|
The crypto space is made up of a lot of blockchain technologies from
Ethereum Blockchain, Solana, Polygon, Avalanche, NEAR etc.
On the public and private fronts, these companies are innovating at breakneck
speed to solve the blockchain trilemma which includes: solving human problems
without compromising on decentralization, scalability and security.
Most implementations of blockchain are sloppy in one or more of those three
areas. Ethereum has been in crypto space news the most for being unable to scale
as quickly as the demand wants it to. Also for the huge gas fees that result
when so many users battle for whose transaction should be processed first. Rival
cryptocurrencies projects have risen to this challenge, attracting developers
and growing rapidly, including NEAR and Avalanche, each with their different
approaches to scaling.
NEAR Blockchain Vs Avalanche Network
NEAR using sharding, Avalanche using subnets and a probabilistic technique.
Let’s dig into what those worlds look like, how adoption is going, and what the
future holds for NEAR, compared to Avalanche.
From inception, NEAR set out to onboard billions of people into the crypto
space, says co-Founder Illia Polosukhin in a recent Twitter space discussion.
This was Ethereum’s goal as well. To bring every device and machine you use onto
the blockchain in order to build a new internet that is free of censorship.
Ethereum faces a bottleneck: the users have begun to rush in but the network is
too slow and too costly to serve them all. NEAR plans to carry billions of
people and solve this issue using sharding.
What is sharding?
Sharding is simply breaking up the ledger spreadsheet that contains transactions
and instructions, and placing them on different servers, with a routing system
on the application layer that directs you to your data on whichever server it
sits on. Having different portions of the record on different servers reduces
latency, increases the speed of transactions because the workload is now spread
into partitions.
For instance, the Ethereum ledger is almost 1 Terabyte in size now. Compare that
with the storage space of most computers and you see why it gets more and more
restricting as to who can be a node to process transactions on the Ethereum
blockchain.
In a blockchain, each node that processes transactions must have critical
information from this ledger such as transaction records and account balances.
That is part of what it means that blockchains are distributed ledger systems.
This has usually meant having the entire network on each node. Sharding asks if
this is necessary, and rather has nodes carrying the partition of data they need
to access.
Sharding is about the most complex of all the scaling solutions, and the
Ethereum network itself intends to implement it as part of their roadmap after
other scaling solutions like layer 2s have been set up at a time when it would
be least disruptive to the system.
NEAR on the other hand is infusing it early on. Sharded systems are prone to
hacks. A shard can be attacked, one shard could attack another, information
could be lost, which is trouble for a blockchain, a system where records should
be unerasable. The process is mostly still theoretical and hasn’t been
stress-tested yet.
NEAR, however, saw tremendous growth in 2021, announcing an $800 million grant
for developers willing to build with them, and almost tripling their number of
full-time developers. This is a real indicator of where the traction could be in
future because it is the solution apps built by developers that solve people’s
problems and end up onboarding them into the crypto space, hence, millions and
billions of users. According to co-Founder Illia Polosukhin, their drive has
been fuelled by the question, “how do we create a developer experience that a
regular engineer can pick up and build an app on day one?” which is why Rust, a
more friendly language than Ethereum’s Solidity, was used in programming NEAR,
and also why NEAR chose to support Typescript.
Avalanche
Emin Gün SIrer agrees that “chains that will do well will be the ones that
can absorb huge user base growth.” However, they take a different approach:
subnets.
A subnet is like a town within a state. Ordinarily, if there were no roads
within a town, and the villages within it are separated by a river or by an
expanse of thick vegetation, people in both villages could live for years
without acknowledging each other. They might not even know that the other
village exists. If they had a major road linking them to the state capital, they
would get all their supplies and do their trade with the state capital. However,
if the state came and built a bridge over the river, cleared the vegetation and
built a road linking the villages, they can trade with each other, and suddenly
they don’t need to go to the state capital as much anymore. Creating a subnet is
building that road, and creating a new separate instance of the blockchain that
communicates with the old blockchain. The subnetwork only communicates with the
old blockchain once in a while, and the majority of transactions happen within
the subnet and happen much faster since the distance for data to travel is much
smaller. You can already go up on the Avalanche network and set up a virtual
machine or a subnetwork, including ones compatible with Ethereum.
Avalanche couples this application of subnetworks with their probabilistic
technique. The probabilistic technique is a fancy name for how their system
makes snap judgments by observing what the majority of validators are ‘saying’
and verifying transactions based on that instead of checking all validators one
by one before adding a transaction to the ledger. Amped up so, Avalanche runs
between 1,000 and 4,500 transactions per second, compared with Ethereum’s 15–30
tps.
The theory is that more and more subnets can be created endlessly, and so the
potential for scaling and catering to billions of users is limitless, but this
is all still science and theory yet to be tested. The hammer of engineering has
not been applied on Avalanche’s network as much as on Ethereum, for instance,
and Gün himself says users should definitely expect adjustments as time goes on.
|
