Lesson 2: Two Different Worlds

Git is content-addressed and repo-authoritative; Salesforce is semantically-addressed and org-authoritative: this lesson builds the precise model that makes Lesson 3 predictable.

Start here

Each topic below covers a building block of this lesson. We recommend following them in sequence as each builds on the last.

Git vs. Salesforce

Contrasts Git's content-addressed, file-level model with Salesforce's semantically-addressed, live-org model: the library catalog vs. the living garden.

Who Each System Was Designed For

Git assumes developer users; Salesforce was designed to eliminate developer mediation: forcing both creates expensive, unscalable translation overhead.

The Metadata API in Detail

Why the Metadata API's XML looks Git-compatible but isn't: partial retrieval, inconsistent serialization, the gap between snapshot and org state.

Content-Addressed vs. Semantically-Addressed

The unifying frame: Git objects are what they contain; Salesforce components are what they mean in the org's data model.

The Topology Insight

Branches descend from main; sandboxes descend from production: the inverted topologies that make standard branching strategies structurally invalid.

Why This Difference Matters

Topology inversion, absent back-sync path, and continuous branch-org divergence: the three consequences that generate every failure mode in Lesson 3.