Does the 2,000-file rule admit an alternative evidence path for Malbolge?

[DannyBaanks]

Hi maintainers,

Question about CONTRIBUTING.md language-extension usage requirements (2,000 files for common extensions, 200 for once-per-repository extensions, last year, forks excluded), and its note that a better assessment method can be proposed in a Discussion.

We have a technically complete Malbolge candidate using only .malbolge: an MIT source.malbolge grammar integrated through the canonical script/add-grammar path, a representative non-Hello MIT sample, language detection, and a reproducible patch. The public evidence branch is https://github.com/DannyBaanks/malbolge-linguist/tree/malbolge-canonical . Key artifacts: audit https://github.com/DannyBaanks/malbolge-linguist/blob/malbolge-canonical/docs/FINAL_AUDIT.md , patch https://github.com/DannyBaanks/malbolge-linguist/blob/malbolge-canonical/evidence/patches/malbolge.patch , and canonical logs https://github.com/DannyBaanks/malbolge-linguist/tree/malbolge-canonical/evidence/canonical .

Honest usage on the proposed extension is limited: extension:malbolge NOT is:fork returned 42 files, with 3 containing the term Malbolge. We do not claim .mal: it is heavily contaminated by unrelated formats. PR #4609 faced the same usage blocker, and we are not reopening its notability argument.

Question: would a reproducible, semantically validated corpus across Malbolge historical but ambiguous extensions be relevant to eligibility assessment, or is the 2,000-file threshold strictly per proposed extension? The corpus would deduplicate user/repo, exclude owner-controlled repositories and forks, distinguish false positives through execution/grammar validation, and avoid using .mal contamination.

We will not open a pull request without maintainer guidance. If the threshold is strictly per extension, we will keep USAGE_GATE=NOT_DEMONSTRATED and POLICY_ELIGIBLE=FALSE.

[lildude]

Question: would a reproducible, semantically validated corpus across Malbolge historical but ambiguous extensions be relevant to eligibility assessment, or is the 2,000-file threshold strictly per proposed extension? The corpus would deduplicate user/repo, exclude owner-controlled repositories and forks, distinguish false positives through execution/grammar validation, and avoid using .mal contamination.

Usage has to be supported per extension or filename being added.

[DannyBaanks]

Thanks, that answers the question clearly.
We’ll keep the Malbolge contribution on hold until usage for the proposed .malbolge extension satisfies the documented requirement. I appreciate the clarification.