Eternal problems of Ethereum, Bitcoin and other sidechains like energy efficiency, transaction latency, computational expenses, fees, decentralisation, security and governance difficulties all occur because of the L1 structure, that is used in all the most popular blockchains.
Obviously, all the core chains are trying to fix the issues they have, using L2 chains on top of the base ones, but the efficiency and reliability of such solutions doesn’t come anywhere near to fiat economical system and that is the main problem of regular people rejecting crypto as their main financial asset. However, heterogeneous chains (what a name though) are said to change everything.
The so-called L0 chains like Polkadot, Cosmos and Avalanche are often mistaken for L2, that add up over Ethereum because of their compatibility with ETH as well as bridges on their EVM (Ethereum Virtual Machine) compatible chains. That gives L0 the ability to connect L1 and build infrastructure between core blockchains.
All the other chains, built on L0 aren’t just a list of smart-contracts, letting them to interact with other chains - each net gains its own virtual machine and uses the L0 to interact with other VMs.
Separating dApps as their own virtual machines allows for two major advantages - the network speed and the predictable fees. Having a divided net for your dApp means having your own blocks that aren’t used by any other dApps as it is with Ethereum as well as not relying on how busy the base network is now. It works the same with controlling the fees: there is no unified system with a pegged fee, calculated by the amount of operations done between all of the dApps - each net has it’s individual pricing, that is calculated specifically for it.
The three L0 networks as of right now are Cosmos, Polkadot and Avalanche. Let’s make a small comparison between those three.
Consensus
Chains, except Avalanche, use PBFT (Practical Byzantine Fault Tolerance) - a type of consensus, where a decision on the block is reached by all the systems nodes interacting with each other. PBFT has one major con and that is the QMC (Quadratic Messaging Complexity) - a mathematical problem, which occurs because of the system’s ability to only process one block at a time and this leads to lower operation speeds.
However, Polkadot only uses a modified version of PBFT for finalising the blocks, whereas validation is done through a modified Ouroboros protocol (An incredibly complicated system, that deserves its own article). Consequently, Polkadot minimises the effect of QMC and has much higher operation speeds than Cosmos because of it’s hybrid consensus.
Avalanche, on the other hand, is pretty different from the first two networks because the messaging used between nodes isn’t linear. DAG (Directed Acyclic Graph) is transitive and allows for more multiple messages between nodes at a time, completely removing the QMC.
Transaction latency
Obviously, Avalanche is the fastest here, achieving operations finality in a second. Followed by Cosmos with it’s not so faultless, but still fast PBFT, achieving finality in 6-7 seconds. And finally, the Polkadot, despite its lower QMC and higher operation speed still has to divide each operation in two, making it the slowest with 12-60 seconds finality.
Fees
Here Avalanche is a winner once again, because all three chains have customisable fees for every net, but only Avalanche burns all the transaction fees, so the price $AVAX increases due to lowering availability.
Decentralization
And again Avalanche with its multiple messages at a time and consequently limitless decentralisation.
Conclusion
We need L0 and that’s it. If L2 becomes the ultimate dApp creator, L0 will allow developers to create full nets and communication between them as well as with the already existing ones. Choosing from the three heterogeneous nets available right we would pick Avalanche, but don’t forget to check information on the other two yourself before making any decisions.
