Dialogue Gavin Wood: Polkadot's parachains may exceed 100, cross-chain bridges are not the future

Original Title: "Gavin Wood: Polkadot's Parachains Could Exceed 100, and a Third Relay Chain May Arise in the Ecosystem"

Written by: The Defiant

Translated by: PolkaWorld

Recently, Polkadot founder Gavin Wood appeared on The Defiant podcast, where he discussed some very interesting topics with host Camila, including:

• What are the differences between Polkadot and Ethereum 2.0 in terms of sharding?
• How many parachains can Polkadot support at most? Where does the limitation come from?
• What is the relationship between the value of DOT and the success of parachains?
• Will there be other relay chains in the Polkadot ecosystem besides Polkadot and Kusama?
• How does Polkadot empower developers and users?
• Is a "bridge" the ultimate solution to connect blockchains?
• What is Web3?

Here is a summary of some content from the podcast by PolkaWorld:

Ethereum 2.0 VS Polkadot

Camila: Ethereum 2.0 is also working on sharding, so can we say that the difference between it and Polkadot is whether it uses WebAssembly? Is WebAssembly a direction you want to experiment with?

Gavin: WebAssembly is indeed one of the main differences, along with several other distinctions. For example, our approach to sharding and our consensus mechanism, as well as specific differences at the implementation level; we proposed sharding earlier.

A significant difference lies in what we call the sharded state machine. Our sharded state machine is primarily based on WebAssembly, meaning you can write any computer program in any language you prefer, as long as it can be compiled to WebAssembly.

You can directly write a chain, but this chain will be secured by Polkadot, and it will use the same security model and staking asset pool as Polkadot. This means that if someone colludes to attack your chain, it is equivalent to attacking Polkadot.

So, it is indeed a sharded structure, where the security of each shard is as high as that of Polkadot, but each shard can do completely different things and run independently in parallel, meaning different tasks can occur simultaneously on each different shard.

Thus, I think the main product-level difference between the two is that in Polkadot, you can rent an entire shard, while Ethereum (2.0) has been trying to stick to the Ethereum model, just making that model more scalable, with scalability primarily coming from having more workflows, allowing some smart contracts to run on this shard and others on another shard.

The idea behind Polkadot arose from the question, "What if these shards don't necessarily have to be smart contracts? What if they could be any kind of chain? What if one focuses solely on domain registration, another on DeFi, and another on NFTs?" Would there be advantages to doing this?

For me, the answer is certainly yes, because there are many high-throughput applications. Sometimes you know you definitely need to handle a lot of transactions for a particular application, like DeFi, so specialization becomes necessary due to the high performance requirements. At the same time, you have the opportunity to experiment; you can try different ideas on different chains without being confined to a smart contract model, and you can even have multiple ways of doing smart contracts.

In fact, there are already different chains promoting various smart contract models; some may be non-Turing complete, some have storage fees, and some do not. I believe experimentation is what makes blockchain great, and Polkadot can conduct around 100 experiments simultaneously, which excites me.

What Determines the Limit of Parachains

Camila: The shards in Polkadot are essentially parachains, right? So, is there a limit to the number of parachains?

Gavin: There is definitely a limit, but we haven't figured out what that limit is yet. We will continue to add parachains until we hit the limit, at which point we will optimize or take necessary measures.

In the early days, we speculated that the limitation would come from messaging. When you add more parachains, each parachain may want to communicate with more chains. For example, when you add a second parachain, there are only two chains in the network, A and B communicating with each other. But if you add the 101st parachain, then each parachain can communicate with the other 100.

This means the total number of connections would be 10,000, increasing exponentially. This is what we call the network effect, which is why Facebook has a snowball effect; every new person joining makes the network more useful than before. Therefore, messaging becomes more expensive as the number of parachains increases because the number of messages is the square of the number of chains, and the cost of transmitting messages increases dramatically with each new chain added.

Now we have an algorithm that helps us avoid this problem, so we can confidently say that we can support more parachains than the originally envisioned 100, possibly several hundred. However, realistically, there will still be a limit, and we can only find that out gradually.

In reality, the bottleneck may not come from messaging but from the block finalization algorithm. Our finalization algorithm is called GRANDPA, which requires all validators to communicate with each other. Although the communication content is simple, the number of validators must be ten times the number of parachains. So if you have 1 million parachains, you need 10 million validators.

Moreover, each validator must communicate with other validators, which means each validator's traffic would be very high, requiring massive bandwidth, leading to centralization because this demands validators to have very specialized equipment and data centers, which is something we do not want to see. We do not want to sacrifice decentralization, so if we reach a point where we must sacrifice decentralization, we will stop adding validators.

If I had to give a number, I would say our goal is to allow Polkadot to support around 100-250 parachains.

Polkadot Ecosystem: Scalability and Specialization

Gavin: Scalability is very important, and we achieve scalability through parachains (sharding).

This includes two aspects: first, we can break down the workload. Previously, one person had to do all the work; now we can assign tasks to 100 people to take home and then return the completed work to the original person. This is a very simple concept—parallelization.

On the other hand, each shard can specialize in a specific area, handling a particular type of work. Just like you wouldn't ask a doctor to do accounting work, the idea of Polkadot is that one shard specializes in accounting, another in NFTs, and another in governance, etc. This specialization allows chains to process their work faster, thus gaining performance advantages from non-generalization.

Ethereum is a very general smart contract platform, which allows it to do everything but not excel at anything. Polkadot, on the other hand, says we can have some all-purpose chains, but we can also have some chains that are very specialized in certain areas.

So scalability is a crucial point, but another important aspect is that when you create a new technology, you can only know its lasting impact after it is built and used. One thing I realized after working on Polkadot is that smart contract platforms are very suitable for smart contracts, but many things happening on these platforms are not actually the work of smart contracts.

Ethereum users mainly fall into two categories: one wants to deploy decentralized applications on the platform using smart contracts, and the other wants to use decentralized applications. The problem is that these users do not want to use Ethereum; they just want to use the applications on it. They do not care what ETH is; if decentralized applications are built on a platform that does not require them to care about the underlying platform token, it is actually better for users.

Similarly, these application developers are not very keen on building applications on smart contracts; sometimes they do not want to structure their applications using smart contracts. Many times, it is the smart contract platform that imposes design decisions on you that violate the way you typically structure software.

At Polkadot, we want to return design freedom to decentralized application developers. Furthermore, we do not want Polkadot to intervene between decentralized applications and their users because that would be foolish and would only reduce the utility of these applications for users. If users still need to interact with the underlying platform's DOT, it becomes a distraction for them.

Furthermore, if you largely restrict the users of decentralized applications to those of the underlying platform, specifically, if you do not already own some Ether, it becomes very difficult to use any smart contract on Ethereum. You not only have to understand the smart contract but also understand Ether; you have to go to an exchange, register, get some Ether, transfer it to an account, and then you can do something related to that decentralized application.

These steps can be cumbersome for someone who is not already familiar with Ethereum. Realistically, your target users are limited to those who already hold Ethereum, have MetaMask, and understand the concept of Ethereum and its accounts.

If we want to reach mainstream users, this is a significant issue. Polkadot aims to ensure that we deliver a decentralized application platform to developers, allowing their applications to interact directly with users without involving any intermediaries.

The Success of Parachains and the Value of DOT

Camila: So for example, users of the platform Acala on Polkadot do not need to hold DOT to pay transaction fees; they can directly hold ACA to use Acala. So where does the value of DOT come from? What is its tokenomics like?

Gavin: We do not want to separate the economics of DOT from the success of parachains, so what we do is allow parachains to be deployed on Polkadot, either as a parachain that serves the public good, providing some useful functions for all DOT holders, or through auctions.

We auction off slots for 100 or 200 parachains; these slots are somewhat like parcels of land in Polkadot. Teams can bid for these slots, and if successful, they can deploy their decentralized applications on Polkadot. Teams can attract users in any way they like without worrying about users needing to understand the DOT token.

The bidding method for these auctions involves locking DOT rather than paying DOT. During the period that the parachain team rents the slot, the DOT will be locked. Currently, the rental period for Polkadot is two years. Suppose a parachain team locks 1 million DOT to rent a slot; after the two-year rental period ends, this 1 million DOT will be unlocked. This process is guaranteed by the Polkadot protocol, ensuring that they can unlock these tokens at that time.

Now, the interesting part is how the team finds 1 million DOT to bid for the slot. The team may have a lot of DOT themselves, or they can seek financing or loans from VCs, or they can use Crowdloan. This concept is somewhat like crowdfunding, except that everyone is just lending their DOT rather than giving it away.

Similarly, this process is also guaranteed by the Polkadot protocol. So teams can gain support from many people through this method, and they may say that as long as you lend DOT to us, we will give you some rewards, which can take various forms depending on the team.

For example, they might offer 5% of the tokens, or provide NFT rewards, where holding this NFT allows you to become a collator node and earn rewards, or offer some discounts on services, etc.

Camila: How many parachains have been rented out through auctions so far?

Gavin: Polkadot currently has 6 parachains that are live.

Will There Be Other Relay Chains Besides Dotsama?

Camila: In the Polkadot ecosystem, besides the canary network Kusama and the main network Polkadot, do you think there will be other relay chains?

Gavin: I guess there might be another one or two. One trend I see is consortium-type chains, which are specifically built for industries, enterprises, and organizations. They may want to collaborate and communicate with each other but do not want to be part of a public chain, perhaps feeling there is no need to join another chain's economic system.

For these chains, using Polkadot is more suitable because Polkadot does not intervene in all transactions. So theoretically, we might see some relay chains connected to the Polkadot relay chain through bridges, which can also be considered part of the broader Polkadot ecosystem.

I think at least from my personal perspective, if a relay chain cannot provide value beyond Polkadot and Kusama, and is merely a clone, we may not welcome such chains.

The Future Shape of the Blockchain Industry

Camila: In the current environment of the blockchain industry, where multiple Layer 1 chains are competing, do you think Polkadot will become one of the major ecosystems, attracting most developers and users? Or will various chains connect with each other through bridges, etc.? What is your view?

Gavin: A few weeks ago, Vitalik published an article stating, "Bridges are declining; stop building bridges." Generally speaking, I agree with this viewpoint. I think bridges may play a relatively unimportant role; they suffer from "negative network effects."

That is to say, they are secure enough for a small amount of financial value flow, but if a bridge becomes very important and the traffic increases, it will become increasingly insecure because the security of the bridge is fixed, while the traffic passing through it may increase significantly. At some point, the value of the traffic exceeds the security of the bridge, and then it is just a matter of time before it gets attacked.

As long as the value flow through the bridge is kept within a reasonable range, meaning the bridge is not used too much, then the bridge solution can work. However, if you build an ecosystem on top of a bridge, that would be self-defeating, so I do not think that will be the trend of the future.

I hope we can maintain two or three major chains, each with different architectures. We always say Polkadot is Layer 0; it does not do much besides ensuring the security of parachains; the parachains are the ones providing applications for users.

So I think it is not far-fetched to imagine a world where only one or two Layer 1 chains are responsible for their own security, while more Layer 0 chains are responsible for the security of their ecosystems.

My main drive is to ensure that we provide the best platform for developers because I believe the most important thing right now is not to acquire users but to acquire developers, thereby ensuring we have the most attractive applications and the most appealing platform for deploying applications.

In my view, decentralized application providers can effectively market their services to users without forcing users to pay for the platform that deploys their applications, which is crucial for mainstream adoption. So I must ensure that this can be achieved in Polkadot.

Because it has not been achieved in other chains, we also need to make APIs and frameworks user-friendly. Moreover, we might create a few demo applications ourselves, partly because it is fun and partly to internally test the things we create to ensure they work well.

The Future of Web 3

Camila: You have been talking about Web3 for a long time, much earlier than many people first heard this term. You have been writing about Web3 since 2014. Can you talk about what Web3 means to you? Also, many people have recently been questioning and opposing NFTs, Web3, etc., believing that tokenized governance merely turns decentralized applications into the rule of the wealthy, and that these are just meaningless investments of large amounts of money and greed into the internet. What is your view on these voices?

Gavin: For me, Web3 is not about cryptocurrencies, nor is it about blockchain or token economics. To me, Web3 is about decentralization, openness, and transparency, ensuring from a technical standpoint that your expectations can be absolutely guaranteed.

I should not have to trust Facebook to send messages to my friends, nor should I have to trust a bank to send money to my dad. It shouldn't be that way. In the past, we didn't need to do this; my dad might have been right next to me, and I could just give him money directly. In that transfer process, I didn't need to trust any bank, or I could just turn around and talk to my friend without trusting Zuckerberg.

This method of embedding trust into everything is a form of laziness. It is simply easier for developers and service providers to give themselves god-like powers in structuring things. If you don't do it this way, it becomes more difficult. But that is what we should be doing in this era.

In this era, we have many capable people working hard to eliminate the concept of a "free society." We have no choice but to embed privacy protections and fundamental rights into the software and technology we use, so these rights cannot be circumvented, ensuring that no individual can break these rules. These elements are essential for the existence of a free world.

What I envision as Web3 is a technology stack that allows us to create web applications that most people use, which are less vulnerable to attacks than applications that do not adopt this technology stack.