What problems do the second layers bring?

Cardano is the first layer (L1) and as a blockchain network does not currently scale well. On the other hand, Hydra is the second layer (L2) that scales very well. But it's not that black and white. Although L2s solve the problem of low scalability of blockchains, they bring many questions to which we do not know the answers. Should we focus our efforts on better blockchain scalability or rather on L2 development?

Why are the second layers actually being built?

Bitcoin and Ethereum are the first and second-generation blockchain networks, and neither scales well. Cardano, as a representative of the third generation, is in a similar situation. Although the IOG team has plans to address this issue and Ouroboros Leios PoS (Input Endorsers) will bring a significant improvement, even Cardano is currently not capable of global adoption. There is no well-scalable blockchain in the top 10 that would simultaneously maintain a high degree of decentralization.

As soon as it became clear that L1s could not handle the transactions of newly arrived users efficiently, it was necessary to quickly come up with some simple solution. People began to complain that transactions were too expensive, slow, and unreliable compared to banking services. Increasing the scalability of L1s is a complex technological problem and it was obvious that the solution could take years.

The reasons why blockchain networks do not scale well are well known and there is no point in discussing them in this article. The problem began to be solved through different types of second layers and the community believes that it is a good solution. Although L2s solve one of the fundamental problems of L2, it brings with it many other problems that we should talk about more.

It can be said that people have started using L2s, so the problem is at least partially solved, or let's say postponed. With the adoption of L2s, however, the problems are more apparent.

What problems come with L2s?

Let's start by looking at the list of problems. Then we will break them down in more detail.

• More complex UX for users
• Splitting users and capital into multiple networks
• Taking over user activity and fees
• Delaying the L1 scalability solution

L2s add additional complexity to the already challenging use of blockchain from the point of view of ordinary users. The reality is that most users hold cryptocurrencies on centralized exchanges because they are afraid to use blockchain wallets. People perceive signing up to a centralized exchange as a safe and familiar process. On the contrary, the direct use of blockchain is a challenge for many. Just handling the passphrase/seed securely is quite a challenge and an activity that people want to avoid.

L2s are often networks with their own address space and different rules. Users are forced to understand not only the basic functioning of blockchain but also how a specific L2 network works. Users must learn how to transfer assets from one network to another and back, and what risks are involved. The room for error and loss of money is relatively large.

The more L2s exist, the more fragmented users and capital will be. There are many L1s and every major project has at least one dominant L2 solution. Ethereum has quite a few large L2s. Fragmentation is a problem not only within one ecosystem but also across the entire blockchain industry.

If Alice uses Lightning Network, Bob Polygon, and Carol Hydra, then these users have no chance to interact with each other. Everyone is in a different ecosystem and on a different L2 network. Users are separated including their capital.

It is also a problem from the point of view of merchants who want to accept cryptocurrency as one of the payment options. They will either support all existing L1s and L2s, which is almost impossible. Alternatively, users have to agree on a unified solution, which also seems unrealistic. If merchants would only adopt the Bitcoin ecosystem, i.e. the Lightning Network, they would lose users for whom the Cardano or Ethereum ecosystem is preferred.

Companies that want to mint NFTs will encounter a similar problem. They have to decide on which specific platform the minting will take place and thus basically only target a part of cryptocurrency users. If minting takes place on Cardano, users loyal to the Ethereum ecosystem cannot interact with these NFTs.

The largest stablecoins USDT and USDC exist in several L1s. Tokenized USD is fungible, so this is not as much of a problem as for NFTs which are non-fungible. If a company were to tokenize shares, what strategy should it choose? Choose a single L1 or several?

It is difficult to decide based on the number of users, as it will very likely change over time. One of the reasons why the adoption of blockchain technology by companies and businesses is slow may be the fragmentation of users in several competing ecosystems. Note that we are talking about fragmentation at the L1s level. Further fragmentation at the L2 level just makes everything more complex. There can be multiple competing L2s in one ecosystem.

Companies and businesses want to target the largest possible number of users and it is naive to expect that their service will be successful if it only targets users of one specific network. Likewise, it is naive to expect people to mass adopt a particular network for a particular service. It turns out that strong communities are created around individual projects that have existed for a long time, and it is impossible to ignore this phenomenon.

L2s take over L1s' user activity and the fees they pay for services.

This is another relatively large problem that threatens the economic sustainability of L1s. In the case of Bitcoin, there is more and more talk about the security budget problem. However, all L1s will face this problem sooner or later. Cardano and Ethereum have the advantage that the PoS consensus is more efficient, thus much less demanding to fund. However, every blockchain network must reward those participants who participate in the security and decentralization of the network or provide computing resources to it.

If users were to permanently switch to L2s without having to return to L1s, then L1s would essentially lose most of the fees. People would pay smaller fees for more user-friendly services on L2s.

L2s usually inherit only part of the security from L1s. L2s are often much more centralized and less secure (admin keys). What is important to keep are the L1s for decentralization. But who will pay for decentralization if users will mostly use L2s?

At the moment, L2 users are forced to pay on-chain transactions when moving assets from L1 to L2 and back again. However, if this were to cease to be the case, it would become an annoying problem for L1. L2s are dependent on L1s for their existence, but they are basically parasites on them economically.

L2s are unable to completely solve the low scalability of L1s. Teams may delay addressing L1 scalability because they believe that it is enough to focus on L2 development. I sometimes come across the opinion that L1s don't need to scale if there is at least one dominant L2 in the ecosystem. I do not agree with this opinion. If L1 doesn't scale well, it will directly affect the quality of the L2s as well. People have to be onboarded to L2, and if it's going to be expensive and slow, then it's going to hinder adoption. Even more so if L2 users are forced to regularly interact with L1 and pay fees.

A successful ecosystem that is to be ready for mass adoption must, first of all, have resolved scalability on L1 and have at least one L2. This solves most of the problems mentioned.

Why does L1 need to scale well?

I see L2s as an option for advanced users. Many users will want to rely solely on L1s. L1s must therefore be economically affordable and reliable enough for a much larger number of users than they are able to handle today.

It is difficult to predict the future, but if there are many L2s in each ecosystem, it is possible that the largest number of users and capital will remain in L1. However, it is also possible that a particular L2 becomes dominant and displaces competing solutions (including L1).

People will only use L1 if fees are low and settlement is fast. From the point of view of the security budget, it is essential that L1 provide similar utility as L2. I wouldn't even be afraid to say that L1 will directly compete with L2 for users, i.e. mainly for their fees. If fees on L2s are several orders of magnitude lower, L1s will only be able to compete with them with difficulty. People will simply be economically forced to learn to use L2 (and permanently leave L1).

It is technologically possible to link blockchains together and enable the transfer of assets. It is already possible to tokenize BTC and transfer it from the Bitcoin blockchain to other blockchains. This kind of blockchain interconnection will be quite common in the next few years. Individual blockchains must scale well if they are to be user-friendly. I see great potential in multi-chain wallets such as Lace. Users will have a single wallet and it will allow them to interact with many blockchains (and eventually L2) in a secure and very easy way.

I can imagine that users will have NFTs that were minted at different L1s in a single wallet. They can have ADA, BTC, ETH, and USDT next to each other and if they want to make a swap between these assets, it will be a matter of a few clicks directly in the wallet. Everything happens in the background and users may not even notice if it was on L1 or L2. The only thing they notice is the fee and the speed of the operation.

Conclusion

There are several well-argued opinions on whether to build L2s or focus more on blockchain scalability. Personally, I think that both are important and that all blockchains, including Bitcoin, should scale better if they are to be economically sustainable in the long term. Some experts even believe that L2s may be a development dead end, and if there was a scalable and at the same time well-decentralized blockchain, it would gradually become the dominant network. I don't think it's possible to achieve this in the next 10 years, but being able to do without L2 entirely would be the best possible solution from various points of view.

Don't get me wrong. I know that there are projects that already have functioning sharding and can be well decentralized at the same time. However, these projects must pass a stress test. People have to start using them and verify that the sharding implementation is of good quality. This has not happened yet.

Anyway, L2s are here to stay, as the level of decentralization and security offered by L1s is not always required. That's why I think it makes sense to focus on L1 scalability while simultaneously building L2s.