Vote on Treasury Withdrawal - IO - Enhancing Plutus

I am voting YES on the governance action with hash 73e171a4c0730b4b59ecae271ab89f12a9d56360b02920e1f95107dbdc1d6762#6.

The metadata for my vote is located here, and this blog serves as a more human-readable form of that justification.

Justification

I came very close to voting NO on this proposal, as I believe many of the things outlined in the roadmap to be a misallocation of resources.

Continuing to evolve Plutus, the underlying runtime and execution model on which all dApps rely, is essential in the long term. It has come a long way since 2021, but still has many glaring gaps, and this proposal does include several of those.

However, it also continues to invest resources in things that I think are superfluous or distractions. Plinth is (through no fault of the team) not a serious contender for the construction of most dApps that have yet to be built. Those serious about performance will reach for lower level primitives like Plutarch or hand-rolled UPLC. And the vast majority will be written in Aiken, Pebble, Opshin, or others.

Additionally, another team has formalized the CEK machine and ledger rules in Lean. Agda is the less-approachable cousin, and to me represents redundant and duplicated work.

Overall, I want to see the new builtins described in the proposal, and they alone (barely) justify the cost. My hope is they will unlock a new era of efficiency, much as reference inputs and inline datums did in the past. I want to see the underlying execution framework evolve and reduce fees for users. And it’s clear to me, from looking at the advocacy on social media, this is where the teams head is really at.

My hope is that, if this proposal passes (as seems likely right now), the team has liberty to evaluate and restructure their roadmap. Stay flexible to the needs of the developers, rather than locking in a full years roadmap up front. If these work items finish faster than expected, devote those resources to the things Phil called for here, instead.