---
title: "Comment by Andrew on \"MVCC Explained: Build Postgres-Style Snapshots in 100 Lines of TypeScript\n\" | Hashnode"
description: "Your insert always pushes a new version and never stamps xmax on the prior live version. A second insert of the same key therefore succeeds instead of waiting o"
url: "https://hashnode.com/posts/mvcc-explained-build-postgres-style-snapshots-in-100-lines-of-typescript/6a8bf1f20127a09e517d0f94/comment/6a8d1459b9707ea7b8e0678f"
---

[Discussion](https://hashnode.com/posts/mvcc-explained-build-postgres-style-snapshots-in-100-lines-of-typescript/6a8bf1f20127a09e517d0f94)

Comment [MVCC Explained: Build Postgres-Style Snapshots in 100 Lines of TypeScript](https://hashnode.com/posts/mvcc-explained-build-postgres-style-snapshots-in-100-lines-of-typescript/6a8bf1f20127a09e517d0f94)

[Andrew](https://hashnode.com/@rizzdev)

[Aug 25](https://hashnode.com/posts/mvcc-explained-build-postgres-style-snapshots-in-100-lines-of-typescript/6a8bf1f20127a09e517d0f94/comment/6a8d1459b9707ea7b8e0678f)

Your insert always pushes a new version and never stamps xmax on the prior live version. A second insert of the same key therefore succeeds instead of waiting or raising a unique violation. A later delete then closes only one of those versions, and the other value remains live under the same key.