docs(dex): census remaining Swap gas vs wallet auto-fee (#1222) #1226

Merged

PlasticDigits merged 1 commit from issue/1222 into main

2026-09-12 02:32:10 +00:00

PlasticDigits commented

2026-09-11 07:34:49 +00:00

Owner

Summary

Research census only: ADR 0004 inventories retail execute families vs Swap/Trade Network fee hint vs wallet confirm vs dated gas_used.
Decision: Stay. No envelope retune, no wallet-adapter patch, no LCD-sim-as-broadcast, no hybrid-off, no mainnet SEC-E08. No implement issue spawned.
Re-fetched columbus-5 AB8BE4F7… on 2026-09-11: gas_wanted 6,785,500 / gas_used 5,026,176 / fee denom uluna. Station WC ~4k LUNC remains G-AUTO-8 (wallet residual).

Fixes #1222

Acceptance

AC1. One ADR covering hint vs broadcast vs wallet vs measured usage.
AC2. Dated measurements or explicit unmeasured rows; #679 hash re-checked against current code + public LCD.
AC3. G0–G7 scored; product decision Stay.
AC4. #123, #546/#618, #1209 named out of this census.
AC5. Docs-only: no production terraGas.ts / wallet-adapter / gas-constant diff.
AC6. No follow-up implement issue (decision is Stay).

Test plan

make verify-issue-1222 (ADR headings, G-CENSUS-1–G-CENSUS-8, production-file git-diff empty, mixed-hybrid + Network fee + inventory Vitest, retest)
Reviewer: open the cited columbus-5 hash and confirm gas_wanted / gas_used / uluna
Optional later copy (not this PR): /trade Market one-liner mirroring swap-wallet-fee-note

## Summary - Research census only: [ADR 0004](docs/adr/0004-terraclassic-retail-gas-census.md) inventories retail execute families vs Swap/Trade **Network fee** hint vs wallet confirm vs dated `gas_used`. - Decision: **Stay**. No envelope retune, no wallet-adapter patch, no LCD-sim-as-broadcast, no hybrid-off, no mainnet SEC-E08. **No implement issue spawned.** - Re-fetched columbus-5 `AB8BE4F7…` on 2026-09-11: `gas_wanted` 6,785,500 / `gas_used` 5,026,176 / fee denom `uluna`. Station WC ~4k LUNC remains **G-AUTO-8** (wallet residual). Fixes #1222 ## Acceptance - [x] AC1. One ADR covering hint vs broadcast vs wallet vs measured usage. - [x] AC2. Dated measurements or explicit **unmeasured** rows; #679 hash re-checked against current code + public LCD. - [x] AC3. G0–G7 scored; product decision **Stay**. - [x] AC4. #123, #546/#618, #1209 named out of this census. - [x] AC5. Docs-only: no production `terraGas.ts` / wallet-adapter / gas-constant diff. - [x] AC6. No follow-up implement issue (decision is Stay). ## Test plan - [x] `make verify-issue-1222` (ADR headings, G-CENSUS-1–G-CENSUS-8, production-file git-diff empty, mixed-hybrid + Network fee + inventory Vitest, retest) - [ ] Reviewer: open the cited columbus-5 hash and confirm `gas_wanted` / `gas_used` / `uluna` - [ ] Optional later copy (not this PR): `/trade` Market one-liner mirroring `swap-wallet-fee-note`

Rows
Columns