Claude Fable 5.1 — 長任務的深度推理

Claude Fable 5.1 實際輸出

一個提示詞，以低思考強度送給 Claude Fable 5.1。未編輯。

Explain the difference between optimistic and pessimistic locking in a database, in two short paragraphs, and say when to use each.

**Pessimistic locking** assumes conflicts are likely, so it prevents them up front: when a transaction reads a row it intends to modify, it acquires a lock (e.g., `SELECT ... FOR UPDATE`) that blocks other transactions from writing (and sometimes reading) that row until the first transaction commits or rolls back. This guarantees no one else can change the data underneath you, but it holds resources for the duration of the transaction, which can reduce concurrency and risks deadlocks if locks are taken in inconsistent orders. **Optimistic locking** assumes conflicts are rare, so it doesn't lock at all. Instead, each row carries a version number (or timestamp), which the transaction reads along with the data. When writing back, it issues an update conditioned on that version still matching (`UPDATE ... WHERE id = ? AND version = ?`); if zero rows are affected, someone else changed the data first, and the transaction must retry or report the conflict. This maximizes throughput and avoids held locks, at the cost of occasional wasted work and the need for retry logic. **When to use each:** Use pessimistic locking when contention is high, conflicts are expensive to redo, or correctness demands that a value not change mid-operation (e.g., decrementing inventory or account balances in a busy system, long-running batch jobs touching hot rows). Use optimistic locking when contention is low and reads dominate writes, transactions are short, or the workload spans user think-time or disconnected clients (e.g., a web form where a user edits a record and submits it minutes later), where holding a database lock would be impractical.

於 2026-09-27 產生。每次執行答案不同。