---
title: "Comment by Damiano Bianda on \"Build a Patch Ledger That Can Survive the Next Deployment\" | Hashnode"
description: "The ledger design ages well because it separates the artifact from the host - and the dotnet out-of-band release this weekend shows why that matters. A self-con"
url: "https://hashnode.com/posts/build-a-patch-ledger-that-can-survive-the-next-deployment/6ad946541877198bcb76c723/comment/6adf27dd3dc566e46d074c42"
---

[Discussion](https://hashnode.com/posts/build-a-patch-ledger-that-can-survive-the-next-deployment/6ad946541877198bcb76c723)

Comment [Build a Patch Ledger That Can Survive the Next Deployment](https://hashnode.com/posts/build-a-patch-ledger-that-can-survive-the-next-deployment/6ad946541877198bcb76c723)

[Damiano Bianda](https://hashnode.com/@damianobianda)

[Oct 26](https://hashnode.com/posts/build-a-patch-ledger-that-can-survive-the-next-deployment/6ad946541877198bcb76c723/comment/6adf27dd3dc566e46d074c42)

The ledger design ages well because it separates the artifact from the host - and the dotnet out-of-band release this weekend shows why that matters. A self-contained app or a container image carries its own runtime, so "host patched" and "application covered" stay different rows even after a successful update. The state that describes missing evidence ("we have not checked what this artifact bundles") is exactly what dotnet --list-runtimes can't give you for those deployments. Curious how you'd represent container rebuilds in the ledger: one row per image digest, or per application with the digest as evidence?
