Repository navigation
chore(release): prepare 2.0.0-dev.2 - #1620
Conversation
Bump TMVersionDefault and the OpenAPI info.version to 2.0.0-dev.2 and add its CHANGELOG section. ABCI (1.5.0), P2P (11) and block (14) protocol versions stay unchanged. <sub>🤖 Co-authored by [Claudius the Magnificent](https://github.com/lklimek/claudius) AI Agent</sub> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WunYYgi2uMsQDcBcKAwLdt
|
Important Review skippedAuto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the
|
|
Waiting for bot review — thepastaclaw not yet. Wait for the missing reviews, or a writer can post |
|
/skip-bots |
|
✅ Final review complete — no blockers (commit 5578d10) · triage: low · Phase 2 only (queue backlog) |
thepastaclaw
left a comment
There was a problem hiding this comment.
Final validation — Phase 1 + Phase 2
The release-version declarations are consistent, and the changelog matches the 15 commits since dev.1. At head 169dd68, however, the blocksync tests cannot compile; current CI confirms that tests (00), go vet, and golangci-lint fail on stale BlockResponse.Block references. This inherited defect blocks validation of the fixes this release promotes and should be repaired before advancing the release; validation here was static, using source inspection and CI logs.
🔴 1 blocking
Review provenance
Source: reviewer 1: glm-5.3-flash (agent: phase1-reviewer, role: general); reviewer 2: gpt-6.1-sol (agent: phase2-reviewer, role: general); final verifier: gpt-6.1-sol (agent: sol-verifier, role: final-verifier)
- Triage:
lowbygpt-6.1-sol(effort low) — The diff is a small, contained prerelease version bump in Go and OpenAPI metadata plus release notes, with no changes to protocol or runtime logic. - Phase 1 reviewers:
glm-5.3-flash— general (completed, effort high); agentphase1-reviewer - Phase 1 model:
glm-5.3-flash— zai quota: 5h 92% left, weekly 59% left; passed overgemini-3.8-flash-high(antigravity quota lookup failed) - Fresh verifier:
gpt-6.1-sol— final-verifier; agentsol-verifier - Phase 2 reviewers:
gpt-6.1-sol— general (completed, effort medium); agentphase2-reviewer
🤖 Prompt for all review comments with AI agents
These findings are from an automated code review (PastaClaw). How to work through them:
- Verify each finding against the code; fix only what is real, as narrowly as the finding.
- If a finding is wrong, decline it on its thread in one or two sentences; the bot answers once and stops.
- To defer one, link a tracking issue on its thread (or reply asking the bot to file one).
- Suggestions don't block the merge, but most are worth fixing in this PR; judge each on its merits.
- Keep the PR's scope; don't grow it to satisfy hardening suggestions.
- Carried items were raised in an earlier round: read the thread first, and don't re-fix what you already fixed, declined or deferred there. Disputed and deferred items are already handled and are not listed.
In `version/version.go`:
- [BLOCKING] version/version.go:12: Repair blocksync test compilation before preparing dev.2
The frozen candidate being promoted here cannot compile its blocksync tests. The internal `BlockResponse` in `internal/blocksync/synchronizer.go:953–970` has `Height` and `BlockBytes`, but `ask_every_peer_test.go:80`, `handover_deadline_test.go:102`, `handover_late_supply_test.go:74`, and several literals in `synchronizer_test.go` still initialize the removed `Block` field. Other cases still read `resp.Block.Height` or `first.Block.Height`.
CI run 43287058351 confirms `unknown field Block in struct literal of type BlockResponse` and `resp.Block undefined`, followed by `FAIL github.com/dashpay/tenderdash/internal/blocksync [build failed]`. The current vet and golangci-lint failures have the same cause. The Linux binary build passes, so this is specifically a test-compilation failure, not evidence that the production binary cannot build.
Although inherited from the base, this is an in-scope release-validation prerequisite: the release explicitly promotes the blocksync and handover fixes, but their package tests cannot execute on this candidate. Migrate the affected tests to `Height` and `BlockBytes`, replace removed-field height accesses with `Height`, and obtain passing test, vet, and lint validation on the resulting candidate before advancing dev.2. The repair can land as a prerequisite on dev followed by refreshed release preparation.
thepastaclaw
left a comment
There was a problem hiding this comment.
Re-review — Final validation — Phase 2 only (queue backlog)
The exact release diff updates Tenderdash, OpenAPI, and changelog metadata consistently to 2.0.0-dev.2; no in-scope correctness, compatibility, security, or documentation defects were found. Build, vet, lint, vulnerability, and relevant integration checks pass; the remaining full test shards, test_apps, and proto check are still pending and do not indicate a failure caused by this diff.
🔴 0 blocking | 🟡 0 suggestion(s) | 💬 0 nitpick(s)
Review provenance
Source: reviewer 1: gpt-6.1-sol (agent: phase2-reviewer, role: general); reviewer 2: gpt-6.1-sol (agent: phase2-reviewer, role: general); final verifier: gpt-6.1-sol (agent: sol-verifier, role: final-verifier)
- Triage:
lowbygpt-6.1-sol(effort low) — This is a small, contained release-version and changelog update with no behavioral or protocol changes in the diff, and its consistency is straightforward to verify. - Phase 1 reviewers: not run (skipped for throughput: 12 PRs queued, above the 10 limit)
- Fresh final gate: an independent Phase-2 review ran after iterative findings were reconciled
- Fresh verifier:
gpt-6.1-sol— final-verifier; agentsol-verifier - Phase 2 reviewers:
gpt-6.1-sol— general (completed, effort medium); agentphase2-reviewer,gpt-6.1-sol— general (completed, effort medium); agentphase2-reviewer
🤖 Prompt for all review comments with AI agents
These findings are from an automated code review. Verify the current code and confirm that no unresolved issues remain.
No unresolved findings remain from the prior review on this head.
|
Policy satisfied — this can merge. |
TL;DR: Prepares the second Tenderdash 2.0 development prerelease,
v2.0.0-dev.2, with the fixes merged todevsincev2.0.0-dev.1.User story
As a Dash Platform developer, I want a versioned prerelease with the latest consensus, block sync and shutdown fixes, to test them against Platform before the stable 2.0.0 release.
Scenario
Base flow
Platform teams pick up Tenderdash through versioned releases and Docker images.
Actual behavior
The fixes merged to
devafterv2.0.0-dev.1are only available as unversioned branch builds.Expected behavior
After merge and tagging,
v2.0.0-dev.2is published as a GitHub prerelease with adashpay/tenderdash:2.0.0-dev.2image. Stable Docker aliases (latest,1,1.8) stay unchanged.Detailed discussion
What was done
version/version.goandrpc/openapi/openapi.yaml:2.0.0-dev.1→2.0.0-dev.2.ABCISemVer(1.5.0),P2PProtocol(11) andBlockProtocol(14) unchanged.CHANGELOG.md: new2.0.0-dev.2section.origin/dev@f49c25d627c7127f7d6a0c68682637d813b5ea66v1.8.2@dd596db0ddbde7e9a23754ddd6dbf6cc482fd9c1(latest stable, reachable)v2.0.0-dev.1(16 commits since)dev, continuation of the 2.0.0 seriesVersion decision (incremental changes since
v2.0.0-dev.1):2.0.0-dev.2. No versioned contract changes incompatibly beyond what 2.0.0 already covers, so the series continues with the next suffix.BlockResponse.blockchanges from an embeddedtendermint.types.Blocktobyteswith the same field number; both encode identically on the wire, so dev.1/1.8 peers and dev.2 peers exchange blocks as before. No new fields.P2PProtocolstays 11.ValidateBlocknow runs beforeProcessProposalinstead of after it; the set of valid blocks is unchanged, invalid ones simply never reach the application. No consensus-rule or block-format change. WAL replay timeliness is local behavior.internal/blocksynctests compile after fix(blocksync): serve blocks as the bytes they were committed with #1595 and fix(blocksync): hand over to consensus at the tip despite unserved peer height claims #1618 met ondev. Not in the changelog.ABCISemVerstays 1.5.0, so Drive compatibility is the same as for dev.1.BlockProtocol14 unchanged.Dash Platform rollout: same as
v2.0.0-dev.1. Additionally, Docker operators should give containers a stop grace period longer than--shutdown-timeout(30s), because the entrypoint now really waits for the shutdown (#1607).After merge: tag
v2.0.0-dev.2at the reviewed merge SHA, create a draft GitHub prerelease from the CHANGELOG section, then publish it (triggersdocker.yml).Testing
go test -count=1 ./version/...(OpenAPI/version equality): OK;gofmt -l version/,go vet ./version/: cleanpython3 -B -m unittest discover -s scripts/release -p test_release.py: OKsh scripts/test-local-gates.sh: passedscripts/release/plan.py --mode next --source f49c25d62 --baseline v1.8.2Not verified here (CI covers the full suite): full race/e2e suites, Platform/Drive integration, mixed-version network upgrade.
Breaking changes
None in the versioned contracts since
v2.0.0-dev.1. Go API: generatedblocksync.BlockResponse.Blockis[]byte.Attribution
Prepared with Claude Code (Claude Opus 5.5) using the
tenderdash-releaseskill.🤖 Generated with Claude Code
https://claude.ai/code/session_01WunYYgi2uMsQDcBcKAwLdt
🤖 Co-authored by Claudius the Magnificent AI Agent
PR Hygiene ·
5578d10CHANGELOG.md,rpc/openapi/openapi.yaml,version/version.go) — approved by lklimekWhen every merge requirement is met, the
PR Hygienecheck passes. Reviewer limits do not block merging; other required GitHub checks and protections still apply.