Two opposing dangers for the eth ecosystem:

1. Opinionated and idealistic developooooors can get stuck solving problems that aren't a real issue. Leads to excess complexity: worse software, shipped late. In bad case, can rise to the level of concern trolling blocking protocol progress.

Real World Ethereum 🌎 
building /daimo ✳️ 

2. Opportunistic growth hackooors can take cheat by shipping theoretically-unsound systems fast. Leads to risk of collapse, damaging the whole space. In the worst case, ponzis.

To balance 1 against 2, we need a crisply articulated threat model.

A curious but nontechnical reader should have a clear picture of WHY:

- Mainnet is trustworthy but expensive
- A stage2 rollup is trustworthy and cheap
- Something like Tron is untrustworthy and cheap
- 1 USDC is highly likely to be redeemable for $1, while USDD... good luck son

If we are clear about exactly which limits we're trying to respect, and which risks we are (and are not) willing to accept,

we have a rational basis for resolving disagreement

and optimizing to ship as fast as possible within those constraints

It sounds dramatic but the base case is failure. The future, by default, belongs to centralized actors because they have a fundamentally easier problem!

Venmo is free and Tron is 0.0001c per transaction because they simply do not care about any of the above.

And if a curious, technical reader should be able dig a bit deeper:
- WHY can't L1 go faster than 12s per block, but L2 can do ~1 or 2s?
- WHY is 4844 shipping with so little blob space?
- WHY is dishonest-majority-proof DA critical? (and why then is it fine that those precious blobs expire after a few weeks?)