https://github.com/systemslibrarian/crypto-lab-oblivious-shelf
https://github.com/systemslibrarian/crypto-lab-oblivious-shelf
Last synced: 1 day ago
JSON representation
- Host: GitHub
- URL: https://github.com/systemslibrarian/crypto-lab-oblivious-shelf
- Owner: systemslibrarian
- Created: 2026-04-08T23:52:20.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2026-04-09T01:48:36.000Z (about 2 months ago)
- Last Synced: 2026-04-09T03:30:11.964Z (about 2 months ago)
- Language: TypeScript
- Size: 44.9 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
[](https://systemslibrarian.github.io/crypto-lab/)
[](https://github.com/systemslibrarian/crypto-lab-oblivious-shelf/actions/workflows/pages.yml)
# crypto-lab-oblivious-shelf
## 1. What It Is
crypto-lab-oblivious-shelf implements the 2-server Information-Theoretic Private Information Retrieval (IT-PIR) scheme from Chor, Goldreich, Kushilevitz, and Sudan (1995). A patron retrieves an entry from a 16-item library catalog by sending XOR-based queries to two non-colluding servers, neither of which learns which entry was requested. The protocol solves the database-query privacy problem: how to fetch a specific record from a remote database without the database operator learning which record was fetched. The security model is information-theoretic — the server provably learns nothing regardless of its computational power, requiring no cryptographic hardness assumptions.
## 2. When to Use It
- Use IT-PIR when patrons must retrieve catalog records without the server logging which record was accessed, and legal or policy protections alone are insufficient.
- Use it when you need privacy guarantees that hold even if the adversary has unlimited computation — computational PIR is not sufficient.
- Use it in multi-server deployments (e.g. mirrored library systems) where servers are genuinely operated independently.
- Do not use it as the sole privacy measure — IT-PIR protects the query, not the physical borrowing transaction, the IP address, or query timing.
- Do not use the 2-server scheme when both servers are operated by the same institution — collusion collapses the guarantee immediately.
## 3. Live Demo
[https://systemslibrarian.github.io/crypto-lab-oblivious-shelf/](https://systemslibrarian.github.io/crypto-lab-oblivious-shelf/)
Select any book from the 16-entry catalog grid to set it as the target index, then click **Generate Query** to run the full 2-server XOR PIR protocol with real arithmetic. Each step — random subset generation, symmetric difference, server-side XOR computation, and client-side recovery — is displayed sequentially in the walkthrough panel. Click **Run Again** to re-run with a fresh random subset S, demonstrating that each query looks different even for the same target. The privacy audit panel below the walkthrough shows exactly what each server sees, confirming that neither server's view reveals the target index.
## 4. What Can Go Wrong
- **Server collusion:** if both servers share logs or are controlled by the same party, the guarantee fails — the scheme provides no protection against a colluding pair.
- **Query linkage:** repeated queries from the same IP for the same item are not protected by PIR alone; traffic analysis can correlate queries over time.
- **O(n) communication:** for a catalog of 1 million items, each query requires transmitting 1 million bits to each server — impractical without the engineering optimizations in systems like RAID-PIR.
- **Implementation bugs in subset generation:** if the random subset S is not uniformly distributed, the privacy proof breaks. Use a cryptographically secure PRNG.
- **Metadata leakage:** PIR protects the index, not the fact that a query occurred, its timing, its size, or the patron's identity.
## 5. Real-World Usage
- **RAID-PIR (Devet and Goldberg, 2014):** A practical IT-PIR system that distributes a large database across multiple servers using RAID-style XOR parity, enabling private queries over databases with millions of records at usable throughput.
- **Percy++ (Ian Goldberg, University of Waterloo):** An open-source C++ library implementing both IT-PIR and CPIR protocols, used in academic privacy research and as a reference implementation for multi-server PIR schemes.
- **PIR-Tor (Mittal et al., 2011):** A proposal to use IT-PIR for private Tor relay selection, allowing clients to fetch relay descriptors from directory servers without revealing which relays they intend to use.
- **Popcorn (Gupta et al., 2016):** A media-streaming system using IT-PIR to let users retrieve video content from a library without the server learning which video was watched, targeting the Netflix-style content selection problem.
- **Checklist (Henry and Goldberg, 2013):** A privacy-preserving messaging system built on IT-PIR, allowing users to retrieve messages from a bulletin board without revealing which messages they are reading.
**Cross-links:**
- [Patron Shield](https://systemslibrarian.github.io/crypto-lab-patron-shield/) — IT-PIR applied to full catalog privacy
- [Silent Tally](https://systemslibrarian.github.io/crypto-lab-silent-tally/) — Shamir-based MPC secure sum
- [Shamir Gate](https://systemslibrarian.github.io/crypto-lab-shamir-gate/) — threshold secret sharing
- [crypto-lab home](https://systemslibrarian.github.io/crypto-lab/)
---
*"So whether you eat or drink or whatever you do, do it all for the glory of God." — 1 Corinthians 10:31*