Thomas pfp
Thomas
@thomasoncrypto
I’d like to rift off this by @bonecondor to a general critique of L[0-9]+ solutions in Ethereum. https://x.com/bonecondor/status/1815759515058372969?s=4 1/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
Most L2/3/4s have no purpose except to replace the Ethereum Mainnet with something faster and less secure. The only natural end-game of an additional EVM that trying to bring all and any applications to itself. https://x.com/0xBalloonLover/status/1795466361143005425 2/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
What an L-Natural-Number does is three things:

a) Removes safety. Or Consensus Layers Running with Scissors This is often good. Ethereum Mainnet is not the appropriate trade-off for everyone. https://x.com/jph98/status/1064953554304212998 3/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
b) Demarcates a space with separate rules and norms. Like stepping into a Medieval market square or a stock exchange. https://x.com/0xDepressionn/status/1747246483949175006 4/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
c) Provides governance by having different (presumably lower) thresholds for forking. (It doesn’t have to, but, in practice, they usually do.) https://x.com/kryptolabPL/status/1825112327718359312 5/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
In reality, most L2 just copy-pastes the OP stack. (Which is naff, as @Optimism are fine with you just spinning up the same code and actually crediting them.) https://x.com/PopPunkOnChain/status/1752780488127762716 https://x.com/Optimism/status/1820950085359857927 6/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
So why have more than one L2 (or L3 or L4) in the long-term? 7/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
They do have a business purpose. Infrastructure projects have been considered more SEC-proof since EOS’s wrist slap in 2019. (Plus $4bn raise. Minus $24m fine.) So you can attach a DEX, a DeFi protocol, a prediction markets, etc, etc to a fresh EVM as a cloaking device. 8/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
But there isn’t really a reason to run these things in their own EVMs. I’d argue that there are better ways of demarcating space and forking assets than creating whole new EVMs too. (Separate thread later on 🙂) 9/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
Speeding up transactions by “hot-spotting” VMs is valid. For example, there is the NASDAQ example, but in a composable system it will be hard to know where these hotspots will occur in advance. 10/17 https://x.com/thomasoncrypto/status/1818946558097789287
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
Which is not so say I’m down on rollups (Optimistic or zK), I think they’re great! I just see them as being go-faster solutions like State Channels. 11/17 https://x.com/statechannels/status/1108072528403156994
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
But do have an L2 per app? City? Currency? Social group?? Take a taxi app example, do we need shared state between *all* the riders? If the chain is for credentials and final payment is that enough to justify a rollup? 12/17 https://x.com/molly0xFFF/status/1629715599038328832
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
Also @functi0nZer0 has a point 🙂 https://x.com/functi0nZer0/status/1772774643671212221 13/17
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
I can see a future where client navigate between rollups using ENS metadata, which solves a lot of issues, especially since shared infra will make rollups cheap. (Not sure how this will work with starless validators.) 14/17 https://x.com/VitalikButerin/status/1742540651797823948
0 reply
0 recast
0 reaction
Thomas pfp
Thomas
@thomasoncrypto
0xPolygon AggLayer feels like a step in the right direction. 15/17 https://x.com/0xRainyReece/status/1823715499890630784
0 reply
0 recast
0 reaction