Store round identity instead of counting emails
Synthetic studio allows two client-change rounds. Emails E1 and E2 both concern proof v1 within open round R1, so together count one round. E3 concerns v2 in R2 and counts another. Correcting the studio’s wrong spelling is tagged error correction and excluded by the supplied rule. This is an invented work allowance, not a contract or fee assertion.
| Constraint | Required record | Example decision rule |
|---|---|---|
| Round grouping | Proof revision + open round ID | E1/E2 → one R1 |
| Change type | Client request or studio error correction | Only client-change rounds consume allowance |
| Allowance state | Two permitted; R1/R2 used | Further round needs a new decision |
Test message count against round count
| Case | Input | Expected decision |
|---|---|---|
| E1 and E2 | Same v1/R1 | One round, not two |
| Spelling correction | Studio error tag | No client-round consumption |
- Identify proof and round
- Group messages within that cycle
- Separate error corrections
- Review new round beyond allowance
A new proof revision does not automatically open an extra permitted round; read the remaining allowance and owner decision.
After R1 and R2, is a proposed R3 inside the two-round allowance?
Show the worked answer
No. Two rounds are already used. Record R3 as a proposed additional round for decision, not an automatically accepted change.
Include repeated emails in one round and an excluded error correction in the demo, exposing a message-count shortcut.