按集体合同生成薪酬工作簿

A is substantially closer: effective-dated rates and percentages feed per-person weekly category totals, with implemented checks for paired-rehearsal exceptions, woodwind designations and engagement-wide sound-check limits. B incorrectly excludes the French-horn premium when trumpet is present, overlooking Article 4.2.c. Its synth-substitute formula also offsets rehearsal and sound-check earnings against the show-specific entitlement, eliminating the sample’s top-up. B checks sound-check frequency only within one week and does not implement the paired-rehearsal waiver. A’s functionality remains uncertain, however: its rendered calculation and check cells repeatedly show Err:507 despite populated cached totals.

B implements substantially better CBA logic: the qualifying paired-rehearsal waiver, engagement-wide sound-check frequency checks, and French horn’s 15% first-local premium even when trumpet is present. A incorrectly forbids that horn premium, always applies a three-hour minimum to each rehearsal, and lets rehearsal/sound-check earnings erase the synthesizer substitute’s show-specific adjustment. B’s effective-dated rate tables and week-filtered attendance ledger also better support changing rates and longer runs. However, B’s rendered calculation/check columns contain pervasive Err:507 values despite numeric, error-free cached results, so its live Excel usability cannot be confirmed.

A applies the CBA more accurately and completely, although it is harder to use and its render shows errors. B is simpler and renders cleanly, but it misapplies several terms, so its pay figures and conflict flags are wrong in common cases. A's strengths: - Rates are effective-dated and include every base rate and percentage, with overlap and coverage checks. - First-local premiums use family, rank and seat. Twenty percent applies only to concertmaster, first trumpet, or first horn when no trumpet is called for. - The first-local French horn correctly gets 15% under 4.2.c even though a trumpet is present. - Woodwind premiums follow employer designation, and the quartet premium requires confirmation. - The synth-sub cap offers two selectable readings, and audits are tracked at two per book across the engagement. - The weekly guarantee top-up offsets only performance base, with premiums and doubling on top. - Rehearsal checks cover 9am and 10am starts, the 3-hour minimum with the 7.4 lunch-pair waiver, 50/10 breaks and 5-hour meals. Sound checks are checked against their anchor call and engagement-length frequency. - Weekly Payroll gives each person Performance, Guarantee, Synth uplift, Rehearsal, Sound check, Audit, Premiums, Doubling, Vacation and Gross. The START review queue summarises flags. - The cached sample total of $30,825.40 reconciles with its documented methods. A's weaknesses: - It is heavy for a contractor: an ID-based attendance ledger plus approval toggles. - Every [@...] structured-reference column shows Err:507 in the render (Checks columns, Level, attendance lookups), so conflict messages do not display there. - It applies the electronic premium to audit pay, which is debatable. B's strengths: - Plain formulas and named ranges. - A clear P/A attendance grid and a CBA Checks sheet marking OK, REVIEW or VIOLATION. - Per-person category totals. - The render works, and the 9:30am rehearsal is visibly flagged. B's errors: - **French horn:** any horn premium is flagged as a VIOLATION whenever a trumpet is present, and the sample horn gets 0%. This contradicts 4.2.c. - **Woodwinds:** the "more than one premium player" check would wrongly flag employer-designated multiple woodwind premiums. - **Twenty percent:** nothing restricts the 20% rate to trumpet, violin or horn. - **Guarantee:** the top-up is measured against a subtotal that includes rehearsals, premiums and doubling, so premiums are absorbed in short weeks. - **Synth sub:** the 150% minimum is netted against rehearsal and sound-check pay, so the sample sub gets no uplift. - **Invented rule:** rehearsals without intermissions, or ending after 6:30pm, are repriced at the per-service rate. - **Missing rules:** there is no 7.4 waiver, and sound-check frequency is checked only within one week.

A is the more reliably usable deliverable, although B has the stronger contract logic. B's advantage only counts if its formulas work, and the rendered evidence leaves that in doubt. **Submission A works as rendered.** - Its formulas recalculate in the rendered pages: the Roster shows doubling/premium/add-on % (e.g., Woodwind 45%+15%), and the Schedule shows $278.35 rehearsal, $76.59 sound check and $251.06 performances. - The 9:30am rehearsal row is highlighted pink with an Art. 7.3 exigent-circumstances flag. - The Payroll sheet gives each musician performance, rehearsal, sound check, audit, premiums, doubling, sub minimum, guarantee top-up, 5.5% vacation and total. - A CBA Checks sheet rolls up 22 rules as OK/REVIEW/VIOLATION. - Rates sit in a year-keyed table with editable percentage parameters. - The P/A attendance grid is quick for a contractor's weekly entry. **Submission A has real logic errors.** - It treats the horn premium as not payable when a trumpet is in the show. That contradicts Art. 4.2.c, which grants French horn 15% regardless of seating, so the sample horn is underpaid and a correct entry would be flagged as a violation. - Nothing stops a 20% first-chair designation on instruments other than trumpet, violin or horn. - The synth-sub minimum is netted against rehearsal and sound-check pay. **Submission B is deeper and more accurate on paper.** - It pays French horn 15% even with a trumpet, validates seats (concertmaster, first trumpet, first horn) and applies the Art. 7.4 lunch-pair waiver. - It checks meal-break placement and links each sound check to its anchor call, with frequency checked across the whole engagement. - Rates are effective-dated and include the percentages; audits are tracked by book; the synth-cap readings are documented, along with a verbatim copy of the CBA. - Cached values compute a weekly gross of $30,825.40. **Submission B's rendered pages show widespread failures.** - Err:507 appears in every Checks column, Premium Chairs Level, and Attendance Date/Type/Musician. - Its formulas store raw `[@[...]]` structured references, which the renderer could not parse and which may not open cleanly in Excel. - It is also heavy to operate: a separate premium-chairs table, service IDs, sound-check anchors, an 8,000-row one-row-per-musician-per-service ledger, and approval gating that leaves the sample "NOT READY". On balance, A's functioning, practical model outweighs B's richer but less verifiable and more burdensome design.