No articles match
Citing a Set of Tables12 days ago
The aim | Set up the project | Version 1: register every input | Version 2: derive the output | Use it | Refresh: new data arrives | Who can read what | Where you are | Teardown
Starting on S312 days ago
Requirements | Load your secrets | Settings | Where the data goes | Build the store | Create the repository | Connect | Step 1: Sync one file | Step 2: Update one file | Step 3: Sync again with nothing new | Step 4: A batch of files | Reading as a reader | Where you are | What next | Governance and migration come later | Teardown
Version SHAs: The Three-SHA Identity Model12 days ago
data_sha: the canonical content hash | Why not just hash the parquet bytes | The datom table contract | Where this bites | Check before you write | The recourse table | The ingestion allowlist | The identity decisions | metadata_sha: the version | The dedup guard is full-history | parquet_sha: stored-object integrity | Change classification | What metadata reveals about your data | How far the cross-language claim goes | One caveat that is not ours to fix | Derived columns: when arithmetic breaks cross-platform identity | The guarantee direction | Which operations are stable | Cross-language note | Practical advice | What you can rely on
Getting Started3 months ago
Requirements | Configure your environment | Build a store | Initialize the data repository | Connect | Step 1: Sync one file | Step 2: Update one file | Step 3: No-op repeat | Step 4: A batch of files | Project health checks | Reading as a reader | Writing a computed table with datom_write() | The datom table contract | Where you are | Teardown
The datom Model: Code in Git, Data in Cloud3 months ago
The split | Property 1: Immutability is automatic | Property 2: The metadata is reviewable like code | Property 3: Readers don't need write access | Property 4: Storage is swappable | What stays in the metadata | What this isn't | Where this leads
Tracing Data Lineage3 months ago
Setup | A simple study: raw imports to derived analysis | Deriving a table | Propagating lineage further downstream | Recomputing lineage consistency | Key points
Serverless and Distributed by Design4 months ago
What datom needs to function | What runs where | Why no server | What the absence of a server costs | The "distributed by design" half | Where this leads