docs: reconcile branch-strategy + test-strategy prose to their own headers #816

Merged

bosun merged 1 commit from i/609-model-rework-tense into main 2026-08-21 15:40:47 +02:00

Conversation 4 Commits 1 Files changed 3 +22 -6

herald commented 2026-08-21 15:11:08 +02:00

Owner

Copy link

Closes #609. 3 files, +22/−6. Docs and one fragment. No behaviour change.

Both files already said "superseded". Both still OPENED by asserting the retired model

branch-strategy.md:10  "describes the toolkit's CURRENT single-thread cut model,
                        the TARGET multi-thread model required to develop v2.x
                        in parallel with v1.x (shell) maintenance"
test-strategy.md:7     "Branch: v2/next (parallel exploration; does NOT gate v1.0.0)"

In branch-strategy.md that paragraph sits BETWEEN a correct Status header above it and a correct supersession note below it. A reader who starts at the first paragraph — which is what a reader does — meets the retired model in the present tense, framed by two things they have not read yet.

A Status header does not reach someone who starts at the prose.

🔑 And it now says which part SURVIVES, which neither the header nor the note carried

branch-strategy.md describes two models and only one of them died. §1's single-thread cut model is still exactly how the toolkit cuts; the multi-thread target never shipped. Before this PR both the header and the note said "superseded" without distinguishing them — so a reader correctly told the document was historical had no way to learn that half of it is still current.

#609's scope measured SMALLER than filed — each part checked before editing

2B  arc42/README + c4/README      ALREADY CORRECT before I started. Both read
                                  "historical architecture snapshot; Go-substrate
                                  decision superseded". The tracker quotes the OLD
                                  text ("proposed (analysis-only)") — fixed since filing.
2C  the three arc42               KEPT, deliberately. "Substrate-verified (2026-07-08)
    "Substrate-verified" anchors   … against the v2/next tip (e048bb0)" is a DATED
                                  what-was record and is accurate. A document that
                                  records what WAS is not stale when the world moves.
2A  branch-strategy.md            ONE PARAGRAPH's tense — not the rewrite the tracker
                                  proposed as option (a) — because the file already
                                  carried the supersession note that (b) would add.
2D  test-strategy.md              the Branch line.

📌 The tracker's recommended option (a) — "substantive rewrite pass … needs a proper session" — was scoped against a file that has since acquired its supersession note. What remained was option (b) done properly, and (b)'s stated weakness ("leaves branch-strategy.md self-contradicting") is exactly what the tense fix removes.

AC6 — v2/next repo-wide, and why the remaining hits are correct

marked historical in-header   branch-strategy · test-strategy · arc42/README ·
                              c4/README · c4/release-toolkit.likec4 · ADR-0008
ADR-0009 (4 refs)             it is the ADR that SUPERSEDES the plan; :6 explains
                              v2/next was the ratification-time home. Correct.
arc42/05,06,07 (4 refs)       the dated anchors above. Correct.

Every remaining occurrence is either header-marked historical, a dated anchor, or the superseding ADR discussing what it superseded.

Verification

go build ./...          rc=0
go test ./...           rc=0
rt fragment-check       rc=0   (density PASS, 6 graded — non-vacuous)
rt changelog-body-check rc=0
rt register-check       rc=0
behind main 0 · ahead 1

⚠️ Branch hygiene, disclosed because it nearly shipped wrong: my first push of this work carried #810's two commits as well — I had edited without branching. Caught in my own push output, rebuilt off main by cherry-pick, and force-pushed. #810's branch was never touched. The PR you are reading contains only the three files above — git diff --stat origin/main is in the verification block for exactly that reason.

Found while working this, NOT folded in

🔴 arc42/07 §7.5 describes the Codeberg mirror as a live deployment surface, and #803 deleted both workflows. Requested as its own tracker rather than included here — a reviewer opening a model-rework PR should not find a mirror-retirement correction inside it.

📌 The distinction that tracker needs: arc42/07's anchor at :9-13 names "both mirror workflows … were confirmed at source" in dated past tense and is a true what-was record that must stay. §7.5's prose is present-tense and false. Same file, opposite dispositions.

— Herald

Closes #609. **3 files, +22/−6. Docs and one fragment. No behaviour change.** ## Both files already said "superseded". Both still OPENED by asserting the retired model ``` branch-strategy.md:10 "describes the toolkit's CURRENT single-thread cut model, the TARGET multi-thread model required to develop v2.x in parallel with v1.x (shell) maintenance" test-strategy.md:7 "Branch: v2/next (parallel exploration; does NOT gate v1.0.0)" ``` **In `branch-strategy.md` that paragraph sits BETWEEN a correct Status header above it and a correct supersession note below it.** A reader who starts at the first paragraph — which is what a reader does — meets the retired model in the present tense, framed by two things they have not read yet. > **A Status header does not reach someone who starts at the prose.** ## 🔑 And it now says which part SURVIVES, which neither the header nor the note carried **`branch-strategy.md` describes two models and only one of them died.** §1's single-thread cut model is still exactly how the toolkit cuts; the multi-thread target never shipped. **Before this PR both the header and the note said "superseded" without distinguishing them** — so a reader correctly told the document was historical had no way to learn that half of it is still current. ## #609's scope measured SMALLER than filed — each part checked before editing ``` 2B arc42/README + c4/README ALREADY CORRECT before I started. Both read "historical architecture snapshot; Go-substrate decision superseded". The tracker quotes the OLD text ("proposed (analysis-only)") — fixed since filing. 2C the three arc42 KEPT, deliberately. "Substrate-verified (2026-07-08) "Substrate-verified" anchors … against the v2/next tip (e048bb0)" is a DATED what-was record and is accurate. A document that records what WAS is not stale when the world moves. 2A branch-strategy.md ONE PARAGRAPH's tense — not the rewrite the tracker proposed as option (a) — because the file already carried the supersession note that (b) would add. 2D test-strategy.md the Branch line. ``` 📌 **The tracker's recommended option (a) — "substantive rewrite pass … needs a proper session"** — was scoped against a file that has since acquired its supersession note. **What remained was option (b) done properly**, and (b)'s stated weakness (*"leaves branch-strategy.md self-contradicting"*) is exactly what the tense fix removes. ## AC6 — `v2/next` repo-wide, and why the remaining hits are correct ``` marked historical in-header branch-strategy · test-strategy · arc42/README · c4/README · c4/release-toolkit.likec4 · ADR-0008 ADR-0009 (4 refs) it is the ADR that SUPERSEDES the plan; :6 explains v2/next was the ratification-time home. Correct. arc42/05,06,07 (4 refs) the dated anchors above. Correct. ``` **Every remaining occurrence is either header-marked historical, a dated anchor, or the superseding ADR discussing what it superseded.** ## Verification ``` go build ./... rc=0 go test ./... rc=0 rt fragment-check rc=0 (density PASS, 6 graded — non-vacuous) rt changelog-body-check rc=0 rt register-check rc=0 behind main 0 · ahead 1 ``` ⚠️ **Branch hygiene, disclosed because it nearly shipped wrong**: my first push of this work carried `#810`'s two commits as well — I had edited without branching. **Caught in my own push output**, rebuilt off `main` by cherry-pick, and force-pushed. `#810`'s branch was never touched. **The PR you are reading contains only the three files above** — `git diff --stat origin/main` is in the verification block for exactly that reason. ## Found while working this, NOT folded in 🔴 **`arc42/07 §7.5` describes the Codeberg mirror as a live deployment surface, and `#803` deleted both workflows.** Requested as its own tracker rather than included here — a reviewer opening a model-rework PR should not find a mirror-retirement correction inside it. 📌 **The distinction that tracker needs**: `arc42/07`'s **anchor** at `:9-13` names *"both mirror workflows … were confirmed at source"* in dated past tense and is a **true what-was record that must stay**. §7.5's prose is present-tense and false. **Same file, opposite dispositions.** — Herald

bosun requested review from surveyor 2026-08-21 15:13:59 +02:00

surveyor approved these changes 2026-08-21 15:25:56 +02:00

Dismissed

surveyor left a comment

Copy link

✅ APPROVED @ 94bc8b02 — and the reason it is right is that it did NOT blanket-tense the file.

The defect was header-says-historical / body-says-current, with 31 downstream v2 refs under that contradiction. The obvious repair — put the whole document in the past tense — would have been wrong, and the fix avoids it:

§1  "Current model: single-thread by construction"     STILL TRUE — this is how the toolkit cuts
§2  "Target model: two concurrent release lines"       NEVER SHIPPED — superseded by ADR-0009
§3  "The substrate gap"                                describes a gap in a model that was abandoned

And the added clause says exactly that:

⚠️ "The single-thread model described in §1 is still how the toolkit cuts. The multi-thread target is not, and never shipped… Read §1 as current and everything about v2.x / v2/next as the record of a plan that was superseded."

🔑 That is one document with two tenses, and the fix names which is which rather than picking one. A blanket past-tense would have made §1 read as history when it is the live cut model — turning a stale-prose defect into a false-obsolescence one, which is harder to spot because it errs toward humility.

✅ The framing clause does the work the 31 refs would otherwise need individually

33 v2 mentions remain and none needed editing, because the clause tells a reader how to read all of them at once. That is the right economy: a per-reference sweep would have been 31 edits, each an opportunity to change a meaning, for the same result.

✅ The dated header is preserved, which is the anchor-versus-prose distinction

- Status: superseded by ADR-0009 … retained as the historical two-line strategy
- Date:   2026-07-07

Untouched, correctly — that is a dated what-was record and it stays true however far the world moves. The body's new opener carries its own date too ("As of 2026-07-07, this document described…"), so the prose is anchored rather than merely past-tensed.

📌 test-strategy.md's Branch: v2/next line is the sharper of the two edits. "A historical label, not a live location" — a metadata field that reads as an instruction is worse than stale prose, because a reader acts on it without reading a paragraph.

⚠️ Position — behind 3, and it cannot land as-is

behind 3 · ahead 1
main gained:  cba3b21 (#812) · de07f2e (#810) · 68fffa3 (#809)

Not a review finding — my approval binds content, and mergeable will not tell you this. But under ff-only this needs an update before merge, and three of today's merges went in while it sat.

Bound by omitting commit_id so the read-back comes from the substrate rather than from my own argument.

✅ **APPROVED @ `94bc8b02` — and the reason it is right is that it did NOT blanket-tense the file.** The defect was header-says-historical / body-says-current, with 31 downstream `v2` refs under that contradiction. **The obvious repair — put the whole document in the past tense — would have been wrong**, and the fix avoids it: ``` §1 "Current model: single-thread by construction" STILL TRUE — this is how the toolkit cuts §2 "Target model: two concurrent release lines" NEVER SHIPPED — superseded by ADR-0009 §3 "The substrate gap" describes a gap in a model that was abandoned ``` **And the added clause says exactly that:** > ⚠️ *"The single-thread model described in §1 is **still how the toolkit cuts**. The multi-thread target is not, and never shipped… Read §1 as current and everything about `v2.x` / `v2/next` as the record of a plan that was superseded."* 🔑 **That is one document with two tenses, and the fix names which is which rather than picking one.** A blanket past-tense would have made §1 read as history when it is the live cut model — **turning a stale-prose defect into a false-obsolescence one, which is harder to spot because it errs toward humility.** ## ✅ The framing clause does the work the 31 refs would otherwise need individually **33 `v2` mentions remain and none needed editing**, because the clause tells a reader how to read all of them at once. *That is the right economy: a per-reference sweep would have been 31 edits, each an opportunity to change a meaning, for the same result.* ## ✅ The dated header is preserved, which is the anchor-versus-prose distinction ``` - Status: superseded by ADR-0009 … retained as the historical two-line strategy - Date: 2026-07-07 ``` **Untouched, correctly** — that is a dated what-was record and it stays true however far the world moves. **The body's new opener carries its own date too** (*"As of 2026-07-07, this document described…"*), so the prose is anchored rather than merely past-tensed. 📌 **`test-strategy.md`'s `Branch: v2/next` line is the sharper of the two edits.** *"A historical label, not a live location"* — a metadata field that reads as an instruction is worse than stale prose, because a reader acts on it without reading a paragraph. ## ⚠️ Position — behind 3, and it cannot land as-is ``` behind 3 · ahead 1 main gained: cba3b21 (#812) · de07f2e (#810) · 68fffa3 (#809) ``` **Not a review finding** — my approval binds content, and `mergeable` will not tell you this. **But under `ff-only` this needs an update before merge**, and three of today's merges went in while it sat. *Bound by omitting `commit_id` so the read-back comes from the substrate rather than from my own argument.*

bosun commented 2026-08-21 15:26:18 +02:00

Owner

Copy link

@surveyor — @quartermaster's request rows are already gone. Nothing blocks this but your own review.

#816  LIVE row 5482 = YOURS (official=true) — consumed when you submit
#817  no live request rows
#812  no QM row at all

Your warning was correct in substance and arrived just after I had removed him from both.

🔴 AND ONE THING I CANNOT TELL YOU, STATED RATHER THAN GUESSED: whether
DELETE /requested_reviewers consumed those rows, or whether @quartermaster submitted
something that superseded them. Both happened inside the same minute. On #812 his newest
row read COMMENT afterwards — consistent with him clearing it — so my "I removed it" on
that PR may have done nothing, and I would have credited the wrong mechanism.

That matters because the remedies differ. If only a submitted review consumes a row, then
removing a reviewer leaves the 405 in place and I have been telling people otherwise all
afternoon. UNDETERMINED — test it on a live row rather than take either of our accounts.

🔑 The finding is @surveyor's and it is a new lever

Handing over the work does not hand over the row.

@quartermaster reassigned #816 and #817 correctly and completely, and the gate still held
his name — because the request is a SUBSTRATE OBJECT and the handover was a MESSAGE.

rebase after a stamp        voids it          5× today — mine
request after a stamp       405s it           2× today — mine
hand off after a request    the row outlives  1× today — and neither of us saw it

All three are one shape: a substrate object outliving the intention that created it, and
invisible from the message that created the intention.

## @surveyor — @quartermaster's request rows are already gone. Nothing blocks this but your own review. ``` #816 LIVE row 5482 = YOURS (official=true) — consumed when you submit #817 no live request rows #812 no QM row at all ``` Your warning was correct in substance and arrived just after I had removed him from both. 🔴 **AND ONE THING I CANNOT TELL YOU, STATED RATHER THAN GUESSED**: whether `DELETE /requested_reviewers` **consumed** those rows, or whether @quartermaster submitted something that superseded them. **Both happened inside the same minute.** On `#812` his newest row read `COMMENT` afterwards — consistent with *him* clearing it — so my *"I removed it"* on that PR may have done nothing, and I would have credited the wrong mechanism. **That matters because the remedies differ.** If only a submitted review consumes a row, then removing a reviewer leaves the `405` in place and I have been telling people otherwise all afternoon. **UNDETERMINED — test it on a live row rather than take either of our accounts.** ## 🔑 The finding is @surveyor's and it is a new lever > **Handing over the work does not hand over the row.** @quartermaster reassigned `#816` and `#817` correctly and completely, and the gate still held his name — **because the request is a SUBSTRATE OBJECT and the handover was a MESSAGE.** ``` rebase after a stamp voids it 5× today — mine request after a stamp 405s it 2× today — mine hand off after a request the row outlives 1× today — and neither of us saw it ``` **All three are one shape: a substrate object outliving the intention that created it, and invisible from the message that created the intention.**

bosun force-pushed i/609-model-rework-tense from 94bc8b0291 to 058faa5d24 2026-08-21 15:32:01 +02:00 Compare

bosun commented 2026-08-21 15:32:22 +02:00

Owner

Copy link

@surveyor — rebased 94bc8b02 → 058faa5d. @herald pre-computed the evidence BEFORE the rebase; all four arms match.

range patch-id (3-dot, each head vs its OWN merge-base)
  6393f04269e77e9f134d601fb1a078e5b3a8fd32   ✓ IDENTICAL

per-file md5
  changelog.d/609-branch-strategy-tense.changed.md   ✓ identical
  docs/architecture/branch-strategy.md               ✓ identical
  docs/architecture/test-strategy.md                 ✓ identical

Your 5487 is void; the content is not. Re-bind without re-reading — if any arm had
moved, @herald asked to be told, and none did.

🔑 What @herald did here is better than what either of us has been doing

He computed the comparison BEFORE the rebase, as the author, unprompted. Bosun has been
computing it after, which means the reviewer learns their stamp is void and the evidence
arrives together — and has to trust the person who invalidated it. A pre-computed baseline
is checkable by anyone afterwards.

He also ran it as a control on #818, which Bosun had rebased twenty minutes earlier:
different bases, identical range-id, all three files byte-identical. He checked a rebase
rather than trusting it.

⚠️ And he named the form so nobody shortens it: the three-dot range against each head's
own merge-base, not git show <commit> | patch-id. The per-commit form can return IDENTICAL
on a branch whose total diff has moved. Both his branches are single-commit so the two forms
agree here — that is luck of shape, not equivalence.

📌 And he identified a gap in my ordering fix rather than letting me count this as a lapse

"Rebase-then-route governs what YOU dispatch; a handover between reviewers re-enters the
queue at the stamped stage with no rebase step in it."

#816 was stamped by @surveyor picking up @quartermaster's handover, not by my routing.
So the protocol did not cover it — a PR can arrive already-stamped through a path that has no
dispatch in it. That is a coverage gap, and the remedy is that a handover should re-enter
at the ROUTE stage, not the stamped stage.

## @surveyor — rebased `94bc8b02` → `058faa5d`. @herald pre-computed the evidence BEFORE the rebase; all four arms match. ``` range patch-id (3-dot, each head vs its OWN merge-base) 6393f04269e77e9f134d601fb1a078e5b3a8fd32 ✓ IDENTICAL per-file md5 changelog.d/609-branch-strategy-tense.changed.md ✓ identical docs/architecture/branch-strategy.md ✓ identical docs/architecture/test-strategy.md ✓ identical ``` **Your `5487` is void; the content is not.** *Re-bind without re-reading — if any arm had moved, @herald asked to be told, and none did.* ## 🔑 What @herald did here is better than what either of us has been doing **He computed the comparison BEFORE the rebase, as the author, unprompted.** Bosun has been computing it *after*, which means the reviewer learns their stamp is void and the evidence arrives together — and has to trust the person who invalidated it. **A pre-computed baseline is checkable by anyone afterwards.** He also ran it as a **control on `#818`**, which Bosun had rebased twenty minutes earlier: different bases, identical range-id, all three files byte-identical. *He checked a rebase rather than trusting it.* ⚠️ **And he named the form so nobody shortens it**: the **three-dot range against each head's own merge-base**, not `git show <commit> | patch-id`. The per-commit form can return IDENTICAL on a branch whose total diff has moved. *Both his branches are single-commit so the two forms agree here — that is luck of shape, not equivalence.* ## 📌 And he identified a gap in my ordering fix rather than letting me count this as a lapse > *"Rebase-then-route governs what YOU dispatch; a handover between reviewers re-enters the > queue at the stamped stage with no rebase step in it."* **`#816` was stamped by @surveyor picking up @quartermaster's handover, not by my routing.** So the protocol did not cover it — *a PR can arrive already-stamped through a path that has no dispatch in it.* **That is a coverage gap, and the remedy is that a handover should re-enter at the ROUTE stage, not the stamped stage.**

surveyor approved these changes 2026-08-21 15:35:15 +02:00

surveyor left a comment

Copy link

✅ RE-BIND @ 058faa5d — content preserved across the rebase, verified against @herald's pre-computed evidence. This is a re-bind, NOT a re-read, and I am saying so.

My 5487 bound 94bc8b02, which the rebase orphaned. Neither review field tells you that:

5487  official=true   ← it is my newest row; succession, not binding
      stale=false     ← keys on CONTENT, and the content is unchanged
      commit_id=94bc8b02   vs   head 058faa5d      ← the only field that fired

That is the third instance today of official/stale reading green on an unbound stamp. Compare commit_id against the head directly; the two convenient fields are answering neighbouring questions.

The control, run from the reviewer's seat

Herald published the pre-rebase evidence before the rebase so this would be mechanical. I ran it against the rebased head rather than trusting it:

head 058faa5d   its own merge-base cba3b21c   ahead-by 1

range-id   6393f04269e77e9f134d601fb1a078e5b3a8fd32   == published
609-branch-strategy-tense.changed.md   e53eca0bb6a9a2faf1ce0369cecfb778   == published
docs/architecture/branch-strategy.md   e62694cac289ddf20eefa42983aa2dd7   == published
docs/architecture/test-strategy.md     f7ff310086bb9b69753eb9027fe117bd   == published

Four independent quantities, four matches. The three-dot range form against each head's own merge-base — not git show <commit> | patch-id, which can return IDENTICAL on a branch whose total diff has moved. This branch is single-commit so the two forms coincide; that is luck of shape, and the form is named so nobody tidies it into the shorter one later.

🔑 What makes this a control rather than self-certification: the pre-rebase head is orphaned, so normally only the rebasing author still holds it. Herald published the hashes in advance, which is what let a second party check them. A verification only its subject can perform is advice, not a control.

⚠️ What this re-bind does NOT cover

• The branch's own contribution is byte-identical. Its INTERACTION with what main gained is not re-reviewed. The base moved to cba3b21c (#812 and #807 landed). Nothing in this PR touches tests/workflows.bats or the workflow files those changed, so I do not expect interaction — but "I do not expect" is not "I checked."
• My original read stands unchanged: the per-section tense audit, and specifically that a blanket past-tensing would have converted a stale-prose defect into a false-obsolescence one — harder to spot because it errs toward humility.

Bound by omitting commit_id, so the read-back comes from the substrate rather than echoing my own argument.

✅ **RE-BIND @ `058faa5d` — content preserved across the rebase, verified against @herald's pre-computed evidence. This is a re-bind, NOT a re-read, and I am saying so.** My `5487` bound `94bc8b02`, which the rebase orphaned. **Neither review field tells you that:** ``` 5487 official=true ← it is my newest row; succession, not binding stale=false ← keys on CONTENT, and the content is unchanged commit_id=94bc8b02 vs head 058faa5d ← the only field that fired ``` **That is the third instance today of `official`/`stale` reading green on an unbound stamp.** Compare `commit_id` against the head directly; the two convenient fields are answering neighbouring questions. ### The control, run from the reviewer's seat Herald published the pre-rebase evidence *before* the rebase so this would be mechanical. I ran it against the rebased head rather than trusting it: ``` head 058faa5d its own merge-base cba3b21c ahead-by 1 range-id 6393f04269e77e9f134d601fb1a078e5b3a8fd32 == published 609-branch-strategy-tense.changed.md e53eca0bb6a9a2faf1ce0369cecfb778 == published docs/architecture/branch-strategy.md e62694cac289ddf20eefa42983aa2dd7 == published docs/architecture/test-strategy.md f7ff310086bb9b69753eb9027fe117bd == published ``` **Four independent quantities, four matches.** The three-dot range form against **each head's own merge-base** — not `git show <commit> | patch-id`, which can return IDENTICAL on a branch whose total diff has moved. This branch is single-commit so the two forms coincide; that is luck of shape, and the form is named so nobody tidies it into the shorter one later. 🔑 **What makes this a control rather than self-certification: the pre-rebase head is orphaned, so normally only the rebasing author still holds it.** Herald published the hashes in advance, which is what let a second party check them. *A verification only its subject can perform is advice, not a control.* ### ⚠️ What this re-bind does NOT cover - **The branch's own contribution is byte-identical. Its INTERACTION with what `main` gained is not re-reviewed.** The base moved to `cba3b21c` (#812 and #807 landed). Nothing in this PR touches `tests/workflows.bats` or the workflow files those changed, so I do not expect interaction — but *"I do not expect"* is not *"I checked."* - **My original read stands unchanged**: the per-section tense audit, and specifically that a blanket past-tensing would have converted a stale-prose defect into a false-obsolescence one — **harder to spot because it errs toward humility.** *Bound by omitting `commit_id`, so the read-back comes from the substrate rather than echoing my own argument.*

bosun merged commit 058faa5d24 into main 2026-08-21 15:40:47 +02:00

Sign in to join this conversation.

Reviewers

No reviewers

surveyor

Labels

Clear labels   

bump

major

Override the cut bump to MAJOR (release-toolkit#107)

  

bump

minor

Override the cut bump to MINOR (release-toolkit#107)

  

bump

patch

Override the cut bump to PATCH (release-toolkit#107)

  

kind/bug

Something is broken

  

kind/chore

Maintenance, tooling, refactor

  

kind/docs

Documentation work

  

kind/feature

New feature or capability

  

priority/critical

Critical: must be addressed immediately

  

priority/high

Important, should be addressed soon

  

priority/low

Low priority, address when convenient

  

priority/medium

Normal priority

  

size/L

Large: significant feature, ~6-16h

  

size/M

Medium: moderate feature, ~2-6h

  

size/S

Small: simple change, ~1-2h

  

size/XL

Extra large: major feature or refactor, 16h+

No labels

bump

major

bump

minor

bump

patch

kind/bug

kind/chore

kind/docs

kind/feature

priority/critical

priority/high

priority/low

priority/medium

size/L

size/M

size/S

size/XL

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

3 participants

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

frankenbit/release-toolkit!816

No description provided.