Repository navigation
fix(pkg)!: re-mint UUID as v8, bump to 0.1.1 - #41
Conversation
Owner ruling 2026-10-08 (extends D320): the patterned placeholder UUID e6a7b8c9-d0e1-4f2a-ab3c-4d5e6f7a8b9c is replaced by a fresh ADR-008 profile-T UUIDv8, 01a11d29-d88f-83f4-a10e-468e4f4e9669, and the package is re-registered in julia-professional-registry at 0.1.1. BREAKING: existing installs must Pkg.rm and Pkg.add again. Hyperpolymath.jl moves to the new UUID in its own PR. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WPSJ7fBhVAMcpSffCBWUDo Signed-off-by: Jonathan D.A. Jewell <6759885+hyperpolymath@users.noreply.github.com>
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configuration
📒 Files selected for processing (1)
Included review availability: This review used your included allowance. Your plan provides up to 1 included review per hour; 0 remain after this review. 📜 Recent review details⏰ Context from checks skipped due to timeout. (11)
🔇 Additional comments (1)
📝 SummarySummary by CodeRabbit
WalkthroughProject.toml changes the package UUID and updates the version from 0.1.0 to 0.1.1. The author declaration is unchanged. ChangesPackage Metadata
Priority: ➖ Normal Estimated code review effort: 1 (Trivial) | ~3 minutes Change: Other Merge Risk: ⚪ Minimal · up to The UUID migration is documented, and no checked-in release or build contract establishes an actionable mismatch. No merge-blocking risk is established. Architecture SummaryArchitecture risk: 🔵 Low · up to The change affects 1 system. Changed systems: Architecture concerns Review detailsSystems and components
Before / after behavior
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. A rabbit checks the package name, Comment |
<!-- SPDX-License-Identifier: CC-BY-SA-4.0 --> ## Summary Points nine `[deps]` entries at the UUIDv8 identities their packages are being re-registered under in `julia-professional-registry` (JPR). Owner ruling 2026-10-08, which extends D320. The old values were patterned placeholders. | dep | old uuid | new uuid | upstream PR | |---|---|---|---| | Causals | `c4a8b6d2-…` | `01a11d29-d897-8383-92d8-041d500f9c95` | hyperpolymath/Causals.jl#68 | | FirmwareAudit | `e6a7b8c9-…` | `01a11d29-d88f-83f4-a10e-468e4f4e9669` | hyperpolymath/FirmwareAudit.jl#41 | | HardwareResilience | `d5f6a7b8-…` | `01a11d29-d880-869f-9c60-029824b42355` | hyperpolymath/HardwareResilience.jl#43 | | LowLevel | `a1b2c3d4-…` | `01a11d29-d884-82e8-9db2-8ff667f451ca` | hyperpolymath/LowLevel.jl#37 | | QuantumCircuit | `b3d4e5f6-…` | `01a11d29-d887-80e0-8c2e-f313792966ce` | hyperpolymath/QuantumCircuit.jl#39 | | ShellIntegration | `c3d4e5f6-…` | `01a11d29-d88a-8008-a50a-9fff81f51eb4` | hyperpolymath/ShellIntegration.jl#35 | | SiliconCore | `c4e5f6a7-…` | `01a11d29-d88c-832a-9ded-976af4b7dc75` | hyperpolymath/SiliconCore.jl#34 | | SMTLib | `7d3f9a2c-…` | `01a11d29-d89a-86da-8e3a-553d114a3537` | hyperpolymath/SMTLib.jl#46 | | ZeroProb | `f9e8c2e0-…` | `01a11d29-d894-81ed-be3f-5f3acc80d59a` | hyperpolymath/ZeroProb.jl#61 | This PR changes nothing except those 9 UUID lines. `[compat]` and this package's own `uuid` are unchanged. ## Type of change - [x] 🐛 Bug fix: the dependency UUIDs follow the re-registration. Until the JPR PR lands, the new UUIDs are not yet registered. - [ ] ✨ New feature: n/a - [ ] 💥 Breaking change: n/a. Only the dependency identities change; this package's own identity and API do not. - [ ] 🕳️ Soundness fix: n/a - [ ] 📖 Documentation: n/a - [ ] 🧹 Refactor: n/a - [ ] ⚡ Performance: n/a - [ ] 🔧 Build/CI: n/a ## 📌 New pins - **Head SHA: `dd1d4589a163cb1c1ac1f4d0857979aedf43fbb7`** (tree `5d81e21b1a467235e7913b966c0366f4b96aa1ce`) - No action, lockfile or container pins are added or changed. No `Manifest.toml` is tracked. ## How has this been verified? - **Cross-checked against the source of each UUID.** For each of the 9 deps, the new value equals the `uuid =` line in that package's re-mint PR worktree. 9 of 9 match. - **No old UUIDs remain.** `git grep` for all 10 old placeholder UUIDs returns nothing in this repo. - **No other dependents were missed.** I grepped every `Project.toml` under `hyper-repos/` and `meta-repos/` (111 files, excluding `julia-ecosystem` copies). The only hits besides the 10 packages themselves were Axiom.jl, InvestigativeJournalism.jl and this repo. - **Why `Pkg.test` was not run locally:** it would not reach this change. `main` already fails at dependency resolution (CI run 37783284204 on `1d0fb8b3`): `Unsatisfiable requirements detected for package Cliometrics … restricted to versions 0.1 by Hyperpolymath — no versions left`. Against JPR `origin/main` 9e4dc83, a compat census shows four independent pre-existing blockers: - `Cliometrics = "0.1"`, but only 0.2.0 is registered. - `SMTLib = "0.1"`, but only 0.2.0 is registered. The re-mint registers 0.2.1, which this bound does not admit either. - `InvestigativeJournalism`: JPR registers UUID `379a7c0a-…` under the name `InvestigativeJournalist`. - `TradeUnionism` is not in JPR at all. They are left out of this PR to keep it reviewable. - **CI on this PR fails earlier than on `main`, by design:** `ERROR: expected package QuantumCircuit [01a11d29] to be registered`. The new UUIDs are not in JPR until the re-registration PR merges. Once they are, the `main` blockers above surface again. ### Deferred red checks (AGENTS.md §5c item 3) - `Julia 1.10 - ubuntu-latest`, `Julia 1.11 - ubuntu-latest`, `Julia 1.11 - macos-latest`: #52 - `governance / Workflow security linter` (pre-existing on `main`; `codeql.yml` has no SPDX header and no top-level `permissions:`): #53 ## Checklist - [x] Commits are signed (`G`) and signed off - [x] Checks run locally (as above) - [x] SPDX headers on new files: no new files - [ ] Docs updated: no doc names these UUIDs - [x] No soundness hole introduced ## Notes for reviewers Merge after the nine upstream PRs above and the JPR re-registration PR. Before those land, the new UUIDs resolve nowhere. LowLevel.jl#37 itself depends on HardwareResilience.jl#43 and SiliconCore.jl#34. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01WPSJ7fBhVAMcpSffCBWUDo Signed-off-by: Jonathan D.A. Jewell <6759885+hyperpolymath@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Summary
Re-mints the package UUID as a v8 so the package can be re-registered in
julia-professional-registry(JPR) at 0.1.1.uuid:e6a7b8c9-d0e1-4f2a-ab3c-4d5e6f7a8b9c(patterned placeholder) →01a11d29-d88f-83f4-a10e-468e4f4e9669(ADR-008 profile T, minted withstandards/scripts/uuid-v8.js t). Owner ruling 2026-10-08, which extends D320.version: 0.1.0 → 0.1.1 (owner ruling: patch bump under the new UUID).⚠ Registration drift: the JPR entry will register this PR's merged tree. That tree carries every change on
mainsince the 0.1.0 registration (registered treea0a83498cb,maintree69a43077c3), not only these lines.Type of change
Pkg.rm("FirmwareAudit")thenPkg.addagain.Hyperpolymath.jldepends on this package and moves to the new UUID in its own PR.📌 New pins
688459958d74c797d5fdd3167503e9000677d640(tree4833940472c2f5f3057b922595ff6ad1d9b5c082)How has this been verified?
Pkg.Types.read_project("Project.toml")returns nameFirmwareAudit, uuid01a11d29-d88f-83f4-a10e-468e4f4e9669, version0.1.1.usinginsrcis declared (Dates,SHA).Pkg.test()(Julia 1.12.6): 39/39 passed,FirmwareAudit tests passed.CIonmainis green (2026-10-08T17:16Z).Checklist
G) and signed offNotes for reviewers
Merge this PR before the JPR PR that re-registers the package. The JPR PR needs BDFL review plus an ADR, per
docs/REGISTRY-POLICY.adoc. After the squash merge, tag the merge commitv0.1.1(signed).🤖 Generated with Claude Code
https://claude.ai/code/session_01WPSJ7fBhVAMcpSffCBWUDo