Skip to content

chore(release): prepare v1.0.11 — version constants, CHANGELOG, PLAN (not tagged) - #65

Merged
Shaan-alpha merged 3 commits into
mainfrom
release/v1.0.11-prep
Jul 31, 2026
Merged

chore(release): prepare v1.0.11 — version constants, CHANGELOG, PLAN (not tagged)#65
Shaan-alpha merged 3 commits into
mainfrom
release/v1.0.11-prep

Conversation

@Shaan-alpha

@Shaan-alpha Shaan-alpha commented Jul 31, 2026

Copy link
Copy Markdown
Owner

Everything the v1.0.11 release ritual needs, stopping before the tag. Tagging and publishing stay operator steps.

The six fixes are already merged and prod-verified — this only labels them.

Version constants — all four

File 1.0.10 → 1.0.11
frontend/package.json
frontend/src/lib/site.ts::APP_VERSION
backend/pyproject.toml
backend/app/settings.py::VERSION

uv.lock relocked — it records the root package version, so leaving it would fail v1.0.10's uv lock --check.

This is the first release where both drift guards actually bite. v1.0.7 added one over the frontend pair; #63 added the backend pair this session, after finding pyproject.toml seven releases behind. Both pass at 1.0.11, which is what makes "all four bumped" a verified claim rather than a careful-editing claim.

CHANGELOG

[Unreleased][1.0.11] — 2026-07-31, with a lead paragraph naming the actual cause: enabling Cache Components in v0.8.6 changed leave-and-return behaviour repo-wide, and one root cause was behind four separate user-visible problems. Two ### Internal entries added for the backend version drift and the console-noise-list audit.

Dry-ran release.yml's own awk extraction against the new heading:

--- extracted 15 lines, 2287 bytes ---

Returns the [1.0.11] section and stops cleanly at [1.0.10]. That is the check the v1.0.6 missing-heading incident would have needed, and the empty [Unreleased] heading above it does not interfere.

PLAN.md

Version-map row plus a full slice section. It records what shipped per PR, what was deliberately not done (tag/release; no cross-stack version guard, with the reasoning), and — kept explicitly — the audit finding that was investigated, fixed, and then withdrawn when Chrome contradicted the jsdom evidence.

Exit criteria are ticked except the last, which reads: CHANGELOG.md [1.0.11]; tag v1.0.11; release published. Paused for operator go-ahead.

Unlike v1.0.8–v1.0.10 there is no design spec or plan doc — this began as a bug report mid-session, not a planned slice. The investigation lives in docs/PROGRESS_LOG.md (2026-07-31) and the six PR bodies.

Verification

Frontend vitest 84 passed / tsc clean / eslint clean. Backend pytest -q 361 passed, 72 skipped / ruff clean / uv lock --check clean.

After merging this

git tag v1.0.11 && git push origin v1.0.11 fires release.yml, which extracts the section verified above and publishes the release. Prod already runs this code; the tag is a label, not a deploy.

Note /_/backend/health will still report 1.0.10 until this merges and redeploys — the six fixes shipped under the old version number, which is exactly the drift the guards now prevent going unnoticed.


⚠️ The failing Vercel check is not this diff

Flagging so it is not read as a code problem. It does not block merge — the PR is MERGEABLE / UNSTABLE, i.e. a non-required check is red.

Evidence:

  • Both attempts (9c8a06b, and an empty retrigger commit 38ff586) failed identically: BUILD_FAILED / Resource provisioning failed, state: ERROR, ~1.2s, and zero build-log eventserrorsOnly returns "No error/stderr/exit events in this build." It never reached a build step, so nothing in the repo was executed.
  • Both are target: nullpreview deployments. Production is a separate target and is healthy: the last production deploy (ee885a8, merge of fix(frontend): cache the session server snapshot (2 more 'known noise' entries were real) #64) is READY, and prod was verified live earlier today (site 200, /_/backend/health 1.0.10 db+cache up, /analyze/octocat 200).
  • GitHub has recorded exactly 2 preview-environment deployments for this repo, ever — both of the above. Every one of the last 20 Vercel deployments is target: production. Preview deploys have never succeeded here because they have never run.
  • The diff is four version strings, uv.lock's root version, and two markdown files. It cannot cause resource provisioning to fail.

Retriggering was deliberate — it distinguishes a transient platform blip from a persistent config problem, and it came back persistent. That is why the branch carries an empty chore: retrigger vercel preview commit; it is kept rather than force-pushed away so the diagnostic is visible.

Likely cause, not confirmed: the Preview environment has never been provisioned for this project. It is a multi-service deploy (framework: services, Node + Python) backed by Marketplace resources (Neon, Upstash), and previews need their own environment/env-var scope. That is a Vercel project-settings matter on your account, not something fixable from the repo — so I stopped here rather than guessing at settings.

Bumps all four version constants -- frontend/package.json,
site.ts::APP_VERSION, backend/pyproject.toml, settings.py::VERSION -- and
relocks uv.lock, which records the root package version. Both drift guards
pass at 1.0.11: the v1.0.7 frontend pair and the backend pair added in #63,
which is the first release where the backend half is actually enforced.

Promotes CHANGELOG [Unreleased] to [1.0.11] and adds the PLAN.md version-map
row plus a slice section recording what shipped, what was deliberately not
done, and the finding that was investigated, fixed and then withdrawn.

Dry-ran release.yml's own awk extraction against the new heading: returns
the [1.0.11] section and stops cleanly at [1.0.10]. That is the check the
v1.0.6 missing-heading incident would have needed.

Stops short of the tag and the release -- both are operator steps.
Copilot AI review requested due to automatic review settings July 31, 2026 08:17

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@vercel

vercel Bot commented Jul 31, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
skill-issue Error Error Jul 31, 2026 8:19am

The preview for 9c8a06b failed in 1.2s with BUILD_FAILED / "Resource
provisioning failed" and produced zero build-log events -- it never
reached a build step. Retriggering to distinguish a transient platform
error from a persistent preview-environment config problem. No file
changes.
@Shaan-alpha
Shaan-alpha merged commit 3ef56b5 into main Jul 31, 2026
5 checks passed
@Shaan-alpha
Shaan-alpha deleted the release/v1.0.11-prep branch July 31, 2026 08:34
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants