Thread by @apolynya

The Polygon PoS & Solana situations are more interesting data points in the transaction quality trilemma discussion.

By optimizing for low fees & censorship resistance, they suffer from spam and very low-quality transactions.

polynya.medium.com/transaction-qu…

In addition to network instability, they lead to state bloat and socializing transactions that may be parasitic to the ecosystems.

I had often conjectured the min transaction fee should be in the $0.01 to $0.10 range. But Polygon PoS's gas fees are now ~500 gwei...

...yet the low quality transactions have kept coming even though AMM swaps have now actually exceeded $0.10. Of course, this will eventually stop, but I think this is important evidence that $0.01 to $0.10 is actually still too low long term. Perhaps $0.50 is a better target?

Fortunately, since my post above, we now have a clearer solution: recursive rollups, or L3s.

L3s can have ultra-low or even zero transaction fees, but use web2-style DDoS protection, filtering out obviously low-quality spam transactions.

(1/3)

medium.com/starkware/frac…

Here's why it's an acceptable tradeoff:

Users have an immediate and very cheap exit available to L2 if they are false positives. Also, it might be possible to submit your transactions directly to L2, you just need to pay a higher fee to get included into L3.

(2/3)

So, by modularizing, you can have censorship resistance alongside zero fees and spam mitigation!

Immutable X already kinda does this at the L2 level, but the alternate CR avenues are too expensive & slow.

L1s may try QoS, but with no CR avenues it's unacceptable IMO.

(3/3)

I'm well aware the popular notion on CT is a race to the bottom with transaction fees, but this is very problematic in various ways. I hope there'll be more nuanced discussions about what the right equilibrium with transaction quality, security, value accrual looks like long term