Michael Schroeder · Preprint · Version 1.0.1 · 6 September 2026 · Revised 14 September 2026 · ORCID 0009-0004-3249-0195

Eight Prime Divisors in Odd Distinct Covering Systems

Abstract

We prove that the least common multiple of the moduli of a finite covering system with distinct odd moduli greater than one has at least eight distinct prime divisors, uniformly in the primes and their exponents. A capped-gain lift converts a normalized one-fibre bound into an increasing supermodular functional on individual congruence labels. Injective reindexing of complete exponent types gives a scalar upper envelope. A certificate of 54 rational inequalities and six affine identities yields 124847/125000<1, with exact affine tails. We also prove a method-specific ceiling: on the first eight odd reference primes, the exact scalar value is at least 1001/1000 for every admissible real threshold vector and block order, and adding admissible blocks cannot restore a value below one. Separate Lean 4 companions verify the covering theorem and this obstruction to the specified scalar architecture, not to other proofs of higher rank bounds.

Read the Paper

Proof Package

The Zenodo edition includes the paper in PDF, Markdown and LaTeX, both Lean 4 formalizations, exact certificate inputs, deterministic generators and reproduction instructions.

This compact archive is byte-identical to the published Zenodo deposit. Start with README.md inside the archive. The scalar project's generated sources must be reconstructed before its Lean build; the included tools verify every regenerated source hash. Compiled objects and dependency caches are not bundled.

Version 1.0.1 corrects an internal draft number mistakenly used as the public version label. The DOI and scientific content are unchanged.

The mathematical argument, Lean proof sources and certificate inputs are unchanged from v6. The companion separates the new covering-project build and audit from the historical scalar cold-replay evidence; it does not claim a new second-machine verification.

The paper and prose are licensed under CC BY 4.0; the original verification code is licensed under MIT. These apply to their respective components, not as alternative licenses for the whole deposit. See licensing and third-party notices.

Earlier version 6: paper, guide and expanded companion

These preserved files belong to the original 6 September 2026 release, not the DOI-bearing version 1.0.1. For the current edition, use the PDF and compact companion above.

Original v6 paper (PDF) · Original v6 paper (Markdown) · Original citation · Original release metadata · Original rights statement

Your download will be verified before it is saved.

Download the compact source ZIP (777 KiB): LaTeX, both Lean projects, exact inputs, deterministic certificate generators and supporting documentation.

Companion guide

The guide describes both formal theorems, the assumptions and verification scope of each Lean project, and how to reproduce the checks. It distinguishes the fresh covering-proof replay from the preserved large scalar replay.

Download details and verification

The ZIP is supplied in five parts to fit the website’s file-size limit. The download button checks every part and the complete archive before offering the original ZIP. The full download needs approximately 101 MiB of network data and several hundred MiB of available browser memory.

Download download_companion.py and run python3 download_companion.py. It uses only Python’s standard library, verifies each part and the finished archive, and will not overwrite an existing output file.

Alternatively, save all five parts in the same folder:

  1. Part 1 · 24,000,000 bytes
  2. Part 2 · 24,000,000 bytes
  3. Part 3 · 24,000,000 bytes
  4. Part 4 · 24,000,000 bytes
  5. Part 5 · 9,789,211 bytes

On macOS or Linux, concatenate the parts in this exact order:

cat eight_prime_support_companion_v6.zip.part01 \
    eight_prime_support_companion_v6.zip.part02 \
    eight_prime_support_companion_v6.zip.part03 \
    eight_prime_support_companion_v6.zip.part04 \
    eight_prime_support_companion_v6.zip.part05 \
    > eight_prime_support_companion_v6.zip
shasum -a 256 eight_prime_support_companion_v6.zip

The result must have SHA-256:

f21fb6f72dbba419adf151243adc02cd91389de430821957309594da7a98b858

Hashes identify the released bytes, not mathematical correctness. Check the extracted archive’s own SHA256SUMS before following its build instructions.

Citation, context and provenance

BibTeX · Citation metadata · Zenodo record · Website checksums · Licensing

Schroeder, Michael. Eight Prime Divisors in Odd Distinct Covering Systems. Version 1.0.1, Zenodo, 2026. DOI: 10.5281/zenodo.22747841.

This work strengthens the author’s rank-seven theorem, credited as antecedent work but not assumed in the rank-eight proof.

Keywords: Erdős–Selfridge odd covering problem; Erdős Problem 7; covering systems; prime divisors; supermodularity; distortion method; formal verification.

Acknowledgments

This work benefited from research assistance by AI systems developed by OpenAI and Anthropic, including support for proof exploration, proof development, and exact computational checks among others.

Explanation of the problem

A congruence class is a regularly spaced set of integers. For example, the class 1 modulo 3 is …, −5, −2, 1, 4, 7, …: all integers with remainder 1 on division by 3. A covering system is a finite collection of such classes whose union contains every integer.

Each spacing is called a modulus. The Erdős–Selfridge odd covering problem asks whether a covering is possible when every modulus is odd and greater than one, and no two classes have the same modulus. Classes may overlap; every integer must be covered at least once.

The rank is the number of different primes appearing across all moduli, equivalently the number of distinct prime divisors of their least common multiple. The set of these primes is called the prime support. For example, moduli 9 and 25 use only the primes 3 and 5; raising them to higher powers does not add new primes.

This paper proves that any odd distinct covering system must use at least eight distinct primes. Equivalently, every finite candidate supported on seven or fewer primes misses some integers, however large the prime powers are. This gives a necessary condition for a covering; the general existence question remains open.