Skip to content

Bitcoin · Ordinals · permanent archive · Dec 2024

Project Spartacus — 76,911 Afghan War Logs inscribed on Bitcoin.

Dev on the inscription pipeline for WikiLeaks' Project Spartacus — permanently archiving the 2010 Afghan War Diary onto Bitcoin via the Ordinals protocol, with OrdinalsBot as the technical partner. Started December 12, 2024: 76,911 files batch-inscribed across many blocks into a permanent, censorship-resistant archive.

The building blocks

76,911

Files inscribed

Bitcoin

Ordinals protocol

Dec 2024

Inscription start

WikiLeaks

Partner · archive owner

The problem

Project Spartacus was WikiLeaks' response to the prosecution of Julian Assange — an effort to cement the documents he published into a permanent, censorship-resistant record, rather than a fundraiser or a legal campaign. The journalism predates the project; the job was to stop anyone from erasing it.

Making a 76,911-file archive permanent on Bitcoin is also an order-of-magnitude jump from a typical Ordinals drop. The constraint isn't novelty — it's endurance. The pipeline has to run over days and weeks of blocks, survive fee spikes, and finish with a verifiable, complete archive.

Every inscription is an independent Bitcoin transaction, so "upload all at once" doesn't exist. You batch, you monitor, you reconcile — and any drift between what the manifest says and what actually made it on-chain is a bug that grows with volume.

The goal was permanence and verifiability: a complete, content-addressed archive of the Afghan War Diary on Bitcoin mainnet, inscribed through OrdinalsBot rails, rechecked against the original source at the end.

What I did

Four layers for a 76,911-file archive.

01

Batched inscription at rare scale

76k+ files meant the pipeline had to ship reliably across many days, not a single drop window. Inscriptions were grouped into fee-tuned batches that re-priced per block — when the mempool ran hot, the batch resized instead of stalling. The rails OrdinalsBot had shipped for high-volume drops were the foundation; this was the first run at a fifty-times-higher scale.

02

Deterministic ordering + content integrity

Every file hashed and manifested before inscription started — a complete list of what had to land, in what order, against which source. As each batch confirmed, the manifest was reconciled against on-chain state. Missing inscriptions, reorgs, and partial confirmations all resolved against the single source of truth, not against whatever the last monitoring script happened to see.

03

Multi-block coordination + recovery

An inscription run that spans weeks is one long series of recovery states: stuck parents, orphaned children, UTXOs drifting on broadcast failures. Operator tooling surfaced each of these states in-flight so nothing slipped. No "we'll check at the end" — the pipeline kept itself honest block by block.

04

Cost control without cutting corners

Inscribing 76k files on Bitcoin mainnet at peak fees would wreck the budget. Mempool-aware batching pushed inscriptions through low-fee windows and held back during spikes, targeting the long-tail average rather than the panic price. The archive is permanent; there was no need to rush it through the expensive blocks.

What transferred

Patterns the Bitcoin stack has been earning for years.

  • Batched inscription + mempool-aware fee engine = the only sane way to ship 10,000+ inscriptions. Fee-blind pipelines die somewhere in the middle.
  • Manifest-first. The manifest isn't an output of the pipeline — it's the contract the pipeline runs against. Reconciliation is cheap when the contract exists upfront.
  • The operator dashboard is the product. When the run spans weeks, the interface that tells you what's happening in the queue is the thing that lets you sleep.
  • Bitcoin doesn't care about your deadline. Let the protocol set the pace — batch size, fee target, block cadence — instead of fighting it.

Stack

  • Bitcoin Core + ord protocol
  • OrdinalsBot inscription rails
  • Batched pipeline · mempool-aware fee engine
  • Content manifest + hash verification
  • Operator tooling for multi-week runs

Free download · 6 pages · PDF

What I learned shipping 1,500 Ordinals on Bitcoin.

Pre-launch, launch-day, and the 90 days after — with the mistakes that cost real BTC. Ordinals, Runes, and beyond.

No spam. You also get the Sunday note — unsubscribe in one click.