Delete README.md
Browse files
README.md
DELETED
|
@@ -1,33 +0,0 @@
|
|
| 1 |
-
# Kaggle sample — EX-NIHILO-CODE-001
|
| 2 |
-
|
| 3 |
-
**This is a sample, not the product.** Strict star-sorted head of the Layer 1 index (n=100 of 339 full): 97 rows are in the verified cut and 3 fall below its 100-merged-PR threshold. No harvested PR diffs.
|
| 4 |
-
|
| 5 |
-
| Locked integer | Value |
|
| 6 |
-
|---|---|
|
| 7 |
-
| Pull date | 2026-09-16 |
|
| 8 |
-
| Full index | 339 |
|
| 9 |
-
| Verified (≥100 merged GitHub PRs) | 319 |
|
| 10 |
-
| Premium (stars ≥ 1000 and estimated_loc ≥ 250000) | 205 |
|
| 11 |
-
| This file | 100 rows, stars descending |
|
| 12 |
-
|
| 13 |
-
## What you are looking at
|
| 14 |
-
|
| 15 |
-
A phone book of public GitHub repositories whose SPDX is MIT / Apache-2.0 / BSD-2-Clause / BSD-3-Clause, linguist-estimated LOC is 100k–10M, and the named language is ≥ 40% of linguist bytes.
|
| 16 |
-
|
| 17 |
-
`estimated_loc` is linguist bytes ÷ bytes-per-line (Rust 42, Go 38, C++ 48, Java 52, COBOL 72, MATLAB 44). `size_kb` is git disk usage. **Do not compute LOC as size_kb × 30.** The chart `size_kb_vs_estimated_loc.svg` is the proof.
|
| 18 |
-
|
| 19 |
-
## What is not in this kernel
|
| 20 |
-
|
| 21 |
-
- Harvested pull-request patches (`files.json`)
|
| 22 |
-
- The verified-only cut
|
| 23 |
-
- The premium harvest worklist
|
| 24 |
-
|
| 25 |
-
Those ship in the paid pack (`listable_harvest=true`, 1,020 PRs). Do not upload `harvest/` here.
|
| 26 |
-
|
| 27 |
-
## Notes
|
| 28 |
-
|
| 29 |
-
- `golang/go` uses Gerrit. GitHub `merged_pr_count` is not contribution volume. It may appear in this mixed sample.
|
| 30 |
-
- COBOL n=3, MATLAB n=1 in the full file. That is coverage.
|
| 31 |
-
- `has_tests` / `contributors` may be empty by contract. Do not advertise “no nulls.”
|
| 32 |
-
|
| 33 |
-
Grain: one repository. License of listed projects stays with those projects.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|