with structured open information
- Stake or trust with your tokens
- Construct a Web3 social app with a completely decentralized tech stack
- Create dApps on RSS3 VSL related to open information
- Deploy local language models on a serving node
- and much more
If you are looking for more technical details, there will be a detailed upgraded version of the white paper coming soon.
If you are looking for the path ahead, stay tuned to the upcoming roadmap, where different stages of the network will be introduced.
If you are a node operator looking for docs to join the network, get ready, as they are coming shortly.
If you are looking for tokenomics coming with the Mainnet, it is also on its way - we’d love to share how the network incentivizes everyone involved.
And if you have more questions…
Is RSS3 transitioning into a blockchain?
No. The RSS3 network contains a layer2 blockchain.
Why doesn’t RSS3 deploy on an existing blockchain but rather build its own?
With a blockchain customized to the needs of the RSS3 network, the overall efficiency and performance increase dramatically. Also, Open Information (social, AI, search, etc.) has the potential to grow into a separate yet thriving ownership and value ecosystem.
Is there going to be a separate token?
No - the current Ethereum $RSS3 will be the Mainnet utility token.
What will happen to the current RSS3 Testnet?
It can go both ways - it gets shut down or stays as a testing environment. But either way, developers should get prepared to switch to Mainnet.
Who can run a node for Mainnet?
Ideally, everyone with an eligible computation device, from a personal computer to a server cluster.
Is the Mainnet the same as RSS3 Network 2.0?
Why did it take so long to develop the Mainnet?
RSS3 is a novel network with few projects to learn from - a lot were built from scratch.
Will there be a new White Paper NFT with the updated version?
Maybe - if so, existing White Paper NFT holders should be able to claim one.
What will the tokenomics be like?
It’s gonna be FUN.
Let’s keep building.