wnab / Wedge Needs A Budget

What it is

wnab (“Wedge Needs A Budget”) was a fork of Actual Budget, an open-source, local-first budgeting engine. The fork was rebuilt into a YNAB-style envelope budgeting app with a coaching layer on top. The plan was to reuse Actual’s loot-core engine, which runs SQLite in the browser through WebAssembly, so the whole app stays static and needs no server. The part to write was what Actual does not have: the onboarding, nudges, and behavioral design that teach the mindset of zero-based budgeting.

What worked

Reusing Actual’s engine was the right decision. Forking @actual-app/web meant the hardest parts already existed and ran offline in the browser: the double-entry ledger, import matching, and multi-account reconciliation. I kept the engine code separate from the view layer, so I could track upstream for the budgeting mechanics while building a different user experience.

What broke

Zero-based envelope budgeting is mechanically sound. Using it every day was exhausting. Every dollar had to be assigned, every category reconciled, and every month re-budgeted. That upkeep took more effort than the insight it gave. People stop opening a budgeting tool that feels like micromanaging, and an unused budgeting tool is worse than having none.

Lessons

• Match the amount of upkeep to the benefit. The envelope method requires constant upkeep to give a kind of control I didn’t actually need day to day. Its successor, WNAC, replaced envelopes with a single “flex number” and a debt-payoff dial. WNAC gives the same insight with much less upkeep.
• Reuse the engine, and design the experience yourself. Forking Actual’s engine was the right decision and saved months. The mistake was adopting Actual’s budgeting philosophy along with its code.
• The reason you start using a tool has to keep you using it every day. Ownership was my reason to build it. Whether I kept using it depended on the daily upkeep, and the daily upkeep is why I stopped.

Status

Retired. The project is archived intact: the data round-trips and the engine still runs. Its successor is WNAC, which gives clean data and AI insight without the upkeep of envelope budgeting.