Some fun ETH math, illustrating the challenges ETH will be faced with pre L2:
As shown in the thread below, block size is effectively full — and there are only 4 apps with greater than 500 DAUs
As shown in the thread below, block size is effectively full — and there are only 4 apps with greater than 500 DAUs
What do I mean? Avg gas / transaction is ~57k right now. This is a weighted avg between the 21k gas for a regular txn and higher gas for more complex smart contracts
But as more dApps launch, the higher gas usage contracts will be weighted more, raising overall gas used per txn
But as more dApps launch, the higher gas usage contracts will be weighted more, raising overall gas used per txn
As gas used rises pre L2, ETH has a *very* hard problem:
(1) Continue to keep block size capped, raising fees, or (2) increase block size raising the burden further on a full node.
@StopAndDecrypt's recent articles details this well
(1) Continue to keep block size capped, raising fees, or (2) increase block size raising the burden further on a full node.
@StopAndDecrypt's recent articles details this well
But ETH *needs* low fees for 2 reasons:
(1) current demand is very elastic b/c there aren’t any mission critical applications
(2) competitors, like EOS, are emerging with dramatically lower, even 0, fees
(1) current demand is very elastic b/c there aren’t any mission critical applications
(2) competitors, like EOS, are emerging with dramatically lower, even 0, fees
Yes, high fees can be indicative of a working system. But block space demand is currently very *elastic*. Most ETH use cases are spurious
B/c of this, fees peaking is going to be a leading indicator of price reversing as people move to other chains or stop usage altogether
B/c of this, fees peaking is going to be a leading indicator of price reversing as people move to other chains or stop usage altogether
Check out the December peak in both txn fees as crypto kitties took over the ETH network. If these dApps actually launch, we're in for part 2 soon.
Let’s look at a hypothetical crypto kitties txn. Back at december recommended gas price of 25 gwei, each birth will cost $3.75-$5.
How long will users be willing to pay these fees on Eth when other alternatives exist?
How long will users be willing to pay these fees on Eth when other alternatives exist?
What also matters is who is paying these fees: IMO, most dApps know that their fees are probably going to be too high for them to have a value prop.
In the case of crypto kitties, they actually decided to *cover* fees for users.dApps are subsidizing users with all their capital
In the case of crypto kitties, they actually decided to *cover* fees for users.dApps are subsidizing users with all their capital
So what will ETH do? They need to increase the gas limit. The gas limit acts as regulation on usage to prevent tragedy of the block space commons which hurts full nodes as the block size exponentially scales
They have to raise this cap to ensure the narrative doesnt break
They have to raise this cap to ensure the narrative doesnt break
Apps also have a concept of DAUs — daily active users. ETH dApps need to enable *daily* usage to ensure virality.
Some back of the envelope math: if 10 apps launch with 50k DAUs, ETH will need to raise gas limit 4.3x to acct for them. That's just 10 modestly successful apps...
Some back of the envelope math: if 10 apps launch with 50k DAUs, ETH will need to raise gas limit 4.3x to acct for them. That's just 10 modestly successful apps...
To save the dApp plaform narrative, ETH will raise blocksize, further centralizing the validating nodes.
The problem is as people realize that ETH isn't architecturally decentralized any longer, the decentralized narrative and their value prop will also start to die...
The problem is as people realize that ETH isn't architecturally decentralized any longer, the decentralized narrative and their value prop will also start to die...
EOS recognized this is where the space is going: many users don't necessarily care about architectural decentralization and "sound money" so why not embrace it, lower fees and pay through inflation.
I dont think this can last, but it can definitely take ETH market share
I dont think this can last, but it can definitely take ETH market share
The last challenge: what happens if miners don’t play ball w. gas increase?
2 scenarios below: The 10mm gas limit has a lower price / gwei, but also reduces miner revenue despite the qty increase. If miners cap gas limit at 8mm, they could hypothetically increase rev...
2 scenarios below: The 10mm gas limit has a lower price / gwei, but also reduces miner revenue despite the qty increase. If miners cap gas limit at 8mm, they could hypothetically increase rev...
Talk about a rock and a hard place. I see a few scenarios
1. dApps launch, fees rise, block size is raised further and nodes continue to centralize
2. dApps launch, fees rise, block size isn't raised (or not fast enough) and apps go elsewhere
3. dApps never launch on ETH
1. dApps launch, fees rise, block size is raised further and nodes continue to centralize
2. dApps launch, fees rise, block size isn't raised (or not fast enough) and apps go elsewhere
3. dApps never launch on ETH
The market woefully misundersands the value prop and "decentralization" of these systems. If architectural and political decentralization both don't exist, what are we left with?
IMO, dApps are going to continue to exploit the lack of user sophistication and ultiamtely just reengineer AWS on a blockchain as they get more and more centralized to lower fees and increase speed.
If EOS can have 21 validators, why not have 2? Why even have a blockchain?!
If EOS can have 21 validators, why not have 2? Why even have a blockchain?!
The value proposition of smart contract platforms is disintegrating
