Something BIG is coming, but it is very small

Are small blockchains becoming the next big thing?

Light clients, snapshots, checkpoints, merkle proofs, and above all, succinctness. From Solana to Ethereum, decentralization of blockchain access is getting faster these days. And if you’ve never heard of these concepts (or never actually thought about how important they are), this is a very good time to start.

This article is about not trusting your blockchain.

Is it really trustless, your dApp?

Short answer, no. Each and every blockchain is very trusty at the very final point that they touch their users: wallets and RPC endpoints. Currently, there are ~10 big RPC companies controlling almost all blockchain connections. Compared to the magnificent 7 stocks, how can you claim blockchain is fundamentally different?

Sure, some blockchains try their best to allow their users to verify the information they receive, but how successful are they? Light clients are nice, with some merkle proofs and checkpoints on their consensus, they offer good-enough guarantees to keep RPC endpoints from deceiving users. But in the end, the consensus itself is trusted to be accurate, and just as if we wait 6 blocks for our Bitcoin to arrive, any information that we receive from them is subject to a hard finality.

Let’s formalize the problem

Take the problem from 2 perspectives: The first is trusting signatures. Verifying a merkle proof can only confirm a client that the current validator root has signed the given state, but how do we know the validator root is right? The only way to be perfectly sure is to check the previous block, to make sure the last consensus actually signed and approved the new one. However, the problem iterates: How can you know the previous root was not compromised, without checking the root before?

This problem is called synchronization. All blockchain validators (if they want full trustlessness) need to get and run all blocks since genesis to make sure the last block is provided to them accurately. In a perfect network connection, where you reach a decent part of the validator set (like more than 50%), getting the final data from all parties you are connected to is secure enough. But in a light client that a single RPC is providing you with a claimed root, you just assume that no double sign has occurred for the current block, and the information you are getting is right.

To summarize, this problem is about the fact that blockchain is not a single entity, and no single signature verification is enough to offer complete confidence.

And this is not all

The second perspective is censorship, as always. Let’s assume your favorite RPC provider is down, or they decided that they don’t like you anymore. Thus, you cannot reach the information in the final block. How can you get the latest information? Or can you even be sure that there is no delay in their serving of data?

At this point, the theory of decentralization asks the question: “But how hard is it for you to run your node?”. In the end, this is the main issue of today’s system. You can very well be aware of how evil your social media platform is, but do you really have a choice to change it? The centralized technology dooms us to keep the current system, as starting a better one has a bigger and bigger cost with each new day and user.

The hardware-heavy aspect of synchronization can be understood as a very similar problem. As running a node is not instantaneous, even if your machine is strong, you are subject to a serious time cost if you choose to self own your RPC. The only choice you have is to run it always, but this is expensive, as blockchain nodes are not simple servers.

But, why is this important?

Now think that you are making some DeFi trades, and watching the price’s sudden changes to decide on your next move. Even a second of delay can be a big problem, right? Or in the case of a country-wide election; is the blockchain ready to promise no censorship or delay on such a massive use?

The problem is not obvious right now because blockchain is not mass adopted. This was true for the early days of the internet or social media as well. Can we keep this system running as decentralized when there are billions of users on it?

Above all, the worst part is that as a user or developer, you do not have any choice. The current blockchain infrastructure is limited, and you cannot prevent these attacks or go outside of the problem with better code. We need a fundamental solution.

Luckily, we do have one.

This article is not just about problems. There is already a lot of very decent effort on this front, as we have cited some most recent at the beginning of the article. Light nodes and succinct blockchains offer a better layer for developers to use, and users to verify.

If reading this made you concerned, there are 3 things you need to do:

1. Go to your favorite search engine (or LLM these days) and ask for these concepts. There is tons of material on these subjects online, and learn what the actual solution is.

2. Ask for the solution to be implemented. As users, we have the choice, and each day that we settle for the inferior tech, we lose something. Fight for your technology and ask for better ideas.

3. And once it is there, care for it. Use the better tech, actually embrace the power that it brings to you.

Decentralization was never about tech or finance, it was always about people. Now, as the next user of blockchain, what do you want to see implemented?