Web3 Data 2023, in retrospect

The flow of new events and big changes in crypto is constant. That's why it's easy to lose track of what even happened a month ago, let alone a year ago. 

As the year nears its end, we're taking a look back at some of the data-related projects that have impacted web3. 

The first one might surprise you. 

Bitcoin Ordinals gone wild 

If you thought that Bitcoin was, at best, a store of Value and there'd be no innovation happening there outside of incremental improvements to the Lighting Network, 2023 proved us all wrong. 

The introduction of Ordinals breathed life into a chain that many had already written off as "just for HOLDing". 

Ordinals allow anyone to attach further data to individual Satoshis, the smallest unit of Bitcoin. When creating tokens on Bitcoin, there is no separate location for metadata. Instead, the data is "inscribed" in the tokens, which is why this type of asset is also called inscription. 

What differentiates Bitcoin-based NFTs and tokens from others is that all data is on-chain, leading some to call them the better NFT (many Ethereum NFTs have their metadata stored off-chain, not necessarily in decentralized databases). 

Ever since the first mint of an inscription, they've steadily grown to hit over 47 million minted by more than 260k users. 

Yet, while some celebrate Ordinals as much-awaited utility and increase demand for block space, not everyone is happy. By inscribing data in tokens, the amount of computation to verify blocks is rising, much to the joy of miners who benefit from higher network fees. They raked in more than $170 million in fees on validating blocks with inscriptions. 

However, the hardcore Bitcoiners who believe that Bitcoin should be, first and foremost, a store and value and a payment network argue that Ordinals are an inefficient use of block space. Some, like Jack Dorsey, are now censoring blocks with inscriptions. That's not very libertarian. 

But inscriptions didn't stay a Bitcoin-only phenomenon. They started taking over other Layer-1 and Layer-2 networks. Savvy developers figured out that people love to speculate and brought inscriptions to Ethereum, Avalanche, Polygon, NEAR, and beyond with impressive results. 

Ethscriptions have reached more than 122k users who inscriped over 2,7 million tokens. More recently, inscriptions have launched on alt-L1s. The launch of inscriptions on NEAR triggered a transaction record of 13.9 million daily transactions, generating more than $170k in transaction fees. On the Gnosis chain, 50% of transactions were inscriptions in early December. However, not all blockchains held up well to the launch of inscriptions. TON, the telegram blockchain, didn't live up to its high scalability claims and slowed down to a halt due to the introduction of Ordinals.

At the time of writing, the Arbitrum sequencer is down under the load of inscriptions, driving up transaction fees on Layer-1 as people seem to try to force their transactions via Ethereum.

Takeaway: If anything, inscriptions have proven to be a great stress test for scalability claims and showcased some of the current pitfalls of blockchain architecture. On chains that support the creation of tokens and NFTs, the question remains: what is the benefit of inscribing vs. using the defined token standards? For Bitcoin, it's a glimpse at what could be a split in the community between those who see Bitcoin solely for payments and those who argue that Ordinals could provide a model for when Mining rewards run out. 2024 will be an interesting year for Bitcoiners, and all the chains who have yet to go through their inscriptions launch. 

The rise of modular blockchains 

Modular blockchains were another narrative that accompanied us all throughout the year. Modular chains, as opposed to monolithic chains, focus on optimizing for one or two features of blockchains while leaving the rest to the developer's choice. 

The most discussed features were data availability and security, the latter brought forward as a "rentable" feature by Cosmos' introduction of shared security and Eigenlayer's re-staking. 

Data availability 

Data availability describes a state in which sufficient data is available to any network participant to reconstruct the state of the chain. Because of Ethereum's pivot to a rollup-centric scaling roadmap, data availability, short DA, quickly advanced to a hot topic as rollups realized that most of their operating cost is defined by the cost of publishing their data on Ethereum. 

Currently, the most popular rollups, such as Arbitrum, Optimism, and Base, are publishing their transaction data alongside proofs to verify honest computation in Ethereum CALLDATA. Since storage space on Ethereum is limited, publishing to the mainnet quickly started becoming expensive. This is the role that DA layers hope to fill: becoming the place for rollups to publish their transactions at a lower cost with similar guarantees. 

In 2023, even though Ethereum has put an update on the roadmap aimed at increasing storage, Proto-dank sharding has now been pushed to 2024. And even when it goes live, chances are that modular chains might prefer a DA layer for storage for more persistent access to the data. 

The most notable modular data availability provider in 2023 was Celestia, whose mainnet went live in October and has since clocked in a monthly rolling average of 12,000 transactions. When looking at the price of its native token, the trend seems to be up only, with the recent announcement of integration into Polygon's CDK leading to a surge to $12.35 (the token started trading at $2.31). 

However, Celestia isn't the only Data availability layer that went into production in 2023. Others include Avail, a spin-out from Polygon Labs that focuses on allowing light clients to verify data availability easily, and Eigenlayer DA, a data availability solution offered by Eigenlayer Labs. NEAR, technically a monolithic blockchain has also started to offer its blockchain as an alternative DA for rollups. 

Takeaway: The DA space seems to continue heating up, with countless chains now competing to become a go-to choice for rollups to publish to. One aspect that's not yet sufficiently discussed when it comes to rollups publishing their data on yet another chain, though, is the question of how that changes the assurances. The biggest risk when relying on a less proven DA could be that if that data layer fails, the entire infrastructure of a dApp will fail with it. 

Decentralized RPC protocols 

2023 was also the year that crypto started talking about censorship resistance again. While initially a core value proposition of crypto, it often got drowned out in the noise of bull markets. With 2023 marking more of a winter season, maybe unsurprisingly, people came to realize that dApps are only as decentralized as their connection to the blockchain, which often wasn't at all. 

One of the events that highlighted this was when Metamask users in various US-sanctioned countries suddenly couldn't execute transactions anymore. This stood in stark contrast with the supposed permissionless nature of the blockchain. Yet, the problem is larger than just a single wallet provider. The part that censored transactions wasn't the wallet itself but the RPC provider, in this case, infura. 

RPC providers connect dApps and wallets to the blockchain nodes. Most dApps don't run their own full nodes but rely on centralized providers like Alchemy and Infura to handle communication with the blockchain. While there is room for centralized RPC providers, and they've arguably been a backbone of the space overall, they, too, face challenges such as regulation, single point of failure, and potential downtime. Additionally, if all of crypto starts relying on a handful of centralized providers, the entire industry is standing on shaky grounds. 

Decentralized RPC protocols address the centralization problem and enhance the security, interoperability, and censorship resistance of the blockchain middleware layer. Since RPCs sit between users and blockchains, dApps are only ever as decentralized as the RPC networks they rely on. 

RPC network architecture combines relayers and a network of nodes to serve RPC requests across multiple chains. By decentralizing the RPC process and incentivizing node uptime, decentralized RPCs increase resilience, redundancy, and user privacy. 

One of the more established RPC networks is Pocket Network, which has seen continued growth this year, hitting over 540 million relays in the last 24 hours, most of them on Ethereum. Lava Network is another popular choice for developers to tap into decentralized RPCs. Initially built on Cosmos, the protocol has expanded to serve more than 25 chains with high-quality RPC services. This means developers can make calls to multiple chains using the same set of credentials, further simplifying the multichain development process. 

Another network that went live in 2023 is DRPC, short for decentralized RPC, a collaborative project that mostly serves EVM dApps and hit more than 5 billion requests this year. 

Takeaway: An increased focus on decentralized middleware reflects the broader awareness of centralization constraints. While initially, questions of access and censorship resistance focused on the consensus and blockchain layer itself, the industry is now taking a more holistic stance where a dApp or service is only as resilient as the weakest of its parts. 

Web3 Data APIs 

In line with decentralizing the entire data stack, decentralized APIs have become more prevalent in crypto in 2023. APIs, application programming interfaces, are how developers access data or how apps interact with each other on the internet. Usually, APIs are offered by centralized parties or companies as part of their service.

If you ever start to learn programming, you'll quickly learn that there's an API for anything from weather data to quote generators to this API where you can get random cute dog pictures. 

Crypto traders often also have some experience using APIs when connecting their exchange accounts with portfolio tracking services or even running trading bots. However, going beyond trading, APIs enable developers to quickly build with access to a wealth of data, leveraging on-chain data such as token distribution or users' NFT collections. 

While in the traditional web2 world, there's often no other option than to use the centralized API from the platform whose data you want to access; things are different in the context of on-chain data where everything is out in the open. 

One of the leading dAPI providers is API3, a project launched three years ago; the developers launched a managed service to facilitate access to its network of quality data providers. API3 serves as a gateway through which it enables dApps to fetch verified data from data providers. Since every data provider signs their submissions on-chain, proving that data hasn't been tampered with. API3 is governed by the API3 DAO and available across 11 chains with feeds for on- and off-chain data. 

But API3 wasn't the only web3 data API winning in 2023. Airstack has quickly become a go-to development platform, especially for builders in the Farcaster, BASE, and broader Ethereum ecosystem. As an API, Airstack provides quick access to on-chain data such as transactions and off-chain data from Farcaster hubs, NFT metadata, and Lens protocol - allowing projects to rely on it for social use cases. Various projects, including Tokenbound, Converse (a messenger using XMTP for intra-wallet comms), and Builder.fi, already rely on Airstack for their data feeds. 

And if you're just a curious on-chain explorer, you can try Airstack's explorer to find out what NFTs someone holds or who they follow on Lens. On-chain stalking has never been easier. 

Takeaway: Oracles might not be the only game in town anymore. Decentralized APIs and teams like Airstack are proving that bringing data on-chain doesn't always require an entire third-party oracle but can be solved by bringing existing APIs on-chain. Either way, for builders, it's a win as they are gaining one more option to build on. 

Blockchain Indexers 

While the sentiment in 2023 has shifted away from building more infrastructure to building for consumers, the need for infrastructure that can keep up with an increasingly fragmented stack hasn't subsided. Modular blockchains are great; nevertheless, they might make it harder to obtain a holistic picture of on-chain activity. 

Similarly, even though Layer-2s have succeeded in reaching a 5x scaling factor for Ethereum and gained meaningful adoption from users, they also fragment the experience and availability of data - especially when combined with a Data availability Layer. 

Making matters more complicated is the introduction of so-called Layer-3s in 2023, such as Momoka by the Lens protocol team, a social graph that requires higher throughput, fast finality, and cheap fees for non-financial interactions such as liking posts.

Successful social networks host plenty of interactions, ideally, none of them requiring the constant signing of transactions. Farcaster, another decentralized social graph, has picked a separate route, storing a significant portion of relevant social data in Farcaster hubs, off-chain storage locations. With now more than 211,000 users and roughly 8,000 monthly active casters, it's one not to be underestimated for any crypto founder or builder. 

All of this puts developers looking to tap into the openly available social data or into Layer-2 activity in a multichain dApp in a difficult spot: they need to access all of that data, often stored across chains, layers, and in off-chain locations - often not in a standard format. 

Blockchain indexers provide a tool that empowers developers to not worry about how they can collect the data and are also a crucial piece in the infrastructure of any API provider. No wonder that indexers have seen high demand this year, with projects looking to build in a multichain, multi-data location paradigm. 

The Graph, the probably best-known indexing protocol, is celebrating its third birthday in space. It has added further support for new chains to its indexing tool kit and enhanced the speed of its query performance. They also successfully migrated the protocol smart contracts from Ethereum mainnet to Arbitrum One and increased their subgraph count to 1,322 subgraphs. While its active Indexers decreased by 41% quarter over quarter in Q3 '23, its delegator count grew nearly 2,5x fold while curators increased by 5%. 

The Graph started as a pretty centralized, hosted service and has taken significant steps to decentralize its network while facilitating access this year. 

At the same time, other more decentralized indexing networks have made significant progress, with Subquery integrating more than 100 networks in 2023, including Ethereum, Polygon, Cosmos, Algorand, NEAR, and Avalanche, while expanding its indexer count to 85. 

And, of course, 2023 has also been an incredible year for Subsquid. Our multichain indexer and data lake now supports over 100 networks, and even though our blockchain network is in testnet, we've already served over 30 billion requests, fuelling apps from social use cases to DeFi and even powering entire protocol's on-chain token discovery. 

Did we mention that more than 40,000 indexers have already been deployed, further increasing the resilience of Subsquid? Our team continues working on increasing data sources we aggregate so that builders can focus on their use case and index entire chains in minutes. 

Takeaway: As we work on decentralizing the entire stack, indexing will be a key pillar of any dApp. What sets the successful services apart will be multichain support, integration with other decentralized layers of the stack, speed, accessibility, and cost. Ultimately, the UX will often just be as good as a dApp can fetch the necessary data. 

Conclusion 

A lot has happened in 2023, with web3 data discussions dominated by the rise of Ordinals, which have added a tremendous amount of data to any chain they were added to (sometimes to the detriment of performance) and tools that enable developers to access a fragmented, ecosystem of chains. 

With more modular chains and Ethereum's introduction of Danksharding, where part of the data won't be stored persistently, how to get data will become a data consumers' challenge. Builders can be optimistic. Countless strong teams are working on optimizing the web3 data economy. 

And if you're interested in just how fast one can start indexing a chain, why don't you try out Subsquid? 

Subsquid docs | GitHub | Twitter | Developer Chat

Article by @Naomi_fromhh