Spec 466: fee cap only — restore 4s block budget - #3182
Conversation
Keep the four-hotkey staking scan fee cap and spec 466. Drop the 12s native block bump and the EVM/scheduler split that existed only for it. Tests that pinned 12s are removed. Co-authored-by: Arbos <unarbos@users.noreply.github.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
@IntiTechnologies spec 466 Propose is the one to approve. Fee cap only. No block-budget change. The earlier 466 train on This merge restores the exact pre-#3180 4s native budget. The 4-hotkey Train: https://github.com/RaoFoundation/subtensor/actions/runs/35292102814 Approve this 466 Propose, not leftover 465 (https://github.com/RaoFoundation/subtensor/actions/runs/35267938998), not leftover 464 Watch (https://github.com/RaoFoundation/subtensor/actions/runs/35262050154), and not the cancelled 12s 466.
After you approve, signers: |
🛡️ AI Review — Skeptic (security review)VERDICT: SAFE LOW contributor risk / baseline scrutiny: write-access maintainer with substantive merged history; no listed Gittensor association. Branch targets main. Reviewed the complete two-file diff. The native block budget returns to 4 seconds; effective scheduler and EVM gas-weight budgets remain unchanged. No dependencies or AI-review trust files changed. Static analysis only; no PR code, builds, tests, or Cargo commands executed. FindingsNo findings. ConclusionNo malicious behavior or introduced security vulnerability found in this rollback. 🔍 AI Review — Auditor (domain review)VERDICT: 👍 Established contributor with write access and substantive merged history; Gittensor association UNKNOWN. Restores the native block budget to 4 seconds while preserving effective scheduler and EVM gas-weight budgets, the staking fee cap, and spec 466. The implementation matches the described rollback. This PR is the better candidate. Recommend closing #3180. Its fee-cap work is already integrated through #3181, while its remaining 12-second budget conflicts with the intended rollback. Other listed overlaps concern separate features.
FindingsNo findings. ConclusionNo substantive correctness issues found in this focused rollback. |
|
🔄 AI review updated — Skeptic: SAFE Auditor: 👍 |
Drops the 12s native block-budget bump from #3181. Keeps the four-hotkey staking scan fee cap and
spec_version466.Native
MAXIMUM_BLOCK_WEIGHT, scheduler, andweight_per_gasare restored to the exact pre-#3180 4s values. The EVM/scheduler split that existed only for the 12s bump is gone.runtime/tests/block_capacity.rs(12s / 65-batch admission pins) is removed.Admission weight is unchanged: a 25-item
transfer_stakebatch still fits; a 65 does not. Clients still chunk. Fees drop.Type of Change
What this is not
mainnet. Propose only.The earlier 466 train on
dfdfba1(12s) is cancelled. Inti should approve this 466 Propose, not leftover 465 or the cancelled 12s 466.