Repository navigation
Gitignore .editorconfig + .claude scaffolding estate-wide #68
Description
Activity
- added a parent issue
on May 16, 2026 Template-side already satisfied: both
rsr-template-repo/.gitignoreandv3-templater/.gitignorecarry.editorconfigand.claude/. Remaining: propagation to consumers (epic-ongoing). Refs #66.- added a commit that references this issue
on May 19, 2026 Progress 2026-05-19 (keeping OPEN):
- Template
.gitignore✓ (rsr-template-repo#49, v3-templater#69 — 2026-05-16) - Enforcement bundled with Eliminate npm lockfiles estate-wide (npm-avoidant) #67 in standards#148.
- Audit (2026-05-19 read-only sweep): 118 repos still track
.editorconfig, 56 still track.claude/, 131 lack the entries entirely. The .gitignore fix prevents new ones; removing existing tracked copies is a separate per-repo owner-gated chore.
Remaining for close: consumer propagation to reach 0 tracked of each. Per-repo, human-gated.
- Template
- added 4 commits that reference this issue
on May 20, 2026 - added a commit that references this issue
on Jun 2, 2026 2026-08-25 — the 2026-07-25 'mass-remediation' claim does not reproduce
standards' own
.gitignoreonorigin/maincontains neither.editorconfignor.claude.5 of 6 sampled repos (standards, laminar, burble, verisimdb, hypatia) lack the
.editorconfigentry, and all 6 still track.editorconfig— includingrsr-template-repo, which manages to gitignore and track it simultaneously.This is a claim-vs-artefact gap of exactly the kind DYADT exists to catch. Worth feeding to #462 as a real test case rather than a synthetic one.
Verified 2026-08-25.
Next step (re-measured 2026-08-27): the template half landed (
rsr-template-repo/.gitignorecarries both entries) — but.gitignoredoes not untrack:.editorconfigis still committed in 408 of 420 repos and.claude/in 193, includingrsr-template-repoitself. Of 417 repos with a root.gitignore, only 17 ignore both.→ Two-part sweep: propagate the ignore lines, and
git rm --cachedacross ~408 repos. Large — this is a campaign, not an edit. (Now free-standing: parent tracker #66 closed 2026-08-28.)Owner ruling 2026-08-28: SUPERSEDED by #343. The estate commits a canonical per-language
.editorconfig(and.gitattributes) rather than gitignoring them. 382/388 repos already commit an.editorconfig, so this issue's local-only approach never reflected reality. This issue will be closed when the #343 canon lands inrsr-template-repo; the.claudescaffolding portion of this issue is unaffected by the ruling.- added a commit that references this issue
on Sep 3, 2026 Closing as superseded by #343.
This issue asked to gitignore
.editorconfigestate-wide. The 2026-08-28 ruling
reversed that:.editorconfigis per-language configuration, not linting logic,
so it is committed rather than ignored. The measurement behind the reversal was
that 382 of 388 repositories already commit one, making the ignore rule a
change against the estate rather than a codification of it.The live work is #343, which treats the same files as a drift problem and lands
a canonical version. Phase 1 of that is now merged into the template repo:
hyperpolymath/rsr-template-repo#67.
Decision (owner, 2026-05-16): GITIGNORE both .editorconfig and .claude/ estate-wide (treat as local-only; do NOT commit agent scaffolding into every repo).
Mechanism: add both to canonical template .gitignore (rsr-template-repo/v3-templater); propagate; stops the untracked-noise wave seen in cafescripto/jaffascript/lucidscript/pseudoscript/standards/rattlescript.NEW etc.
Acceptance: both ignored via template-propagated .gitignore estate-wide; no more untracked .editorconfig/.claude churn.
Parent: #66