planner/optimizer — `cost_model.rs` and `solution.rs` use math-notation variable names carried over from the design doc: `rho_g`, `n`, `g`, `a`, `f_a`, and the term `N(s,g)` in comments. These read fine next to the design doc's formulas but are opaque on their own (e.g. `what is n? what is N(s, g)?` came up directly in review).
Before renaming, we need agreement on what each symbol means in plain terms (arrival rate, subpopulation count, candidate config, AQE, query frequency, etc. — several of these already have partial doc comments, e.g. `n` is noted as "N(s,g) = 1 if subpopulation_aware else N_g" in `cost_model.rs`). Once confirmed, do a mechanical rename pass to descriptive names.
Found during code review of PR #407.
planner/optimizer — `cost_model.rs` and `solution.rs` use math-notation variable names carried over from the design doc: `rho_g`, `n`, `g`, `a`, `f_a`, and the term `N(s,g)` in comments. These read fine next to the design doc's formulas but are opaque on their own (e.g. `what is n? what is N(s, g)?` came up directly in review).
Before renaming, we need agreement on what each symbol means in plain terms (arrival rate, subpopulation count, candidate config, AQE, query frequency, etc. — several of these already have partial doc comments, e.g. `n` is noted as "N(s,g) = 1 if subpopulation_aware else N_g" in `cost_model.rs`). Once confirmed, do a mechanical rename pass to descriptive names.
Found during code review of PR #407.